Split (1)

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

identifier stringlengths 11 32	pdf_url stringlengths 17 4.62k ⌀	lang stringclasses 120 values	error stringclasses 1 value	title stringlengths 2 500 ⌀	source_name stringlengths 1 435 ⌀	publication_year float64 1.9k 2.02k	license stringclasses 3 values	word_count int64 0 1.64M	text stringlengths 1 9.75M
https://openalex.org/W1480549966	http://www.zurnalai.vu.lt/problemos/article/download/5290/3448	Greek, Modern	null	Rinktiniai gnominiai fragmentai	Problemos	2,016	cc-by	1,084	ISSN 1392-1126. PROBLEMOS 2015 87 ISSN 1392-1126. PROBLEMOS 2015 87 IŠ PIRMOSIOS OLIMPINĖS ODĖS (Ol. 1. 28–35) ἦ θαύματα πολλά, καί πού τι καὶ βροτῶν 28b φάτις ὑπὲρ τὸν ἀλαθῆ λόγον δεδαιδαλμένοι ψεύδεσι ποικίλοις ἐξαπατῶντι μῦθοι. Βʹ Χάρις δ’, ἅπερ ἅπαντα τεύχει τὰ μείλιχα θνατοῖς, 31 ἐπιφέροισα τιμὰν καὶ ἄπιστον ἐμήσατο πιστόν ἔμμεναι τὸ πολλάκις· ἁμέραι δ’ ἐπίλοιποι μάρτυρες σοφώτατοι. 35 ἔστι δ’ ἀνδρὶ φάμεν ἐοικὸς ἀμφὶ δαι- μόνων καλά· μείων γὰρ αἰτία. ἦ θαύματα πολλά, καί πού τι καὶ βροτῶν ἦ θαύματα πολλά, καί πού τι καὶ βροτῶν Ir taip daug kas stulbina, bet kur tik marieji šneka, ten, man regis, margaspalviais melais išpuošti pasakojimai suvedžioja, Ir taip daug kas stulbina, bet kur tik marieji šneka, ten, man regis, margaspalviais melais išpuošti pasakojimai suvedžioja, ἦ θαύματα πολλά, καί πού τι καὶ βροτῶν 28b φάτις ὑπὲρ τὸν ἀλαθῆ λόγον δεδαιδαλμένοι ψεύδεσι ποικίλοις ἐξαπατῶντι μῦθοι. Βʹ Χάρις δ’, ἅπερ ἅπαντα τεύχει τὰ μείλιχα θνατοῖς, 31 ἐπιφέροισα τιμὰν καὶ ἄπιστον ἐμήσατο πιστόν ἔμμεναι τὸ πολλάκις· ἁμέραι δ’ ἐπίλοιποι μάρτυρες σοφώτατοι. 35 ἔστι δ’ ἀνδρὶ φάμεν ἐοικὸς ἀμφὶ δαι- μόνων καλά· μείων γὰρ αἰτία. peržengdami tiesą atitinkantį žodį. peržengdami tiesą atitinkantį žodį. Mat Charitė, kuri sumeistrauja visa, kas miela mirtingiesiems, atnešdama garbę, net ir tikėti nevertą dažnukart sumaniai padaro tikėtiną; Βʹ Χάρις δ’, ἅπερ ἅπαντα τεύχει τὰ μείλιχα θνατοῖς, Βʹ Χάρις δ’, ἅπερ ἅπαντα τεύχει τὰ μείλιχα θνατοῖς, 31 ἐπιφέροισα τιμὰν καὶ ἄπιστον ἐμήσατο πιστόν ἔμμεναι τὸ πολλάκις· ἁμέραι δ’ ἐπίλοιποι μάρτυρες σοφώτατοι. ateinančios dienos – išmintingiausi liudytojai. išmintingiausi liudytojai. 35 ἔστι δ’ ἀνδρὶ φάμεν ἐοικὸς ἀμφὶ δαι- μόνων καλά· μείων γὰρ αἰτία. Tinka juk žmogui apie dievybes gerai kalbėti – tada menkesnis kaltinimas. IŠ FRAGMENTŲ KORPUSO 175 61 Stob. ecl. 2. 1. 8 (2. 4 Wachsmuth-Henze) Πίνδαρος παιάνων· τί ἔλπεαι σοφίαν ἔμμεν, ἃν ὀλίγον τοι ἀνὴρ ὑπὲρ ἀνδρὸς ἴσχει; οὐ γὰρ ἔσθ’ ὅπως τὰ θεῶν βουλεύματ’ ἐρευνάσει βροτέᾳ φρενί· 5 θνατᾶς δ’ ἀπὸ ματρὸς ἔφυ. 108ab Epist. Socr. 1. 7 (p. 610 Hercher = 11. 6 Köhl) ἀπειθεῖν δὲ αὐτῷ ὀκνῶ, καὶ τὸν Πίνδαρον ἡγούμενος εἰς τοῦτο εἶναι σοφόν, ὅς φησι· 61 Stob. ecl. 2. 1. 8 (2. 4 Wachsmuth-Henze): iš Pindaro pajanų: Negi išmintimi laikai tai, kuo tiek nedaug vyras vyrą pralenkia? Juk neįmanoma, kad dievų sumanymus mariuoju protu ištirtų tas, kurs kilęs iš mirtingos motinos. 108ab Epist. Socr. 1. 7 (p. 610 Hercher = 11. 6 Köhl): netikėti juo (sc. dievu) nesiryžtu, šiuo klausimu laikydamas išmintingu ir Pindarą, kuris sako: (a) θεοῦ δὲ δείξαντος ἀρχὰν ἕκαστον ἐν πρᾶγος εὐθεῖα δὴ κέλευθος ἀρετὰν ἑλεῖν, τελευταί τε καλλίονες. σχεδὸν γὰρ οὕτω που αὐτῷ ἔχει τὸ ὑπόρχημα. Clem. Alex. strom. 5. 14. 101 (2. 393 Stählin) ὁ μελοποιὸς δέ· (b) θεῷ δὲ δυνατὸν μελαίνας ἐκ νυκτὸς ἀμίαντον ὄρσαι φάος, κελαινεφέι δὲ σκότει καλύψαι σέλας καθαρόν 5 ἁμέρας. (a) Dievui parodžius bet kurio reikalo pradžią, kelias tikrai tiesus sugriebti šaunumą ir atbaiga esti tauresnė. Mat taip maždaug byloja hiporchema. Clem. Alex. strom. 5. 14. 101 (2. 393 Stählin): o giesmių kūrėjas [taip sako]: (b) O dievui įmanoma iš juodos nakties nepalytėtą šviesą iškelti ir juodamigle tamsybe tyrą dienos spindesį paslėpti. 61 Stob. ecl. 2. 1. 8 (2. 4 Wachsmuth-Henze) Πίνδαρος παιάνων· 61 Stob. ecl. 2. 1. 8 (2. 4 Wachsmuth-Henze): iš Pindaro pajanų: (b) 175 131b Plut. consol. ad Apoll. 35. 120c: 131b Plut. consol. ad Apoll. 35. 120c 5 137 ιππορατει? αθηναιωι Clem. Alex. strom. 3. 3. 17 (2. 203 Stählin) Πίνδαρος περὶ τῶν ἐν Ἐλευσῖνι μυστηρίων λέγων ἐπιφέρει· ὄλβιος ὅστις ἰδὼν κεῖν’ εἶσ’ ὑπὸ χθόν’· οἶδε μὲν βίου τελευτάν, οἶδεν δὲ διόσδοτον ἀρχάν. 137 137 hipokratui atėniečiui (?) Clem. Alex. strom. 3. 3. 17 (2. 203 Stählin): kalbėdamas apie Eleusino misterijas, Pindaras priduria: Palaimingas tasai, kurs tai išvydęs eina po žeme: žino gyvenimo atbaigą, žino ir Dzeuso duotą pradžią. 140d Clem. Alex. strom. 3. 3. 17 (2. 203 Stählin) Ir giesmių kūrėjas Pindaras, tarytum bakchiško siautulio apimtas, tiesiai sako: Kas yra dievas? Visa kas. 159 Dion. Hal. de or. ant. 2 (1. 4. IŠ FRAGMENTŲ KORPUSO 20 Usener- Radermacher): mat juk ne tik geriausias teisiųjų vyrų gelbėtojas – Laikas, pasak Pindaro, bet ir menų […] ir bet kurio kito rimto dalyko. 140d Clem. Alex. strom. 3. 3. 17 (2. 203 Stählin) Πίνδαρός τε ὁ μελοποιὸς οἷον ἐκβακχεύεται, ἄντικρυς εἰπών· τί θεός; τὸ πᾶν. 159 Dion. Hal. de or. ant. 2 (1. 4. 20 Usener- Radermacher) ἀλλὰ γὰρ οὐ μόνον ἀνδρῶν δικαίων Χρόνος σωτὴρ ἄριστος κατὰ Πίνδαρον, ἀλλὰ καὶ τεχνῶν … καὶ παντὸς ἄλλου σπουδαίου χρήματος. 176 209 Stob. ecl. 2. 1. 21 (2. 7 Wachsmuth-Henze) Πινδάρου· τοὺς φυσιολογοῦντας ἔφη Πίνδαρος· ἀτελῆ σοφίας καρπὸν δρέπ(ειν). 210 Plut. de cohib. ira 8. 457b χαλεπώτατοι δὲ (κατὰ Πίνδαρον) ἄγαν φιλοτιμίαν μνώμενοι ἐν πόλεσιν ἄνδρες· ἱστᾶσιν ἄλγος ἐμφανές. 217 Clem. Alex. paedag. 3. 72. 1 (1. 275 Stählin) Πίνδαρος· γλυκύ τι κλεπτόμενον μέλημα Κύπριδος. 209 Stob. ecl. 2. 1. 21 (2. 7 Wachsmuth-Henze) Pindaro: gamtos tyrinėtojai, pasak Pindaro, neprinokusį išminties vaisių raško. 210 Plut. de cohib. ira 8. 457b o nepakenčiamiausi (pasak Pindaro) Garbės iš miestų pernelyg trokštą vyrai didžiai pavojingi: jie sukelia akivaizdų skausmą. 217 Clem. Alex. paedag. 3. 72. 1 (1. 275 Stählin) Pindaras [sako]: Nežinia kodėl vagiamas Kipridės numylėtinis yra saldus. 225 Schol. Pind. Ol. 2. 42e πρὸ τῶν ἀγαθῶν τοῖς ἀνθρώποις τὰ κακά· ὅπερ καὶ ἐν ἑτέρῳ (-ροις v.l.) φησίν· ὁπόταν θεὸς ἀνδρὶ χάρμα πέμψῃ, πάρος μέλαιναν καρδίαν ἐστυφέλιξεν. 233 Clem. Alex. paedag. 3. 12. 92. 4 (1. 286 Stählin) (κατὰ Πίνδαρον) πιστὸν δ’ ἀπίστοις οὐδέν. 225 Schol. Pind. Ol. 2. 42e prieš gerus dalykus žmonėms nutinka blogi, kaip ir kitame kūrinyje sako [Pindaras]: Kai tik dievas vyrui siunčia džiaugsmą, pirm to smogia juodon širdin. 233 Clem. Alex. paedag. 3. 12. 92. 4 (1. 286 Stählin) (pasak Pindaro) Nepatikimiems nėra nieko patikimo. Graikiškas tekstas parengtas pagal: Pindar, Vol. 1: Olympian Odes, Pythian Odes (Loeb Classical Library 56); Vol. 2: Nemean Odes, Isthmian Odes, Fragments (Loeb Classical Library 485), edited and translated by William H. Race, Cambridge, MA, London: Harvard University Press, 1997. Graikiškas tekstas parengtas pagal: Pindar, Vol. 1: Olympian Odes, Pythian Odes (Loeb Classical Library 56); Vol. 2: Nemean Odes, Isthmian Odes, Fragments (Loeb Classical Library 485), edited and translated by William H. Race, Cambridge, MA, London: Harvard University Press, 1997. Iš sen. graikų kalbos vertė filosofinio teksto vertimo seminaro dalyviai: Vytautas Ališauskas, Viktoras Bachmetjevas, Paulius Garbačiauskas, Elzė Sigutė Mikalonytė, Jonas Skarulskis, Julita Slipkauskaitė, Gendvilė Svirskaitė ir Gintarė Vaičiulytė Mokslinis redaktorius Paulius Garbačiauskas Iš sen. IŠ FRAGMENTŲ KORPUSO graikų kalbos vertė filosofinio teksto vertimo seminaro dalyviai: Vytautas Ališauskas, Viktoras Bachmetjevas, Paulius Garbačiauskas, Elzė Sigutė Mikalonytė, Jonas Skarulskis, Julita Slipkauskaitė, Gendvilė Svirskaitė ir Gintarė Vaičiulytė 177
https://openalex.org/W2997569865	https://ieeexplore.ieee.org/ielx7/6287639/8600701/08926351.pdf	English	null	Practical Automated Video Analytics for Crowd Monitoring and Counting	IEEE access	2,019	cc-by	8,534	Received November 6, 2019, accepted December 2, 2019, date of publication December 6, 2019, date of current version December 27, 2019. Received November 6, 2019, accepted December 2, 2019, date of publication December 6, 2019, date of current version December 27, 2019. Digital Object Identifier 10.1109/ACCESS.2019.2958255 Digital Object Identifier 10.1109/ACCESS.2019.2958255 KANG HAO CHEONG 1,2, (Member, IEEE), SANDRA POESCHMANN 3, JOEL WEIJIA LAI 1,2, JIN MING KOH 1, U. RAJENDRA ACHARYA 4, SIMON CHING MAN YU 5, AND KENNETH JIAN WEI TANG 1 KANG HAO CHEONG 1,2, (Member, IEEE), SANDRA POESCHMANN 3, JOEL WEIJIA LAI 1,2, JIN MING KOH 1, U. RAJENDRA ACHARYA 4, SIMON CHING MAN YU 5, AND KENNETH JIAN WEI TANG 1 1Science and Math Cluster, Singapore University of Technology and Design (SUTD), Singapore 487372 2SUTD-MIT International Design Centre, Singapore 487372 3Engineering Cluster, Singapore Institute of Technology, Singapore 138683 4Department of Electronics and Computer Engineering, Ngee Ann Polytechnic, Singapore 599489 5Interdisciplinary Division of Aeronautical and Aviation Engineering, The Hong Kong Polytechnic University, Hong Kong Corresponding author: Kang Hao Cheong (kanghao_cheong@sutd.edu.sg) This work was supported by the SUTD-MIT International Design Centre (IDC) under Grant IDG21900101 and G 1Science and Math Cluster, Singapore University of Technology and Design (SUTD), Singapore 487372 2SUTD-MIT International Design Centre, Singapore 487372 3Engineering Cluster, Singapore Institute of Technology, Singapore 138683 4Department of Electronics and Computer Engineering, Ngee Ann Polytechnic, Singapore 599489 5Interdisciplinary Division of Aeronautical and Aviation Engineering, The Hong Kong Polytechnic University, Hong Kong Corresponding author: Kang Hao Cheong (kanghao_cheong@sutd.edu.sg) This work was supported by the SUTD MIT International Design Centre (IDC) under Grant IDG21900101 and This work was supported by the SUTD-MIT International Design Centre (IDC) under Grant IDG21900101 and Grant IDIN190 ABSTRACT Video surveillance is gaining popularity in numerous applications, including facility man- agement, trafﬁc monitoring, crowd analysis, and urban security. Despite the increasing demand for closed- circuit television (CCTV) and related infrastructure in public spaces, there remains a notable lack of readily- deployable automated surveillance systems. In this study, we present a low-cost and efﬁcient approach that integrates the use of computational object recognition to perform fully-automated identiﬁcation, tracking, and counting of human trafﬁc on camera video streams. Two software implementations are explored and the performance of these schemes is compared. Validation against controlled and non-controlled real-world environments is also demonstrated. The implementation provides automated video analytics for medium crowd density monitoring and tracking, eliminating labor-intensive tasks traditionally requiring human operation, with results indicating great reliability in real-life scenarios. INDEX TERMS Crowd monitoring, counting, trafﬁc monitoring, data analytics, background subtraction, security. his work is licensed under a Creative Commons Attribution 4.0 License. For more information, see http://creativecommons.org/licenses/by/4.0/ The associate editor coordinating the review of this manuscript and approving it for publication was Dongxiao Yu. reative Commons Attribution 4.0 License. For more information, see http://creativecommons.org/licenses/by/4.0/ VOLUME 7, 2019 I. INTRODUCTION This stands to yield operational cost reductions and produc- tivity improvements. Video surveillance is an integral component of modern urban security, and when coupled with computational analytics, can have greatly expanded functionality including facial recog- nition, motion detection, trafﬁc and crowd monitoring, and automated hazard alarms [1]–[7]. The continued advance- ment in computational tools and machine learning has in principle enabled automation of a wide variety of practical analyses on image and video inputs [8]–[14]; more advanced machine intelligence systems are also increasingly capable of fulﬁlling traditionally human-controlled tasks that require real-time complex decisions, for instance initiating mitiga- tion measures for severe trafﬁc congestion or the dispatch- ing of emergency services [15]–[18]. In general, automated surveillance eliminates the need for round-the-clock manual monitoring, thereby reducing manpower requirements [19]. Nonetheless, amidst the progressing state-of-the-art, inte- gration of automated analytics in commercial video surveil- lance for crowd monitoring and counting is an area that can be further explored [19]; and there is at present lim- ited literature on demonstrated effective low-cost systems for deployment. In security and management sectors, there remains a great reliance on traditional manual monitoring of CCTV footage [20], [21], and human patrols to conduct crowd monitoring and tracking. Utilizing computer vision and real-time automated analytics in replacement of manual labour not only reduces operational costs but also eliminates human errors and lapses [22]–[24] —we seek to develop a viable deployment-ready implementation in this study. In this paper, we examine several viable approaches to automated crowd monitoring and tracking in indoor and outdoor scenarios, ultimately selecting a statistical back- ground subtraction (BGS) scheme and a convolutional neural The associate editor coordinating the review of this manuscript and approving it for publication was Dongxiao Yu. 183252 VOLUME 7, 2019 VOLUME 7, 2019 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting Existing studies have shown excellent performance in the identiﬁcation of key markers, such as lane boundaries, traf- ﬁc signs, and pedestrians on systems intended for driver- less vehicles [36]–[39] under a wide range of lighting and driving conditions. CNN-based image recognition has also been applied very successfully to facial identiﬁcation tasks [40]–[44], achieving large reductions in error rate when compared to non-CNN methods. These types of CNN-based methods are presently employed for automated user iden- tiﬁcation and tagging systems in prominent social media platforms [45]. network-based single shot detector (SSD). III. METHODOLOGY h f k The software package developed in this study comprises a video processing back-end encompassing human subject recognition and tracking, and a front-end graphical interface for operators. The software implementation is broadly dis- cussed in Section III-A, with object recognition methods in Sections III-B1–III-B2, and lastly tracking and counting tech- niques in Sections III-C–III-D. A block diagram summarizing the video tracking and counting process is given in Figure 1. I. INTRODUCTION These methods are easily deployable in the real world. A software solution is developed for use in general public spaces, and we validate the performance of the platform through indoor controlled tests in a shopping mall, and outdoor non-controlled tests in a public transport hub with considerable human trafﬁc. It is noted that the implementation of this video analytics system can also be applied to a broader range of scenarios— for instance, in factories to detect personnel in restricted places or in dangerous proximity to machinery, or in high-rise buildings to detect crowd densities that exceed safe thresholds for timely evacuation in case of emergencies. Time-oriented data collected from such deployments can also be logged and transmitted to a dashboard for data-informed planning and predictive analytics [25]. In general, CNN-based systems are hugely robust to changing background conditions and object appearances, but are typically computational expensive to train and run. In comparison to explicit rule-based or statistical methods, the employed CNN architectures are also more akin to black boxes and offer limited tractability—troubleshooting and tun- ing the systems for speciﬁc environments can therefore be challenging. There is also recent evidence attributing the efﬁ- cacy of neural network deep-learning solutions to ﬁne-tuning rather than a fundamental architectural advantage, suggesting that a properly tuned classical method may be able to achieve similar performance in certain scenarios [46]–[50]. The structure of the paper is as follows—a technical review is ﬁrst provided (Section II), followed by a discussion on the methodology employed and the development of the soft- ware solution (Section III), and ﬁnally validation test results (Section IV) and concluding remarks (Section V). II. TECHNICAL REVIEW II. TECHNICAL REVIEW We ﬁrst provide an overview of object recognition frame- works, the challenges associated with achieving satisfactory performance, and the application of these frameworks in auto- mated video analytics. A fundamental operational require- ment of automated video surveillance analytics is the ability to identify and track different objects within the recorded footage, hence the need for object recognition; in the current context of crowd analysis, recognition of human subjects is critically relevant. Indeed, non-CNN methods have been deployed to perform similar tasks. A real-time system for pedestrian tracking using gray-scale images from stationary cameras has been demon- strated [51], with satisfactory robustness to visual occlu- sions and ambiguities in perceived subject shape proﬁles. The implementation relied on Gaussian-mixture foreground masking followed by contour detection through a princi- ple component analysis (PCA) model. Numerous studies on pedestrian and trafﬁc tracking, motion detection and analy- sis, and object classiﬁcation using non-CNN methods have also been presented to date [30], [52]–[55], suggesting good viability in these approaches. Non-CNN methods may hence be preferred in some scenarios. While visual recognition and classiﬁcation of objects is intuitive to human perception, robust computational imple- mentation is challenging [26]. Varying exposure to outdoor conditions, changing illumination levels and direction, inter- mittent and sustained visual obstruction, and unpredictable movement of tracked subjects must all be overcome for reliable operation of recognition systems, oftentimes with limitations on available computational power [27]–[30]. The resolution and clarity of available video footage is also typ- ically non-ideal, limiting the effectiveness of pre-processing techniques aimed at compensating for variance in image con- ditions. In our context of recognition and tracking of human subjects, additional complexities arise from the wide range of possible dynamical behaviour—for instance, two persons in physical contact may be detected as a single entity, and the shape proﬁle of a person may change drastically because of carried items or differing attire. B. OBJECT RECOGNITION We examine two categories of object recognition methods in detail—background subtraction and CNN-based image classiﬁers. A comparison of the performance between cho- sen variants of these two methods in controlled and non- controlled test environments is later presented in Section IV. A. SOFTWARE IMPLEMENTATION Advanced machine vision systems are already being devel- oped for security- and safety-critical applications, such as driverless vehicles and autonomous drones [31]–[35]. These systems typically employ convolutional neural net- work (CNN)-based solutions that are trained on massive datasets of numerous modalities, including infrared and vis- ible video input, lidar data, sound pick-ups from micro- phones, and navigational data from GPS or inertial guidance. Our software package is implemented on Python with the Open Source Computer Vision (OpenCV) library. OpenCV supports machine deep-learning frameworks, and provides image manipulation, object identiﬁcation, and motion track- ing tools that are greatly relevant for the development of soft- ware in our context [56], [57]. Our speciﬁc implementation assumes a pre-existing video surveillance system that writes to a centralized storage pool, from which footage may be 183253 183253 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting FIGURE 1. Block diagram depicting data flow in the adopted video analytics pipeline. The input undergoes either BGS or SSD, yielding human personnel identification and count. FIGURE 1. Block diagram depicting data flow in the adopted video analytics pipeline. The input undergoes either BGS o SSD, yielding human personnel identification and count. pulled in real-time for analysis; as such all functionalities are developed and tested in a stream-based format. The soft- ware implementation accommodates video streams of general frame size and rate, but we use footage of 720p at 30 fps for illustrative purposes in this paper, unless otherwise stated. can all diminish the ability of the reference frame subtraction to separate background and foreground elements. A number of methods to overcome these problems have been utilized to date. Pre-processing of footage frames to remove glare and illumination changes can be utilized; major changes in background and ambient conditions can be detected through regression against a history of frames, and the background reference either adjusted or recaptured at opportune times; a comprehensive set of background references can also be captured a priori against possible ambient conditions, selec- tions of which are subtracted from footage frames on a trial basis until a sufﬁciently clean output is produced. Movement patterns of detected objects across numerous frames can also be used as an additional ﬁlter against false positives— for instance, human subjects must realistically be in con- tact, or otherwise close proximity, with the ground at all times. 1) BACKGROUND SUBTRACTION Pilot study results comparing the performance of YOLO and SSD object recognition schemes, suggesting vastly cheaper computation FIGURE 3. Frame comparison of the MOG, MOG2 and GMG background subtraction schemes, showing a cleaner result from MOG2. FIGURE 2. Illustration of the functioning of BGS. (a) An empty scene as the background mask; (b) a scene with human subjects in the foreground; and (c) the scene after background subtraction, separating the human subjects. FIGURE 3. Frame comparison of the MOG, MOG2 and GMG background subtraction schemes, showing a cleaner result from MOG2. FIGURE 2. Illustration of the functioning of BGS. (a) An empty scene as the background mask; (b) a scene with human subjects in the foreground; and (c) the scene after background subtraction, separating the human subjects. FIGURE 4. Pilot study results comparing the performance of YOLO and SSD object recognition schemes, suggesting vastly cheaper computation using SSD. a background model, which is then used for backgound subtraction. BGS is not completely successful due to background ﬂuc- tuations, and differentiation between human and non-human subjects may not be ideal. We note that GMG is limited in suitability for outdoor scenes with constant motion of objects, as there is often no dedicated time periods available for background capture. Deploying GMG in the intended use case of automated crowd analysis in busy public spaces is therefore not viable. On the other hand, there is evidence in existing literature that MOG2 provides good results in practice [64], [65], and is also sufﬁciently computationally cheap to deploy in large- scale video analysis applications. Our own pilot programme also suggests that MOG2 consistently produces better human subject identiﬁcation and segmentation results than MOG and GMG in indoor use cases (example in Figure 3). We therefore employ MOG2 in our BGS-based implementation. 1) BACKGROUND SUBTRACTION A widely used method for detecting moving objects from a stationary camera placement is background subtraction (BGS) [58]. In general, the operation of such a method relies on a known background frame with no present objects. This background reference is then subtracted from each frame of the video footage, or subset of frames to reduce com- putational cost, therefore yielding frames containing only foreground objects. Appropriate contour detection or region segmentation models can then be applied to isolate distinct objects in these frames. An example illustrating the operation of BGS is presented in Figure 2. BGS-based methods are presently applied in commercial video surveillance systems for malls and public spaces. An early variant of a BGS-based object recognition frame- work is the Mixture of Gaussians (MOG) method introduced in 2001, utilizing a Gaussian mixture background/foreground segmentation algorithm [59], [60]. An improved version, named MOG2, was later presented, with a signiﬁcant improvement being an automatic selection scheme for the number of Gaussian kernels used for each pixel, in place of the constant number of distribution kernels in the orig- inal MOG [61], [62]. As a result, MOG2 provides better adaptability to changing illumination conditions in scenes. A more recent algorithm is the GMG [63], named after its founders, which combines statistical background image esti- mation and per-pixel Bayesian segmentation. GMG uses the ﬁrst few hundred frames of the input footage to construct Simplistic implementations of BGS typically suffer from limited reliability, due mostly to changing background condi- tions. In an outdoor environment, volatile weather, illumina- tion changes, and reﬂections from surfaces on moving objects 183254 183254 VOLUME 7, 2019 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting FIGURE 3. Frame comparison of the MOG, MOG2 and GMG background subtraction schemes, showing a cleaner result from MOG2. FIGURE 4. Pilot study results comparing the performance of YOLO and SSD object recognition schemes, suggesting vastly cheaper computation using SSD. FIGURE 2. Illustration of the functioning of BGS. (a) An empty scene as the background mask; (b) a scene with human subjects in the foreground; and (c) the scene after background subtraction, separating the human subjects. FIGURE 3. Frame comparison of the MOG, MOG2 and GMG background subtraction schemes, showing a cleaner result from MOG2. FIGURE 4. C. TEMPORAL TRACKING counts are recorded together with cumulative counts (start- ing from a speciﬁed time, say, the start of each day). This mode is useful if on-scene subject counts are of interest, for instance, to monitor the number of people in an enclosed space, or congestion conditions along corridors and passage- ways. Alternatively, counting can be set to be portal-wise, that is, subjects are added to a cumulative count only when they cross a speciﬁed boundary in the captured scene. This mode is useful to monitor crowd inﬂux or outﬂux through key doorways or area perimeters, for instance, in tracking the boarding of public buses or commuter movement through security checkpoints. Analyzing incoming video streams frame-wise does not guar- antee temporal continuity in the identiﬁed subjects, and there- fore cannot immediately support counting functions. In order to support reliable counting of moving subjects, a temporally- consistent labelling of subjects must be achieved between frames, such that distinct objects are not misidentiﬁed as being identical (leading to under-counting), and identical objects are not misidentiﬁed as being distinct (leading to over-counting). In essence, objects undergoing movement has to be continuously tracked across all frames in which they appear. We implement this by comparing the centroids of iden- tiﬁed bounding boxes for each frame against those of the previous, and labelling pairs as identical on a nearest-distance basis. Centroids in the previous frame that have no matching counterpart in the current are deemed to have left the scene, and centroids in the current frame that have no match in the previous are deemed as new subjects that have entered. This is illustrated in Figure 6. To cope with subject occlusions and intermittent image quality issues, a loss-of-visibility thresh- old of nlv = 18 frames is set, such that if a subject disappears from view within the scene and reappears within nlv frames in a location deemed to be matching by the nearest-distance scheme, a new subject identiﬁcation will not be assigned and the reappeared subject will be considered identical to the previous. A movement rate threshold can also be set (say, to the typical human running speed), such that subjects that move faster than expected between frames are identiﬁed as distinct. In the scene-wise mode, extremal boundaries can be set by the user on the video scene, only within which count- ing is active. 2) CONVOLUTIONAL NEURAL NETWORKS We consider CNN-based recognition frameworks You-Only- Look-Once (YOLO) [66], a state of the art real-time object detection system shown to be capable of identifying and classifying objects effectively, and the Single Shot Detector (SSD) [67], also a highly-established method. Both of these frameworks run the full incoming frames through a CNN in a region-wise manner to yield bounding boxes and class proba- bilities on identiﬁed objects. These CNNs are pre-trained on large imagery datasets. A pilot programme had been carried out to compare the computational running cost of YOLO and SSD on preliminary hardware (Figure 4), revealing that SSD is considerably faster in processing incoming frames, and is therefore more viable in achieving real-time stream-based automated video analysis with. We choose SSD as the pre- ferred CNN-based solution. We utilize MobileNet supported on the deep neural network (DNN) module of OpenCV for SSD implementation. MobileNet provides pre-trained CNNs for image classiﬁcation, and can robustly handle frames of different aspect ratios and sizes. A three-step processing pipeline is utilized for incoming video frames. First, a brightness and colour-correction ﬁlter is applied to adjust for under- or over-exposure, or changing daylight conditions. Noise reduction is also used to improve image quality. Next, a background subtraction mask is com- puted and applied through MOG2. Lastly, contours on the masked image are identiﬁed through contrast segmentation, to extract the bounding boxes and positions of the human subjects. This process is computationally cheap, but face potential limitations—detection accuracy is compromised if 183255 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting FIGURE 6. Illustration of the subject tracking mechanism. (a) Two subjects are recognized in the scene and labelled with distinct identification numbers, and the centroids of their bounding boxes are computed; (b) the subjects have moved in the next frame, and the new positions are compared against the previous; (c) identifications are made on a nearest-distance basis; (d) identification numbers are consistently applied, with a new subject deduced to have entered the scene. Images adapted from [70]. FIGURE 5. Recognition of a moving human subject through SSD. FIGURE 5. Recognition of a moving human subject through SSD. Speciﬁcally, the model utilized in this paper is an imple- mentation of Google’s MobileNet SSD [68], which was ini- tially trained on the Common Objects in Context (COCO) dataset [69]. The model was further reﬁned on PASCAL VOC0712 [68]. E. DATA PROCESSING It is important to keep the analysis low-cost, as practical deployment hardware may be limited in speed, especially when there are multiple video streams sharing compute time. To reduce computational load, the program can be conﬁgured to perform BGS or SSD object identiﬁcation only every N0 frames; in our implementation, subject tracking and counting is set to occur every N0 = 6 frames, corresponding to a ∼ 200 ms refresh rate. These limits were found to be sufﬁcient in providing good tracking results for typical pedestrian traf- ﬁc encountered in our test environments (Section IV), and can 2) CONVOLUTIONAL NEURAL NETWORKS A sample snapshot of the execution of this implementation on real-world footage is presented in Fig- ure 5, on which a moving human subject with headwear is identiﬁed with > 99% conﬁdence. FIGURE 6. Illustration of the subject tracking mechanism. (a) Two subjects are recognized in the scene and labelled with distinct identification numbers, and the centroids of their bounding boxes are computed; (b) the subjects have moved in the next frame, and the new positions are compared against the previous; (c) identifications are made on a nearest-distance basis; (d) identification numbers are consistently applied, with a new subject deduced to have entered the scene. Images adapted from [70]. C. TEMPORAL TRACKING Subject tracking is maintained throughout the entire scene regardless of boundary settings, however, so that identical subjects repeatedly crossing the set boundaries will not be misidentiﬁed as distinct instances, so long as they remain within the scene throughout. In the portal-wise mode, the user may select from a default preset of 10 regularly spaced boundary lines, or freely draw a desired boundary. Movement direction ﬁltering can also be set, such that only subjects moving to the left or right are counted. A. CONTROLLED ENVIRONMENT An indoor area in a shopping mall, in proximity to a lift lobby, was used as an indoor controlled environment. A number of subjects, dressed in varying attire, was sent to walk across the captured space at varying speeds, ranging from a slow walk typical of the elderly to fast jogs. These subjects are of mixed genders and of varied heights. The controlled studies were conducted with steady artiﬁcial lighting and primarily constant environmental parameters; each test lasted a dura- tion of 150 seconds with both the BGS and SSD methods, and a manual on-site count was performed simultaneously to match the results against. Snapshots of the controlled validation tests are shown in Figure 7, and subject counting results are presented in Figure 8. These results indicate sat- isfactory counting accuracy for both BGS and SSD methods, with BGS notably achieving perfect accuracy in the idealized single-subject scenario, but is ultimately outperformed by SSD in more realistic multiple-subject scenarios. The SSD method yielded a maximum of a single miscount in these controlled tests, suggesting good deployment viability in the signiﬁcantly more demanding non-controlled outdoor envi- ronments. FIGURE 8. Subject counting results in the controlled environment comparing BGS and SSD methods, in (a) the constrained case of a single subject in-scene at any point in time, and (b) with multiple subjects in-scene simultaneously. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. FIGURE 8. Subject counting results in the controlled environment comparing BGS and SSD methods, in (a) the constrained case of a single subject in-scene at any point in time, and (b) with multiple subjects in-scene simultaneously. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. be adjusted for different hardware capabilities. While such an approach effectively reduces the imposed computational load per video stream, the trade-off between practicality and accuracy stands to be further characterized; an important line of development for future work is also in studying alternative approaches that does not impact analysis frame rate. be adjusted for different hardware capabilities. B. NON-CONTROLLED ENVIRONMENT Non-controlled validation tests were performed at a public transport hub with considerable human trafﬁc. The camera placements were chosen for the purpose of tracking com- muter volume boarding public buses at various terminals, suitable for assessing the ability of the system to cope with massive crowd surges. The tests were conducted over typ- ical bus-boarding durations of approximately 20 seconds, with both BGS and SSD used, and a simultaneous manual count to match results against. Sample snapshots showing the recorded environments are presented in Figure 9, and a comparison of subject counting results is given in Figure 10. It was observed that SSD yields > 92% accuracy (mean square error), illustrating the robustness of SSD in handling large crowd densities and volatile outdoor illumination con- ditions. The obvious failure of BGS in comparison to the con- trolled tests can be attributed to its inability to handle rapidly Non-controlled validation tests were performed at a public transport hub with considerable human trafﬁc. The camera placements were chosen for the purpose of tracking com- muter volume boarding public buses at various terminals, suitable for assessing the ability of the system to cope with massive crowd surges. The tests were conducted over typ- ical bus-boarding durations of approximately 20 seconds, with both BGS and SSD used, and a simultaneous manual count to match results against. Sample snapshots showing the recorded environments are presented in Figure 9, and a comparison of subject counting results is given in Figure 10. D. SUBJECT COUNTING The counting of identiﬁed subjects can be set to be scene- wise—subjects are counted the moment they enter the scene captured by the video camera, and real-time on-scene subject 183256 VOLUME 7, 2019 183256 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting FIGURE 7. Video snapshots of the controlled environment tests, for (a) a constrained case of a single subject in-scene at any point in time, and (b) multiple subjects in-scene simultaneously. FIGURE 7. Video snapshots of the controlled environment tests, for (a) a constrained case of a single subject in-scene at any point in time, and (b) multiple subjects in-scene simultaneously. FIGURE 8. Subject counting results in the controlled environment comparing BGS and SSD methods, in (a) the constrained case of a single subject in-scene at any point in time, and (b) with multiple subjects in-scene simultaneously. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. (Section IV-B) for validation. In these validation tests, portal- wise subject counting mode was utilized. (Section IV-B) for validation. In these validation tests, portal- wise subject counting mode was utilized. A. CONTROLLED ENVIRONMENT While such an approach effectively reduces the imposed computational load per video stream, the trade-off between practicality and accuracy stands to be further characterized; an important line of development for future work is also in studying alternative approaches that does not impact analysis frame rate. F. COMPUTATIONAL RESOURCE Reasonable computational cost is a requisite for viable deployability in the real-world—a considerably modest workstation was hence used for testing purposes. The work- station was a laptop equipped with an Intel I5 8250U quad- core processor at 1.6 GHz (base), 8 GB of RAM, and an Intel UHD Graphics 620 graphics processor, running the Ubuntu (Linux) operating system. On this platform, the SSD method runs in real-time on the CPU. It was observed that SSD yields > 92% accuracy (mean square error), illustrating the robustness of SSD in handling large crowd densities and volatile outdoor illumination con- ditions. The obvious failure of BGS in comparison to the con- trolled tests can be attributed to its inability to handle rapidly IV. RESULTS & DISCUSSION IV. RESULTS & DISCUSSION The developed video analysis software was tested in two types of environments—ﬁrst a controlled environment (Section IV-A), and then a non-controlled environment 183257 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting FIGURE 9. Video snapshots of the non-controlled environment tests. (a) Full scene captured by the camera; (b) zoomed snapshot of boarding queue onto a public bus; and (c) zoomed snapshot of the queue at a later point in time. FIGURE 9. Video snapshots of the non-controlled environment tests. (a) Full scene captured by the camera; (b) zoomed snapshot of boarding queue onto a public bus; and (c) zoomed snapshot of the queue at a later point in time. FIGURE 9. Video snapshots of the non-controlled environment tests. (a) Full scene captured by the camera; (b) queue onto a public bus; and (c) zoomed snapshot of the queue at a later point in time. FIGURE 10. Subject counting results in the non-controlled environment, comparing BGS and SSD methods. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. FIGURE 10. Subject counting results in the non-controlled environment, comparing BGS and SSD methods. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. FIGURE 11. Separated foreground elements by BGS, corresponding to the input frame shown in Figure 9(c). The fragmented masking of human subjects is clearly observed. implementing BGS and SSD methods suitable for deploy- ment for crowd monitoring in public spaces. Real-world validation of our solution has been carried out with both controlled and non-controlled tests, and the results strongly indicate good accuracy of the SSD system, even in outdoor conditions. This yields great conﬁdence in expanding the deployment of the developed system into other venues. Our proposed automated video analytics for crowd mon- itoring and tracking will enable signiﬁcant manpower sav- ings, especially in key security-sensitive installations such as public transport facilities and protected areas, where CCTV monitoring is oftentimes performed by human operators. Data collection of crowd density and movement can be performed more consistently and with better accuracy than otherwise achievable with manual monitoring. REFERENCES [1] A. Hampapur, L. Brown, J. Connell, A. Ekin, N. Haas, M. Lu, H. Merkl, and S. Pankanti, ‘‘Smart video surveillance: Exploring the concept of multiscale spatiotemporal tracking,’’ IEEE Signal Process. Mag., vol. 22, no. 2, pp. 38–51, Mar. 2005. [2] T. Ko, ‘‘A survey on behavior analysis in video surveillance for homeland security applications,’’ in Proc. 37th IEEE Appl. Imag. Pattern Recognit. Workshop, Oct. 2008, pp. 1–8. [3] A. C. Caputo, Digital Video Surveillance and Security. Oxford, U.K.: Butterworth-Heinemann, 2014. [4] P. Remagnino, G. A. Jones, N. Paragios, and C. S. Regazzoni, Video- Based Surveillance Systems: Computer Vision and Distributed Processing. Boston, MA, USA: Springer, 2002. IV. RESULTS & DISCUSSION It is noted that the software solution developed here can accept multiple video streams from a centralized storage location, suitable for operation in facilities management or public spaces with multiple installed security cameras. Other useful applications of the current framework may include utilization in factories to detect personnel in restricted places or unsafe proximity to equipment. FIGURE 10. Subject counting results in the non-controlled environment, comparing BGS and SSD methods. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. FIGURE 10. Subject counting results in the non-controlled environment, comparing BGS and SSD methods. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. FIGURE 10. Subject counting results in the non-controlled environment, comparing BGS and SSD methods. The actual counts were obtained through manual counting by watching the same video footage, matched against an on-site surveyor for consistency. Further extensions of the current framework may include layering facial identiﬁcation on top of the current object recognition and tracking for enhanced surveillance capabil- ities, or to conﬁgure recognition for potentially dangerous items such as knives or ﬁrearms in the ﬁght against terrorism, to enhance commuter safety and security. FIGURE 11. Separated foreground elements by BGS, corresponding to the input frame shown in Figure 9(c). The fragmented masking of human subjects is clearly observed. varying backgrounds. We illustrate this in Figure 11, in which the imprecise and fragmented masking of overlapping sub- jects by BGS can be seen. The MOG2 implementation of BGS statistically constructs a background mask from a sub- sample of video frames, and is thus theoretically able to re- adjust for changing background conditions; but in this real- world deployment in a transport hub, there is insufﬁcient time for such a mechanism to work as intended. With background masks of inadequate quality, multiple subjects in close prox- imity are frequently misidentiﬁed as a single subject, hence resulting in the severe under-counting. It is thus obvious that the SSD implementation is greatly more suitable for use, especially in places of signiﬁcant human trafﬁc. V. CONCLUSION V. CONCLUSION I hi d [5] T. P. Chen, H. Haussecker, A. Bovyrin, R. Belenov, K. Rodyushkin, A. Kuranoc, and V. Eruhimov, ‘‘Computer vision workload analysis: Case study of video surveillance systems,’’ Int. Technol. J., vol. 9, no. 2, pp. 109–118, 2005. In this study, we have considered a number of classical and CNN-based object recognition techniques for real-time video analytics, and have developed a software platform 183258 VOLUME 7, 2019 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting [6] S. A. M. Saleh, S. A. Suandi, and H. Ibrahim, ‘‘Recent survey on crowd density estimation and counting for visual surveillance,’’ Eng. Appl. Artif. Intell., vol. 41, pp. 103–114, May 2015. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0952197615000081 [30] S. A. Velastin, ‘‘CCTV video analytics: Recent advances and limita- tions,’’ in Proc. Int. Vis. Informat. Conf. Berlin, Germany: Springer, 2009, pp. 22–34. [31] E. D. Dickmanns, ‘‘The development of machine vision for road vehicles in the last decade,’’ in Proc. Intell. Vehicle Symp., vol. 1, Jun. 2002, pp. 268–281. [7] L. Zhang, M. Shi, and Q. Chen, ‘‘Crowd counting via scale-adaptive convolutional neural network,’’ in Proc. IEEE Winter Conf. Appl. Comput. Vis. (WACV), Mar. 2018, pp. 1113–1121. [32] B. Ranft and C. Stiller, ‘‘The role of machine vision for intelligent vehi- cles,’’ IEEE Trans. Intell. Vehicles, vol. 1, no. 1, pp. 8–19, Mar. 2016. [8] H. Liu, S. Chen, and N. Kubota, ‘‘Intelligent video systems and analytics: A survey,’’ IEEE Trans. Ind. Informat., vol. 9, no. 3, pp. 1222–1233, Aug. 2013. [33] M. Slade and F. Marmoiton, ‘‘Toward smart autonomous cars,’’ in Intel- ligent Transportation Systems. Boca Raton, FL, USA: CRC Press, 2016, pp. 94–120. [9] G. Ananthanarayanan, P. Bahl, P. Bodík, K. Chintalapudi, M. Philipose, L. Ravindranath, and S. Sinha, ‘‘Real-time video analytics: The killer app for edge computing,’’ Computer, vol. 50, no. 10, pp. 58–67, 2017. [34] V. Moskalenko, A. Moskalenko, A. Korobov, O. Boiko, S. Martynenko, and O. Borovenskyi, ‘‘Model and training methods of autonomous naviga- tion system for compact drones,’’ in Proc. IEEE 2nd Int. Conf. Data Stream Mining Process. (DSMP), Aug. 2018, pp. 503–508. [10] A. A. Adams and J. M. Ferryman, ‘‘The future of video analytics for surveillance and its ethical implications,’’ Secur. J., vol. 28, no. 3, pp. 272–289, 2015. [35] A. A. Zhilenkov and I. R. V. CONCLUSION 2007, pp. 35–42. [41] H. Jiang and E. Learned-Miller, ‘‘Face detection with the faster R-CNN,’’ in Proc. 12th IEEE Int. Conf. Autom. Face Gesture Recognit. (FG), May/Jun. 2017, pp. 650–657. [18] S. Kim, R. Maciejewski, K. Ostmo, E. J. Delp, T. F. Collins, and D. S. Ebert, ‘‘Mobile analytics for emergency response and training,’’ Inf. Vis., vol. 7, no. 1, pp. 77–88, 2008. [42] S. Guo, S. Chen, and Y. Li, ‘‘Face recognition based on convolutional neural network and support vector machine,’’ in Proc. IEEE Int. Conf. Inf. Automat. (ICIA), Aug. 2016, pp. 1787–1792. [19] M. Zabocki, K. Gościewska, D. Frejlichowski, and R. Hofman, ‘‘Intelligent video surveillance systems for public spaces—A survey,’’ J. Theor. Appl. Comput. Sci., vol. 8, no. 4, pp. 13–27, 2014. [43] X. Sun, P. Wu, and S. C. Hoi, ‘‘Face detection using deep learn- ing: An improved faster RCNN approach,’’ Neurocomputing, vol. 299, pp. 42–50, Mar. 2018, doi: 10.1016/j.neucom.2018.03.030. [20] H. Kruegle, CCTV Surveillance: Video Practices and Technology. Amsterdam, The Netherlands: Elsevier, 2011. [44] S. Balaban, ‘‘Deep learning and face recognition: The state of the art,’’ Proc. SPIE, vol. 9457, May 2015, Art. no. 94570B, doi: 10.1117/12.2181526. [21] H. Keval and M. Sasse, ‘‘Man or gorilla? Performance issues with CCTV technology in security control rooms,’’ in Proc. 16th World Congr. Ergonom. Conf., Int. Ergonom. Assoc., 2006, pp. 10–14. [45] K. Hazelwood, S. Bird, D. Brooks, S. Chintala, U. Diril, D. Dzhulgakov, M. Fawzy, B. Jia, Y. Jia, A. Kalro, J. Law, K. Lee, J. Lu, P. Noordhuis, M. Smelyanskiy, L. Xiong, and X. Wang, ‘‘Applied machine learning at facebook: A datacenter infrastructure perspective,’’ in Proc. IEEE Int. Symp. High Perform. Comput. Archit. (HPCA), Feb. 2018, pp. 620–629. [22] A. Şentaş, I. Tashiev, F. Küçükayvaz, S. Kul, S. Eken, A. Sayar, and Y. Becerikli, ‘‘Performance evaluation of support vector machine and convolutional neural network algorithms in real-time vehicle type classi- ﬁcation,’’ in Proc. Int. Conf. Emerg. Internetworking, Data Web Technol. Cham, Switzerland: Springer, 2018, pp. 934–943. [46] W. Brendel and M. Bethge, ‘‘Approximating CNNs with bag-of-local- features models works surprisingly well on imagenet,’’ in Proc. Int. Conf. Learn. Represent., 2019. [Online]. Available: https://openreview.net/ forum?id=SkfMWhAqYQ [23] S. Kul, S. Eken, and A. Sayar, ‘‘Evaluation of real-time performance for BGSLibrary algorithms: A case study on trafﬁc surveillance video,’’ in Proc. 6th Int. Conf. IT Converg. Secur. (ICITCS), 2016, pp. 1–4. [47] F. Hu, G.-S. Xia, J. Hu, and L. V. CONCLUSION Epifantsev, ‘‘The use of convolution arti- ﬁcial neural networks for drones autonomous trajectory planning,’’ in Proc. IEEE Conf. Russian Young Res. Elect. Electron. Eng. (EIConRus), Jan./Feb. 2018, pp. 1044–1047. [11] O. Javed and M. Shah, ‘‘Tracking and object classiﬁcation for automated surveillance,’’ in Computer Vision, A. Heyden, G. Sparr, M. Nielsen, and P. Johansen, Eds. Berlin, Germany: Springe, 2002, pp. 343–357. [12] J. Barthélemy, N. Verstaevel, H. Forehead, and P. Perez, ‘‘Edge-computing video analytics for real-time trafﬁc monitoring in a smart city,’’ Sensors, vol. 19, no. 9, p. 2048, 2019. [36] Y. Satılmiş, F. Tufan, M. Şara, M. Karslı, S. Eken, and A. Sayar, ‘‘CNN based trafﬁc sign recognition for mini autonomous vehicles,’’ in Proc. Int. Conf. Inf. Syst. Archit. Technol., J. Świątek, L. Borzemski, and Z. Wilimowska, Eds. Cham, Switzerland: Springer, 2019, pp. 85–94. [13] M. A. Uddin, A. Alam, N. A. Tu, M. S. Islam, and Y.-K. Lee, ‘‘SIAT: A distributed video analytics framework for intelligent video surveillance,’’ Symmetry, vol. 11, no. 7, p. 911, 2019. [37] M. Hirz and B. Walzel, ‘‘Sensor and object recognition technologies for self-driving cars,’’ Comput.-Aided Des. Appl., vol. 15, no. 4, pp. 501–508, 2018. [14] X. Kang, B. Song, and F. Sun, ‘‘A deep similarity metric method based on incomplete data for trafﬁc anomaly detection in IoT,’’ Appl. Sci., vol. 9, no. 1, p. 135, 2019. [38] A. Shustanov and P. Yakimov, ‘‘CNN design for real-time trafﬁc sign recognition,’’ Procedia Eng., vol. 201, pp. 718–725, Dec. 2017. [15] M. Naphade, G. Banavar, C. Harrison, J. Paraszczak, and R. Morris, ‘‘Smarter cities and their innovation challenges,’’ Computer, vol. 44, no. 6, pp. 32–39, Jun. 2011. [39] H. Wang, B. Wang, B. Liu, X. Meng, and G. Yang, ‘‘Pedestrian recognition and tracking using 3D LiDAR for autonomous vehicle,’’ Robot. Auton. Syst., vol. 88, pp. 71–78, Feb. 2017. [16] R. Sundar, S. Hebbar, and V. Golla, ‘‘Implementing intelligent trafﬁc con- trol system for congestion control, ambulance clearance, and stolen vehicle detection,’’ IEEE Sensors J., vol. 15, no. 2, pp. 1109–1113, Feb. 2015. [40] S. Lawrence, C. L. Giles, A. C. Tsoi, and A. D. Back, ‘‘Face recognition: A convolutional neural-network approach,’’ IEEE Trans. Neural Netw., vol. 8, no. 1, pp. 98–113, Jan. 1997. [17] S. Y. Kim, Y. Jang, A. Mellema, D. S. Ebert, and T. Collinss, ‘‘Visual analytics on mobile devices for emergency response,’’ in Proc. IEEE Symp. Vis. Anal. Sci. Technol., Oct./Nov. V. CONCLUSION Zivkovic and F. van der Heijden, ‘‘Efﬁcient adaptive density estimation per image pixel for the task of background subtraction,’’ Pattern Recognit. Lett., vol. 27, no. 7, pp. 773–780, 2006. [63] A. B. Godbehere, A. Matsukawa, and K. Goldberg, ‘‘Visual tracking of human visitors under variable-lighting conditions for a responsive audio art installation,’’ in Proc. Amer. Control Conf. (ACC), 2012, pp. 4305–4312. JIN MING KOH received the NUS High School Diploma (High Distinction), in 2016. Since 2017, he has been undertaking research projects offered by K. H. Cheong. He is currently with the Califor- nia Institute of Technology. JIN MING KOH received the NUS High School Diploma (High Distinction), in 2016. Since 2017, he has been undertaking research projects offered by K. H. Cheong. He is currently with the Califor- nia Institute of Technology. JIN MING KOH received the NUS High School Diploma (High Distinction), in 2016. Since 2017, he has been undertaking research projects offered by K. H. Cheong. He is currently with the Califor- nia Institute of Technology. [64] T. Trnovszký, P. Sýkora, and R. Hudec, ‘‘Comparison of background sub- traction methods on near infra-red spectrum video sequences,’’ Procedia Eng., vol. 192, pp. 887–892, Jun. 2017. [Online]. Available: http://www. sciencedirect.com/science/article/pii/S1877705817327005, doi: 10.1016/ j.proeng.2017.06.153. [65] I. Benraya and N. Benblidia, ‘‘Comparison of background subtraction methods,’’ in Proc. Int. Conf. Appl. Smart Syst. (ICASS), 2018, pp. 1–5. [66] J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, ‘‘You only look once: Uniﬁed, real-time object detection,’’ in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., Jun. 2016, pp. 779–788. [67] W. Liu, D. Anguelov, D. Erhan, C. Szegedy, S. Reed, C.-Y. Fu, and A. C. Berg, ‘‘SSD: Single shot multibox detector,’’ in Proc. Eur. Conf. Comput. Vis. Cham, Switzerland: Springer, 2016, pp. 21–37. [68] Chuanqi305 Mobilenet-SSD. Accessed: Oct. 29, 2019. [Online]. Available: https://github.com/chuanqi305/MobileNet-SSD U. RAJENDRA ACHARYA received the Ph.D. degree from the National Institute of Technol- ogy Karnataka, Surathkal, India, and the D.Eng. degree from Chiba University, Japan. He is cur- rently a Senior Faculty Member with Ngee Ann Polytechnic, Singapore. He is also an Adjunct Professor with Taylor’s University, Malaysia, an Adjunct Faculty with the Singapore Institute of Technology–University of Glasgow, Singapore, and an Associate Faculty with the Singapore Uni- versity of Social Sciences, Singapore. V. CONCLUSION degree (Hons.) in sustainable infrastructure engi- neering (building services) from the Singapore Institute of Technology, in 2018. She is cur- rently an Associate Geospatial Specialist with the Government Technology Agency (GovTech), Singapore. SANDRA POESCHMANN received the B.Eng. degree (Hons.) in sustainable infrastructure engi- neering (building services) from the Singapore Institute of Technology, in 2018. She is cur- rently an Associate Geospatial Specialist with the Government Technology Agency (GovTech), Singapore. [53] M. Bilal, A. Khan, M. U. K. Khan, and C.-M. Kyung, ‘‘A low- complexity pedestrian detection framework for smart video surveillance systems,’’ IEEE Trans. Circuits Syst. Video Technol., vol. 27, no. 10, pp. 2260–2273, Oct. 2017. [54] R. Lucas, A. Rowlands, A. Brown, S. Keyworth, and P. Bunting, ‘‘Rule- based classiﬁcation of multi-temporal satellite imagery for habitat and agricultural land cover mapping,’’ ISPRS J. Photogramm. Remote Sens., vol. 62, no. 3, pp. 165–185, Aug. 2007. [55] Z. Miao, S. Zou, Y. Li, X. Zhang, J. Wang, and M. He, ‘‘Intelligent video surveillance system based on moving object detection and tracking,’’ in Proc. Int. Conf. Inf. Eng. Commun. Technol., 2016, pp. 1–4. [56] K. Pulli, A. Baksheev, K. Kornyakov, and V. Eruhimov, ‘‘Real-time com- puter vision with OpenCV,’’ Commun. ACM, vol. 55, no. 6, pp. 61–69, Jun. 2012. [57] I. Culjak, D. Abram, T. Pribanic, H. Dzapo, and M. Cifrek, ‘‘A brief introduction to OpenCV,’’ in Proc. 35th Int. Conv. MIPRO, 2012, pp. 1725–1730. JOEL WEIJIA LAI received the B.Sc. degree (Hons.) in physics with a second major in math- ematical sciences from Nanyang Technological University (NTU), Singapore, in 2017. He is currently pursuing the Ph.D. degree with the Singapore University of Technology and Design (SUTD), Singapore. He stayed on NTU to develop technology enhanced learning tools for the Divi- sion of Physics and Applied Physics. He went to become a Research Assistant with SUTD. [58] S. H. Shaikh, K. Saeed, and N. Chaki, Moving Object Detection Using Background Subtraction. Cham, Switzerland: Springer, 2014, pp. 15–23. [59] P. KaewTraKulPong and R. Bowden, ‘‘An improved adaptive back- ground mixture model for real-time tracking with shadow detection,’’ in Video-Based Surveillance Systems. Boston, MA, USA: Springer, 2002, pp. 135–144. pp [60] E. Grevelink, ‘‘A closer look at object detection, recognition and tracking,’’ Intel, Santa Clara, CA, USA, Tech. Rep., 2017. [Online]. Available: https://software.intel.com/en-us/articles/a-closer- look-at-object-detection-recognition-and-tracking [61] Z. Zivkovic, ‘‘Improved adaptive Gaussian mixture model for background subtraction,’’ in Proc. ICPR, 2004, pp. 28–31. [62] Z. V. CONCLUSION Zhang, ‘‘Transferring deep convolu- tional neural networks for the scene classiﬁcation of high-resolution remote sensing imagery,’’ Remote Sens., vol. 7, no. 11, pp. 14680–14707, 2015. [24] S. Kul, S. Eken, and A. Sayar, ‘‘Distributed and collaborative real-time vehicle detection and classiﬁcation over the video streams,’’ Int. J. Adv. Robot. Syst., vol. 14, no. 4, pp. 1–12, 2017. [25] M. Ulvi, S. Eken, and A. Sayar, ‘‘Service oriented visual interpretation tool for time series data,’’ Anadolu Univ. J. Sci. Technol.-Appl. Sci. Eng., vol. 14, pp. 191–198, Jan. 2013. [48] X. Ma, Z. Dai, Z. He, J. Ma, Y. Wang, and Y. Wang, ‘‘Learning trafﬁc as images: A deep convolutional neural network for large-scale transportation network speed prediction,’’ Sensors, vol. 17, no. 4, p. 818, 2017. [26] E. R. Davies, Machine Vision: Theory, Algorithms, Practicalities. Amsterdam, The Netherlands: Elsevier, 2004. [49] R. A. Minhas, A. Javed, A. Irtaza, M. T. Mahmood, and Y. B. Joo, ‘‘Shot classiﬁcation of ﬁeld sports videos using alexnet convolutional neural network,’’ Appl. Sci., vol. 9, no. 3, p. 483, 2019. [27] L. Tyapi and K. Sowmya, ‘‘Real time human detection from video surveillance,’’ Int. J. Innov. Res. Comput. Commun. Eng., vol. 3, no. 5, pp. 4413–4417, 2015. [50] M. He, H. Luo, B. Hui, and Z. Chang, ‘‘Pedestrian ﬂow tracking and statistics of monocular camera based on convolutional neural network and Kalman ﬁlter,’’ Appl. Sci., vol. 9, no. 8, p. 1624, 2019. [28] J. C. S. J. Junior, S. R. Musse, and C. R. Jung, ‘‘Crowd analysis using computer vision techniques,’’ IEEE Signal Process. Mag., vol. 27, no. 5, pp. 66–77, Sep. 2010. [51] V. Abrishami, A. Rezaee, H. Baherzadeh, and H. Abrishami, ‘‘Real-time pedestrian detecting and tracking in crowded and complicated scenario,’’ in Proc. 3rd Int. Conf. Imag. Crime Detection Prevention (ICDP), Dec. 2009, pp. 1–6. [29] C. Regazzoni, A. Cavallaro, Y. Wu, J. Konrad, and A. Hampapur, ‘‘Video analytics for surveillance: Theory and practice,’’ IEEE Signal Process. Mag., vol. 27, no. 5, pp. 16–17, Sep. 2010. VOLUME 7, 2019 183259 VOLUME 7, 2019 VOLUME 7, 2019 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting [52] A. Dore, M. Soto, and C. S. Regazzoni, ‘‘Bayesian tracking for video analytics,’’ IEEE Signal Process. Mag., vol. 27, no. 5, pp. 46–55, Sep. 2010. SANDRA POESCHMANN received the B.Eng. JIN MING KOH received the NUS High School Diploma (High Distinction), in 2016. Since 2017, he has been undertaking research projects offered by K. H. Cheong. He is currently with the Califor- nia Institute of Technology. V. CONCLUSION He has published more than 400 arti- cles in refereed international SCI-IF journals (345), international conference proceedings (42), books (17) with more than 20 000 citations in Google Scholar (with H-index of 75), and Research Gate (RG) score of 47.1. He holds three patents. His major academic interests include biomedical signal processing, biomedical imaging, data mining, visualization, and bio- physics for better healthcare design, delivery, and therapy. He is ranked in the top 1% of the Highly Cited Researchers for the last four consecutive years (2016–2019) in computer science according to the Essential Science Indicators of Thomson. [69] N/A, Common Objects in Context. Accessed: Oct. 29, 2019. [Online]. Available: http://cocodataset.org/#home [70] A. Rosebrock. (2018). Simple Object Tracking With OpenCV. Accessed: Aug. 8, 2019. [Online]. Available: https://www.pyimagesearch. com/2018/07/23/simple-object-tracking-with-opencv/ KANG HAO CHEONG (M’18) received the B.Sc. degree (Hons.) from the Department of Mathemat- ics and University Scholars Programme, National University of Singapore (NUS), in 2007, the Post- graduate diploma in education from the National Institute of Education, Singapore, and the Ph.D. degree from the Department of Electrical and Computer Engineering, NUS, in 2015. He was an Assistant Professor with the Engineering Cluster, Singapore Institute of Technology, from 2016 to 2018. He is currently an Assistant Professor with the Science and Math Cluster, Singapore University of Technology and Design (SUTD). He is also afﬁliated with the SUTD-MIT International Design Centre. VOLUME 7, 2019 VOLUME 7, 2019 183260 183260 K. H. Cheong et al.: Practical Automated Video Analytics for Crowd Monitoring and Counting KENNETH JIAN WEI TANG received the B.Eng. degree (Hons.) in sustainable infras- tructure engineering (building services) and the M.Eng.Tech. degree from the Singapore Institute of Technology (SIT), in 2018 and 2019, respec- tively. He is currently pursuing the Ph.D. degree with the Science and Math Cluster, Singapore University of Technology and Design. SIMON CHING MAN YU received the B.Eng. degree (Hons.) in ACGI and the Ph.D. degree in DIC from the Department of Mechanical Engi- neering, Imperial College, London, in 1987 and 1991, respectively. He is currently a Professor with the Interdisciplinary Division of Aeronautical and Aviation Engineering, The Hong Kong Polytech- nic University. He joined Nanyang Technological University, Singapore, immediately after his grad- uation, as a Lecturer. He became a Senior Lecturer, in 1996, and an Associate Professor, in 2000. V. CONCLUSION He has been involved in the University Administration, since 2000: a Principal Staff Ofﬁcer (Pres- ident’s Ofﬁce), from 2000 to 2003, a University Council Member, from 2001 to 2003, the Vice Dean, Admission Ofﬁce, from 2003 to 2006, and the Head of the Division of Aerospace Engineering, School of Mechanical and Aerospace Engineering, from 2008 to 2013. He moved over to the Singapore Institute of Technology as a Professor and a Programme Director, in 2013, to establish one of the ﬁrst engineering degree programmes offered solely by the University. He has published more than 200 research articles in archive journals and conferences. He managed to secure more than SGD 50 million external grant during his tenure in the Nanyang Technological University, especially during the period as the Head of the Division of Aerospace Engineering. KENNETH JIAN WEI TANG received the B.Eng. degree (Hons.) in sustainable infras- tructure engineering (building services) and the M.Eng.Tech. degree from the Singapore Institute of Technology (SIT), in 2018 and 2019, respec- tively. He is currently pursuing the Ph.D. degree with the Science and Math Cluster, Singapore University of Technology and Design. 183261 VOLUME 7, 2019
https://openalex.org/W2341757677	https://www.journals.vu.lt/nonlinear-analysis/article/download/13511/12429	English	null	Compound method of time series classification	Nonlinear analysis	2,015	cc-by	6,013	Received: July 13, 2014 / Revised: January 17, 2015 / Published online: September 9, 2015 Abstract. Many real phenomenona preserves the properties of chaotic dynamics. However, unambiguous determination of belonging to a group of chaotic systems is difﬁcult and complex problem. The main purpose of this paper is to present compound method of time series classiﬁcation which is basically directed to the detection of chaotic behaviors. The method has been designed for differentiation of three types of time series: chaotic, periodic and random. Our approach assumes, that more reliable information about the dynamics of the system will provide the compilation of several methods, than any individual. This paper focuses on choosing a good set of methods and analysis of their results. In our investigation, we used the following methods and indicators: time delay embedding, mutual information, saturation of system invariants, the largest Lyapunov exponent and Hurst exponent. We checked the validity of the methods applying them to three kinds of basic systems which generate chaotic, periodic and random time series. As a summary of this paper, all selected methods and indicators computed for generated times series have been summarized in the table, which gives the authors a possibility to conclude about type of observed behavior. Keywords: deterministic chaos, time series analysis, Takens theorem. ISSN 1392-5113 ISSN 1392-5113 ISSN 1392-5113 http://dx.doi.org/10.15388/NA.2015.4.6 Nonlinear Analysis: Modelling and Control, Vol. 20, No. 4, 545–560 Compound method of time series classiﬁcation Łukasz Korusa, Michał Piórekb aDepartment of Control Systems and Mechatronic Wroclaw University of Technology, Janiszewskiego 11/17, 50-372 Wroclaw, Poland Łukasz Korusa, Michał Piórekb aDepartment of Control Systems and Mechatronics, Wroclaw University of Technology, Janiszewskiego 11/17, 50-372 Wroclaw, Poland lukasz.korus@pwr.edu.pl aDepartment of Control Systems and Mechatronics, Wroclaw University of Technology, Janiszewskiego 11/17, 50-372 Wroclaw, Poland lukasz.korus@pwr.edu.pl bDepartment of Computer Engineering, Wroclaw University of Technology, Janiszewskiego 11/17, 50-372 Wroclaw, Poland michal.piorek@pwr.edu.pl Received: July 13, 2014 / Revised: January 17, 2015 / Published online: September 9, 2015 c⃝Vilnius University, 2015 2.1 Time delay embedding Many of real phenomenon are the results of nonlinear systems dynamics evolution [1]. Observation of such kind of systems provides many difﬁculties due to the fact that only a limited set of information is accessible outside as a measurement of time series. The basic assumption of this paper is that one time series is formed by samples measured in the output of the system s(n) = x(t0 + n∆t), (1) (1) where s(n) is the value of nth sample of measurement of physical process x in time t0 + n∆t and t0 is initial time. According to Takens’ embedding theorem, it is possible to reconstruct the state trajectory from single time series using below algorithm: y(n) = s(n), s(n + T), . . . , s n + (d −1) T , (2) (2) where T is time delay and d is embedding dimension, which estimates a real dimension of the observed system. The main point of the state space reconstruction method is T and d estimation. To estimate time delay T, an average mutual information I has been used, while for embedding dimension the saturation of system invariants method [1]. In our investigations, methods described below were used. In our investigations, methods described below were used. 1 Introduction In 1981, Floris Takens presents his theorem about Embedding Nonlinear Dynamical systems, which conceive an opportunity to reconstruct attractors of chaotic dynamical system. Reconstruction is based on a time series given by observations of one state of dynamical system [37]. Over the next few years many of embedding approaches and theirs applications were described, for example, by Sauer, Yorke and Casdagli [34]. Stark applied Taken’s theorem to Forced Systems [36]. Abarbanel described the analysis of observed chaotic data in physical systems [1]. 546 Ł. Korus, M. Piórek There are also many of papers about observing chaos in real systems [4, 5, 6, 12, 18, 19,21,22,26,28,32,38,41,42] and about controlling chaos [3,8,8,15,25,31,39]. Such a large number of methods from listed papers proves, that the ﬁeld of analysis of chaotic systems is heavily explored. This is due to the difﬁculty and complexity of the subject. The approach described in this article was the result of previous experimental studies. In Section 2, a set of time series analysis methods were chosen and described. Selection was preceded by a review of the literature. Section 3 describes a way for generating test time series for further experiments. Section 4 presents the numerical results of experi- ments carried out on the generated time series. Conclusions are described in Section 5. 2.2 Time delay estimation: average mutual information method There are three types of criterions of time delay Td selection [29]: There are three types of criterions of time delay Td selection [29]: • series correlation approaches (autocorrelation, mutual information [17] or high- order correlations [2]), • approaches of phase space extension (ﬁllfactor [11], wavering product [10] or av- erage displacement [14]), • multiple autocorrelation and nonbias multiple autocorrelation [27]. http://www.mii.lt/NA http://www.mii.lt/NA 547 Compound method of time series classiﬁcation For the purposes of this paper, mutual information method was selected. This ap- proach is based on information theory and transformation of linear autocorrelation to non- linear systems. More precisely, this method consists of 2-dimensional adaptive histogram. Let’s assume that there are two nonlinear systems: A and B. Outputs of these systems are denoted as a and b, while values of these outputs are represented by ai and bk. Mutual information factor describes how many bits of bk could be predicted, where ai is known: IAB(ai, bk) = log2 PAB(ai, bk) PA(ai)PB(bk). (3) (3) Here PA(ai) is probability that a = ai and PB(bk) is probability that b = bk and PAB(ai, bk) is join probability that a = ai and b = bk. Average mutual information factor can be described by IAB(T) = X ai,bk PAB(ai, bk)IAB(ai, bk). (4) (4) In order to use this method to assess correlation between different samples in the same time series, the average mutual information factor is ﬁnally described by the equation I(T) = N X n=1 P S(n), S(n + T) log2 P(S(n), S(n + T)) P(S(n))P(S(n + T)). (5) (5) Fraser and Swinney [17] propose that Tm lag Td, where the ﬁrst minimum of I(T) occurs as a useful selection of time. This selection guarantees that the measurements are somewhat independent, but not statistically independent. In the case of absence of the average mutual information clear minimum, this criterion needs to be replaced by choosing Td as the time for which the average mutual information reaches 4/5 of its initial value: ( ) Fraser and Swinney [17] propose that Tm lag Td, where the ﬁrst minimum of I(T) occurs as a useful selection of time. This selection guarantees that the measurements are somewhat independent, but not statistically independent. 2.2 Time delay estimation: average mutual information method In the case of absence of the average mutual information clear minimum, this criterion needs to be replaced by choosing Td as the time for which the average mutual information reaches 4/5 of its initial value: I(Td) I(0) ≈4 5. (6) (6) Nonlinear Anal. Model. Control, 20(4):545–560 2.3 Embedding dimension estimation: saturation of system invariants As it was mentioned, delay coordinates are used to construct d-dimensional vector to state space reconstruction. Takens theorem guarantees that if chosen embedding dimen- sion is large enough, properties of the attractor of dynamical system will be the same when computed on lagged coordinates and when computed in the physical coordinates. The main goal of state space reconstruction is to provide a Euclidean space Rd large enough so that the set of points of dimension d can be unfolded without ambiguity. When all ambiguities are resolved, one says that the space Rd provides an embedding of the attractor of dynamical system and dimension d is an embedding dimension dE = d. There are several methods of selecting embedding dimension. One of them is the method of false neighbors [9]. Very promising seems to be neural network method [30]. In this paper, saturation of system invariants method was used [1]. Nonlinear Anal. Model. Control, 20(4):545–560 548 Ł. Korus, M. Piórek Ł. Korus, M. Piórek For properly reconstructed attractor, every its property depending on distances be- tween points in the state space should become independent of the value of the embedding dimension once the large enough dE has been reached. Appropriate necessary embedding dimension can be established by computing such property for dE = 1, 2 . . . until variation with dE cases. Mentioned property can be the correlation integral [20]. The average number of points on the attractor within a radius of r of points x in the state space, n(r, x), is deﬁned by n(r, x) = 1 N N X i=1 θ r − x(i) −x , (7) (7) where θ(u) is the Heaviside function: θ(u) = 0, u < 0; θ(u) = 1, u > 0. The average over all points in attractor of n(r, x) is called correlation integral and is described below: C(r) = 1 M M X j=1 n r, x(j) = 1 M 1 N M X j=1 N X i=1 θ r − x(i) −x(j) . (8) (8) We can evaluate C(r) as a function of dE and determine when the slope of its logarithm as function of log r becomes independent of dE. In this method, the authors perceive the possibility to improve its efﬁciency through the use of correlation dimension. 2.4 Correlation dimension saturation Fractal geometry and fractal dimension conception provide a general framework for the study of such irregular sets, like strange attractors. Fractals, which are irregular geometric objects, require a special meaning of dimension. Very roughly, fractal dimension provide a description of how much space a set ﬁlls [7, 16]. Of the wide variety of fractal dimen- sions in use, the deﬁnition of Hausdorff is the most important: dimH F = inf s: Hs(F) = 0 = sup s: Hs(F) = ∞ . (9) (9) Here Hs(F) is s-dimensional Hausdorff measure of set F deﬁned as follows: Here Hs(F) is s-dimensional Hausdorff measure of set F deﬁned as follows: Hs(F) = lim σ→0 Hs σ(F), (10) (10) where Hs σ(F) is described by equation where Hs σ(F) is described by equation Hs σ(F) = inf ( ∞ X i=1 \|Ui\|s: Ui is σ-cover of F ) . (11) (11) The attractor of chaotic system has a ﬁxed fractal dimension determined by Hausdorff. Because of difﬁculties in its computations, several other fractal dimension estimators were introduced. http://www.mii.lt/NA Compound method of time series classiﬁcation 549 Correlation dimension D2 is one of the fractal dimension estimators particularly suited for relatively easy experimental determination. It is connected with correlation integral C(r) ≈rD2. (12) (12) Above equation after some transformations can be written as D2 = lim r→0 C(r) ln r , (13) D2 = lim 1 M 1 N PM j=1 PN i=1 θ(r −\|x(i) −x(j)\|) l . (14) D2 = lim r→0 C(r) ln r , (13) 1 1 PM PN θ( \| (i) (j)\|) D2 = lim r→0 C(r) ln r , (13) D2 lim r→0 ln r , (13) D2 = lim r→0 1 M 1 N PM j=1 PN i=1 θ(r −\|x(i) −x(j)\|) ln r . (14) (14) For large N, Eq. (14) is useful estimator of D2. For large N, Eq. (14) is useful estimator of D2. ln C(r) ≈ln rD2 ≈D2 ln r. (15) (15) In conclusion, estimation of D2 value is a slope factor of regression function chart tangent to the most linear part of ln(C(r)) = f(ln r) dependency. The authors propose to saturate correlation dimension D2 instead to correlation in- tegral in saturation of system invariants method. As will be shown in Section 4.2, this improvement allows for a more clear and intuitive embedding dimension selection. 2.4 Correlation dimension saturation Cor- relation dimension was chosen because of its intuitive implementation. There are several other fractal dimension estimators, such as box-counting dimension, information dimen- sion or Higuchi dimension. Interesting approaches to fractal dimension estimation and its applications was presented [33,35]. 2.5 The largest Lyapunov exponent Lyapunov exponents describe velocity of distance increasing between two initially neigh- boring orbits in the state space. The largest Lyapunov exponent describes the mean diver- gence between initially neighboring trajectories by the following formula: d(t) = DeL1t, (16) (16) where d(t) is distance between orbits in time, D is initial separation between neighboring points and L1 is the largest Lyapunov exponent. In practice, value of the largest Lyapunov exponent is commonly used chaos detection factor, because its positive value clearly indicates chaotic behavior in the system. Wolf in his article [40] proposed to estimate the largest Lyapunov’s exponent for time series based on below equation: where d(t) is distance between orbits in time, D is initial separation between neighboring points and L1 is the largest Lyapunov exponent. In practice, value of the largest Lyapunov exponent is commonly used chaos detection factor, because its positive value clearly indicates chaotic behavior in the system. Wolf in his article [40] proposed to estimate the largest Lyapunov’s exponent for time series based on below equation: L1 = 1 t m X j=1 log2 L′(tj+1) L(tj) , (17) (17) where L(tj) is a distance (in Euclidean sence) between pairs of trajectory points in time t1 and L′(tj+1) is a distance between pairs of trajectory points in time tj+1. ( j+ ) j+ Alternative methods of the largest Lyapunov exponent estimation are Rossenstein’s algorithm [13] or Kantz’s algorithm [24]. Nonlinear Anal. Model. Control, 20(4):545–560 550 Ł. Korus, M. Piórek 2.6 Hurst exponent Hurst exponent H can be counted by using R/S analysis based on the following algo- rithm [23]: 1. N-length time series divide into k n-length subsets, where k · n = N. 1. N-length time series divide into k n-length subsets, where k · n = N. 2. For each subset m = 1, . . . , k: 2. For each subset m = 1, . . . , k: 2. For each subset m = 1, . . . , k: 2. For each subset m = 1, . . . , k: • count average Em and standard deviation Sm, • count average Em and standard deviation Sm, • count average Em and standard deviation Sm, • rescale each value in subset xi,m by cutting of the average value counted for this subset • rescale each value in subset xi,m by cutting of the average value counted for this subset zi,m = xi,m −Em for i = 1, . . . , n, (18) (18) • construct cumulated subset of rescaled values • construct cumulated subset of rescaled values • construct cumulated subset of rescaled values yi,m = i X j=1 zj,m for i = 1, . . . , n, (19) (19) • calculate range Rm • calculate range Rm Rm = max{y1,m, . . . , yn,m} −min{y1,m, . . . , yn,m}, (20) (20) • rescale range Rm Sm . (21) 3. The average value of rescaled range for n-length subset can be described by 3. The average value of rescaled range for n-length subset can be described by R S n = 1 k k X m=1 Rm Sm . (22) (22) Hurst exponent can be calculated from below equation: Hurst exponent can be calculated from below equation: R S n = (cn)H, (23) (23) where: R/S – rescaled range, n – number of measurements in subset, c – constant value. Finally, after logarithmization of both sides above equation can be rewritten as follows: where: R/S – rescaled range, n – number of measurements in subset, c – constant value. Finally, after logarithmization of both sides above equation can be rewritten as follows: ln R S = H ln n + H ln c. (24) (24) Slope factor value of linear regression function tangent to the most linear part of ln(R/S) = f(ln n) dependency is an estimation of H. 2.6 Hurst exponent Slope factor value of linear regression function tangent to the most linear part of ln(R/S) = f(ln n) dependency is an estimation of H. ( / ) ( ) It can be observed that, Hurst exponent values can be divided into three sets: • H = 0.5 – samples in examined time series is random and not correlated (i.i.d. series), • 0 < H < 0.5 – examined time series is antipersistent and ergodic, which means that the distribution parameters are constant (if the system trajectory in some period of time increases, then it is highly probable that it will decrease in upcoming period, http://www.mii.lt/NA http://www.mii.lt/NA http://www.mii.lt/NA Compound method of time series classiﬁcation 551 • 0.5 < H < 1 – examined time series is persistant, which means that samples create some trends (if the system trajectory in some period of time follows a particular direction, it is highly probable that in the upcoming period the trajectory will keep the same direction). 3.1 Lorenz system Considered Lorenz system is deﬁned by differential equations dx(t) dt = σ y(t) −x(t) , dy(t) dt = −x(t)z(t) + rx(t) −y(t), dz(t) dt = x(t)y(t) −bz(t). (25) (25) For σ = 10, r = 28 and b = 8/3, this system generates chaotic time series. Figure 1 presents Lorenz attractor drawn for 214 points. Figure 2 presents time series generated by state variable x. Fig. 2. Time series generated by variable x of Lorenz’s system. Fig. 2. Time series generated by variable x of Lorenz’s system. Fig. 1. Lorenz’s system attractor. Fig. 2. Time series generated by variable x of Lorenz’s system. Fig. 1. Lorenz’s system attractor. Fig. 1. Lorenz’s system attractor. Fig. 2. Time series generated by variable x of Lorenz’s system. 3 Test time series In this article, times series generated by four systems described below are examined. Nonlinear Anal. Model. Control, 20(4):545–560 3.3 Random and periodic systems Random time series are generated by uniform distributed system. Periodic time series are generated by the system described by the following equations: x(n) = sin(n) + cos(n). (27) (27) 3.2 Henon system Henon system is described by differential equations Henon system is described by differential equations dx(t) dt = 1.4 + 0.3y(t) −x(t)2, dy(t) dt = x(t). (26) (26) (26) Figure 3 presents Henon atractor drawn for 214 points. Figure 4 presents time series generated by state variable x. Nonlinear Anal. Model. Control, 20(4):545–560 552 Ł. Korus, M. Piórek Fig. 3. Henon’s system attractor. Fig. 4. Time series generated by variable x of Henon’s system. Fig. 4. Time series generated by variable x of Henon’s system. Fig 3 Henon’s system attractor Fig. 3. Henon’s system attractor. Fig. 3. Henon’s system attractor. Fig. 4. Time series generated by variable x of Henon’s system. 4.1 Time delay estimation Based on I = f(T) dependency charts presented below, time delay values T for test systems were estimated. Optimal time delay value Topt is estimated either as a ﬁrst minimum of dependency I = f(T) or from the equation I(Topt)/I(0) ≈4/5. Analyzing Fig. 5, it can be assumed that Topt ≈13, because the ﬁrst minimum is located in this area. In order to estimate Fig. 5. I = f(T) dependency chart for Lorenz’s system. Fig. 5. I = f(T) dependency chart for Lorenz’s system. Fig. 6. I = f(T) dependency chart for Henon’s system. Fig. 6. I = f(T) dependency chart for Henon’s system. Fig. 5. I = f(T) dependency chart for Lorenz’s system. Fig. 5. I = f(T) dependency chart for Lorenz’s system. Fig. 6. I = f(T) dependency chart for Henon’s system. http://www.mii.lt/NA Compound method of time series classiﬁcation 553 Fig. 7. I = f(T) dependency chart for random system. Fig. 8. I = f(T) dependency chart for periodic system. Topt for the Henon system, the second criteria was used. It means that the optimal time delay value was taken as 80% of initial value of mutual information I(0). In this case, I(0) ≈7.62, I(Topt) = 6.1 and in consequence Topt ≈1. Analyzing dependency I = f(T) for random system, it can be assumed that all samples for T > 0 are not correlated. For periodic system, a mutual information value I is high in wide range of T. It conﬁrms d i i i f hi Fig. 8. I = f(T) dependency chart for periodic system. Fig. 7. I = f(T) dependency chart for random system. Fig. 7. I = f(T) dependency chart for random system. Fig. 7. I = f(T) dependency chart for random system. Fig. 8. I = f(T) dependency chart for periodic system. Topt for the Henon system, the second criteria was used. It means that the optimal time delay value was taken as 80% of initial value of mutual information I(0). In this case, I(0) ≈7.62, I(Topt) = 6.1 and in consequence Topt ≈1. Analyzing dependency I = f(T) for random system, it can be assumed that all samples for T > 0 are not correlated. For periodic system, a mutual information value I is high in wide range of T. It conﬁrms a deterministic nature of this system. Nonlinear Anal. Model. Control, 20(4):545–560 4.2 Embedding dimension estimation In order to estimate embedding dimension of examined systems, saturation of system invariants have been used. Above charts express dependency between correlation integral and radius used to calculate this invariant. This method was described in Section 2.3 and it assumes control of correlation integral C2 value for increasing value of radius e. Analyzing Fig. 9, it can be stated that the biggest change can be observed for di- mensions lower than dE ⩽3. Further increase of embedding dimension value doesn’t have signiﬁcant impact on examined invariant C2. In conclusion estimated embedding dimension value was taken as dE = 3. In Fig. 10 it can be noticed that the correlation integral C2 is saturated for dE ⩾2. For random time series, correlation integral is never saturated which can be observed in Fig. 11. Analyzing Fig. 12, it can be stated that saturation is present for wide range of dE ⩾1. In the second approach, system’s dynamics can be described by correlation dimension which is one of the fractal dimension and can be also treated as an invariant in saturation method. For instance, analyzing Fig. 13 for Lorenz system, it can be observed that if embed- ding dimension is greater than three (dE ⩾3), correlation dimension value C2 seems to be saturated and doesn’t change to much. For Henon time series saturation is reached for dE ⩾2, which can be easily seen in Fig. 14. 554 Ł. Korus, M. Piórek Fig. 9. Correlation integral C2 versus radius e for Lorenz time series. Fig. 10. Correlation integral C2 versus radius e for Henon time series. Fig. 11. Correlation integral C2 versus radius e for random time series. Fig. 12. Correlation integral C2 versus radius e for periodic time series. Fig. 13. D2 values versus dE for Lorenz time series. Fig. 14. D2 values versus dE for Henon time series. Fig. 9. Correlation integral C2 versus radius e for Lorenz time series. Fig. 10. Correlation integral C2 versus radius e for Henon time series. Fig. 10. Correlation integral C2 versus radius e for Henon time series. Fig. 9. Correlation integral C2 versus radius e for Lorenz time series. Fig. 10. Correlation integral C2 versus radius e for Henon time series. Fig. 9. Correlation integral C2 versus radius e for Lorenz time series. Fig. 11. Correlation integral C2 versus radius e for random time series. Fig. 12. 4.2 Embedding dimension estimation Correlation integral C2 versus radius e for periodic time series. Fig. 11. Correlation integral C2 versus radius e for random time series. Fig. 12. Correlation integral C2 versus radius e for periodic time series. Fig. 14. D2 values versus dE for Henon time series. Fig. 13. D2 values versus dE for Lorenz time series. As it was mentioned before, for random time series saturation is never observed, which can be also assumed from Fig. 15. It means that the distance between pairs of points in state space never became a constant value. Figure 16 presents that saturation is reached for dE = 1. http://www.mii.lt/NA Compound method of time series classiﬁcation 555 Fig. 15. D2 values versus dE for random time series. Fig. 16. D2 values versus dE for periodic time series. Fig. 15. D2 values versus dE for random time series. Fig. 16. D2 values versus dE for periodic time series. 4.3 Hurst exponent analysis Below ﬁgures allow to estimate Hurst exponent using linear regression method. Fig. 17. ln(R/S) versus ln(n) for Lorenz time series. Fig. 18. ln(R/S) versus ln(n) for Henon time series. Fig. 19. ln(R/S) versus ln(n) for random time series. Fig. 20. ln(R/S) versus ln(n) for periodic time series. e Fig. 18. ln(R/S) versus ln(n) for Henon time series. Fig. 17. ln(R/S) versus ln(n) for Lorenz time series. Fig. 17. ln(R/S) versus ln(n) for Lorenz time series. Fig. 18. ln(R/S) versus ln(n) for Henon time series. Fig. 18. ln(R/S) versus ln(n) for Henon time series. Fig. 20. ln(R/S) versus ln(n) for periodic time series. Fig. 19. ln(R/S) versus ln(n) for random time series. Nonlinear Anal. Model. Control, 20(4):545–560 556 Ł. Korus, M. Piórek Analyzing Fig. 17, it can be stated that examined time series is persistent and H ≈0.8. Based on results of Hurst analysis presented in Fig. 18, it can be stated that Henon time series is antipersistent and H ≈0.42. Results of R/S analysis for random time series can be found in Fig. 19. Analyzing this ﬁgure, it can be concluded that H ≈0.5, which conﬁrms randomness of this time series. Figure 20 conﬁrms a deterministic character of examined time series (H ≈1). 4.4 Lyapunov exponents Below ﬁgures present the largest Lyapunov exponent estimations for the test time series computed by using Wolf’s algorithm [40]. Based on below ﬁgures it can be stated that for chaotic times series generated by Lorenz and Henon system, the largest Lyapunov exponent are positive. It means that these systems are highly sensitive to initial condi- tions. Small change of initial state, generates exponential increasing distance between trajectories. Unfortunately, Lyapunov exponents can not be used to differ chaotic and random time series, because both of them give positive values. The largest Lyapunov values for random times series are presented on Fig. 23. Fig. 22. The largest Lyapunov exponent for Henon time series. Fig. 24. The largest Lyapunov exponent for Periodic time series. http://www mii lt/NA Fig. 21. The largest Lyapunov exponent for Lorenz time series. Fig. 22. The largest Lyapunov exponent for Henon time series. Fig. 23. The largest Lyapunov exponent for Random time series. Fig. 24. The largest Lyapunov exponent for Periodic time series. Fig. 22. The largest Lyapunov exponent for Henon time series. Fig. 21. The largest Lyapunov exponent for Lorenz time series. Fig. 22. The largest Lyapunov exponent for Henon time series. Fig. 22. The largest Lyapunov exponent for Henon time series. Fig. 23. The largest Lyapunov exponent for Random time series. Fig. 24. The largest Lyapunov exponent for Periodic time series. Fig. 23. The largest Lyapunov exponent for Random time series. http://www.mii.lt/NA 557 Compound method of time series classiﬁcation As it can be seen on Fig. 24, the largest Lyapunov exponent for periodic system is negative, which conﬁrms the existence of non-chaotic attractor. 5 Conclusions In this article, three types of times series were examined: chaotic, random and periodic. To differ type of behavior, the following methods were used: mutual information to estimate time delay Topt, saturation of system’s invariants for estimation of embedding dimension dE, correlation dimension to estimate fractal dimension D2, Hurst exponent to examinate memory effect H and the largest Lyapunov exponent to check dependency on initial conditions. Values of all of mentioned above parameters for all examined systems are gathered in Table 1. Analyzing Topt for random system, it can be stated that information in the system is lost very quickly and it is exactly opposite to periodic system, where this dependency is continously observable. Analyzing fractal dimension, it can be noticed that for chaotic time series, saturated values are non-integer, while for periodic are very close to integer and inﬁnity for random. Hurst analysis allows to differ mainly random and periodic behaviors. Lyapunov exponents are positive for chaotic time series (Lorenz and Henon) and random. For periodic time series the largest Lyapunov exponent is nega- tive. Based on above summary, we can conclude that below set of indicators are properly selected to differ behaviors in time series. If there is a need to distinguish chaotic and periodic behavior in time series, the largest Lyapunov estimation is suggested method. On the other hand, if the goal is to differ random and periodic times series, Hurst exponent estimation could be useful. Fractal dimension estimations can provide information about attractor, which in consequence allows simple method to differ chaotic from periodic systems. To sum it up, it can be stated that there is none common method which gives a possibility to fully differ three types of behavior: chaotic, random and periodic. It is necessary to use wider set of methods, which provide more complex information about different properties of examined times series. Additional conclusion which can be drawn from this paper, is that correlation dimen- sion seems to be very efﬁcient invariant in the saturation of system’s invariant method. Saturation of correlation dimension instead of correlation integral is more precise and in- tuitive, which can be observed on Figs. 13–16. This approach allows to deﬁne embedding dimension more clearly. Table 1. Parameters estimation results for examined time series. Nonlinear Anal. Model. Control, 20(4):545–560 References 1. H. Abarbanel, Analysis of Observed Chaotic Data, Springer-Verlag, New York, 1996. 2. A.M. Albano et al., Using high-order correlations to deﬁne an embedding window, Physica D, 54:85–97, 1991. 3. B.R. Andrievskii, A.L. Fradkov, Control of chaos: Methods and applications, Autom. Remote Control, 64(5):673–713, 2003. 4. R.G. Andrzejak, K. Lehnertz, F. Mormann, C. Rieke, P. David, C.E. Elger, Indications of nonlinear deterministic and ﬁnite-dimensional structures in time series of brain electrical activity: Dependence on recording region and brain state, Phys. Rev. E, 64(1):1–8, 2001. 5. F. Argoul, A. Arneodo, P. Richetti, J.C. Roux, From quasiperiodicity to chaos in the Belousov– Zhabotinskii reaction. I. Experiment, J. Chem. Phys., 86(6):3325–3339, 1987. 6. S. Banerjee, A.P. Misra, P.K. Shukla, L. Rondoni, Spatiotemporal chaos and the dynamics of coupled langmuir and ion-acoustic waves in plasmas, Phys. Rev. E, 81(1):1–10, 2010. 7. M. Barnsley, Fractals Everywhere, Academic Press, New York, 1988. 8. A. Boukabou, N. Mansouri, Predictive control of higher dimensional chaos, Nonlinear Phenom. Complex Syst., Minsk, 8(3):258–265, 2005. 9. R. Brown, M. Kennel, H. Abarbanel, Determinig embedding dimension for phase-space reconstruction using a geometrical construction, Phys. Rev. A, 45(6):3403–3411, 1992. 10. T. Buzug, G. Pﬁster, Comparison of algorithms calculating optimal embedding parameters for delay time coordinates, Physica D, 58:127–137, 1992. 11. T. Buzug, G. Pﬁster, Optimal delay time and embedding dimension for delay-time coordinates by analysis of the global static and local dynamical behaviour of strange attractors, Phys. Rev. A, 45:7073–7084, 1992. 12. S.T. Chui, K.B. Ma, Nature of some chaotic states for Dufﬁng’s equation, Phys. Rev. A, 26(4):2262–2265, 1982. 13. J.J. Colins, M.T. Rossenstein, C.J. de Luca, A practical method for calculating largest Lyapunov exponents from small data sets, Physica D, 65:117–134, 1993. 14. J.J. Colins, M.T. Rossenstein, C.J. de Luca, Reconstruction expansion as a geometry-based framework for choosing proper delay times. Physica D, 73:82–98, 1994. 15. A. Córdoba, M.C. Lemos, F. Jiménez-Morales, Periodical forcing for the control of chaos in a chemical reaction, J. Chem. Phys., 124, 014707, 6 pp., 2006. 16. K. Falconer, Fractal Geometry, Wiley, New York, 1990. 16. K. Falconer, Fractal Geometry, Wiley, New York, 1990. 17. A.M. Fraser, H.L. Swinney, Independent coordinates for strange attractors from mutual information, Phys. Rev. A, 33(2):1134–1140, 1986. 18. T. Gautama, D.P. Mandic, M.M. Van Hulle, Indications of nonlinear structures in brain electrical activity, Phys. Rev. E, 67(4), 046204, 5 pp., 2003. 19. R.B. Govindan, K. Narayanan, M.S. 5 Conclusions Parameters Lorenz Henon Random Periodic Topt 13 1 NA NA dE 3 2 ∞ 1 D2 1.98 1.19 ∞ 1 H 0.80 0.42 0.5 1 L + + + − Nonlinear Anal. Model. Control, 20(4):545–560 Table 1. Parameters estimation results for examined time series. Parameters Lorenz Henon Random Periodic Topt 13 1 NA NA dE 3 2 ∞ 1 D2 1.98 1.19 ∞ 1 H 0.80 0.42 0.5 1 L + + + − Nonlinear Anal. Model. Control, 20(4):545–560 Table 1. Parameters estimation results for examined time series. Parameters Lorenz Henon Random Periodic Topt 13 1 NA NA dE 3 2 ∞ 1 D2 1.98 1.19 ∞ 1 H 0.80 0.42 0.5 1 L + + + − 558 Ł. Korus, M. Piórek Ł. Korus, M. Piórek References Gopinathan, On the evidence of deterministic chaos in ECG: Surrogate and predictability analysis, Chaos, 8(2):495–502, 1998. http://www.mii.lt/NA 559 Compound method of time series classiﬁcation 20. P. Grassberger, I. Procaccia, Measuring the strangeness of strange attractors, Physica D, 9(1–2):189–208, 1983. 21. G.H. Gunaratne, P.S. Lisnay, M.J. Vinson, Chaos beyond onset: A comparison of theory and experiment, Phys. Rev. Lett., 63(1), 4 pp., 1989. 22. G.A. Held, C. Jeffries, E.E. Haller. Observation of chaotic behavior in an electron-hole plasma in ge. Phys. Rev. Lett., 52(12):1037–1040, 1984. 23. H.E. Hurst, Long-Term Storage Capacity of Reservoirs, Transactions of the American Society of Civil Engineers, Vol. 116, American Society of Civil Engineers, New York, 1951, pp. 770– 808. 24. H. Kantz, A robust method to estimate the maximal Lyapunov exponent of a time series, Phys. Lett. A, 185:77–87, 1994. 25. L. Korus, Simple environment for developing methods of controllinig chaos in spatially distributed systems, Int. J. Appl. Math. Comput. Sci., 21(1):149–159, 2011. 26. J. Langenberg, G. Pﬁster, J. Abshagen, Chaos from Hopf bifurcation in a ﬂuid ﬂow experiment, Phys. Rev. E, 70(1), 5 pp., 2004. 27. J. Lin, Y. Wang, Z. Huang, Z. Shen, Selection of proper time-delay in phase space reconstruction of speech signals, Signal Process., 15(2):220–225, 1999. 28. B. Luciano, Geometry of dynamical systems and topological stability: From bifurcation, chaos and fractals to dynamics in natural and life sciences, Int. J. Bifurcation Chaos Appl. Sci. Eng., 21(3):815–867, 2011. 29. H.-G. Ma, C.-Z. Han, Selection of embedding dimension and delay time in phase space reconstruction, Front. Electr. Electron. Eng. China, 1:111–114, 2006. 30. A. Maus, J.C. Sprott, Neural network method for determining embedding dimension of a time series, Commun. Nonlinear Sci. Numer. Simul., 16:3294–3302, 2011. 31. P. Parmananda, Controlling turbulence in coupled map lattice systems using feedback techniques, Phys. Rev. E, 56(1):239–244, 1997. 32. I. Procaccia, E. Meron, Low-dimensional chaos in surface waves: Theoretical analysis of an experiment, Phys. Rev. A, 34(4):3221–3237, 1986. 33. E. Rafajłowicz, Testing (non-)existence of input-output relationships by estimating fractal dimensions, IEEE Trans. Signal Process., 52(11):3151–3159, 2004. 34. T. Sauer, J.A. Yorke, M. Casdagli, Embedology, J. Stat. Phys., 65(3–4):579–616, 1991. 35. E. Skubalska-Rafajłowicz, A new method of estimation of the box counting dimension of multivariate objects using space-ﬁlling curves, Nonlinear Anal., Theory Methods Appl., Ser. A, Theory Methods, 63(5–7):1281–1287, 2005. 36. J. Stark, Delay embeddings for forced systems. I: Deterministic forcing, J. Nonlinear Sci., 9(3):255–332, 1999. 37. F. 42. G. Yim, J. Ryu, Y. Park, S. Rim, S. Lee, W. Kye, C. Kim, Chaotic behaviors of operational ampliﬁers, Phys. Rev. E, 69(1), 045201(R), 4 pp., 2004. 41. T. Yamada, R. Graham, Chaos in a laser system under a modulated external ﬁeld, Phys. Rev. Lett., 45(16):1322–1324, 1980. References Takens, Detecting Strange Attractors in Turbulence, Springer-Verlag, Berlin, 1981. 38. J. Used, J.C. Martin, Multiple topological structures of chaotic attractors ruling the emission of a driven laser, Phys. Rev. E, 82(1), 016218, 7 pp., 2010. Nonlinear Anal. Model. Control, 20(4):545–560 560 Ł. Korus, M. Piórek 39. W. Wei, L. Zonghua, H. Bambi, Phase order in chaotic maps and in coupled map lattices, Phys. Rev. Lett., 84(12):2610–2613, 2000. 40. A. Wolf, J.B. Swift, H.L. Swinney, J.A Vastano, Determining Lyapunov exponents from a time series, Physica D, 16:285–317, 1985. 41. T. Yamada, R. Graham, Chaos in a laser system under a modulated external ﬁeld, Phys. Rev. Lett., 45(16):1322–1324, 1980. http://www.mii.lt/NA http://www.mii.lt/NA
https://openalex.org/W3180780402	https://hal.science/hal-03458198/document	English	null	DMS emissions from the Arctic marginal ice zone	Elementa	2,021	cc-by	26,252	DMS emissions from the Arctic marginal ice zone Martí Galí, Martine Lizotte, David Kieber, Achim Randelhoff, Rachel Hussherr, Lei Xue, Julie Dinasquet, Marcel Babin, Eric Rehm, Maurice Levasseur To cite this version: Martí Galí, Martine Lizotte, David Kieber, Achim Randelhoff, Rachel Hussherr, et al.. DMS emissions from the Arctic marginal ice zone. Elementa: Science of the Anthropocene, 2021, 9 (1), 10.1525/ele- menta.2020.00113. hal-03458198 Distributed under a Creative Commons Attribution 4.0 International License 1Takuvik International Research Laboratory, Laval University (Canada), CNRS, France 2De´partement de biologie et Que´bec-Oce´an, Universite´ Laval, Que´bec City, QC, Canada 3Barcelona Supercomputing Center (BSC) 4Department of Chemistry, College of Environmental Science and Forestry, State University of New York, Syracuse, NY, USA 5Marine Biology Research Division, Scripps Institution of Oceanography, La Jolla, CA, USA 6Sorbonne Universite´, Laboratoire d’Oce´ anographie Microbienne, Banyuls-sur-Mer, France * Corresponding author: Email: marti.gali.tapias@gmail.com DMS emissions from the Arctic marginal ice zone Martı´ Galı´1,2,3,, Martine Lizotte1,2, David J. Kieber4, Achim Randelhoff1,2, Rachel Hussherr1,2, Lei Xue4, Julie Dinasquet5,6, Marcel Babin1,2, Eric Rehm1,2, and Maurice Levasseur1,2 Phytoplankton blooms in the Arctic marginal ice zone (MIZ) can be prolific dimethylsulfide (DMS) producers, thereby influencing regional aerosol formation and cloud radiative forcing. Here we describe the distribution of DMS and its precursor dimethylsulfoniopropionate (DMSP) across the Baffin Bay receding ice edge in early summer 2016. Overall, DMS and total DMSP (DMSPt) increased towards warmer waters of Atlantic origin concurrently with more advanced ice-melt and bloom stages. Relatively high DMS and DMSPt (medians of 6.3 and 70 nM, respectively) were observed in the surface layer (0–9 m depth), and very high values (reaching 74 and 524 nM, respectively) at the subsurface biomass maximum (15–30 m depth). Microscopic and pigment analyses indicated that subsurface DMS and DMSPt peaks were associated with Phaeocystis pouchetii, which bloomed in Atlantic-influenced waters and reached unprecedented biomass levels in Baffin Bay. In surface waters, DMS concentrations and DMS:DMSPt ratios were higher in the MIZ (medians of 12 nM and 0.15, respectively) than in fully ice-covered or ice-free conditions, potentially associated with enhanced phytoplanktonic DMSP release and bacterial DMSP cleavage (high dddP:dmdA gene ratios). Mean sea–air DMS fluxes (mmol m–2 d–1) increased from 0.3 in ice-covered waters to 10 in open waters (maximum of 26) owing to concurrent trends in near-surface DMS concentrations and physical drivers of gas exchange. Using remotely sensed sea-ice coverage and a compilation of sea–air DMS flux data, we estimated that the pan-Arctic DMS emission from the MIZ (EDMS, MIZ) was 5–13 Gg S yr–1. North of 80N, EDMS, MIZ might have increased by around 10 + 4% yr–1 between 2003 and 2014, likely exceeding open-water emissions in June and July. We conclude that EDMS, MIZ must be taken into account to evaluate plankton-climate feedbacks in the Arctic. Keywords: Dimethylsulfide, Dimethylsulfoniopropionate, Arctic, Marginal ice zone, Microbial plankton, Phaeocystis pouchetii HAL Id: hal-03458198 https://hal.science/hal-03458198v1 Submitted on 8 Apr 2022 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License Galı´, M, et al. 2021. DMS emissions from the Arctic marginal ice zone. Elem Sci Anth, 9: 1. DOI: https://doi.org/10.1525/ elementa.2020.00113 RESEARCH ARTICLE DMS emissions from the Arctic marginal ice zone Martı´ Galı´1,2,3,, Martine Lizotte1,2, David J. Kieber4, Achim Randelhoff1,2, Rachel Hussherr1,2, Lei Xue4, Julie Dinasquet5,6, Marcel Babin1,2, Eric Rehm1,2, and Maurice Levasseur1,2 Keywords: Dimethylsulfide, Dimethylsulfoniopropionate, Arctic, Marginal ice zone, Microbial plankton, Phaeocystis pouchetii 1. Introduction These pulses of sympagic and pelagic algal growth, largely dominated by diatom species, account for a major portion of annual primary production in polar waters (Perrette et al., 2011; Wassmann and Re- igstad, 2011; Renaut et al., 2018) and result in the emis- sion of a wide diversity of biogenic particles and gases to the atmosphere (Levasseur, 2013; Gabric et al., 2018; Ab- batt et al., 2019). The emission of DMS from seawater results from a com- plex network of biotic and abiotic processes (Simo´, 2001). The primary precursor of DMS is dimethylsulfoniopropio- nate (DMSP), a compound synthesized mainly by marine algae (Stefels et al., 2007) and also by some bacteria (Cur- son et al., 2017). In phytoplankton, DMSP serves different physiological functions such as osmoregulation, antioxida- tion or cryoprotection (Stefels et al., 2007), and its intra- cellular concentrations vary widely across taxonomic groups, with most high DMSP producers found among the haptophytes and dinoflagellates (Keller, 1989; McParland and Levine, 2019). Conversion of DMSP to DMS proceeds mostly through the enzymatic cleavage by algal (Alcolom- bri et al., 2015) and bacterial (Curson et al., 2011) DMSP- lyase enzymes. The cleavage pathway typically accounts for <30% of the microbial DMSP consumption (Kiene and Linn, 2000; Galı´ and Simo´, 2015), and the remainder is transferred to bacteria and protists and diverted away from DMS production. However, the efficiency of the con- version of DMSP to DMS can vary widely in response to food-web processes, such as grazing (Simo´ et al., 2018), bacterial metabolism (Kiene et al., 2000), and environ- mental stressors, such as nutrient limitation (Sunda et al., 2007), high irradiance (Galı´ et al., 2013), and ocean acidification (Be´nard et al., 2021). In the upper mixed layer of the ocean, dissolved DMS is readily transformed to dissolved non-volatile sulfur species by DMS-consuming bacteria (Kiene and Bates, 1990; del Valle et al., 2009; Lidbury et al., 2016) and photolysis (Kieber et al., 1996; Galı´ et al., 2016). In consequence, typically less than 15% of the DMS produced in the upper mixed layer is vented to the atmosphere in ice-free waters (Galı´ and Simo´, 2015). Biogenic emissions play an important role in the Arctic climate in late spring and summer, when biological activ- ity is maximal and the atmosphere is depleted of aerosols (both natural and anthropogenic; Abbatt et al., 2019). 1. Introduction known as the marginal ice zone (MIZ) will move every year from its southernmost extent in late winter to its north- ernmost extent prior to complete melt in late summer. Therefore, biogeochemical processes occurring within the MIZ, including the emission of biogenic gases and parti- cles to the atmosphere, have the potential to reshape marine ecosystems and geosystems across the entire Arctic Ocean. The shrinking and thinning of Arctic sea ice are among the most striking consequences of anthropogenic global warming (Arctic Monitoring & Assessment Programme [AMAP], 2017; Intergovernmental Panel on Climate Change [IPCC], 2019). Model projections indicate that the Arctic could become entirely ice free in summer by 2040– 2050 (Thackeray and Hall, 2019). In this new regime, devoid of multiyear ice, the belt of retreating sea ice The MIZ is an ephemeral environment, in this study operationally defined as the zone where the sea-ice cover decreases from nearly total (>85%) to nearly absent (<15%) before permanent ice opening in spring and sum- mer. Several tens or a few hundred kilometers wide (Strong, 2012; Strong and Rigor, 2013) and with a duration of days to weeks (Randelhoff et al., 2019), the MIZ hosts intense exchanges of freshwater, heat and momentum between the ocean and the atmosphere (McPhee, 2008). From a biogeochemical standpoint, the MIZ period marks the transition between two major blooms of microalgae in the Arctic. Prior to the melt season, biological activity generally concentrates in the bottom layer of the sea ice, where sympagic algae grow massively as soon as sufficient light penetrates through the ice or snow surface (Leu et Art. 9(1) page 2 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Ghahremaninezhad et al., 2019), and also the episodic nature of new particle formation events (Collins et al., 2017; Dall’Osto et al., 2017; Heintzenberg et al., 2017). These findings stress the need to better understand and quantify DMS emissions from ephemeral environments such as the MIZ that can produce intense but short-lived emission events. al., 2015). When sufficient sunlight crosses the sea ice or the free ocean surface, phytoplankton can also bloom in the water column, taking advantage of nutrient stocks replenished over the winter and stable stratification caused by ice melt. 1. Introduction The low aerosol baseline, caused by limited transport from lower latitudes and efficient scavenging (Heintzenberg et al., 2015; Croft et al., 2016), favors secondary aerosol for- mation from local gaseous emissions (Leaitch et al., 2013; Collins et al., 2017). The gas dimethylsulfide (DMS) is, quantitatively, the main volatile organic compound emit- ted by the activity of marine and sea-ice microbes (Simo´, 2011; Carpenter et al., 2012). In the clean Arctic atmo- sphere, DMS oxidation end-products (mostly, sulfuric and methanesulfonic acids) readily form new aerosol particles that scatter sunlight (Dawson et al., 2012; Leaitch et al., 2013; Hodshire et al., 2019; Veres et al., 2020; Brean et al., 2021). Moreover, DMS derivatives contribute to the con- densational growth of aerosols into cloud condensation nuclei (CCN; Ghahremaninezhad et al., 2019), altering the balance between cloud shortwave forcing (increasing cloud albedo) and longwave forcing (increasing heat retention) (Andreae and Rosenfeld, 2008; Carslaw et al., 2013; Mahmood et al., 2019). In aerosol-depleted areas and over the pack ice, this radiative balance is particularly sensitive to local CCN sources (Mauritsen et al., 2011). Thus, ongoing changes in DMS emission patterns and their underlying drivers (Six et al., 2013; Galı´ et al., 2019) can influence the ensemble of climate feedbacks, known as Arctic amplification, that cause the Arctic to warm faster than the global average (Serreze and Barry, 2011). In turn, changes in atmospheric radiative transfer and ice dynamics can modulate phytoplankton productiv- ity, potentially establishing a feedback mechanism (Charl- son et al., 1987) that has yet to be verified (Quinn and Bates, 2011). Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on The multiple biogeochemical pathways that mediate between DMSP synthesis and DMS emission generally tend to buffer DMS concentrations in open ocean settings (Galı´ and Simo´, 2015), where sea-surface DMS concentra- tions above 10 nM are rare (Lana et al., 2011). 1. Introduction In the MIZ, however, seawater DMS concentrations and sea–air DMS fluxes (FDMS) can greatly increase when some factors co- occur, namely: the formation of massive phytoplankton blooms (Perrette et al., 2011; Arrigo et al., 2014; Renaut et al., 2018) depending on the type of ice edge (multi-year versus first-year sea ice; Lizotte et al., 2020); the presence of high DMSP and DMS producers (Levasseur, 2013); the enhancement of DMS production by food-web interac- tions or physiological stress (Matrai et al., 1995; Sunda et al., 2007; Galı´ et al., 2013); and weak biological DMS removal during early bloom stages (del Valle et al., 2009). Among these factors, blooms of the colony-forming hap- tophyte Phaeocystis are thought to play a prominent role. This genus has an intracellular DMSP content of about 5%–10% of cell carbon, roughly one order of magnitude higher than diatoms and, unlike diatoms, displays high DMS emission (EDMS) in the Arctic originates from both surface seawater and ice environments such as the ice bottom (Galindo et al., 2014) and saline melt ponds (Gourdal et al., 2018; Park et al., 2019). Current estimates obtained with different approaches indicate that ice-free waters largely dominate EDMS regionally at the pan-Arctic level (Lana et al., 2011; Galı´ et al., 2019; Hayashida et al., 2020). Yet, DMS has an atmospheric lifetime of a few days, at most, which limits atmospheric transport and magnifies the effects of local DMS sources on aerosols in conjunction with local meteorological conditions. Recent studies high- lighted the broad variability and patchiness in tropo- spheric DMS concentrations (Mungall et al., 2016; Art. 9(1) page 3 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Figure 1. Maps of the study area and sea-surface DMS concentrations and fluxes (FDMS). (a) Bathymetric map of southern Baffin Bay, Davis Strait and northern Labrador Sea. Dots show the grid of Green Edge cruise stations, colored according to the “Arctic N-P relationship” (ANP) at 20 m, which distinguishes waters of Arctic (blue) and Atlantic (red) origin, whose approximate circulation is shown by arrows; (b) close-up of the study area. The size of empty and filled circles indicates the magnitude of sea-surface DMS concentrations and sea–air DMS fluxes, respectively, and the color of filled circles indicates whether a station is classified as ICE, MIZ, or OW (Table 1). 1. Introduction Contours of satellite-retrieved sea-ice fractional coverage are shown in different textures and shades of blue for different periods of 5 days of the year: d176–d180 (0.10 contour), d181–d185 (0.10 contour), and d186–d190 (0.10, 0.50, 0.80, and 0.85 contours). Sampled stations and the start dates of sampled transects (Julian day, d) are also indicated. Stations not sampled for sulfur compounds are marked with crosses. DOI: https://doi.org/10.1525/elementa.2020.00113.f1 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by gues Figure 1. Maps of the study area and sea-surface DMS concentrations and fluxes (FDMS). (a) Bathymetric map of southern Baffin Bay, Davis Strait and northern Labrador Sea. Dots show the grid of Green Edge cruise stations, colored according to the “Arctic N-P relationship” (ANP) at 20 m, which distinguishes waters of Arctic (blue) and Atlantic (red) origin, whose approximate circulation is shown by arrows; (b) close-up of the study area. The size of empty and filled circles indicates the magnitude of sea-surface DMS concentrations and sea–air DMS fluxes, respectively, and the color of filled circles indicates whether a station is classified as ICE, MIZ, or OW (Table 1). Contours of satellite-retrieved sea-ice fractional coverage are shown in different textures and shades of blue for different periods of 5 days of the year: d176–d180 (0.10 contour), d181–d185 (0.10 contour), and d186–d190 (0.10, 0.50, 0.80, and 0.85 contours). Sampled stations and the start dates of sampled transects (Julian day, d) are also indicated. Stations not sampled for sulfur compounds are marked with crosses. DOI: https://doi.org/10.1525/elementa.2020.00113.f1 DMSP-lyase activity (Sheehan and Petrou, 2019; Stefels et al., 2007). These factors, together with the global distribu- tion of bloom-forming Phaeocystis species, make Phaeocys- tis blooms a major source of atmospheric DMS globally (Schoemann et al., 2005; Wang et al., 2015). transmissometer (0.25-m pathlength). The CTD data were processed according to standard Amundsen Science Team procedures (P Guillot, personal communication). Transmis- someter profiles were used to calculate the beam attenu- ation coefficient at 650 nm due to particles, cp (m–1), following standard procedures (Loisel and Morel, 1998). Particulate organic carbon (POC; units of mg m–3) was estimated from cp using a conversion factor of 500 mg POC m–2 (Cetinic´ et al., 2012, and references therein). 1. Introduction Water samples were collected at several depths using Ni- skin bottles mounted on the CTD rosette for the determi- nation of dimethylated sulfur compounds (Section 2.4), phytoplankton pigments and abundance (Section 2.6), bacterial abundance and sulfur metabolism (Section 2.7), and inorganic nutrients (nitrate, nitrite, phosphate, and silicate; Koroleff and Hansen, 1999; see details in Randelh- off et al., 2019). In this article we describe spatial DMS and DMSP con- centration patterns across a receding first-year ice edge in Baffin Bay, seeking to disentangle the role of Pacific- derived Arctic water masses and Atlantic water masses, meltwater stratification, and phytoplankton bloom pro- gression, with particular attention to Phaeocystis pouchetii. Finally, we compare our findings to a compilation of in situ studies, and upscale in situ observations using remote sensing to provide pan-Arctic estimates of DMS emission from the MIZ (EDMS, MIZ) between 2003 and 2016. 2.2. Satellite sea-ice concentration data sets DOI: https://doi.org/10.1525/elementa.2020.00113.t1 Characteristic Ice-Cover Category ICE MIZ OW Descriptors of the stations OWD thresholds (d) OWD < –3 –3 OWD 3 OWD > 3 Mean observed OWD –12 0 13 SIC thresholds (%) (days –2, –1, 0)a SIC > 85 85 SIC 15 SIC < 15 Mean observed SIC (%) (days –2, –1, 0)a 94, 93, 90 52, 45, 15 0, 1, 0 Arctic-domain stations 400, 403, 406, 600, 719 519, 603, 604.5, 605, 716 608, 700, 703, 707, 713 Atlantic-domain stations 409, 412, 413 507, 512, 515 418, 506, 612, 615 Total number of stations 8 8 9 Descriptors of the water column physics and biogeochemistry: mean (min–max) MLD0.03 (m)b 6 (2–22) 3 (1–5) 7 (1–15) hBD (m)c 20 (15–32) 15 (11–17) 18 (12–31) Isolume 0.415 (m) 26 (11–41) 35 (21–47) 42 (29–60) Nitracline (m) 1 (0–7) 4 (0–15) 21 (1–43) SBM (m)d 8 (3–21) 12 (7–16) 24 (16–31) a Satellite-retrieved sea-ice concentration (SIC); (days –2 –1 0) refers to the 3-day period ending on the sampling day Characteristic Ice-Cover Category ICE MIZ OW Descriptors of the stations OWD thresholds (d) OWD < –3 –3 OWD 3 OWD > 3 Mean observed OWD –12 0 13 SIC thresholds (%) (days –2, –1, 0)a SIC > 85 85 SIC 15 SIC < 15 Mean observed SIC (%) (days –2, –1, 0)a 94, 93, 90 52, 45, 15 0, 1, 0 Arctic-domain stations 400, 403, 406, 600, 719 519, 603, 604.5, 605, 716 608, 700, 703, 707, 713 Atlantic-domain stations 409, 412, 413 507, 512, 515 418, 506, 612, 615 Total number of stations 8 8 9 Descriptors of the water column physics and biogeochemistry: mean (min–max) Descriptors of the water column physics and biogeochemistry: mean (min–max) MLD0.03 (m)b 6 (2–22) 3 (1–5) 7 (1–15) hBD (m)c 20 (15–32) 15 (11–17) 18 (12–31) Isolume 0.415 (m) 26 (11–41) 35 (21–47) 42 (29–60) Nitracline (m) 1 (0–7) 4 (0–15) 21 (1–43) SBM (m)d 8 (3–21) 12 (7–16) 24 (16–31) a Satellite-retrieved sea-ice concentration (SIC); (days –2, –1, 0) refers to the 3-day period ending on the sampling b Shallowest mixed layer depth where sigma-t exceeded by 0.03 kg m–3 the value at 1 m. c Equivalent mixed layer depth (as defined by Randelhoff et al., 2017). d Subsurface biomass maximum. 2.2. Satellite sea-ice concentration data sets An additional SIC dataset for the entire Arctic between 2003 and 2016 was downloaded from the National Snow and Ice Data Center (http://nsidc.org), and consisted of images obtained by the Special Sensor Microwave/Imager (2003–2007) and the Special Sensor Microwave Imager/ Sounder (2008–2016). Daily SIC images at 25-km resolu- tion were re-projected onto a 4.64-km sinusoidal grid to better match ocean color remote sensing products, and then bin-averaged into 6 6 macropixels (27.84 km) and 8-day periods. This dataset was used to compute pan- Arctic estimates of EDMS, MIZ (Section 2.9) in a format that was directly comparable to previous satellite-based esti- mates of EDMS from open waters (Galı´ et al., 2019). Thus, the mean irradiance during the 24-h prior to sam- pling at each station was computed from the hourly SBDART output. p Downwelling PAR was measured in the water column down to 100 m using a Compact Optical Profiling System (C-OPS, Biospherical Instruments Inc., San Diego, CA). Underwater measurements were referenced to simulta- neous above-surface measurements to calculate the verti- cal PAR transmittance profile for a given C-OPS cast. To correct for the effect of variable ice cover, the mean PAR transmittance profile at each station was computed as the SIC-weighted average of the closest ice-free and ice- covered C-OPS casts. To obtain the mean daily PAR profile, the mean transmittance profile was multiplied by the mean daily PAR0þ (see details about the measurements and calculations in Randelhoff et al. 2019). 2.3. Irradiance lli h Downwelling photosynthetically active radiation (PAR) immediately above the sea surface (PAR0þ) was estimated with the Santa Barbara DISORT Atmospheric Radiative Transfer (SBDART) model (Ricchiazzi et al., 1998) using precomputed look-up tables (Laliberte´ et al., 2016). Hourly SBDART time series showed good agreement with in situ data recorded on the ship’s meteorological tower (Kipp and Zonen PAR Lite; sampling frequency 1 min–1), with R2 ¼ 0.88 and mean bias ¼ –6%. Measurements taken on the ship were deemed not representative of the daily mean PAR0þ at any fixed station because they reflected rapid variations in ice albedo and fog conditions as the CCGS Amundsen traversed back and forth across the MIZ. 2.2. Satellite sea-ice concentration data sets The Green Edge expedition sampled seven transects across the ice edge in southern Baffin Bay, north of the Davis Strait (approximately at 70N and 60W) aboard the CCGS Amundsen research icebreaker. Here we report measure- ments from four East–West transects, named T4, T5, T6 and T7, sampled during the second leg of the cruise between June 24 and July 10, 2016 (Figure 1). At each station along a given transect, vertical profiles of water column properties were measured down to at least 300 m using a sampling rosette with attached Seabird SBE-911 plus conductivity-temperature-depth (CTD) sensors, a Sea- point chlorophyll fluorometer, and a WetLabs C-Star Sea-ice fractional coverage was retrieved from microwave satellite sensors, and will be referred to as sea-ice concen- tration (SIC) hereafter, following the conventional name of the remote sensing products. Daily SIC images from the Advanced Microwave Scanning Radiometer 2 (AMSR2) on a 3.125-km grid (Spreen et al., 2008) were downloaded from ttp://www.iup.uni-bremen.de:8084/amsr2data/asi_ daygrid_swath/n3125/ and used to describe ice dynamics over the entire melt season in the Green Edge study area (Figure 1). From this dataset, a SIC time series for each station was extracted from the closest satellite pixel. Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 4 of 31 Table 1. Physical and biogeochemical characteristics of the stations across the ice edge, classified according to the open water days (OWD) metric. DOI: https://doi.org/10.1525/elementa.2020.00113.t1 Table 1. Physical and biogeochemical characteristics of the stations across the ice edge, classified according to the open water days (OWD) metric. Galı´ et al: DMS Emissions from the Arctic MIZ Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 5 of 31 Samples for DMS analysis were collected into 120-mL precleaned borosilicate vials leaving no head space, allow- ing some overflow and taking care to avoid bubbling. The vials were filled using a precleaned silicone tube con- nected to the Niskin bottle spigot, fitted with a 5-mm Nitex mesh at its tip to gently screen out large phyto- plankton cells and colonies (Kinsey et al., 2016). Occasion- ally, additional vials were filled using a 100-mm mesh Nylon screen to test for DMS measurement artifacts asso- ciated with Phaeocystis cell breakage upon filtration. DMS samples were stored in the dark at 4C and processed within 2 h of collection. Processed DMSP samples were analyzed at Laval Uni- versity (UL, Que´bec City) and at the State University of New York, College of Environmental Science and Forestry (SUNY-ESF, Syracuse, NY) within 30 months of collection. Storage tests confirmed the good preservation of samples (>95% recovery), and excellent agreement between labo- ratories was found (mean difference of 1.2 + 10%). At both laboratories, DMSP was analyzed as DMS after stoi- chiometric base (NaOH) cleavage. At UL, the evolved DMS was analyzed in a GC-MS (GC: Agilent Intuvo 9000; MS: Agilent 5977B), and peak areas determined with the Mas- sHunter software were converted to concentration against a 7-point calibration of a diluted DMSP standard solution (R2 > 0.99; Lizotte et al., 2020). At SUNY-ESF, DMSP was quantified following the procedure outlined in Kinsey and Kieber (2016) on a Shimadzu GC-14A with flame photo- metric detector. Further details on DMSP analysis are pro- vided in Text S2. For DMS analysis, 10–20 mL were withdrawn from the vial, spiked with a DMS-d6 internal standard (4.5 nM final concentration), sparged with ultra high purity helium for 5 min and trapped in a glass tube filled with Tenax-TA (kept below –10C) using a custom-made purge-and-trap device (Gourdal et al., 2018). Trapped DMS was injected into a gas chromatograph-mass spectrometer (GC-MS; GC: Agilent 7890A; MS: Agilent 5975C) for quantification. The GC-MS was calibrated at the beginning of each transect by injections of DMS-d3 and DMS-d6 at 4–6 concentrations ranging between 0.5 and 45 nM, and a single calibration curve was used for the whole dataset (R2 of the log-log linear calibration > 0.99). 2.5. Sea-air DMS flux (FDMS) calculation We estimated FDMS (mmol m–2 d–1) as the product of the sea–air gas transfer coefficient (K), the DMS gradient across the sea–air interface, and the ice-free water fraction (1 – SIC): FDMS ¼ K DMSw DMSa ð Þ 1 SIC ð Þ: ð1Þ ð1Þ K was computed using the wind speed–based gas exchange parameterization of Woolf (1997) and Woolf (2005), taking into account air- and water-side resistance (1/K ¼ 1/ka þ 1/kw) and the effects of SST and salinity on DMS diffusivity and solubility, and assuming DMSa ¼ DMSw/253 based on field measurements (Land et al., 2014). Wind speed measurements were acquired at 1 min–1 frequency by the anemometer installed 16 m above sea level on the ship’s meteorological tower (Burgers et al., 2017), and converted to 10-m wind speeds according to a standard wind profile for neutral atmospheric stability (Hsu et al., 1994). At each station we calculated FDMS using one single value for DMSw and SST and the 1 min–1 time series of wind speed during the 24 h prior to sampling, and then calculated the daily mean FDMS. The reader is referred to Galı´ et al. (2019) for further details on gas exchange calculations. Samples for DMSP analysis were collected in triplicate and included a set of unfiltered samples for total DMSP (DMSPt) and a second set of filtered samples for dissolved DMSP (DMSPd). This sampling plan enabled the calcula- tion of particulate DMSP (DMSPp) as DMSPt – DMSPd. Unfiltered seawater samples were collected into 20-mL pre-cleaned borosilicate scintillation vials screw-capped with a solid cap containing a Teflon-lined silicone insert. Samples were microwaved with the cap loosened, then bubbled with purified air to remove DMS, acidified with hydrochloric acid, and stored in the dark at room temper- ature (Kinsey and Kieber, 2016). Filtered samples were obtained by the small volume, gravity drip filtration method (Kiene and Slezak, 2006) using prebaked (550C, 8 h) 25-mm diameter glass fiber filters (Whatman GF/F); samples were bubbled with purified air for several minutes to remove DMS followed by addition of hydro- chloric acid and storage at room temperature in the dark. Galı´ et al: DMS Emissions from the Arctic MIZ The peak areas of endogenous DMS and internal standards, determined with the Mas- sHunter software, were converted to concentrations after correcting for blank signals and internal standards. No analytical replicates were analyzed for the majority of the samples. The mean propagated uncertainty of DMS mea- surements was estimated at 10%, which includes the rep- lication uncertainty estimated from duplicate analyses (n ¼ 16) and the calibration uncertainty. In addition, pre- filtration tests suggested that DMS was 19% higher on average (n ¼ 19) in samples filtered through a 100-mm mesh, compared to their 5-mm-filtered counterparts. Although this test was statistically inconclusive, it indi- cates the importance of sample pre-screening in the pres- ence of Phaeocystis sp. (del Valle et al., 2009). Further details regarding the analysis of DMS are provided in the Supplemental Materials (Text S1). 2.4. DMS and DMSP sampling and analysis DMS and DMSP samples were collected during the morn- ing CTD cast at 8 AM local time from 5 or 6 depths between the surface and 100 m; additional samples were taken occasionally from the CTD casts performed around noon. At several intermediate hydrographic stations, located between full stations, samples were collected only from the surface (1 m) and the subsurface chlorophyll maximum (SCM). The SCM depth was estimated from fluo- rescence depth profiles. A total of 25 stations (Tables 1 and S1) and 35 CTD casts were sampled. 2.6. Phytoplankton pigments, abundance, carbon pools, and DMSP-biomass Samples for phytoplankton pigment analyses were col- lected from Niskin bottles on the CTD rosette. Up to 2.7 L of seawater were filtered on Whatman GF/F filters to collect phytoplankton cells, and the filters contained in cryotubes were stored in liquid nitrogen until analysis. Back on land, filters were soaked in 100% methanol, dis- rupted by sonication, and the resulting solution of ex- tracted pigments was clarified by filtration (GF/F) and analyzed using High Performance Liquid Chromatography (HPLC) on an Agilent Technologies HPLC 1200 system equipped with a diode array detector (Ras et al., 2008). Based on multivariate analyses (see Section 2.8), we selected the following pigments as predictors of dimethy- lated sulfur compounds: total chlorophyll a (TChl a; Chl Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 6 of 31 a conservative conversion factor of 1.5 kg C (mol leu- cine)–1 (Simon and Azam, 1989). a plus allomers and epimers), chlorophyll b (Chl b), chlo- rophyll c2 (Chl c2; in practice, Chl c1þc2), chlorophyll c3 (Chl c3), fucoxanthin (Fuco), 190-hexanoyloxyfucoxanthin (Hex-fuco), 190-butanoyloxyfucoxanthin (But-fuco), 190- butanoyloxyfucoxanthin-like (But-fuco-like), peridinin (Peri), prasinoxanthin (Pras), b-carotene (in practice, a þ b-carotene; Ras et al., 2008), diadinoxanthin þ diatox- anthin (DD), violaxanthin þ antheraxanthin þ zeaxanthin (VAZ) and pheophorbide a (Pheo a). Pigments were grouped as either chlorophylls, photosynthetic (light-har- vesting) carotenoids (PSC), photoprotective carotenoids (PPC) or degradation products (Table 2) and used as phys- iological and taxonomic indicators. ) ( ) The functional genes dddP and dmdA, associated with different DMSP catabolism pathways in bacteria, were quantified by q-PCR following previously described proto- cols (Levine et al., 2012; Zeng et al., 2016). Briefly, univer- sal primers dddP_874F/971R and dmdA_282F/591R were used to quantify dddP and dmdA, respectively. Quan- tification was performed on an AriaMx (Agilent) using SsoAdvanced Sybr Green universal supermix (BioRad). Triplicate seven-fold serially diluted standard curves, using genomic DNA from a Phaeobacter sp. strain presenting both genes (Genbank accession number MW504624; de- tails will be provided elsewhere) was run on each plate. Controls with no DNA template were also run in triplicate on each plate. Single amplifications were confirmed by a melt curve for each analysis. At nine (of 25) stations, samples for identification and counting of micro- and nanophytoplankton were taken at the surface and at the depth of the fluorescence maxi- mum, preserved in Lugol’s solution and kept at 4C. 2.8. Data analysis and statistics f d y Metrics of ice cover and water mass origin were used to depict variations along the horizontal sea-surface axis. We calculated the number of open water days (OWD) at each station using the time series of remotely sensed SIC. Pos- itive OWD indicated the days elapsed since the station became permanently ice-free (SIC < 15%), whereas nega- tive OWD indicated the days remaining before permanent ice opening (Randelhoff et al., 2019). Stations were classi- fied into three categories according to SIC and OWD me- trics (Tables 1 and S1). Ice-covered (ICE) stations were defined by OWD < –3d, which generally corresponded to SIC persistently higher than 85% during the 3 days ending on the day of sampling. MIZ stations were defined by –3d OWD 3d. Open water (OW) stations were defined by OWD > 3, which corresponded to SIC persis- tently below 15% during the 3 days ending on the day of sampling. The abundances of phytoplankton cells were converted to carbon biomass using allometric relationships (Menden-Deuer and Lessard, 2000) and subsequently used to estimate their contribution to the DMSPp stock using literature values for the DMSP-carbon proportion of each taxonomic group (Stefels et al., 2007). For this calculation we assumed, based on measurements performed at sta- tions 713 and 719, that DMSPp represented 90% of the total DMSP at all stations. This result is consistent with literature values (Kiene and Slezak, 2006; Kinsey et al., 2016). We also made order-of-magnitude estimates for the DMSPp bound to detrital particles, Phaeocystis colonies and DMSP-producing bacteria. Further details are given in Text S3 and Table S2. Water masses were depicted using the “Arctic N-P re- lationship” (ANP), a tracer that distinguishes the nutrient signature of Atlantic Water from that of the Pacific-derived Arctic Water. Essentially, an ANP of zero means that the NO3 ––PO4 3– pairs fall on the regression line for Atlantic Water, whereas for an ANP value of one they fall on the regression line for Pacific-derived Water (Randelhoff et al., 2019, and references therein). Stations were further clas- sified into contiguous Arctic and Atlantic domains using a fuzzy c-means clustering algorithm based on three vari- ables: ANP at 20 m, maximum temperature in the Atlantic Water layer, and salinity at the estimated winter convec- tion depth (Randelhoff et al., 2019). Clustering coefficients smaller (greater) than 0.4 were indicative of the Arctic (Atlantic) domain. 2.6. Phytoplankton pigments, abundance, carbon pools, and DMSP-biomass Cells were counted under an inverted microscope after sedi- mentation of 25–50 mL following standard procedures. In addition to traditional microscopy, an Imaging FlowCy- tobot (IFCB; Olson and Sosik, 2007) was used to take pictures of phytoplankton and fluorescing detritus, which were subsequently sized, counted and identified with the software EcoTaxa (https://ecotaxa.obs-vlfr.fr/) (Grondin, 2019). Eukaryotic picophytoplankton were enumerated in live samples by flow cytometry (Accuri C6, Becton Dick- inson) (Marie et al., 1999). 2.7. Bacterial abundance, production, and DMSP catabolism genes Samples (1.5 mL) withdrawn from Niskin bottles were preserved with glutaraldehyde (1% final concentration) and stored at –80C. Bacterial abundances were deter- mined on a flow cytometer (FACSCanto, BD Biosciences) after SYBR1 Green I (ThermoFisher Scientific) staining as described by Gasol and del Giorgio (2000). Bacterial pro- duction rates were measured by [H3]-leucine incorpora- tion (Smith and Azam, 1992). Triplicate 1.7-mL aliquots were incubated in the dark with a mixture of 50/50 (v/ v) [H3]-leucine (Perkin Elmer) and nonradioactive leucine for 4 h at in situ temperature (1.5C). Samples with 5% trichloroacetic acid added prior to isotope addition served as a control. Saturation and time-course experiments were performed beforehand to determine the concentration of leucine and minimum incubation time. Leucine incorpo- ration was converted to carbon production using Along the vertical water-depth axis, the biogeochemical setting was characterized using descriptors of vertical mix- ing, light penetration, biological nutrient utilization and phytoplankton biomass maxima. The mixed layer depth was calculated as the shallowest depth where sigma-t ex- ceeded by 0.03 kg m–3 the value at 1 m (MLD0.03). Ver- tical mixing was also characterized using the “equivalent mixed layer depth” hBD (Randelhoff et al., 2017). This met- ric was developed for meltwater-influenced surface waters Art. 9(1) page 7 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Table 2. Spearman’s rank correlations between selected phytoplankton pigments and total carbon biomass proxied by the beam attenuation coefficient (cp), DMSPt and DMS. DOI: https://doi.org/10.1525/elementa.2020.00113.t2 Table 2. Spearman’s rank correlations between selected phytoplankton pigments and total carbon biomass proxied by the beam attenuation coefficient (cp), DMSPt and DMS. DOI: https://doi.org/10.1525/elementa.2020.00113.t2 Pigmentsa Phytoplankton Classb Correlation Coefficientsc cp DMSPt DMS Chlorophylls Total chlorophyll a (TChl a) All 0.85 0.87 0.72 Chlorophyll b (Chl b) Chlorophytes, prasinophytes, dinoflagellates (type 5) 0.65 0.75 0.39 Chlorophyll c2 (Chl c2) Diatoms, haptophytes (type-7), most dinoflagellates, cryptophytes, chrysophytes, pelagophytes 0.79 0.79 0.68 Chlorophyll c3 (Chl c3) Haptophytes, diatoms, dinoflagellates (type 2), raphidophytes, dictyochophytes 0.45 0.50 0.32 Photosynthetic carotenoids (PSC) Fucoxanthin (Fuco) Diatoms, haptophytes, chrysophytes, dinoflagellates (types 2–3) 0.69 0.68 0.64 190- hexanoyloxyfucoxanthin (Hex-fuco) Haptophytes (type-7, e.g., Chrysochromulina spp.), dinoflagellates (type 2) 0.60 0.75 0.39 190-butanoyloxyfucoxanthin (But-fuco) Haptophytes, chrysophytes, pelagophytes, dinoflagellates (type 2) 0.53 0.59 0.36 190- butanoyloxyfucoxanthin- like (But-fuco-like)d Haptophytes (type 8; e.g. 2.7. Bacterial abundance, production, and DMSP catabolism genes Phaeocystis sp.) 0.53 0.53 0.54 Peridinin (Peri) Dinoflagellates (type 1) 0.57 0.74 0.46 Alloxanthin (Allo) Cryptophytes, dinoflagellates (type 4) 0.53 0.69 0.36 Prasinoxanthin (Pras)e Prasinophytes (type 3) 0.60 0.71 0.39 Photoprotective carotenoids (PPC) b carotenef All except cryptophytes 0.84 0.90 0.71 Diadinoxanthin þ Diatoxanthin (DD) Diatoms, prymnesiophytes, dinoflagellates, chrysophytes 0.65 0.77 0.50 Violaxanthin þ Antheraxanthin þ Zeaxanthin (VAZ) Chlorophytes, chrysophytes, prasinophytes 0.47 0.73 0.31 Degradation products Pheophorbide a (Pheo a) Protozoan fecal pellets 0.43 0.40 0.40 a Pigments are grouped according to functional and chemotaxonomic criteria following Zapata et al. (2004), Jeffrey et al. (2012), Coupel et al (2015) and Schanke et al (2020) and their abbreviations follow the Scientific Council for Oceanic Research (Jeffrey et al a Pigments are grouped according to functional and chemotaxonomic criteria following Zapata et al. (2004), Jeffrey et al. (2012), Coupel et al. (2015) and Schanke et al. (2020), and their abbreviations follow the Scientific Council for Oceanic Research (Jeffrey et al., 2012). b For chlorophylls and PSC, major pigments of each algal class are indicated by highlighting the taxa in bold. b For chlorophylls and PSC, major pigments of each algal class are indicated by highlighting the taxa in bold. c All correlations are significant with P < 0.05, and bold face distinguishes correlations with P < 10–4. Sample sizes are 80–118, except for But-fuco-like (n ¼ 58) and Peri (n ¼ 72). Underlining denotes pigments more closely associated to either DMSPt or DMS, whereas no underlining indicates similarly strong association with both compounds, according to a redundancy analysis (Figure 2). c All correlations are significant with P < 0.05, and bold face distinguishes correlations with P < 10–4. Sample sizes are 80–118, except for But-fuco-like (n ¼ 58) and Peri (n ¼ 72). Underlining denotes pigments more closely associated to either DMSPt or DMS, whereas no underlining indicates similarly strong association with both compounds, according to a redundancy analysis (Figure 2). d A 190-butanoyloxyfucoxanthin-like pigment was found in Phaeocystis by Rowan (1989), and Zapata et al. (2004) reported the occurrence of an unknown 190-butanoyloxyfucoxanthin-like pigment in haptophytes, characterized by absorption peaks shifted þ1 nm from those of 190-butanoyloxyfucoxanthin. d A 190-butanoyloxyfucoxanthin-like pigment was found in Phaeocystis by Rowan (1989), and Zapata et al. (2004) reported the occurrence of an unknown 190-butanoyloxyfucoxanthin-like pigment in haptophytes, characterized by absorption peaks shifted þ1 nm from those of 190-butanoyloxyfucoxanthin. 2.9. Satellite-based estimates of DMS emission from the MIZ (EDMS, MIZ) We computed pan-Arctic EDMS, MIZ as the product of MIZ ice-free area, estimated from satellite images, and mean FDMS per unit of ice-free ocean area in the MIZ, <FDMS>, based on a compilation of in situ studies (Table 3). First, pixels were flagged as MIZ or non-MIZ using the SIC time series for each pixel (see below). Second, <FDMS> was mul- tiplied by (1 – SIC) in flagged pixels to obtain maps of FDMS, MIZ (8-day and 28-km resolution). Third, the resulting flagged pixels were summed over 5 latitude bands between 65N and 85N and over time to obtain annual EDMS, MIZ estimates (Gg S y–1). , ( g y ) To obtain a plausible range for EDMS, MIZ, we used two different criteria (“A” and “B”) for flagging pixels as MIZ. Both criteria rely on the date of permanent ice opening calculated from daily SIC time series (Renaut et al., 2018), but they differ in the duration of the MIZ period. In case A, the MIZ period was defined as the 8 days that comprised the ice opening date in a given pixel, and the correspond- ing mean (median) SIC was 22% (21%) at pan-Arctic scale. These two metrics (MIZ duration and corresponding ice cover) compare well with those observed during our study (Table 1). However, we observed that in certain areas SIC had decreased below 85% several days before permanent ice breakup. To account for the corresponding DMS emis- sions, in case B the duration of the MIZ period was defined by a backward search starting with the ice opening 8-day period and ending when SIC exceeded 85%, but never exceeding 32 days (to ensure that only the melt period was included). The mean (median) duration in case B was 16 (12) days, and the corresponding mean (median) SIC during the MIZ period was 36% (42%) at the pan-Arctic scale. Thus, case B allowed for geographic and interannual variations in MIZ duration, reflecting different ice-edge dynamics. ) Data analyses were performed with Matlab (R2013b), python (3.6) and R (3.3.3; R Core Team, 2017) software. Prior to computing statistics and multivariate analyses, we tested variables for normality with the Shapiro–Wilk test. Given that none of the biogeochemical variables were normally distributed, here we generally report non- parametric statistics: median, interquartile range (IQR) and Spearman’s rank correlation coefficient (rS). Reference (Region) 2.3 (0.2–6.0), n ¼ 10 7.3 (1.1–17.1), n ¼ 13 10.0 (1.7–25.7), n ¼ 12 a FDMS reported here were not corrected by the fraction of ice-free water surface (1 – SIC). The effect of sea ice is factored in afterward, using satellite-observed SIC (Section 2 9) a FDMS reported here were not corrected by the fraction of ice-free water surface (1 – SIC). The effect of sea ice is factored in afterward, using satellite-observed SIC (Section 2.9). c Continuous underway measurements; sample size cannot be compared to other studies. equations linking DMSPt, DMS and FDMS to their main environmental drivers. Further details on multivariate analysis are provided in Text S4. biomass maximum (SBM) was estimated from cp profiles smoothed with a 5-point running mean to remove spikes. To describe biogeochemical patterns along the hori- zontal and vertical dimensions, we calculated median profiles of selected variables for the three OWD cate- gories (ICE, MIZ, and OW) after additional binning of the measurements into four depth (z) layers: z 9, 9 < z 21, 21 < z 41 and 41 < z 81 m. Vertically integrated stocks were also calculated for variables mea- sured from CTD casts (cp) and discrete Niskin bottle profiles (DMS, DMSPt, pigments). To this end, profiles were first binned (cp) or interpolated (Niskin bottle vari- ables) to 1-m resolution between the surface and 100 m. Discrete Niskin bottle profiles were closed by impos- ing a concentration of 0 at 100 m. Profiles were inte- grated by summation between 0 and 60 m, a range that comprised most of the phytoplankton and dimethylated sulfur stocks and sufficient discrete measurements (pro- files with n < 4 between 0–60 m depth were not considered). 2.7. Bacterial abundance, production, and DMSP catabolism genes The effect of sea ice is factored in afterward, using satellite-observed SIC (Section 2.9). b Not available. c Continuous underway measurements; sample size cannot be compared to other studies. 2.7. Bacterial abundance, production, and DMSP catabolism genes remely similar correlations were found for neoxanthin, which occurs in chlorophytes and prasinophytes. e Extremely similar correlations were found for neoxanthin, which occurs in chlorophytes and prasinophytes. e Extremely similar correlations were found for neoxanthin, which occurs in chlorophytes and pr f Includes a þ b carotene because of challenges in quantifying them separately, but a carotene is assumed to occur at much lower concentration (e.g., Ras et al., 2008). f Includes a þ b carotene because of challenges in quantifying them separately, but a carotene is assumed to occur at much lower concentration (e.g., Ras et al., 2008). referred to as the 0.415 isolume (Letelier et al., 2004). The nitracline depth was defined as the shallowest depth where nitrate concentration exceeded 1 mM, based on a linear interpolation of discrete nutrient measurements to a vertical spacing of 1 m. The depth of the subsurface in the MIZ and found to constrain the extent of vertical mixing better than criteria based on density or tempera- ture thresholds (Randelhoff et al., 2019). The lower depth limit of phytoplankton growth was estimated as the depth where daily PAR was 0.415 mol photons m–2 d–1, here Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 8 of 31 Art. 9(1) page 8 of 31 Table 3. Compilation of means (and ranges) of sea–air DMS flux (FDMS) estimatesa (mmol m–2 d–1) in the Arctic sorted by ice-cover categories. DOI: https://doi.org/10.1525/elementa.2020.00113.t3 Reference (Region) Ice-Cover Category ICE MIZ OW Matrai and Vernet (1997; Barents Sea) n.a.b 17 (6–33), n ¼ 3 33, n ¼ 1 Galı´ and Simo´ (2010; Greenland Sea, Fram Strait) n.a. 6.5 (0.5–22.5), n ¼10 n.a. Jarnı´kova´ et al. (2018; Baffin Bay, Canadian Arctic Archipelago)c n.a. (<1–20) (<1–80) This study (Baffin Bay) 2.3 (0.2–6.0), n ¼ 10 7.3 (1.1–17.1), n ¼ 13 10.0 (1.7–25.7), n ¼ 12 Reference (Region) Ice-Cover Category ICE MIZ OW Matrai and Vernet (1997; Barents Sea) n.a.b 17 (6–33), n ¼ 3 33, n ¼ 1 Galı´ and Simo´ (2010; Greenland Sea, Fram Strait) n.a. 6.5 (0.5–22.5), n ¼10 n.a. Jarnı´kova´ et al. (2018; Baffin Bay, Canadian Arctic Archipelago)c n.a. (<1–20) (<1–80) This study (Baffin Bay) 2.3 (0.2–6.0), n ¼ 10 7.3 (1.1–17.1), n ¼ 13 10.0 (1.7–25.7), n ¼ 12 a FDMS reported here were not corrected by the fraction of ice-free water surface (1 – SIC). 3. Results stages of development in ice-covered western stations, as indicated by the near-surface nitracline and shallow SBM depths (Table 1; Figure 3). At the eastern side, the phy- toplankton bloom had further developed into a mature or even post-bloom stage with considerable nutrient deple- tion at the time of sampling, causing a deepening of the nitracline to around 20 m (T4–T6) or even 40 m (T7). The 0.415 mE m–2 d–1 isolume was deeper than the nitracline regardless of ice cover and surface plankton biomass (Figure 3), resulting in the formation of a prominent SBM between 10 and 40 m at some eastern-side stations. 3.1. Oceanographic and biogeochemical setting Transect hydrography was characterized by east–west gra- dients in water masses and gradual ice retreat (Figure 1; Table 1). At the eastern end of the station grid, close to the West Greenland shelf, the clear signature of warm Atlantic-derived water flowing in through the Davis Strait was observed. The western end was influenced by less dense and colder waters of Pacific origin coming from the Canadian Archipelago (Tang et al., 2004). Consequently, ice retreat started in the east, with the ice edge steadily moving west at around 4 km d–1. This retreat led to a par- allel gradient in surface layer properties: eastern stations had sea-ice concentrations at or close to 0%, and near- surface layer (<50 m) hydrography showed substantial surface warming, whereas western stations were mostly ice-covered and showed no surface warming (Randelhoff et al., 2019). The Atlantic- and Pacific-derived water masses also had distinct nutrient signatures. In the west, the pre- bloom nutrient supply was restricted by the lower nutri- ent concentrations characteristic of the Arctic outflow and the strong stratification in the upper 200 m of the water column. Distinct phytoplankton communities were found across the ice-edge transects, with biomass usually domi- nated by diatoms (Lafond et al., 2019). Pennate diatoms (Fragilariopsis spp., Pseudo-nitzschia spp., Ceratoneis clos- terium) were numerically dominant in the western side, whereas colonial centric diatoms (Thalassiosira spp., Chae- toceros spp.) dominated in the eastern side. Due to their larger biovolume, centric diatoms generally dominated phytoplankton biomass (accompanied by the ice diatom Melosira arctica and its spores at some MIZ stations). The haptophyte Phaeocystis pouchetii was present at back- ground levels at different bloom stages, with a median abundance of 1.7105 solitary cells L–1. 2.9. Satellite-based estimates of DMS emission from the MIZ (EDMS, MIZ) How- ever, in some instances we also report parametric statis- tics because they convey complementary information. In particular, means rather than medians are used to esti- mate fluxes at large scales because they include the effect of extreme values (see next section), and Pearson correlations (rP) are used to evaluate linearity. We per- formed a redundancy analysis (RDA) with variance parti- tioning (Borcard et al., 2011; Oksanen et al., 2019; R package vegan) to depict the correlation structure between sulfur compounds and two sets of predictors simultaneously: physical variables and phytoplankton pigments. Finally, we used multiple linear regression with stepwise variable selection to develop empirical Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 9 of 31 Figure 2. Redundacy analysis (RDA) between dimethylated sulfur compounds and two different sets of predictor variables. Variance partitioning plot (top), RDA triplot with physical variables only (bottom left), and RDA triplot with phytoplankton pigments only (bottom right). In the bottom plots, dots correspond to individual samples colored according to the OWD classification as in previous figures (blue for ICE, green for MIZ, orange for OW); samples with either DMS > 20 nM or DMSPt > 200 nM are identified (station_depth). DOI: https://doi.org/10.1525/ elementa.2020.00113.f2 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 Figure 2. Redundacy analysis (RDA) between dimethylated sulfur compounds and two different sets of predictor variables. Variance partitioning plot (top), RDA triplot with physical variables only (bottom left), and RDA triplot with phytoplankton pigments only (bottom right). In the bottom plots, dots correspond to individual samples colored according to the OWD classification as in previous figures (blue for ICE, green for MIZ, orange for OW); samples with either DMS > 20 nM or DMSPt > 200 nM are identified (station_depth). DOI: https://doi.org/10.1525/ elementa.2020.00113.f2 3. Results Yet, this species formed massive proliferations below the pycnocline at some ice-free stations along the eastern side (stations The development of the phytoplankton bloom tracked the east–west gradient in ice retreat and water mass dis- tribution. Overall, phytoplankton growth was in the early Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 10 of 31 -62 -61 -60 -59 -58 Longitude 60 40 20 0 Depth (m) hBD Isolume NO3-cline SBM 0 5 10 15 20 0 100 200 -20 400 -17 403 -15 406 -16 409 -8 412 -6 413 21 418 0 100 Ice (%) -64 -63 -62 -61 -60 -59 80 60 40 20 0 Depth (m) 0 10 20 30 40 DMS (nM) 0 100 200 300 DMSPt (nM) -11 600 -2 603 2 604.5 -1 605 8 608 9 612 13 615 Open Water Days Arctic Atlantic 0 100 Ice (%) -62 -61 -60 -59 60 40 20 0 Depth (m) 0 20 40 60 0 200 400 600 -3 519 -2 515 2 512 3 507 17 506 0 100 Ice (%) -63 -62 -61 -60 -59 -58 80 60 40 20 0 Depth (m) 0 5 10 0 50 100 150 200 -4 719 3 716 5 713 9 707 20 703 32 700 0 100 Ice (%) a) T6 (70.5°N) b) T5 (70°N) c) T7 (69.5°N) d) T4 (68°N) Open Water Days Open Water Days Open Water Days e 3. Vertical profiles of DMS and DMSPt and their physicochemical context. Transects are ordered by latitu h transect, the vertical and horizontal variability of DMS and DMSPt is illustrated with four similarly org ts. The left bubble plots show depth-longitude transects of DMS (filled circles) and total DMSPt (empty h black outlines). The area of the circles is proportional to the concentration; the DMSPt circles are scaled t 0.10 proportion (i.e., DMS and DMSPt have the same circle radius when DMS:DMSPt ¼ 0.10 in a given sa superimposed lines indicate the depths of the equivalent mixed layer (hBD), the PAR isolume of 0.41 tons m–2 d–1, the nitracline, and the subsurface biomass maximum (see text for definitions). 3. Results Other taxa that made smaller but relevant contributions to phytoplankton bio- mass were haptophytes of the genus Chrysochromulina, athecate dinoflagellates of the genera Gymnodinium and Gyrodinium (more abundant in the eastern side), crypto- phytes, chrysophytes, prasinophytes (Pyramimonas sp. and likely Micromonas sp.), dictyochophytes, and unidentified autotrophic nanoflagellates (see abundances in Figure S1). Median DMSPt profiles were remarkably similar at ICE and MIZ stations, with highest concentrations at the surface (around 70 nM) and a sharp decrease with depth that approximately followed planktonic biomass (as depicted by cp). Median DMSPt profiles at OW stations differed slightly from this pattern, with decreased concentrations in the surface layer and increased concentrations between 10 and 41 m, indicating that subsurface DMSPt maxima became more widespread with bloom progression. Median DMS profiles showed some distinct features com- pared to those of DMSPt: (1) a less pronounced decrease with depth at ICE stations, (2) a marked peak in median DMS concentration (12 nM) at the MIZ, and (3) a general increase throughout the water column at OW stations. Differences in plankton community functioning were also observed across the ice edge. Concomitant with the more advanced phytoplankton bloom stages, the ice-free, eastern-side stations with Atlantic influence were charac- terized by increased organic carbon recycling via the microbial loop (bacteria and microzooplankton; Saint-Be´at et al., 2020). These stations generally displayed higher bacterial abundances and production (Figure S2), and hosted greater relative abundances of copiotrophic bacte- rial taxa (bloomers) like Rhodobacteraceae (a-proteobac- teria) and Polaribacter (Bacteroidetes) and lower relative abundances of oligotrophic taxa like SAR11 (a-proteobac- teria; J Dinasquet, unpublished data). g Further insights were obtained by plotting the concen- trations ofplanktonic biomass and sulfur compounds along a continuous OWD axis (Figure 5).The highest near-surface DMS concentrations were generally observed between –3 and 9 OWD, and concentrations exceeding 10 nM were found in both Arctic- and Atlantic-influenced stations. Ver- tically integrated stocks of DMSPt and DMS peaked slightly later, between 2 and 13 OWD. At ICE stations, vertically integrated DMSPt and DMS showed little variability in both the Arctic (Pacific-influenced) or Atlantic domains. Con- versely, at MIZ and OW stations the largest vertically inte- grated stocks of DMSPt (up to 8 mmol m–2) and DMS (up to 1.4 mmol m–2) were invariably found in the Atlantic- influenced stations. 3.2. Spatiotemporal patterns of DMSPt and DMS 3.2. Spatiotemporal patterns of DMSPt and DMS Concentrations of dimethylated sulfur compounds were usually highest in the surface layer (0–9 m), with median DMSPt and DMS concentrations of 70 and 6.3 nM, respec- tively. However, marked subsurface concentration maxima occurred between 10 and 30 m at some stations, and important horizontal variability and gradients were found across the four transects sampled (T4–T7), as depicted in detail in Figure 3. The earliest and southernmost transect, T4 (68N), had the heaviest ice cover and the lowest DMS (<2.5 nM) and DMSPt (<90 nM) concentrations among all transects, with the stark exception of its westernmost sta- tion (400), where high DMSPt (255 nM) and DMS (22 nM) were measured just beneath the sea ice. In contrast with T4, the northernmost transects T5 (70N) and T6 (70.5N), which covered mostly the MIZ, showed the highest DMSPt and DMS concentrations throughout the water column. Within T5 and T6, prominent subsurface maxima were observed at 15 m at station 512 in the MIZ (524 nM DMSPt; 74 nM DMS), at 30 m at station 615 in the OW (159 nM DMSPt; 45 nM DMS) and at 20 m at station 612 (302 nM DMSPt; no marked DMS maximum). Transect T7 (69.5N) was sampled mostly during OW conditions except for its western end. This transect showed interme- diate DMS (<12 nM) and DMSPt (<150 nM) concentrations and no prominent subsurface features. The analyses shown in Figures 4 and 5 offer comple- mentary views of how physical and biological processes shaped DMSPt and DMS concentrations over the horizon- tal and vertical axes. To further understand how these processes differently shaped sulfur compound distribu- tions, we performed an RDA with variance partitioning (Figure 2). Physical variables and phytoplankton pigments explained, jointly, 78% of the variance in DMS and DMSPt concentrations in log10 space, and the total variance was partitioned as follows: 57% explained only by pigments, 21% explained jointly by pigments and physical variables, and 1% explained only by physical variables. The first axis of the physics RDA explained 25% of the log10-variance of sulfur compounds and accounted for the vertical variation in PAR and salinity, whereas the first axis of the pigments RDA explained 75% of the log10-variance of sulfur com- pounds and accounted for phytoplankton biomass. 3. Results Large vertically integrated stocks were associated with sharp subsurface biomass maxima at sta- tions 512 and 615, but not at stations 507, 707 and 713. Potentially different temporal evolutions of DMSPt and DMS in the Arctic versus Atlantic domains (e.g., Saint-Be´at et al., 2020) could not be resolved statistically owing to limited spatiotemporal coverage. Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 3. Results The plots hBD Isolume NO3-cline SBM -64 -63 -62 -61 -60 -59 80 60 40 20 0 Depth (m) 0 10 20 30 40 DMS (nM) 0 100 200 300 DMSPt (nM) -11 600 -2 603 2 604.5 -1 605 8 608 9 612 13 615 Open Water Days Arctic Atlantic 0 100 Ice (%) 17 100 ) a) T6 (70.5°N) b) T5 (70°N) Open Water Days Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 0 0 5 10 0 50 100 150 200 d) T4 (68°N) n Water Days -62 -61 -60 -59 -58 Longitude 60 40 20 0 Depth (m) 0 5 10 15 20 0 100 200 -20 400 -17 403 -15 406 -16 409 -8 412 -6 413 21 418 0 100 Ice (%) -63 -62 -61 -60 -59 -58 d) T4 (68°N) Open Water Days Figure 3. Vertical profiles of DMS and DMSPt and their physicochemical context. Transects are ordered by latitude. For each transect, the vertical and horizontal variability of DMS and DMSPt is illustrated with four similarly organized plots. The left bubble plots show depth-longitude transects of DMS (filled circles) and total DMSPt (empty circles with black outlines). The area of the circles is proportional to the concentration; the DMSPt circles are scaled to DMS in a 0.10 proportion (i.e., DMS and DMSPt have the same circle radius when DMS:DMSPt ¼ 0.10 in a given sample). The superimposed lines indicate the depths of the equivalent mixed layer (hBD), the PAR isolume of 0.415 mol photons m–2 d–1, the nitracline, and the subsurface biomass maximum (see text for definitions). The plots above Art. 9(1) page 11 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ 512 and 615), either dominating phytoplankton biomass (station 615; 6.6106 cells L–1) or co-dominating with dia- toms (station 512; 1.2107 cells L–1). Other taxa that made smaller but relevant contributions to phytoplankton bio- mass were haptophytes of the genus Chrysochromulina, athecate dinoflagellates of the genera Gymnodinium and Gyrodinium (more abundant in the eastern side), crypto- phytes, chrysophytes, prasinophytes (Pyramimonas sp. and likely Micromonas sp.), dictyochophytes, and unidentified autotrophic nanoflagellates (see abundances in Figure S1). 512 and 615), either dominating phytoplankton biomass (station 615; 6.6106 cells L–1) or co-dominating with dia- toms (station 512; 1.2107 cells L–1). 3.2. Spatiotemporal patterns of DMSPt and DMS These results indicate that bloom dynamics (i.e., phytoplankton Grouping the DMSPt and DMS profiles by ice-cover categories (Figure 4) revealed some general patterns. the bubble plots depict the percent sea-ice coverage on the day of sampling (gray bars) and corresponding open water days (black line and numbers). Filled circles distinguish stations classified as MIZ; vertical dashed lines separate Arctic and Atlantic domains (see text). The center and right plots show DMS and DMSPt vertical profiles, respectively. For all transects, each station is distinguished with a different color in the bubble and profile plots. No scale is given for DMS and DMSPt concentrations in the bubble plots, as concentrations can be seen in the profile plots. DOI: https:// doi.org/10.1525/elementa.2020.00113.f3 Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 12 of 31 Figure 4. Median vertical profiles grouped according to ice cover and binned into 4 depth horizons. (a) Total chlorophyll a; (b) beam attenuation; (c) total DMSP; (d) DMS; (e) chlorophyll c3; (f) 19-butanoyloxyfucoxanthin-like; (g) 19- hexanoyloxyfucoxanthin; (h) pheophorbide a; (i) temperature; (j) Brunt-Va¨isa¨la¨ (buoyancy) frequency; (k) Arctic N-P relationship (unitless); (l) photosynthetically active radiation. Note that the near-surface sample at station 400 was removed owing to suspected influence of ice-derived materials. DOI: https://doi.org/10.1525/ elementa.2020.00113.f4 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 oaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Figure 4. Median vertical profiles grouped according to ice cover and binned into 4 depth horizons. (a) Total chlorophyll a; (b) beam attenuation; (c) total DMSP; (d) DMS; (e) chlorophyll c3; (f) 19-butanoyloxyfucoxanthin-like; (g) 19- hexanoyloxyfucoxanthin; (h) pheophorbide a; (i) temperature; (j) Brunt-Va¨isa¨la¨ (buoyancy) frequency; (k) Arctic N-P relationship (unitless); (l) photosynthetically active radiation. Note that the near-surface sample at station 400 was removed owing to suspected influence of ice-derived materials. DOI: https://doi.org/10.1525/ elementa.2020.00113.f4 3.3. Biogeochemical factors affecting DMSPt concentrations 3.3. Biogeochemical factors affecting DMSPt concentrations Vertical integrals correspond to (a) beam attenuation at 650 nm (cp), a proxy of particulate organic carbon; (b) total chlorophyll a (TChla a); (c) total DMSP (DMSPt); and (d) DMS. Panel (e) shows the mean near-surface (0–5 m) DMS concentration, and panel (f) the sea–air DMS flux. Station numbers are reported to the right of each symbol for data points above the first tercile (to avoid clutter). DOI: https://doi.org/10.1525/elementa.2020.00113.f5 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 Figure 5. Evolution of biological stocks and sea–air DMS fluxes across the ice edge. Stations are classified into ice-cover categories, defined by open water days (OWD), and Arctic versus Atlantic domains. Vertical integrals correspond to (a) beam attenuation at 650 nm (cp), a proxy of particulate organic carbon; (b) total chlorophyll a (TChla a); (c) total DMSP (DMSPt); and (d) DMS. Panel (e) shows the mean near-surface (0–5 m) DMS concentration, and panel (f) the sea–air DMS flux. Station numbers are reported to the right of each symbol for data points above the first tercile (to avoid clutter). DOI: https://doi.org/10.1525/elementa.2020.00113.f5 layers, DMSPt:TChl a ratios decreased to around 40 nmol mg–1. The higher DMSPt:TChl a ratios found closer to the surface and especially in OW can be attributed, at least partially, to the reduction of phytoplankton TChl a content caused by photoacclimation to high PAR (Figure 4l; see also Figure 8b, k, and l). layers, DMSPt:TChl a ratios decreased to around 40 nmol mg–1. The higher DMSPt:TChl a ratios found closer to the surface and especially in OW can be attributed, at least partially, to the reduction of phytoplankton TChl a content caused by photoacclimation to high PAR (Figure 4l; see also Figure 8b, k, and l). correlations were found for pennate diatoms at the near- surface and for centric diatoms in both the SCM and sur- face layers. Diatoms (pennate þ centric) were likely the main DMSP producers in some samples with moderately high DMSPt (72–94 nM) in the MIZ, such as in the surface and SCM at station 507 (up to 80% of DMSPp) and the surface at station 512 (up to 30% of DMSPp). In the re- maining samples, diatoms accounted for around 3% of the DMSPp (Figure S3; Text S3). 3.3. Biogeochemical factors affecting DMSPt concentrations Median Fuco:TChl a ratios in the top 21 m were highest in the MIZ, probably indic- ative of enhanced diatom biomass (Coupel et al., 2015), supporting the important role of diatoms as DMSP produ- cers in shallow MIZ waters. Phytoplankton cell counts, available at nine stations, revealed a strong positive correlation between DMSPt and the abundance of Phaeocystis pouchetii solitary cells (rS ¼ 0.82; P < 0.05; rP ¼ 0.94; P < 10–3; Figure 7a) at the SCM. The biomass of single-celled Phaeocystis pouchetii ac- counted for up to 46% of the DMSPp (159 nM DMSPt) in the SCM at the OW station 615, and 24% (524 nM DMSPt) at the MIZ station 512. For all remaining available samples, solitary Phaeocystis accounted for <17% and <7% of the DMSPp at the SCM and the sea surface, respec- tively (Figure S3; Text S3). Prominent subsurface DMSPt peaks in the MIZ and OW stations co-occurred with ele- vated concentrations of Chl c3 and a But-fuco-like pig- ment. Chlorophyll c3 was identified by Lafond et al. (2019) as a diagnostic compound for Phaeocystis pouchetii in the Green Edge cruise. The But-fuco-like pigment, which occurred at lower concentrations, was suggested as a marker for Phaeocystis pouchetii (Rowan, 1989) or, more broadly, type-8 haptophytes (Zapata et al., 2004). Although both pigments peaked in the 21–41 m layer, they showed different distributions (Figure 4e and f) and ratios to TChl a (Figure 8e and f) across the ice edge (see next section). Concentrations of DMSPt were also signifi- cantly correlated to the abundance of pennate diatoms at the SCM (rP ¼ 0.85; P < 0.01; rS ¼ 0.62; P < 0.10). Weaker Whereas evidence suggests that Phaeocystis and dia- toms dominated the DMSPp stock in high-biomass sam- ples, other taxa were dominant DMSPp producers in samples with moderate-to-low biomass (Figure S3), which was especially true of dinoflagellates. Athecate dinoflagel- lates of the Gyrodinium/Gymnodinium complex contrib- uted 30%–50% of the DMSPp at the SCM at stations 605, 707, 713, and 719. In addition, a moderate correla- tion was also found between DMSPt and peridinin (Table 2; Figure 2), a marker for some DMSP-rich dinoflagellates (Caruana and Malin, 2014) that reached its highest con- centrations in subsurface waters at the ICE and OW sta- tions (Figure 4g). However, the athecate dinoflagellates that dominated microscopic counts lacked peridinin, pointing to the role of other, unidentified dinoflagellates. 3.3. Biogeochemical factors affecting DMSPt concentrations and their covariates) exerted proximate control on the sulfur compounds. The RDA also showed that phytoplank- ton pigments, with the exception of TChl a and DD, clus- tered with either DMS or DMSPt, suggesting that different taxonomic groups (or their biogeochemical covariates) influenced DMS and DMSPt differently. DMSPt concentrations were strongly correlated to cp, a proxy of POC and phytoplankton biomass in the eupho- tic layer, which explained 56% of the DMSPt variance (Figure 6a). A slightly lower fraction of the DMSPt vari- ance, 52%, was explained by TChl a (which, indeed, was highly redundant with cp; Table 2). Up to 80% of the DMSPt variance was explained by an optimal linear regres- sion model that used as predictors five photosynthetic pigments of different taxonomic specificity plus photopro- tective b carotene (Text S4.2). The DMSPt:cp ratio changed little across the ice edge or along the vertical profiles (Figure 8a), except for the high median ratio observed in subsurface waters at the OW stations, driven by high Phaeocystis pouchetii biomass. Compared to DMSPt:cp, median DMSPt:TChl a ratios showed greater vertical vari- ation. In the top 21 m, median DMSPt:TChl a ratios were around 60 nmol mg–1 at MIZ and ICE stations and 120 nmol mg–1 at the OW stations (Figure 8b). In deeper In the following two sections, we analyze in greater detail the biogeochemical factors that affect DMSPt (Sec- tion 3.3) and DMS (Section 3.4). This detailed analysis relies mostly on: (1) the correlation patterns among phy- toplankton pigments, biomass, and sulfur compounds (Table 2; Figures 2, 6, and 7); (2) the distributions of the different variables (Figure 4) and their ratios (Figure 8) along the vertical profiles and across the ice edge; and (3) the vertical profiles of bacterial sulfur metabolism genes in the MIZ (Figure 9). Further analyses are presented in the supplemental materials, namely: DMSP partitioning among phytoplankton groups and other compartments (Text S3) and stepwise multiple linear regressions (Text S4.2 and Table S3). Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 13 of 31 Figure 5. Evolution of biological stocks and sea–air DMS fluxes across the ice edge. Stations are classified into ice-cover categories, defined by open water days (OWD), and Arctic versus Atlantic domains. 3.3. Biogeochemical factors affecting DMSPt concentrations Photosynthetic pico- and nanoeukaryotes contributed around 10–20% of the DMSPp in some low-biomass sam- ples. Some of them could be identified as chrysophytes Art. 9(1) page 14 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Figure 6. Scatterplots of (a) DMSPt versus cp and (b) DMS versus DMSPt. Points are colored by ice-cover categories, and symbols represent depth intervals (in meters). Outliers, defined as points beyond the first or third quartile + 1.5 times the interquartile range, are highlighted with outer black lines and annotated with station numbers. Reference straight lines on the background mark different values of the y/x ratio (gray) and the median y/x ratio (black). Pearson and Spearman correlations (rp and rs; untransformed data) and sample sizes (n) are also shown. DOI: https://doi.org/ 10.1525/elementa.2020.00113.f6 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by gue Figure 6. Scatterplots of (a) DMSPt versus cp and (b) DMS versus DMSPt. Points are colored by ice-cover categories, and symbols represent depth intervals (in meters). Outliers, defined as points beyond the first or third quartile + 1.5 times the interquartile range, are highlighted with outer black lines and annotated with station numbers. Reference straight lines on the background mark different values of the y/x ratio (gray) and the median y/x ratio (black). Pearson and Spearman correlations (rp and rs; untransformed data) and sample sizes (n) are also shown. DOI: https://doi.org/ 10.1525/elementa.2020.00113.f6 and prasinophytes (Pyramimonas sp. and likely Micromo- nas sp.), and pigments characteristic of these groups (Chl b, But-fuco, Pras, and photoprotective VAZ pigments) clus- tered together in the RDA along with the cryptophyte- specific alloxanthin (Figure 2). Pico- and nano-sized hap- tophytes other than Phaeocystis likely contributed signifi- cantly to the DMSP pool in low-biomass conditions, as supported by the close association between Hex-fuco and DMSPt (Table 2; Figure 2) and the high median Hex- fuco:TChl a and DMSPt:TChl a ratios in the top 21 m in the OW (Figure 8b and g). Figure 7. Relationships between Phaeocystis abundance and sulfur compounds in the near-surface and SCM layers. Phaeocystis pouchetii solitary cell abundance plotted against the concentration of (a) DMSPt and (b) DMS; the color scale depicts the mean daily irradiance, distinguishing near-surface and SCM samples. Spearman rank correlations for each sample subset are also shown. 3.3. Biogeochemical factors affecting DMSPt concentrations DOI: https://doi.org/10.1525/elementa.2020.00113.f7 The mosaic of different phytoplankton classes with con- trasting intracellular DMSP concentrations resulted in lim- ited spatiotemporal variability of the median DMSPt:cp ratio (Figures 6a and 8a). Positive outliers in the DMSPt versus cp scatterplot were found at the subsurface of the OW sta- tions 612 and 615, concurrent with very high Phaeocystis pouchetii biomass, and just beneath the ice at station 400, probably indicating the release of bottom-ice biomass. Tak- ing advantage of the linear relationship between cp and POC, we estimated that DMSPp-carbon (CDMSPp) accounted for a median of 2.5% of the POC (IQR of 1.8–3.4%). Exclud- ing an outlier (station 400 near-surface), the CDMSPp:POC fraction ranged approximately between 1 and 10%, which is entirely compatible with mixed phytoplankton popula- tions of high and low DMSP producers (Stefels et al., 2007). Indeed, this range must also encompass the DMSP bound to bacteria, detritus and zooplankton. Figure 7. Relationships between Phaeocystis abundance and sulfur compounds in the near-surface and SCM layers. Phaeocystis pouchetii solitary cell abundance plotted against the concentration of (a) DMSPt and (b) DMS; the color scale depicts the mean daily irradiance, distinguishing near-surface and SCM samples. Spearman rank correlations for each sample subset are also shown. DOI: https://doi.org/10.1525/elementa.2020.00113.f7 Our DMSP partitioning model was only partly success- ful at accounting for the totality of DMSPp. Whereas, on average, we could account for 77%–99% of the DMSPp at Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 15 of 31 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Figure 8. Vertical profiles of various median concentration ratios by ice cover and depth horizon. (a) DMSPt:cp, a proxy for DMSPt:POC; (b) DMSPt:TChl a; (c) DMS:DMSPt; (d) pheophorbide a to TChl a ratio, a proxy for phytodetritus degradation linked to grazing activity; (e) Chl c3:TChl a, a proxy for Phaeocystis pouchetii relative abundance; (f) 190- butanoyloxyfucoxanthin-like to TChl a ratio, a putative proxy for Phaeocystis pouchetii; (g) 190- hexanoyloxyfucoxanthin to TChl a ratio (haptophytes and dinoflagellates); (h) peridinin:TChl a (type-1 dinoflagellates); (i) Chl c2:TChl a (chryso- and pelagophytes and other groups); fucoxanthin to TChl a (mostly diatoms); (k) photoprotective carotenoids:TChl a, a generic indicator of photoprotection; and (l) (diadinoxanthin þ diatoxanthin):TChl a (photoprotection through the xanthophyll cycle; see Table 2). DOI: https://doi.org/10.1525/ elementa.2020.00113.f8 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 Figure 8. 3.3. Biogeochemical factors affecting DMSPt concentrations Vertical profiles of various median concentration ratios by ice cover and depth horizon. (a) DMSPt:cp, a proxy for DMSPt:POC; (b) DMSPt:TChl a; (c) DMS:DMSPt; (d) pheophorbide a to TChl a ratio, a proxy for phytodetritus degradation linked to grazing activity; (e) Chl c3:TChl a, a proxy for Phaeocystis pouchetii relative abundance; (f) 190- butanoyloxyfucoxanthin-like to TChl a ratio, a putative proxy for Phaeocystis pouchetii; (g) 190- hexanoyloxyfucoxanthin to TChl a ratio (haptophytes and dinoflagellates); (h) peridinin:TChl a (type-1 dinoflagellates); (i) Chl c2:TChl a (chryso- and pelagophytes and other groups); fucoxanthin to TChl a (mostly diatoms); (k) photoprotective carotenoids:TChl a, a generic indicator of photoprotection; and (l) (diadinoxanthin þ diatoxanthin):TChl a (photoprotection through the xanthophyll cycle; see Table 2). DOI: https://doi.org/10.1525/ elementa.2020.00113.f8 the SCM (depending on the assumptions), we were able to explain only 37%–52% of the DMSPp at the near- surface (Figure S3). Although these estimates are obvi- ously sensitive to the choice of intracellular DMSP con- tent of each phytoplankton group, the large disagreement at the surface indicates that our recon- struction missed relevant DMSPp pools. IFCB measure- ments suggest that the unaccounted fraction might correspond to Phaeocystis colonies, fluorescing detrital particles and spores of the ice diatom Melosira arctica. In addition, order-of-magnitude estimates suggest a neg- ligible contribution by DMSP-producing bacteria (likely <0.1%). Comparison between microscopic counts of single-celled Phaeocystis and the diagnostic pigment ratios (which should account for both colonial and single-celled Phaeocystis) can also shed light on this issue. According to microscopic data, the abundance of solitary Phaeocystis increased by 20-fold from the near- surface (median 4.3104 cells L–1) to the SCM (median 8.6105 cells L–1). In contrast, the Chl c3:TChl a and But-fuco-like:TChl a ratios increased approximately threefold from the surface to the SCM in the same sam- ple subset, supporting the notion that our DMSP esti- mates based on single-cell counts underestimated Phaeocystis-bound DMSP at the surface. Unfortunately, IFCB data were not available at most stations concur- rently with traditional microscopy, and their conversion into DMSPp is uncertain. Further details on DMSPp parti- tioning estimates are provided in Text S3. 3.4. Biogeochemical factors affecting DMS concentrations These patterns cannot be explained by abiotic DMS removal pro- cesses controlling spatial distributions, because removal rates will be maximal in the surface layer and would likely increase towards the OW stations owing to stronger gas exchange and light penetration (Figure 4l). Therefore, var- iations in the net balance between biological production and consumption likely controlled the spatial distribution of the DMS:DMSPt ratio. Figure 9. Inferred bacterial DMSP metabolism in the MIZ. Vertical profiles at stations 507 and 519 in the MIZ are shown for (a) DMSPt and (b) DMS concentrations, (c) abundance of the bacterial DMSP cleavage gene dddP, and (d) the ratio between dddP and the DMSP demethylation gene dmdA. DOI: https://doi.org/ 10.1525/elementa.2020.00113.f9 Strong relationships were also found between DMS and cp (R2 ¼ 0.60) or TChl a (R2 ¼ 0.53; Figure S4).The highest DMS concentrations were clearly associated with a high Phaeocys- tis pouchetii abundance at the SCM, as found for DMSPt. In the subset of nine stations where microscopy counts were available,theabundanceofthisspecieswashighlycorrelated to DMS (rS ¼ 0.98, P < 10–4; Figure 7b) and explained virtually all of the DMS variance (99%) at the SCM. The rela- tionship between Phaeocystis and DMS was relatively weaker in the surface layer (rS ¼ 0.80, P < 0.05; Figure 7b), pointing toa morecomplex interactionbetween variousDMS produc- tion and removal pathways. To further examine the role of biological processes, we explored genes associated with bacterial DMSP metabo- lism at several stations. Unfortunately, quantifications were performed for the first leg of the Green Edge Expe- dition (transects T1–T3), where DMS and DMSP were not measured, and only for stations 507 and 519 in the MIZ during the second leg described here (T4–T7; Figure 1). The dddP gene encodes for a widespread bacterial DMSP- lyase, while dmdA encodes for the only known DMSP de- methylation enzyme, which diverts DMSP away from DMS production. Hence, the dddP:dmdA ratio should reflect the relative importance of these competing pathways, that is, it should serve as a proxy of bacterial DMS yield from dissolved DMSP. Overall, the abundance of both dddP and dmdA increased towards the top 20 m, following the ver- tical trend of bacterial abundance and production (Figure S2). 3.4. Biogeochemical factors affecting DMS concentrations The concentration of DMS was strongly correlated to its pre- cursor DMSPt, explaining 59% of its variance (Figure 6b). Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 16 of 31 Figure 9. Inferred bacterial DMSP metabolism in the MIZ. Vertical profiles at stations 507 and 519 in the MIZ are shown for (a) DMSPt and (b) DMS concentrations, (c) abundance of the bacterial DMSP cleavage gene dddP, and (d) the ratio between dddP and the DMSP demethylation gene dmdA. DOI: https://doi.org/ 10.1525/elementa.2020.00113.f9 DMS (Figure 8g and h; see also stepwise regression results in Text S4.1 and Table S3). In the absence of DMSP and DMS cycling rate measure- ments, the net DMSPt-to-DMS conversion efficiency can be approximated using the DMS:DMSPt ratio (Galı´ et al., 2018), with higher ratios implying relatively fast DMS pro- duction, relatively weak DMS removal, or both. The median DMS:DMSPt ratio was 0.10, but values were scat- tered across two orders of magnitude (0.01–1; Figures 3 and 6b). The most striking feature of the DMS:DMSPt ratio was its marked increase from 0.10 at the surface to 0.30 at 60 m at the MIZ and OW stations, but not under ice (Figure 8c). The increase of DMS:DMSPt with depth closely paralleled proxies of Phaeocystis relative abun- dance (Figure 8e and f) and phytoplankton biomass decay and grazing intensity (Figure 8d), suggesting that both factors enhanced DMS yields away from the sea surface. Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on y y Beyond their clear increase with depth, DMS:DMSPt ratios showed more subtle vertical and horizontal variations (Figure 8c) that were not matched by Phaeocystis or phy- todetritus proxies. First, in the MIZ and OW, median DMS:DMSPt ratios were higher in the 0–9 m layer than in the 9–21 m layer, and the difference was significant in the MIZ (P < 0.01; Bonferroni-adjusted Wilcoxon test). Second, median DMS:DMSPt ratios were maximal in the MIZ at all depths, even in the top 5 m of the water column (Figure 10g). Third, for a given Phaeocystis abundance, DMS was invariably higher at the surface than at the SCM (Figure 7b), which was not observed in the case of DMSPt. 3.4. Biogeochemical factors affecting DMS concentrations (a) Sea–air DMS flux; (b) fractional sea-ice coverage (SIC); (c) wind speed; (d) sea-surface temperature; (e) near sea-surface DMS concentration (0–5 m); (f) DMS ventilation rate constants (kvent) in the equivalent mixed layer (hBD; see text); (g) near sea-surface DMS:DMSPt ratio (0–5 m); and (h) near sea-surface DMSPt concentration (0–5 m). Box and whisker plots distinguish between ice-covered waters (ICE), the marginal ice zone (MIZ), and open waters (OW; Table 1). Thick horizontal lines represent group medians and empty squares represent group means. Boxes encompass the interquartile range (IQR), and whiskers encompass measurements within +1.5 IQR from the median. Individual measurements outside of this range are displayed as outliers. DOI: https://doi.org/10.1525/elementa.2020.00113.f10 > 15 nM (Figure 9). In the OW, the highest dddP:dmdA ratios occurred in the 9–41 m layer with a median of 3–4 (data from leg 1a), in good accordance with enhanced DMS below the surface (Figure 4d). The dddP:dmdA ratios reported here provide a lower bound for the ratio between bacterial cleavage and demethylation genes, because as many as eight genes encoding for bacterial DMSP lyases have been described to date (Curson et al., 2011; Li et al., 2020), of which three (dddK, dddP, and dddD) appear to be abundant in the global surface ocean (Landa et al., 2019). sea–air exchange was typically <0.03 d–1 (>30 d) at the ICE stations and 0.04–0.10 d–1 (10–25 d) at the MIZ and OW stations. Assuming instead that ventilation affected only the shallow layer defined by MLD0.03 (Table 1), we obtained shorter turnover times, especially in the MIZ where the median and maximum were 0.3 d–1 (3 d) and 1 d–1, respectively. Wind speed during the 24 h prior to sampling did not show significant correlations to the residual DMS variance, that is, the DMS variance not ex- plained by phytoplankton biomass. Thus, ventilation did not deplete near-surface DMS appreciably. > 15 nM (Figure 9). In the OW, the highest dddP:dmdA ratios occurred in the 9–41 m layer with a median of 3–4 (data from leg 1a), in good accordance with enhanced DMS below the surface (Figure 4d). 3.4. Biogeochemical factors affecting DMS concentrations Yet, the dddP:dmdA ratio varied widely, from median ratios 1 under ice to ratios > 3 in the MIZ and OW, suggesting that changes in bacterial DMSP catabolism could partly explain spatial patterns of DMS and DMS:DMSPt. In the MIZ, the highest dddP:dmdA ratios were found in the surface layer, with a median of 3 and a maximum of 7 at station 507, concurrent with high DMS Pigment data confirmed the results based on phyto- plankton counts, with the highest DMS concentrations (>20 nM; Figure 3) matched by the highest concentra- tions of the But-fuco-like pigment (>0.03 mg m–3) observed at stations 400 (surface), 512 (15 m) and 615 (30 m). Up to 81% of the DMS variance could be explained with an optimal multilinear regression model using the same subset of predictor pigments as for DMSPt (except for TChl a; Table S3). However, unlike DMSPt, a striking 94% of DMS variance was explained by adding the But- fuco-like pigment to the regression model in the subset of samples that contained it. In the RDA (Figure 2), pigment markers of Phaeocystis and other haptophytes, diatoms and grazing clustered with DMS. Markers for potential DMS producers such as type-1 dinoflagellates (Peri) and pico- or nanohaptophytes (Hex-fuco, But-fuco) clustered with DMSPt, perhaps related to their higher relative abun- dance in samples with lower-than-average biomass and Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 17 of 31 Figure 10. Sea–air DMS fluxes and corresponding controlling factors across the ice edge. (a) Sea–air DMS flux; (b) fractional sea-ice coverage (SIC); (c) wind speed; (d) sea-surface temperature; (e) near sea-surface DMS concentration (0–5 m); (f) DMS ventilation rate constants (kvent) in the equivalent mixed layer (hBD; see text); (g) near sea-surface DMS:DMSPt ratio (0–5 m); and (h) near sea-surface DMSPt concentration (0–5 m). Box and whisker plots distinguish between ice-covered waters (ICE), the marginal ice zone (MIZ), and open waters (OW; Table 1). Thick horizontal lines represent group medians and empty squares represent group means. Boxes encompass the interquartile range (IQR), and whiskers encompass measurements within +1.5 IQR from the median. Individual measurements outside of this range are displayed as outliers. DOI: https://doi.org/10.1525/elementa.2020.00113.f10 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 Figure 10. Sea–air DMS fluxes and corresponding controlling factors across the ice edge. 3.4. Biogeochemical factors affecting DMS concentrations The dddP:dmdA ratios reported here provide a lower bound for the ratio between bacterial cleavage and demethylation genes, because as many as eight genes encoding for bacterial DMSP lyases have been described to date (Curson et al., 2011; Li et al., 2020), of which three (dddK, dddP, and dddD) appear to be abundant in the global surface ocean (Landa et al., 2019). 3.5. Sea-air DMS flux (FDMS) and its drivers Sea–air fluxes ranged between <0.05 and 26 mmol m–2 d–1. When plotted on a continuous OWD axis, FDMS peaked between 0 and 20 OWD, wherein it usually exceeded 10 mmol m–2 d–1. The average FDMS was highest at the OW stations, with 10.0 mmol m–2 d–1 (Figure 10a). Correcting for the daily fractional ice cover (i.e., multiplying the FDMS from ice-free waters by 1 – SIC) decreased the mean FDMS from 2.3 to 0.3 mmol m–2 d–1 in the ICE stations and from 7.3 to 6.9 mmol m–2 d–1 in the MIZ stations. However, the main driver of the increasing FDMS towards OW was not the decreasing sea-ice cover, but the concomitant increase in the sea-surface DMS concentration, wind speed and SST (Figure 10). Among these four factors, the sea-surface DMS concentration and wind speed exerted the strongest con- trol on FDMS, together explaining 75% of the total variance in the flux data. 4.1. DMS cycling regimes and pathways 4.1.1. Phaeocystis pouchetii subsurface 4.1.1. Phaeocystis pouchetii subsurface maxima Phaeocystis was present at fairly high background levels at most stations according to cell counts (Figures 7 and S1) and marker pigments (Figure 8e and f), and its abun- dance typically increased towards the SCM. Massive Phaeo- cystis growth, however, was confined to Atlantic- influenced waters below the fresh surface layer, where the highest DMSPt and DMS concentrations were recorded (Figures 3 and 5). The highest Phaeocystis abundance found in our dataset (1.2107 cells L–1; station 512) is, to our knowledge, a record high for Baffin Bay, and similar to the highest abundances reported so far in the Arctic (1.2107 cells L–1 in Kongsfjorden, Eilertsen et al., 1989; 1.8107 cells L–1 over the Chukchi Plateau, Sherr et al., 2003; 8.7107 cells L–1 over the Yermak Plateau in the Fram Strait, Assmy et al., 2017). Our estimates of Phaeo- cystis biomass are conservative because they do not include colonies (see Section 3.4 and Text S3). Wassmann et al. (2005) found that colonies made up, on average, 16% of the total Phaeocystis biomass in the euphotic layer in the Barents Sea and adjacent areas. In our data- set, colonial Phaeocystis may have represented a larger fraction of the total Phaeocystis biomass in the surface layer, given the large fraction of DMSPp that could not be reconstructed from available phytoplankton cell counts (Section 3.3). Understanding the dispersion mechanisms underpin- ning Arctic Phaeocystis blooms is key to assessing its role in the present and future Arctic. Abundant recent litera- ture supports the association of Phaeocystis pouchetii with Atlantic waters at high northern latitudes (e.g., Galı´ and Simo´, 2010; Metfies et al., 2016; Assmy et al., 2017; Engel et al., 2017; Kubiszyn et al., 2017; Simo-Matchim et al., 2017; Krawczyk et al., 2018; Ardyna et al., 2020), and its preference for growing below the pycnocline, in more saline and less irradiated waters (Lasternas and Agustı´, 2010; Simo-Matchim et al., 2017; this study). Despite its well-established Atlantic origin, however, the median rel- ative Phaeocystis abundance during Green Edge was sim- ilar across east–west transects according to marker pigments (Figure 8e and f). We hypothesize that seed populations carried by the Atlantic inflow (West Green- land Current) mixed with Arctic water masses (Figure 1), a process likely enhanced by submesoscale motions at the ice edge (Castro et al., 2017; Manucharyan and Thompson, 2017). 4. Discussion h di ib i The distribution of DMSPt and DMS across the Baffin Bay receding ice edge largely followed the evolution of the phytoplankton bloom. The maximum observed DMSPt (524 nM) and DMS (74 nM) concentrations are, by far, the highest reported in the Atlantic and Canadian sectors of the Arctic to our knowledge (see compilations by Jarnı´ko- va´ et al., 2018, and Matrai et al., 2007; Lizotte et al., 2020). Observed near-surface DMS concentrations and sea–air FDMS were within the range of previous studies (Table 3). In the first part of this section we analyze the relation- ship between the physicochemical environment and the spatiotemporal patterns of DMSPt and DMS concentra- tions, with the goal of inferring the main processes con- trolling DMS cycling and emissions across the ice edge. This analysis draws on the comprehensive description of the ecosystem (and, foremost, phytoplankton dynamics) provided by the Green Edge datasets (Lafond et al., The distribution of DMSPt and DMS across the Baffin Bay receding ice edge largely followed the evolution of the phytoplankton bloom. The maximum observed DMSPt (524 nM) and DMS (74 nM) concentrations are, by far, the highest reported in the Atlantic and Canadian sectors of the Arctic to our knowledge (see compilations by Jarnı´ko- va´ et al., 2018, and Matrai et al., 2007; Lizotte et al., 2020). Observed near-surface DMS concentrations and sea–air FDMS were within the range of previous studies (Table 3). I h fi f hi i l h l i In the first part of this section we analyze the relation- ship between the physicochemical environment and the spatiotemporal patterns of DMSPt and DMS concentra- tions, with the goal of inferring the main processes con- trolling DMS cycling and emissions across the ice edge. This analysis draws on the comprehensive description of the ecosystem (and, foremost, phytoplankton dynamics) provided by the Green Edge datasets (Lafond et al., Assuming DMS ventilation losses affected the entire mixed layer represented by hBD, DMS turnover due to Art. 9(1) page 18 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ 2019; Randelhoff et al., 2019; Saint-Be´at et al., 2020). In the second part, we present pan-Arctic estimates of present-day EDMS, MIZ, and discuss the implications of changing ice cover for future DMS emissions. 4.1. DMS cycling regimes and pathways 4.1.1. Phaeocystis pouchetii subsurface In this regard, marked patchiness was evident in high resolution hydrographic data recorded during Green Edge, and also in the large variation of di- methylated sulfur concentrations registered at some sta- tions between the morning and afternoon CTD casts (e.g., station 615, 30 m sample in Figure 3). Thus, water-mass mixing possibly allowed for the early initiation of the bloom under ice, perhaps connected to the colonization of bottom ice by Phaeocystis (see Section 4.1.3), eventually leading to a massive subsurface bloom under optimal environmental conditions. Large-scale under-ice Phaeocys- tis blooms have been observed near the Fram Strait (Ass- my et al., 2017). These blooms could eventually produce very high FDMS if they were rapidly advected away from the ice cover and suddenly exposed to high irradiance (Galı´ et al., 2013; Vance et al., 2013; Galindo et al., 2016). Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 Phaeocystis single-cell abundance explained an over- whelming fraction of the DMS variance in the SCM layer (R2 ¼ 0.99, n ¼ 9; Figure 7b). Such a strong relationship is unprecedented in the literature, to our knowledge, and suggests that Phaeocystis DMSP-lyase activity controlled DMS production in the SCM layer, even at stations where this species did not dominate the DMSPp pool (Figure S3). Our finding is akin to the very strong association between Phaeocystis globosa biomass and potential DMSP-lyase activity (R2 ¼ 0.97) found by Stefels et al. (2007) during a coastal phytoplankton bloom. Still, the observed corre- lation does not exclude the contribution of bacterial DMSP cleavage (Figure S2). Although the positive correla- tion between Phaeocystis abundance and DMS was not as strong in the near-surface, it was significant and suggests that this species made a substantial contribution to DMS production in the surface layer. 4. Discussion h di ib i subsurface DMS stocks was probably the consumption by specialized methylotrophic bacteria (Vila-Costa et al., 2006; del Valle et al., 2007), capable of rapid biological DMS turnover (0.5–3 d–1) in subsurface Arctic waters (Galı´ and Simo´, 2010). Events of rapid DMS outgassing from the SCM could occur if storms were powerful enough to erode the stable pycnocline (Le Clainche et al., 2006). However, such events are unlikely to happen under high ice-melt rates typically present at the ice edge in summer (Randelh- off et al., 2019). 4.1.2. Phytoplankton-bacterial connections modu- lated by stress in MIZ surface waters In the surface layer of the MIZ, high DMS concentrations and relatively high DMS:DMSPt ratios possibly resulted from the multiplicative effects of enhanced DMSP and DMS release by phytoplankton (Sunda et al., 2007; Galı´ et al., 2013) and enhanced bacterial DMS yields from the microbial consumption of DMSP (Slezak et al., 2007). Uti- lization of DMSP by the free-living bacterial community clearly increased towards the surface, and the highest ratios between the potential cleavage and demethylation activities (dddP:dmdA) were found in the top 10 m of the water column in several MIZ and some OW stations (Fig- ures 9 and S4). High bacterial DMS yields are a response Despite the sharp vertical DMS gradients encountered, upward diffusion of DMS from the subsurface maxima likely made a minor contribution to sea–air DMS fluxes, compared to DMS cycling in the upper mixed layer. Our order-of-magnitude calculations (Text S5) indicate that diapycnal DMS transport was generally smaller than other DMS budget terms. An exception was found at station 512, where the diapycnal DMS flux may have been com- parable in magnitude to the sea–air flux, though smaller than biological and photochemical DMS turnover rates. According to current knowledge, the main fate of Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 19 of 31 Art. 9(1) page 19 of 31 (Toole et al., 2006; Galı´ and Simo´, 2010). In any case, the net effect of interacting DMS production and consump- tion pathways was a peak in DMS concentrations in the stably stratified surface layer of the MIZ (Table 1; Figure 4). to the oversupply of reduced sulfur, allowing bacteria to retain the carbon moiety of DMSP while releasing the sulfur as DMS (Kiene et al., 2000; Varaljay et al., 2015; Galindo et al., 2015). , ) Phytoplanktonic DMSP has to be released to the extra- cellular medium to be usable by bacteria. Because zwitter- ions like DMSP cannot passively cross healthy cell membranes, DMSP release must result from cell breakage by grazers, cell membrane damage, or active exudation through membrane transporters or secretory vesicles (Ste- fels et al., 2007; Orellana et al., 2011). Release of DMSP through zooplankton grazing was likely heightened in mature bloom stages in OW (Saint-Be´at et al., 2020), prob- ably accompanied by viral lysis (Malin et al., 1998) and autolysis. 4.1.3. Ice-water exchange Very high concentrations of DMSPt (255 nM) and DMS (22 nM) were found beneath the ice at station 400, concur- rently with an anomalously high DMSPt:cp ratio (Figure 6a) and But-fuco-like concentration. This outlier observa- tion (see Figure 3) suggests that Phaeocystis grew in and was released from bottom sea ice, adding to scarce previ- ous reports of Phaeocystis colonizing sea-ice environments (Ferna´ndez-Me´ndez et al., 2018; Selz et al., 2018). Spores of the ice diatom Melosira were also found in the water column, in particular at the SCM of stations 512 (MIZ) and 703 (OW), and potentially contributed to the DMSP pool. Galindo et al. (2014; Galindo et al., 2015) observed that, at the onset of the melt season, snow melting caused ice drainage events that flushed ice-bottom biomass and seeded the pelagic diatom bloom, boosting bacterial DMSP cycling beneath the sea ice. If Phaeocystis was pres- ent in the sea ice, such a process could further enhance DMS production and eventual outgassing. Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 0 p g g To further explore the role of ice melt in supplying dimethylated sulfur compounds to the upper water col- umn, we analyzed the correlation between several vari- ables and salinity, repeating the analysis in progressively thicker layers from 0–5 m to 0–41 m (Figure S5). The correlation between DMS and salinity became more neg- ative as shallower layers were considered, reaching rS ¼ – 0.49 (P < 0.05) in the 0–5 m layer. Other variables (DMSPt, TChl a and cp) showed similar but nonsignificant patterns, whereas Phaeocystis diagnostics (Chl c3, But-fuco-like) were positively correlated to salinity but only if a deeper layer (0–41 m) was considered. The significantly negative correlation between salinity and near-surface DMS may be seen as supportive of the role of ice-released organic mat- ter; however, this correlation was clearly weaker than that found between DMS and Phaeocystis (Figure 7b). More- over, an important role of sea ice in directly supplying DMS or its precursors to near-surface waters during Green Edge is at odds with other observations. First, surveys covering a longer period of the melt season indicated that the majority of ice-bound DMSP (Galindo et al., 2014) and POC (Amiraux et al., 2019) was released to the water col- umn well before permanent ice breakup. 4.1.2. Phytoplankton-bacterial connections modu- lated by stress in MIZ surface waters In contrast, release of DMSP upon radiative dam- age was likely important in the MIZ surface layer, as pre- viously observed experimentally by Galindo et al. (2016). In support of this hypothesis, Alou-Font et al. (2016) observed that around 20% of the phytoplankton cells in the Arctic surface layer had permeable (compromised) membranes for daily PAR between 20–40 mol photons m–2 d–1, and this fraction exceeded 50% for daily PAR > 50 mol photons m–2 d–1. Median (maximum) daily PAR levels of 31 (35) and 38 (52) mol photons m–2 d–1, respec- tively, were recorded in MIZ and OW stations during our cruise (Figure 4l). Despite the much higher light levels in the MIZ and OW surface layers, the major photoprotective pigments were not upregulated in the MIZ compared to ICE stations (Figure 8k and l), whereas a distinct photo- protective response was found in the OW stations. Thus, insufficient photoprotection might have favored DMSP release by MIZ phytoplankton. Finally, the enhanced avail- ability of algal substrates can also be inferred from the high bacterial abundance and production in the 0–21 m layer in the MIZ (Figure S4). 4.1.3. Ice-water exchange In the coastal Baffin Bay, the fraction of ice-derived POC was typically less than 5% in the top 10 m of the water column during the three weeks preceding ice opening (Amiraux et al., 2019). Second, model results (Hayashida et al., 2020) also indicate that ice-released DMSP is unlikely to fuel ocean- emitted DMS during the advanced melt season; yet, it could have fueled DMS emission in May, during the bottom-ice algal bloom. y ( g ) High DMSP availability in the MIZ may have also re- sulted from upregulated DMSP synthesis in diatoms as a consequence of nutrient and radiative (ultimately oxida- tive) stress (Sunda et al., 2002; McParland and Levine, 2019). Diatoms made up a substantial fraction of the phy- toplankton community in the 0–21 m layer in the MIZ according to pigment ratios (Figure 8i and j). As described in Text S3, a DMSP content of 0.4% of cell carbon in diatoms (Stefels et al., 2007) was insufficient to recon- struct DMSP concentrations during Green Edge. A higher DMSP-carbon content of 2% provided a better fit, and an even higher content was compatible with observed near- surface DMSPt concentrations. On the other hand, diatoms may also obtain antioxidant DMSP through uptake (The- seira et al., 2020), a pathway that is enhanced under high irradiance (Ruiz-Gonza´lez et al., 2012) and that diverts DMSP from DMS production. In addition to boosting DMS production through the coupled activities of phytoplankton and bacteria, nutrient and radiative stress in the MIZ and OW surface layer likely enhanced direct DMS release by phytoplankton (Sunda et al., 2002; Sunda et al., 2007). UVR may have partially offset the increase in bacterial DMS production via photo- inhibition of DMSP uptake (Slezak et al., 2007), while simultaneously inhibiting bacterial DMS consumption 4.1.4. Detrital matter 4.1.4. Detrital matter The increase in DMS:DMSPt with depth paralleled the increase in pigment markers for both Phaeocystis and Galı´ et al: DMS Emissions from the Arctic MIZ Art. 9(1) page 20 of 31 Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 phytodetritus (Figure 8d–f). The latter is consistent with the high abundance of large marine snow aggregates in the MIZ and OW found during Green Edge by Trudnowska et al. (2021). Our observation suggests that (1) efficient DMSP-to-DMS conversion occurred in suspended and sink- ing particles below the euphotic layer (depth > 40 m), and (2) this DMS production was not fully balanced by increased biological DMS consumption. Increased DMSP cleavage may have resulted from both phytoplanktonic and bacterial DMSP-lyase activities, combined with parti- cle processing by zooplankton. The few available studies indicate that large aggregates and marine snow may be a hotspot for both DMSP cleavage (Scarratt et al., 2000) and demethylation (Steiner et al., 2019) and for DMS bio- consumption (del Valle et al., 2009) in productive marine settings where particle-attached lifestyles tend to be more prevalent (Teeling et al., 2012). Moreover, bacterial groups with affinity to particles such as Bacteroidetes, gammapro- teobacteria and some alphaproteobacteria (Teeling et al., 2012; Ferna´ndez-Go´mez et al., 2013) can harbor several DMSP lyase genes (Curson et al., 2011; Zeng et al., 2016). Particle sinking removes DMSP from the surface layer where it could potentially fuel DMS emissions. The impor- tance of this process depends on the sinking-speed spec- trum of DMSP-bearing particles, relative to biological DMSP turnover in the surface layer. Lizotte et al. (2008) found that less than 2% d–1 of DMSPp sank from the bottom of the euphotic zone during the decay of the Northwest Atlantic diatom bloom. Studies of Phaeocystis pouchetii blooms generally found low vertical export effi- ciency (Reigstad and Wassmann, 2007; Assmy et al., 2017), although export occasionally may be enhanced by pro- cesses other than gravitational sinking, e.g. water-mass subduction (Rellinger et al., 2009). During Green Edge, the vertical patterns of DMS:DMSPt and marker pigments suggest relatively slow sinking and rapid breakdown of DMSP-bearing aggregates, which altogether would atten- uate the gravitational DMSP flux. Figure 11. Evolution of Arctic MIZ DMS emission during summer (May–August) aggregated by 5 latitude bands. 4.1.4. Detrital matter During Green Edge, the vertical patterns of DMS:DMSPt and marker pigments suggest relatively slow sinking and rapid breakdown of DMSP-bearing aggregates, which altogether would atten- uate the gravitational DMSP flux. Figure 11. Evolution of Arctic MIZ DMS emission during summer (May–August) aggregated by 5 latitude bands. Bars represent DMS emission from the MIZ (EDMS, MIZ; left axis); lines represent the additional percentage of EDMS from the MIZ compared to OW (right axis). Colors represent methods A (light cyan) and B (dark cyan) used to compute MIZ extent (see Section 4.2). The left y-axis scale is the same in all panels to facilitate comparisons of the magnitude of EDMSMIZ among latitude bands, whereas right y axes scales differ because the percentages span two orders of magnitude. DOI: https://doi.org/10.1525/elementa.2020.00113.f11 explain the episodic spikes in DMS concentrations just above the sea surface reported in Baffin Bay and the Cana- dian Arctic (Mungall et al. 2016; Ghahremaninezhad et al., 2019). 4.1.4. Detrital matter Bars represent DMS emission from the MIZ (EDMS, MIZ; left axis); lines represent the additional percentage of EDMS from the MIZ compared to OW (right axis). Colors represent methods A (light cyan) and B (dark cyan) used to compute MIZ extent (see Section 4.2). The left y-axis scale is the same in all panels to facilitate comparisons of the magnitude of EDMSMIZ among latitude bands, whereas right y axes scales differ because the percentages span two orders of magnitude. DOI: https://doi.org/10.1525/elementa.2020.00113.f11 phytodetritus (Figure 8d–f). The latter is consistent with the high abundance of large marine snow aggregates in the MIZ and OW found during Green Edge by Trudnowska et al. (2021). Our observation suggests that (1) efficient DMSP-to-DMS conversion occurred in suspended and sink- ing particles below the euphotic layer (depth > 40 m), and (2) this DMS production was not fully balanced by increased biological DMS consumption. Increased DMSP cleavage may have resulted from both phytoplanktonic and bacterial DMSP-lyase activities, combined with parti- cle processing by zooplankton. The few available studies indicate that large aggregates and marine snow may be a hotspot for both DMSP cleavage (Scarratt et al., 2000) and demethylation (Steiner et al., 2019) and for DMS bio- consumption (del Valle et al., 2009) in productive marine settings where particle-attached lifestyles tend to be more prevalent (Teeling et al., 2012). Moreover, bacterial groups with affinity to particles such as Bacteroidetes, gammapro- teobacteria and some alphaproteobacteria (Teeling et al., 2012; Ferna´ndez-Go´mez et al., 2013) can harbor several DMSP lyase genes (Curson et al., 2011; Zeng et al., 2016). Downloaded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 Particle sinking removes DMSP from the surface layer where it could potentially fuel DMS emissions. The impor- tance of this process depends on the sinking-speed spec- trum of DMSP-bearing particles, relative to biological DMSP turnover in the surface layer. Lizotte et al. (2008) found that less than 2% d–1 of DMSPp sank from the bottom of the euphotic zone during the decay of the Northwest Atlantic diatom bloom. Studies of Phaeocystis pouchetii blooms generally found low vertical export effi- ciency (Reigstad and Wassmann, 2007; Assmy et al., 2017), although export occasionally may be enhanced by pro- cesses other than gravitational sinking, e.g. water-mass subduction (Rellinger et al., 2009). 5. Conclusions and outlook g g Ice melting and water-column stratification are poised to play an increasingly critical role in controlling Arctic EDMS. The stability of stratification has numerous biogeo- chemical ramifications as it controls, at least: (1) phyto- plankton productivity, through turbulent nutrient supply (Randelhoff et al., 2020), and ensuing nutrient stress that can boost DMSP production by low DMSP producers such as diatoms (McParland and Levine, 2019); (2) mean PAR and UVR exposure of the entire food web in the upper water column, and therefore radiative stress-driven DMSP and DMS cycling (Galı´ and Simo´, 2010; Galı´ et al., 2013; Vance et al., 2013); and (3) CO2 solubility and invasion of deeper waters, and thus the carbonate system parameters and pH, with potentially important effects on sulfur cycling (Hussherr et al., 2017; Hopkins et al., 2020; Be´nard et al., 2021). In addition, better understanding is needed of ice-water biogeochemical fluxes and biogeochemical transformations of ice-released organic matter (Galindo et al., 2014; Hayashida et al., 2020). Physicochemical pro- cesses that alter the near-surface turbulence can further amplify the uncertainty in MIZ DMS emissions. Examples include the turbulence generated by ice drag, and the formation of freshwater lenses and sea-surface microlayers enriched in surfactants, damping near-surface turbulence (Carpenter et al., 2012). The Arctic MIZ can produce strong DMS emission pulses (Table 3), which can influence aerosol formation and properties in the clean Arctic summer atmosphere (Abbatt et al., 2019). Although their pan-Arctic magnitude is still uncertain (likely 5–13 Gg S yr–1), these emissions cannot be overlooked and the biogeochemical factors that control them deserve further study. Productive polar ice edges like the one described here are paradigmatic of the bloom regime, whereby phytoplankton biomass is a strong pre- dictor of dimethylated sulfur dynamics (Toole and Siegel, 2004). Here we attempted to dissect the bloom regime, which allowed us to infer a number of co-occurring DMS cycling pathways linked to different autotrophic and he- trotrophic activities. Their spatiotemporal distribution was related to processes that shape the development of the late spring bloom in Baffin Bay, namely, the W–E gradient between Arctic (Pacific-derived) and Atlantic water masses, ice melting, and their combined effects on the diverging vertical gradients of light and nutrient availability (Lafond et al., 2019; Randelhoff et al., 2019; Saint-Be´at et al., 2020). ) Below we summarize the main findings of our study: Water masses. 4.2. Present and future EDMS from the MIZ: Physical and biological constraints The DMS fluxes measured in the MIZ during Green Edge are within the range of previous studies in the Arctic MIZ (<1–33 mmol m–2 d–1), and the compilation in Table 3 suggests that the mean FDMS from ice-free waters in the MIZ is typically between 5 and 10 mmol m–2 d–1. This range is consistent with the FDMS diagnosed from satellite ocean color data immediately after ice opening in several Arctic subregions (Galı´ et al., 2019). Given that estimated mean FDMS from ice-free waters between May and August at latitudes >70N was 2 mmol m–2 d–1 (Galı´ et al., 2019), phytoplankton blooms in the Arctic MIZ can be qualified as DMS emission hotspots. In the aerosol-poor summer Arctic atmosphere, episodes of high FDMS, MIZ can easily exceed aerosol nucleation thresholds (Leaitch et al., 2013; Collins et al., 2017; Dall’Osto et al., 2017). The epi- sodic and patchy nature of DMS flux from the MIZ is an important attribute regarding aerosol nucleation events, as recently highlighted by Webb et al. (2019) in a study of the West Antarctic Peninsula. These attributes may partly No published estimates of pan-Arctic EDMS, MIZ are available, to our knowledge, in spite of their importance for high Arctic aerosols. Assuming a mean FDMS in ice-free MIZ waters of 5 mmol m–2 d–1, pan-Arctic EDMS, MIZ between 2003 and 2014 would amount to a mean (+standard deviation, n ¼ 12) of 5.4 (+0.6) Gg S yr–1 in scenario A (where the duration of the MIZ status is fixed at 8 days for any given pixel) and 12.9 (+1.6) Gg S yr–1 in scenario B (which allows for variable duration of the “MIZ period” for any given pixel and year; Figure 11). In both scenarios, the 70–80N latitude band would account for 75% of the total EDMS, MIZ north of 65N. We also computed the weight of EDMS, MIZ relative to open water EDMS (EDMS, OW) in June and July as 100 x EDMS, MIZ/ EDMS, OW. The additional contribution of EDMS, MIZ increases with latitude, from a modest 3%–6% (means of A–B sce- narios) in the 65–70N band to 60%–162% in the 80– Art. 9(1) page 21 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ 85N band. 4.2. Present and future EDMS from the MIZ: Physical and biological constraints Our estimates also suggest a temporal increase in the magnitude of EDMS, MIZ of about 10 + 4% yr–1 (P < 0.05) between 2003 and 2014 north of 80N, with highest EDMS, MIZ in years with record-low sea-ice extent (2007 and 2012; Figure 11a). This increas- ing trend is consistent with the progressive widening of the MIZ in summer detected until 2011 (Strong and Rigor, 2013). Our estimates are admittedly crude and should be refined as new field measurements become available. Large uncertainty arises from assigning a fixed FDMS to all pixels flagged as MIZ, and future statistical upscaling ex- ercises should take into account spatial variations in sea- surface DMS concentrations across contrasting Arctic bio- geochemical regimes. 2008), no silicate requirement (Ardyna et al., 2020), the protection from zooplankton predators provided by muci- laginous colonies (Long et al., 2007; Verity et al., 2007), and even allelopathy (Hansen and Eilertsen, 2007) may explain the ability of Phaeocystis to outcompete diatoms and form massive blooms under certain conditions. The northward expansion of Phaeocystis along with Atlantic waters in the Barents Sea detected by remote sensing (Orkney et al., 2020), analogous to that of its temperate relative Emiliania huxleyi in subpolar waters, indicates ongoing floristic shifts are impacting Arctic EDMS. Still, what factors control Phaeocystis abundance and eventual dominance elsewhere in the Arctic remains unknown. 5. Conclusions and outlook Extremely high biomass of Phaeocystis pouchetii in the subsurface bio- mass maximum, along with very high DMSPt and DMS, were associated unambiguously with Atlantic-influenced waters. Marked spa- tial patchiness in planktonic biomass and biogenic sulfur concentrations were observed in the mixing zone between the Arctic and Atlantic water masses. Over large spatiotemporal scales, uncertainty in present-day and future estimates of EDMS, MIZ may also arise from changes in plankton biogeography. In this re- gard, knowledge of the ecological niches occupied by the main blooming groups, that is, diatoms and Phaeocystis pouchetii, is key. Available data tie P. pouchetii to the Atlan- tic inflow at high northern latitudes, with a growth opti- mum between 1C and 5C (Brun et al., 2015), likely warmer than that of some Arctic diatoms (Lacour et al., 2017). Others have suggested that P. pouchetii does not grow well in meltwater-influenced surface waters (Laster- nas and Agustı´, 2010) and, interestingly, decreased growth at low salinities was reported for its austral relative P. antarctica (Kameyama et al., 2020). Despite its apparent preference for relatively low irradiance (Simo-Matchim et al., 2017; this study), other field data suggest that this species features extremely wide photophysiological plas- ticity, as it can grow efficiently at both low (2 mol photons m–2 d–1; Assmy et al., 2017) and high (40 mol photons m–2 d–1; Cota et al., 1994) irradiance. This plasticity, the ability to use regenerated nitrogen forms (Sanderson et al., Temporal evolution. A single space-for-time axis, based on the open water days (OWD) metric, captured major features of the tem- poral evolution of the planktonic biomass and dimethylated sulfur concentrations, from bloom initiation under the ice to its peak in the MIZ and decay in open waters. DMSP producers and particulate frac- tions. Pigment markers for phytoplankton taxonomy and photoacclimation explained Art. 9(1) page 22 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Phytoplankton biogeography and bio- geochemical regimes. Further field and lab work is required to understand the ecological niche occupied by Phaeocystis pouchetii to predict its fitness and prevalence in the future Arctic compared to diatoms. Dinoflagellates and picoeukaryotes should be included in global compilations and ecological niche studies (e.g., Brun et al., 2015). 5. Conclusions and outlook The latter groups may contribute disproportionately to the background (non-bloom) levels of DMSPt and DMS production in polar environments, and some can play a dual role acting as gra- zers (Stoecker et al., 2017). The response of plankton communities to decadal- or centennial-scale environmental changes (AMAP, 2017; IPCC, 2019; Inness et al., 2020) is poorly understood. 80% of DMSPt variance. The contribution of different DMSP producers changed progres- sively from samples with very high DMSPt (Phaeocystis-dominated), moderate DMSPt (diatoms, dinoflagellates) and low DMSPt (di- noflagellates, non-Phaeocystis pico- and na- noeukaryotes). Particulate DMSP represented 1–10% of the POC. Detrital matter, zoo- plankton biomass and perhaps ice-released particles may have accounted for a substantial DMSPt fraction in some samples. Microbial DMS cycling pathways. Meltwa- ter stratification caused the vertical segrega- tion of sulfur cycling pathways linked to different auto- and hetetrotrophic processes. In the surface layer of the MIZ (and to some extent OW) we inferred high DMSP-to-DMS conversion via phytoplankton-bacterial inter- actions, likely favored by nutrient and light stress, ultimately leading to high median DMS:DMSPt ratios (0.15) and DMS concentra- tions (12 nM). Conversely, Phaeocystis pou- chetii drove DMS production in subsurface biomass maxima. Efficient DMSP-to-DMS conversion was inferred in suspended or sinking aggregates below the euphotic layer, concurrent with a clear pigment signature of grazing. Microbial DMS cycling pathways. Meltwa- ter stratification caused the vertical segrega- tion of sulfur cycling pathways linked to different auto- and hetetrotrophic processes. In the surface layer of the MIZ (and to some extent OW) we inferred high DMSP-to-DMS conversion via phytoplankton-bacterial inter- actions, likely favored by nutrient and light stress, ultimately leading to high median DMS:DMSPt ratios (0.15) and DMS concentra- tions (12 nM). Conversely, Phaeocystis pou- chetii drove DMS production in subsurface biomass maxima. Efficient DMSP-to-DMS conversion was inferred in suspended or sinking aggregates below the euphotic layer, concurrent with a clear pigment signature of grazing. Repeat surveys and exploration. Continu- ation of survey programs in areas with long historical records is key to detecting trends driven by climate change. Studies in under- sampled regions (Central Basin and Siberian shelves, e.g., Uhlig et al., 2019; Schanke et al., 2020) are needed to upscale in situ mea- surements and generalize current knowledge. 5. Conclusions and outlook The transition toward a first-year Arctic ice cap may favor a high-DMS MIZ such as the one described here in contrast to the low DMS concentrations associated with multi-year ice margins (Lizotte et al., 2020). Given the wide range of environmental conditions sam- pled during the Green Edge cruise, we argue that our findings, and the new questions they have raised, are rel- evant for other regions in the Arctic. Although our study is limited by the descriptive nature of the dataset, it provides a good template for upcoming process studies and mod- eling efforts and will be useful to refine existing concep- tual models. Below we conclude by listing potential research priorities for future studies: Modeling. Biogeochemical modeling studies should place emphasis on the ability to reproduce (1) the large-scale distribution pat- terns of key phytoplankton groups (Wang et al., 2015), (2) the vertical segregation of plankton communities and sulfur-cycling re- gimes, and (3) biogeochemical fluxes between sea ice and water (Hayashida et al., 2020). These modeling efforts will better constrain the EDMS, MIZ estimates presented here (Figure 11) and allow for future projections, regional and pan-Arctic, of DMS emissions. Modeling. Biogeochemical modeling studies should place emphasis on the ability to reproduce (1) the large-scale distribution pat- terns of key phytoplankton groups (Wang et al., 2015), (2) the vertical segregation of plankton communities and sulfur-cycling re- gimes, and (3) biogeochemical fluxes between sea ice and water (Hayashida et al., 2020). These modeling efforts will better constrain the EDMS, MIZ estimates presented here (Figure 11) and allow for future projections, regional and pan-Arctic, of DMS emissions. Sulfur budgets. Comprehensive measure- ments of biological DMS production and consumption rates are needed to constrain DMS budgets and sea–air fluxes in relation to plankton and ice-melt dynamics. Special emphasis should be placed on (1) the micro- bial processes occurring in and around parti- cles across the entire size spectrum (cells, excretion vesicles, colonies, metazoan zoo- plankton and large detrital aggregates), (2) ice–water mass fluxes, and (3) the modulation of sulfur budgets by interacting environmen- tal stressors (Galı´ et al., 2013; Hopkins et al. 2020). Competing interests Arrigo, KR, Perovich, DK, Pickart, RS, Brown, ZW, van Dijken, GL, Lowry, KE, Mills, MM, Palmer, MA, Balch, WM, Bates, NR, Benitez-Nelson, CR. 2014. Phytoplankton blooms beneath the sea ice in the Chukchi sea. Deep Sea Research Part II: Topical Studies in Oceanography 105: 105–117. DOI: http:// dx.doi.org/10.1016/j.dsr2.2014.03.018. The authors declare no conflict of interest. ed from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Amiraux, R, Smik, L, Ko¨seog˘lu, D, Rontani, JF, Galindo, V, Grondin, PL, Babin, M, Belt, ST. 2019. Temporal evolution of IP25 and other highly branched isopre- noid lipids in sea ice and the underlying water col- umn during an Arctic melting season. Elementa: Science of the Anthropocene 7: 38. DOI: http://dx. doi.org/10.1525/elementa.377. Funding k Andreae, MO, Rosenfeld, D. 2008. Aerosol–cloud–pre- cipitation interactions. Part 1. The nature and sources of cloud-active aerosols. Earth-Science Re- views 89(1–2): 13–41. DOI: http://dx.doi.org/10. 1016/j.earscirev.2008.03.001. g We acknowledge funding from AGAUR (Generalitat de Catalunya) Beatriu de Pino´s postdoctoral fellowship pro- gram (MG), the Canada Excellence Research Chair in Remote Sensing of Canada’s New Arctic Frontier (MB), the Canada Research Chair on Ocean Biogeochemistry and Climate and an NSERC Discovery Grant Program and Northern Research Supplement Program (MLe), NETCARE (NSERC Climate Change and Atmospheric Research pro- gram, MLe), the U.S. National Science Foundation (NSF- OCE 1756907, DJK), Marie Curie Actions-International Outgoing Fellowship (PIOF-GA-2013-629378, JD), and Arc- ticNet (The Network of Centres of Excellence of Canada). This is a contribution to the research program of Que´bec- Oce´an and the Takuvik Joint International Laboratory (CNRS-France & Universite´ Laval-Canada). Arctic Monitoring & Assessment Programme. 2017. Snow, Water, Ice and Permafrost in the Arctic (SWIPA) 2017. Oslo, Norway: Arctic Monitoring & Assessment Programme. Ardyna, M, Mundy, CJ, Mills, MM, Oziel, L, Lacour, L, Verin, G, Van Dijken, G, Ras, J, Alou-Font, E, Ba- bin, M, Gosselin, M, Tremblay, J-E, Raimbault, P, Assmy, P, Nicolaus, M, Claustre, H, Arrigo, KR. 2020. Environmental drivers of under-ice phyto- plankton bloom dynamics in the Arctic Ocean. Ele- menta: Science of the Anthropocene 8: 30. DOI: http://dx.doi.org/10.1525/elementa.430. References Abbatt, JPD, Leaitch, WR, Aliabadi, AA, Bertram, AK, Blanchet, J, Boivin-Rioux, A, Bozem, H, Burkart, J, Chang, RYW, Charette, J, Chaubey, JP. 2019. Overview paper: New insights into aerosol and cli- mate in the Arctic. Atmospheric Chemistry and Phys- ics 19: 2527–2560. DOI: http://dx.doi.org/10. 5194/acp-2018-995. Supplemental files The supplemental files for this article can be found as follows: This article is accompanied by a Supplemental Material file that contains five subsections of text (S1–5), five fig- ures (S1–S5) and three tables (S1–S3). Alcolombri, U, Ben-Dor, S, Feldmesser, E, Levin,Y, Taw- fik, DS, Vardi, A. 2015. Identification of the algal dimethyl sulfide-releasing enzyme: A missing link in the marine sulfur cycle. Science 348(6242): 1466– 1469. DOI: http://dx.doi.org/10.1126/science. aab1586. Data accessibility statement Data are deposited in the LEFE CYBER Database (http:// www.obs-vlfr.fr/proof/) and can also be provided by the corresponding author M.G. on request. Scientific code for the analysis and plotting of DMS(P) data is publicly avail- able at https://github.com/mgali/GreenEdge. Scientific code for the analysis of physical data made by Randelhoff Art. 9(1) page 23 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ et al. (2019) is available at https://www.doi.org/10.5281/ zenodo.2653855. et al. (2019) is available at https://www.doi.org/10.5281/ zenodo.2653855. Acknowledgments We are thankful to the officers and crew of the CCGS Amundsen, Jean-E´ric Tremblay and Flavienne Bruyant for assuring the smooth development of the Green Edge expedition. We are indebted with Guillaume Masse´ for providing the GC-MS instrument and technical advice, and with the entire Green Edge for providing the core dataset: Jose´phine Ras and Annick Bricaud (HPLC), Pierre-Luc Grondin, Philippe-Israel Morin and Joannie Ferland (IFCB), Augustin Lafond (diatom biomass), Dominique Marie and Daniel Vaulot (bacteria and picoeukaryote abundances), Fabien Joux (bacterial production), Brent Else and Tonya Burgers (meteorological data), and Philippe Massicotte (data synthesis). Thanks are extended to Guillaume Le´guen for fieldwork support. We thank Alison Webb, two anony- mous reviewers, and editor Jody W. Deming for construc- tive criticisms that improved the paper. Alou-Font, E, Roy, S, Agustı´, S, Gosselin, M. 2016. Cell viability, pigments and photosynthetic performance of Arctic phytoplankton in contrasting ice-covered and open-water conditions during the spring summer transition. Marine Ecology Progress Series 543: 89–106. DOI: http://dx.doi.org/10.3354/ meps11562. Author contributions MG collected and analyzed DMS(P) samples onboard, ana- lyzed data, and wrote the manuscript with significant input from MLi, DJK, and AR, and contributions from all coauthors. Assmy, P, Ferna´ndez-Me´ndez, M, Duarte, P, Meyer, A, Randelhoff, A, Mundy, CJ, Olsen, LM, Kauko, HM, Bailey, A, Chierici, M, Cohen, L. 2017. Leads in Arctic pack ice enable early phytoplankton blooms below snow-covered sea ice. Scientific Re- ports 7: 40850. DOI: http://dx.doi.org/10.1038/ srep40850. Designed the DMS(P) sampling strategy: MG, MLi, MLe. Analyzed DMSP on land: LX, DJK, RH, MLi. Provided the analysis of the physical setting: AR. Conducted the sampling and analysis of bacterial abun- dance and activities: JD. Provided satellite estimates of ice opening dates: ER. Provided satellite estimates of ice opening dates: ER. Led the Green Edge project: MB. Art. 9(1) page 24 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Be´nard, R, Lizotte, M, Levasseur, M, Scarratt, M, Mi- chaud, S, Starr, M, Tremblay, J-E´, Kiene, RP, Ka- meyama, S. 2021. Impact of anthropogenic pH perturbation on dimethylsulfide cycling: A peek into the microbial black box. Elementa: Science of the Anthropocene 9(1): 00043. ultrafine particle formation and growth in the Cana- dian Arctic marine environment. Atmospheric Chem- istry and Physics 17: 13119–13138. DOI: http://dx. doi.org/10.5194/acp-2017-411. Cota, GF, Smith WO, Mitchell, BG. 1994. Photosynthesis of Phaeocystis in the Greenland Sea. Limnology and Oceanography 39(4): 948–953. DOI: http://dx.doi. org/10.4319/lo.1994.39.4.0948 Borcard, D, Gillet, F, Legendre, P. 2011. Canonical ordi- nation, in Numerical ecology with R. Springer: 153– 225. DOI: http://dx.doi.org/10.1007/978-1-4419- 7976-6. Coupel, P, Matsuoka, A, Ruiz-Pino, D, Gosselin, M, Marie, D, Tremblay, JE, Babin, M. 2015. Pigment signatures of phytoplankton communities in the Beaufort Sea. Biogeosciences 12(4): 991–1006. DOI: http://dx.doi.org/10.5194/bg-12-991-2015. Brean, J, Dall’Osto, M, Simo´, R, Shi, Z, Beddows, DCS, Harrison, RM. 2021. Open ocean and coastal new particle formation from sulfuric acid and amines around the Antarctic Peninsula. Nature Geoscience. DOI: http://dx.doi.org/10.1038/s41561-021- 00751-y. Croft, B, Martin, R V, Leaitch, WR, Tunved, P, Breider, TJ, Andrea, SDD, Pierce JR. 2016. Processes con- trolling the annual cycle of Arctic aerosol number and size distributions. Atmospheric Chemistry and Physics 16(6): 3665–3682. DOI: http://dx.doi.org/ 10.5194/acp-16-3665-2016. y Brun, P, Vogt, M, Payne, MR, Gruber, N, O’brien, CJ, Buitenhuis, ET, Le Que´re´, C, Leblanc, K, Luo, YW. 2015. Ecological niches of open ocean phytoplank- ton taxa. Limnology and Oceanography 60(3): 1020– 1038. DOI: http://dx.doi.org/10.1002/lno.10074. Author contributions Curson, A, Liu, J, Bermejo Martı´nez, A, Green, RT, Chan, Y, Carrio´n, O, Williams, BT, Zhang, S-H, Yang, G-P, Page, PCB, Zhang, X-H, Todd, JD. 2017. Dimethylsulfoniopropionate biosynthesis in marine bacteria and identification of the key gene in this process. Nature Microbiology 2: 17009. DOI: https://doi.org/10.1038/nmicrobiol.2017.9. Burgers, TM, Miller, LA, Thomas, H, Else, BGT, Gosse- lin, M, Papakyriakou, T. 2017. Surface water pCO2 variations and sea–air CO2 fluxes during summer in the Eastern Canadian Arctic. Journal of Geophysical Research Oceans 122(12): 9663–9678. DOI: http:// dx.doi.org/10.1002/2017JC013250. Curson, ARJ, Todd, JD, Sullivan, MJ, Johnston, AWB. 2011. Catabolism of dimethylsulphoniopropionate: Microorganisms, enzymes and genes. Nature reviews Microbiology 9(12): 849–859. DOI: http://dx.doi. org/10.1038/nrmicro2653. Carpenter, LJ, Archer, SD, Beale, R. 2012. Ocean–atmo- sphere trace gas exchange. Chemical Society Reviews 41: 6473–6505. DOI: http://dx.doi.org/10.1039/ c2cs35121h. Carslaw, KS, Lee, LA, Reddington, CL, Pringle, KJ, Rap, A, Forster, PM, Mann, GW, Spracklen, DV, Wood- house, MT, Regayre, LA, Pierce, JR. 2013. Large contribution of natural aerosols to uncertainty in indirect forcing. Nature 503(7474): 67–71. DOI: http://dx.doi.org/10.1038/nature12674. Dall’Osto, M, Beddows, DCS, Tunved, P, Krejci, R, Stro¨m, J, Hansson, H-C, Yoon, YJ, Park, K-T, Beca- gli, S, Udisti, R, Onasch, T. 2017. Arctic sea ice melt leads to atmospheric new particle formation. Scien- tific Reports 7(1): 3318. DOI: http://dx.doi.org/10. 1038/s41598-017-03328-1. Caruana, AMN, Malin, G. 2014. The variability in DMSP content and DMSP lyase activity in marine dinofla- gellates. Progress in Oceanography 120: 410–424. DOI: http://dx.doi.org/10.1016/j.pocean.2013.10. 014. Dawson, ML, Varner, ME, Perraud, V, Ezell, MJ, Gerber, RB, Finlayson-Pitts, BJ. 2012. Simplified mecha- nism for new particle formation from methanesul- fonic acid, amines, and water via experiments and ab initio calculations. Proceedings of the National Academy of Sciences 109(46): 18719–18724. DOI: http://dx.doi.org/10.1073/pnas.1211878109. Castro, SL, Emery, WJ, Wick, GA, Tandy, W. 2017. Sub- mesoscale sea surface temperature variability from UAV and satellite measurements. Remote Sensing 9(11): 1089. DOI: http://dx.doi.org/10.3390/ rs9111089. del Valle, D, Kieber, DJ, Kiene, RP. 2007. Depth- dependent fate of biologically-consumed dimethyl- sulfide in the Sargasso Sea. Marine Chemistry 103(1–2): 197–208. DOI: http://dx.doi.org/10. 1016/j.marchem.2006.07.005. Cetinic´, I, Perry, MJ, Briggs, NT, Kallin, E, D’Asaro, EA, Lee, CM. 2012. Particulate organic carbon and inherent optical properties during 2008 North Atlantic bloom experiment. Journal of Geophysical Research Oceans 117(6). DOI: http://dx.doi.org/10. 1029/2011JC007771. del Valle, D, Kieber, DJ, Toole, DA, Brinkley, J, Kiene, RP. 2009. Biological consumption of dimethylsul- fide (DMS) and its importance in DMS dynamics in the Ross Sea, Antarctica. Author contributions Limnology and Oceanogra- phy 54(3): 785–798. Charlson, RJ, Lovelock, JE, Andreae, MO, Warren, SG. 1987. Oceanic phytoplankton, atmospheric sulphur, cloud albedo and climate. Nature 326: 655–661. DOI: http://dx.doi.org/10.1038/326655a0. Eilertsen, HC, Taasen, JP, Wesiawski, JM. 1989. Phyto- plankton studies in the fjords of West Spitzbergen: Physical environment and production in spring and summer. Journal of Plankton Research 11(6): 1245– 1260. DOI: http://dx.doi.org/10.1093/plankt/11.6. 1245. Collins, DB, Burkart, J, Chang, RY-W, Lizotte, M, Boivin-Rioux, A, Blais, M, Mungall, EL, Boyer, M, Irish, VE, Masse´, G, Kunkel, D. 2017. Frequent Art. 9(1) page 25 of 31 Art. 9(1) page 25 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Engel, A, Piontek, J, Metfies, K, Endres, S, Sprong, P, Peeken, I, Ga¨bler-Schwarz, S, No¨thig, EM. 2017. Inter-annual variability of transparent exopolymer particles in the Arctic Ocean reveals high sensitiv- ity to ecosystem changes. Scientific Reports 7(1): 1–9. DOI: http://dx.doi.org/10.1038/s41598-017- 04106-9. Tremblay, J-E´, Lizotte, M. 2016. Contrasted sensi- tivity of DMSP production to high light exposure in two Arctic under-ice blooms. Journal of Experimental Marine Biology and Ecology 475: 38–48. DOI: http://dx.doi.org/10.1016/j.jembe.2015.11.009. Galindo, V, Levasseur, M, Mundy, CJ, Gosselin, M, Tremblay, J-E´, Scarratt, MG, Gratton, Y, Papakir- iakou, T, Poulin, M, Lizotte, M. 2014. Biological and physical processes influencing sea ice, under- ice algae, and dimethylsulfoniopropionate during spring in the Canadian archipelago. Journal of Geo- physical Research Oceans 119: 3746–3766. DOI: http://dx.doi.org/10.1002/2013JC009497. Ferna´ndez-Go´mez, B, Richter, M, Schu¨ler, M, Pinhassi, J, Acinas, SG, Gonza´lez, JM, Pedro´s-Alio´, C. 2013. Ecology of marine Bacteroidetes: A comparative genomics approach. ISME Journal 7(5): 1026– 1037. DOI: http://dx.doi.org/10.1038/ismej.2012. 169. Ferna´ndez-Me´ndez, M, Olsen, LM, Kauko, HM, Meyer, A, Ro¨sel, A, Merkouriadi, I, Mundy, CJ, Ehn, JK, Johansson, AM, Wagner, PM, Ervik, Å. 2018. Algal hot spots in a changing Arctic Ocean: Sea-ice ridges and the snow-ice interface. Frontiers in Marine Sci- ence 5: 75. DOI: http://dx.doi.org/10.3389/fmars. 2018.00075. Galindo, V, Levasseur, M, Scarratt, M, Mundy, C, Gos- selin, M, Kiene, R, Gourdal, M, Lizotte, M. 2015. Under-ice microbial dimethylsulfoniopropionate metabolism during the melt period in the Canadian Arctic Archipelago. Marine Ecology Progress Series 524: 39–53. DOI: http://dx.doi.org/10.3354/ meps11144. Gabric, A, Matrai, P, Jones, GM, Middleton, J. 2018. The nexus between sea ice and polar emissions of marine biogenic aerosols. Bulletin of the American Meteorological Society 99(1): 61–82. DOI: http://dx. doi.org/10.1175/BAMS-D-16-. Gasol, JM, del Giorgio, PA. 2000. Author contributions Using flow cytometry for counting natural planktonic bacteria and under- standing the structure of planktonic bacterial com- munities. Scientia Marina 64(2): 197–224. DOI: http://dx.doi.org/10.3989/scimar.2000.64n2197. Ghahremaninezhad, R, Gong, W, Galı´, M, Norman, A- L, Beagley, SR, Akingunola, A, Zheng, Q, Lupu, A, Lizotte, M, Levasseur, M, Leaitch, WR. 2019. Dimethyl sulfide and its role in aerosol formation and growth in the Arctic summer - a modelling study. Atmospheric Chemistry and Physics 19(23): 14455–14476. DOI: http://dx.doi.org/10.5194/ acp-19-14455-2019. Galı´, M, Devred, E, Babin, M, Levasseur, M. 2019. Decadal increase in Arctic dimethylsulfide emission. Proceedings of the National Academy of Sciences 116(39): 19311–19317. DOI: http://dx.doi.org/10. 1073/pnas.1904378116. Galı´, M, Kieber, DJ, Romera-Castillo, C, Kinsey, JD, Devred, E, Pe´rez, GL, Westby, GR, Marrase´, C, Babin, M, Levasseur, M, Duarte, CM. 2016. CDOM sources and photobleaching control quantum yields for oceanic DMS photolysis. Environmental Science & Technology 50: 13361–13370. DOI: http://dx.doi. org/10.1021/acs.est.6b04278. Gourdal, M, Lizotte, M, Masse´, G, Gosselin, M, Scar- ratt, M, Levasseur, M. 2018. Dimethylsulfide dynamics in first-year sea ice melt ponds in the Canadian Arctic Archipelago. Biogeosciences 15: 3169–3188. DOI: http://dx.doi.org/10.5194/bg- 2017-432. Galı´, M, Levasseur, M, Devred, E, Simo´, R, Babin, M. 2018. Sea-surface dimethylsulfide (DMS) concentra- tion from satellite data at global and regional scales. Biogeosciences 15: 3497–3519. DOI: http://dx.doi. org/10.5194/bg-15-3497-2018. Grondin, PL. 2019. Microalgae species succession and its drivers during under-ice spring blooms in the Arctic Ocean [MSc thesis]. Laval University. Galı´, M, Ruiz-Gonza´lez, C, Lefort, T, Gasol, JM, Carde- lu´s, C, Romera-Castillo, C, Simo´, R. 2013. Spectral irradiance dependence of sunlight effects on plank- ton dimethylsulfide production. Limnology and Oceanography 58(2): 489–504. DOI: http://dx.doi. org/10.4319/lo.2013.58.2.0489. Hansen, E, Eilertsen, HC. 2007. Do the polyunsaturated aldehydes produced by Phaeocystis pouchetii (Hariot) Lagerheim influence diatom growth during the spring bloom in Northern Norway? Journal of Plank- ton Research 29(1): 87–96. DOI: http://dx.doi.org/ 10.1093/plankt/fbl065. Galı´, M, Simo´, R. 2010. Occurrence and cycling of di- methylated sulfur compounds in the Arctic during summer receding of the ice edge. Marine Chemistry 122: 105–117. DOI: http://dx.doi.org/10.1016/j. marchem.2010.07.003. Hayashida, H, Carnat G, Galı´ M, Monahan AH, Mor- tenson E, Sou T, Steiner NS. 2020. Spatio-temporal variability in modelled bottom-ice and sea-surface dimethylsulfide concentrations and fluxes in the Arctic during 1979–2015. Global Biogeochemical Cy- cles 34(10): e2019GB006456. DOI: http://dx.doi. org/10.1029/2019GB006456. Galı´, M, Simo´, R. 2015. A meta-analysis of oceanic DMS and DMSP cycling processes: Disentangling the sum- mer paradox. Global Biogeochemical Cycles 29: 496– 515. DOI: http://dx.doi.org/10.1002/2014GB 004940. Author contributions Heintzenberg, J, Leck C, Tunved P. 2015. Potential source regions and processes of aerosol in the sum- mer Arctic. Atmospheric Chemistry and Physics 15(11): 6487–6502. DOI: http://dx.doi.org/10. 5194/acp-15-6487-2015. Galindo,V, Levasseur, M, Mundy, CJ, Gosselin, M, Scar- ratt, M, Papakyriakou, T, Stefels, J, Gale, MA, Art. 9(1) page 26 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ photosynthetic activity of the haptophyte phaeocys- tis antarctica. Journal of Phycology 56(3): 761–774. Heintzenberg, J, Tunved P, Galı´ M, Leck C. 2017. New particle formation in the Svalbard region 2006– 2015. Atmospheric Chemistry and Physics 17: 6153–6175. DOI: http://dx.doi.org/10.5194/acp- 2016-1073. Keller, MD. 1989. Dimethyl sulfide production in marine phytoplankton, in Saltzman, ES, Cooper, WJ eds., Biogenic sulfur in the environment. New York, NY (ACS Symposium series): 167–182. Available at https://pubs.acs.org/doi/abs/10.1021/bk-1989- 0393.ch011. Hodshire, AL, Campuzano-Jost, P, Kodros, JK, Croft, B, Nault, BA, Schroder, JC, Jimenez, JL, Pierce, JR. 2019. The potential role of methanesulfonic acid (MSA) in aerosol formation and growth and the asso- ciated radiative forcings. Atmospheric Chemistry and Physics 19(5): 3137–3160. DOI: http://dx.doi.org/ 10.5194/acp-19-3137-2019. Kieber, DJ, Jiao, J, Kiene, RP, Bates, TS. 1996. Impact of dimethylsulfide photochemistry on methyl sulfur cycling in the equatorial Pacific Ocean. Journal of Geophysical Research 101(C2): 3715–3722. DOI: http://dx.doi.org/10.1029/95JC03624. Hopkins, FE, Nightingale PD, Stephens JA, Moore CM, Richier S, Cripps GL, Archer SD. 2020. A meta- analysis of microcosm experiments shows that dimethyl sulfide (DMS) production in polar waters is insensitive to ocean acidification. Biogeosciences 17(1): 163–186. DOI: http://dx.doi.org/10.5194/ bg-17-163-2020. Kiene, RP, Bates, TS. 1990. Biological removal of dimethyl sulphide from sea water. Nature 345: 702–705. DOI: http://dx.doi.org/10.1038/345702a0. Kiene, RP, Linn, L, Bruton, J. 2000. New and important roles for DMSP in marine microbial communities. Journal of Sea Research 43: 209–224. DOI: http:// dx.doi.org/10.1016/S1385-1101(00)00023-X. Hsu, SA, Meindl, EA, Gilhousen, DB. 1994. Determining the power-law wind-profile exponent under near- neutral stability conditions at sea. Journal of Applied Meteorology 33(6): 757–765. DOI: http://dx.doi. org/10.16309/j.cnki.issn.1007-1776.2003.03.004. ´ Kiene, RP, Linn, LJ. 2000. The fate of dissolved dimethyl- sulfoniopropionate (DMSP) in seawater: tracer stud- ies using 35S-DMSP. Geochimica et Cosmochimica Acta 64(16): 2797–2810. DOI: http://dx.doi.org/ 10.1016/S0016-7037(00)00399-9. Hussherr, R, Levasseur, M, Lizotte, M, Tremblay, J-E´, Mol, J, Thomas, H, Gosselin, M, Starr, M, Miller, LA, Jarnikova´, T, Schuback, N, Mucci, A. 2017. Impact of ocean acidification on Arctic phytoplank- ton blooms and dimethyl sulfide concentration under simulated ice-free and under-ice conditions. Biogeosciences 14(9): 2407–2427. Kiene, RP, Slezak, D. 2006. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Li, C-Y, Wang, X-J, Chen, X-L, Sheng, Q, Zhang, S,Wang, P, Quareshy, M, Rihtman, B, Shao, X, Gao, C, Li, F, Li, S. 2020. Novel enzyme for dimethyl sulfide- releasing in bacteria reveals a missing route in the marine sulfur cycle. bioRxiv preprint. DOI: http://dx. doi.org/10.1101/2020.10.29.360743. Land, PE, Shutler, JD, Bell, TG,Yang, M. 2014. Exploiting satellite earth observation to quantify current global oceanic DMS flux and its future climate sensitivity. Journal of Geophysical Research Oceans 119(11): 7725–7740. DOI: http://dx.doi.org/10.1002/ 2014JC010104. Lidbury, I, Kro¨ber, E, Zhang, Z, Zhu, Y, Murrell, JC, Chen, Y, Scha¨fer, H. 2016. A mechanism for bacte- rial transformations of DMS to DMSO: A missing link in the marine organic sulfur cycle. Environmental Microbiology 18(8): 2754–2766. DOI: http://dx. doi.org/10.1111/1462-2920.13354. Landa, M, Burns, AS, Durham, BP, Esson, K, Nowinski, B, Sharma, S, Vorobev, A, Nielsen T, Kiene, RP, Moran, MA. 2019. Sulfur metabolites that facilitate oceanic phytoplankton–bacteria carbon flux. ISME Journal 13: 2536–2550. DOI: http://dx.doi.org/10. 1038/s41396-019-0455-3,2019. Lizotte, M, Levasseur, M, Galindo, V, Gourdal, M, Gos- selin, M, Tremblay, JE, Blais, M, Charette, J, Hus- sherr, R. 2020. Phytoplankton and dimethylsulfide dynamics at two contrasting Arctic ice edges. Bio- geosciences 17: 1557–1581. DOI: http://dx.doi. org/10.5194/bg-17-1557-2020. Lasternas, S, Agustı´, S. 2010. Phytoplankton commu- nity structure during the record Arctic ice-melting of summer 2007. Polar Biology 33(12): 1709– 1717. DOI: http://dx.doi.org/10.1007/s00300- 010-0877-x. Lizotte, M, Levasseur, M, Scarratt, M, Michaud, S, Mer- zouk, A, Gosselin, M, Pommier, J. 2008. Fate of dimethylsulfoniopropionate (DMSP) during the decline of the northwest Atlantic Ocean spring dia- tom bloom. Aquatic Microbial Ecology 52(August): 159–173. DOI: http://dx.doi.org/10.3354/ ame01232. Leaitch, WR, Sharma, S, Huang, L, Toom-Sauntry, D, Chivulescu, A, Macdonald, AM, von Salzen, K, Pierce, JR, Bertram, AK, Schroder, JC, Shantz, NC, Chang, RY-W, Norman, A-L. 2013. Dimethyl sulfide control of the clean summertime Arctic aero- sol and cloud. Elementa: Science of the Anthropocene 1: 000017. DOI: http://dx.doi.org/10.12952/ journal.elementa.000017. Loisel, H, Morel, A. 1998. Light scattering and chloro- phyll concentration in case 1 waters: A reexamina- tion. Limnology and Oceanography 43(5), 847–858. DOI: http://dx.doi.org/10.4319/lo.1998.43.5.0847. Long, JD, Smalley, GW, Barsby, T, Anderson, JT, Hay, ME. 2007. Chemical cues induce consumer-specific defenses in a bloom-forming marine phytoplankton. Proceedings of the National Academy of Sciences USA 104(25): 10512–7. DOI: http://dx.doi.org/10.1073/ pnas.0611600104. Le Clainche, Y, Levasseur, M, Ve´zina, A, Bouillon, R-C, Merzouk, A, Michaud, S, Scarratt, M, Wong, C, Rivkin, RB, Boyd, PW, Harrison, PJ. 2006. Author contributions Low dissolved DMSP concen- trations in seawater revealed by small volume grav- ity filtration and dialysis sampling. Limnology and Oceanography: Methods 4: 80–95. Kinsey, JD, Kieber, DJ. 2016. Microwave preservation method for DMSP, DMSO, and acrylate in unfiltered seawater and phytoplankton culture samples. Lim- nology and Oceanography Methods 14(3): 196–209. DOI: http://dx.doi.org/10.1002/lom3.10081. Inness, A, Chabrillat, S, Flemming, J, Huijnen, V, Lan- genrock, B, Nicolas J, Polichtchouk, I, Razinger, M. 2020. Exceptionally low Arctic stratospheric ozone in spring 2020 as seen in the CAMS reanaly- sis. Journal of Geophysical Research-Atmospheres 125(23): e2020JD033563. DOI: http://dx.doi.org/ 10.1029/2020JD033563. Kinsey, JD, Kieber, DJ, Neale, PJ. 2016. Effects of iron limitation and UV radiation on Phaeocystis antarcti- ca growth and dimethylsulfoniopropionate, di- methylsulfoxide and acrylate concentrations. Environmental Chemistry 13(2): 195–211. DOI: http://dx.doi.org/10.1071/EN14275. Intergovernmental Panel on Climate Change. 2019. IPCC Special Report on the Ocean and Cryosphere in a Changing Climate [H.-O. Po¨rtner, D.C. Roberts,V. Masson-Delmotte, P. Zhai, M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegrı´a, M. Nicolai, A. Okem, J. Petzold, B. Rama, N.M. Weyer (eds.)]. Available at https://www.ipcc.ch/srocc/. Koroleff, F, Hansen HP. 1999. Determination of nutri- ents, in Grasshoff, K, Kremling, K, Ehrhardt, M eds., Methods of seawater analysis. New York, NY: Wiley: 159–228. DOI: http://dx.doi.org/10.1002/ 9783527613984.ch10. Jarnı´kova´, T, Dacey, JW, Lizotte, M, Levasseur, M, Tor- tell, PD. 2018. The distribution of methylated sulfur compounds, DMS and DMSP, in Canadian Subarctic and Arctic waters during summer 2015. Biogeos- ciences 15: 2449–2465. DOI: http://dx.doi.org/10. 5194/bg-15-2449-2018. Krawczyk, DW, Meire, L, Lopes, C, Juul-Pedersen, T, Mortensen, J, Li, CL, Krogh, T. 2018. Seasonal suc- cession, distribution, and diversity of planktonic pro- tists in relation to hydrography of the Godthåbsfjord system (SW Greenland). Polar Biology 41(10): 2033– 2052. DOI: http://dx.doi.org/10.1007/s00300-018- 2343-0. Jeffrey, SW, Wright, SW, Zapata, M. 2012. Microalgal classes and their signature pigments, in Roy, S, Lle- wellyn, CA, Egeland, ES, Johnsen, G eds., Phytoplank- ton pigments: Characterization, chemotaxonomy and applications in oceanography. Cambridge University Press: 3–77. DOI: http://dx.doi.org/10.1017/ cbo9780511732263.004. Kubiszyn, AM, Wiktor, JM, Wiktor, JM, Griffiths, C, Kristiansen, S, Gabrielsen, TM. 2017. The annual planktonic protist community structure in an ice- free high Arctic fjord (Adventfjorden, West Spitsber- gen). Journal of Marine Systems 169: 61–72. DOI: http://dx.doi.org/10.1016/j.jmarsys.2017.01.013. Kameyama, S, Otomaru, M, McMinn, A, Suzuki, K. 2020. Ice melting can change DMSP production and Lacour,T, Larivie`re, J, Babin, M. 2017. Growth, Chl a con- tent, photosynthesis, and elemental composition in Art. 9(1) page 27 of 31 Art. Author contributions 9(1) page 27 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ 49(2): 508–519. DOI: http://dx.doi.org/10.4319/lo. 2004.49.2.0508. polar and temperate microalgae. Limnology and Oceanography 62(1): 43–58. DOI: http://dx.doi. org/10.1002/lno.10369. Leu, E, Mundy, CJ, Assmy, P, Campbell, K, Gabrielsen, TM, Gosselin, M, Juul-Pedersen, T, Gradinger, R. 2015. Arctic spring awakening–Steering principles behind the phenology of vernal ice algal blooms. Progress in Oceanography 139: 151–170. DOI: http://dx.doi.org/10.1016/j.pocean.2015.07.012. Lafond, A, Leblanc, K, Que´guiner, B, Moriceau, B, Ley- naert, A, Cornet, V, Legras, J, Ras, J, Parenteau, M. 2019. Late spring bloom development of pelagic diatoms in Baffin Bay. Elementa: Science of the Anthropocene 7(44). DOI: http://dx.doi.org/10. 1525/elementa.382. Levasseur, M. 2013. Impact of Arctic meltdown on the microbial cycling of sulphur. Nature Geoscience 6(9): 691–700. DOI: http://dx.doi.org/10.1038/ ngeo1910. Laliberte´, J, Be´langer, S, Frouin, R. 2016. Evaluation of satellite-based algorithms to estimate photosynthet- ically available radiation (PAR) reaching the ocean surface at high northern latitudes. Remote Sensing of Environment 184: 199–211. DOI: http://dx.doi.org/ 10.1016/j.rse.2016.06.014. Levine, NM, Varaljay, VA, Toole, DA, Dacey, JWH, Do- ney, SC, Moran, MA. 2012. Environmental, bio- chemical and genetic drivers of DMSP degradation and DMS production in the Sargasso Sea. Environ- mental Microbiology 14(5): 1210–1223. DOI: http:// dx.doi.org/10.1111/j.1462-2920.2012.02700.x. j Lana, A, Bell, TG, Simo´, R, Vallina, SM, Ballabrera-Poy, J, Kettle, AJ, Dachs, J, Bopp, L, Saltzman, ES, Stefels, J, Johnson, JE. 2011. An updated climatol- ogy of surface dimethlysulfide concentrations and emission fluxes in the global ocean. Global Biogeo- chemical Cycles 25: GB1004. DOI: http://dx.doi.org/ 10.1029/2010GB003850. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Model- ing analysis of the effect of iron enrichment on dimethyl sulfide dynamics in the NE Pacific (SERIES experiment). Journal of Geophysical Research 111(C1): 1–15. DOI: http://dx.doi.org/10.1029/ 2005JC002947. Mahmood, R, von Salzen, K, Norman, A-L, Galı´, M, Levasseur, M. 2019. Sensitivity of Arctic sulfate aerosol and clouds to changes in future surface sea- water dimethylsulfide concentrations. Atmospheric Chemistry and Physics 19: 6419–6435. DOI: http://dx.doi.org/10.5194/acp-2018-876. Letelier, RM, Karl, DM, Abbott, MR, Bidigare, RR. 2004. Light driven seasonal patterns of chlorophyll and nitrate in the lower euphotic zone of the North Pacific Subtropical Gyre. Limnology and Oceanography Art. 9(1) page 28 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ Liggio, J, Abbatt, JPD. 2016. Dimethyl sulfide in the summertime Arctic atmosphere: measurements and source sensitivity simulations. Atmospheric Chemistry and Physics 16: 6665–6680. DOI: http://dx.doi.org/10.5194/acp-16-6665-2016. Malin, G, Wilson, WH, Bratbak, G, Liss, PS, Mann, NH. 1998. Elevated production of dimethylsulfide result- ing from viral infection of cultures of Phaeocystis pouchetii. Limnology and Oceanography 43(6): 1389–1393. DOI: http://dx.doi.org/10.4319/lo. 1998.43.6.1389. Oksanen, J, Blanchet, FG, Friendly, M, Kindt, R, Le- gendre, P, McGlinn, D, Minchin, PR, O’Hara RB, Simpson, GL, Solymos, P, Stevens, MHM, Szoecs, E, Wagner, H. 2019. vegan: Community Ecology Package. R package version 2.5-6. Available at https://CRAN.R-project.org/package¼vegan. Manucharyan, GE, Thompson, AF. 2017. Submesoscale sea ice-ocean interactions in marginal ice zones. Journal of Geophysical Research Oceans 122(12): 9455–9475. DOI: http://dx.doi.org/10.1002/ 2017JC012895. Marie, D, Brussaard, C, Partensky, F, Vaulot, D. 1999. Flow cytometric analysis of phytoplankton, bacteria and viruses, in Robinson, JP ed., Vol. 11.11. Current protocols in cytometry. New York, NY: John Wiley and Sons: 1–15. Olson, RJ, Sosik, HM. 2007. A submersible imaging-in- flow instrument to analyze nano- and microplank- ton: Imaging FlowCytobot. Limnology and Oceanog- raphy Meth 5: 195–203. DOI: http://dx.doi.org/10. 4319/lom.2007.5.195. Orellana, MV, Matrai, PA, Janer, M, Rauschenberg, CD. 2011. Dimethylsulfoniopropionate storage in Phaeo- cystis (Prymnesiophyceae) secretory vesicles. Journal of Phycology 47(1): 112–117. DOI: http://dx.doi. org/10.1111/j.1529-8817.2010.00936.x. Matrai, P, Vernet, M, Wassmann, P. 2007. Relating tem- poral and spatial patterns of DMSP in the Barents Sea to phytoplankton biomass and productivity. Journal of Marine Systems 67(1-2): 83–101. DOI: http://dx.doi.org/1016/j.jmarsys.2006.10.001. Matrai, PA, Vernet, M. 1997. Dynamics of the vernal bloom in the marginal ice zone of the Barents Sea: Dimethyl sulfide and dimethylsulfoniopropionate budgets. Journal of Geophysical Research: Oceans 102(C10): 22965–22979. Orkney, A, Platt, T, Narayanaswamy, BE, Kostakis, I, Bouman, HA. 2020. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Bio-optical evidence for increasing Phaeocystis dominance in the Barents Sea. Philosophical Transactions of the Royal Society A 378(2181): 20190357. DOI: http://dx.doi.org/10. 1098/rsta.2019.0357. Matrai, PA, Vernet, M, Hood, R, Jennings, A, Brody, E, Saemundsdo´ttir, S. 1995. Light-dependence of car- bon and sulfur production by polar clones of the genus Phaeocystis. Marine Biology 124(1): 157– 167. DOI: http://dx.doi.org/10.1007/BF00349157. Park, K, Kim, I, Choi, JO, Lee,Y, Jung, J, Ha, SY, Kim, JH, Zhang, M. 2019. Unexpectedly high dimethyl sul- fide concentration in high-latitude Arctic sea ice melt ponds. Environmental Science: Processes & Im- pacts 21(10): 1642–1649. DOI: http://dx.doi.org/ 10.1039/c9em00195f. Mauritsen, T, Sedlar, J, Tjernstro¨m, M, Leck, C, Martin, M, Shupe, M, Sjogren, S, Sierau, B, Persson, POG, Brooks, IM, Swietlicki, E. 2011. An Arctic CCN- limited cloud-aerosol regime. Atmospheric Chemistry and Physics 11: 165–173. DOI: http://dx.doi.org/10. 5194/acp-11-165-2011. Perrette, M, Yool, A, Quartly, GD, Popova, EE. 2011. Near-ubiquity of ice-edge blooms in the Arctic. Bio- geosciences 8(2): 515–524. DOI: http://dx.doi.org/ 10.5194/bg-8-515-2011. p McParland, EL, Levine, NM. 2019. The role of differential DMSP production and community composition in predicting variability of global surface DMSP con- centrations. Limnology and Oceanography 64(2): 757–773. DOI: http://dx.doi.org/10.1002/lno. 11076. Quinn, PK, Bates, TS. 2011. The case against climate reg- ulation via oceanic phytoplankton sulphur emis- sions. Nature 480: 51–56. DOI: http://dx.doi.org/ 10.1038/nature10580. Randelhoff, A, Fer, I, Sundfjord, A. 2017. Turbulent upper-ocean mixing affected by meltwater layers during Arctic summer. Journal of Physical Oceanog- raphy 47(4): 835–853. DOI: http://dx.doi.org/10. 1175/JPO-D-16-0200.1. McPhee, M. 2008. Air-ice-ocean interaction: Turbulent ocean boundary layer exchange processes. New York, NY: Springer. DOI: http://dx.doi.org/10.1007/978- 0-387-78335-2. Randelhoff, A, Holding, J, Janout, M, Sejr, MK, Babin, M, Tremblay, JE´, Alkire, MB. 2020. Pan-Arctic ocean primary production constrained by turbulent nitrate fluxes. Frontiers in Marine Science 7: 150. DOI: http://dx.doi.org/10.3389/fmars.2020.00150. Menden-Deuer, S, Lessard, EJ. 2000. Carbon to volume relationships for dinoflagellates, diatoms, and other protist plankton. Limnology and Oceanography 45(3): 569–579. DOI: http://dx.doi.org/10.4319/ lo.2000.45.3.0569. Randelhoff, A, Oziel, L, Massicotte, P, Be´cu, G, Galı´, M, Lacour, L, Dumont, D, Vladoiu, A, Marec, C, Bruyant, F, Houssais, MN. 2019. The evolution of light and vertical mixing across a phytoplankton ice- edge bloom. Elementa: Science of the Anthropocene 7(1). DOI: http://dx.doi.org/10.1525/elementa.357. Metfies, K, Von Appen, WJ, Kilias, E, Nicolaus, A, No¨thig, EM. 2016. Biogeography and photosyn- thetic biomass of Arctic marine pico-eukaroytes dur- ing summer of the record sea ice minimum 2012. PLoS One 11(2): 1–20. DOI: http://dx.doi.org/10. 1371/journal.pone.0148512. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Ras, J, Claustre, H, Uitz, J. 2008. Spatial variability of phytoplankton pigment distributions in the Sub- tropical South Pacific Ocean: Comparison between Mungall, EL, Croft, B, Lizotte, M, Thomas, JL, Murphy, JG, Levasseur, M, Martin, RV, Wentzell, JJB, Art. 9(1) page 29 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ in situ and predicted data. Biogeosciences 5(2): 353– 369. DOI: http://dx.doi.org/10.5194/bg-5-353- 2008. Schanke, NL, Bolinesi, F, Mangoni, O, Katlein, C, An- haus, P, Hoppmann, M, Lee, PA, DiTullio, GR. 2020. Biogeochemical and ecological variability dur- ing the late summer–early autumn transition at an ice-floe drift station in the Central Arctic Ocean. Limnology and Oceanography 66:.S363–S382. DOI: http://dx.doi.org/10.1002/lno.11676. R Core Team. 2017. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Available at https://www. R-project.org/. Schoemann, V, Becquevort, S, Stefels, J, Rousseau, V, Lancelot, C. 2005. Phaeocystis blooms in the global ocean and their controlling mechanisms: A review. Journal of Sea Research 53(1–2): 43–66. DOI: http://dx.doi.org/10.1016/j.seares.2004.01.008. Reigstad, M, Wassmann, P. 2007. Does Phaeocystis spp. contribute significantly to vertical export of organic carbon? Biogeochemistry 83(1–3): 217–234. DOI: http://dx.doi.org/10.1007/s10533-007-9093-3. Rellinger, AN, Kiene, RP, del Valle, DA, Kieber, DJ, Slezak, D, Harada, H, Bisgrove, J, Brinkley, J. 2009. Occurrence and turnover of DMSP and DMS in deep waters of the Ross Sea, Antarctica. Deep Sea Research Part I 56(5): 686–702. DOI: http://dx.doi. org/10.1016/j.dsr.2008.12.010. Selz, V, Lowry, KE, Lewis, KM, Joy-Warren, HL, Van De Poll, W, Nirmel, S, Tong, A, Arrigo, KR. 2018. Dis- tribution of Phaeocystis antarctica-dominated sea ice algal communities and their potential to seed phy- toplankton across the western Antarctic Peninsula in spring. Marine Ecology Progress Series 586: 91–112. DOI: http://dx.doi.org/10.3354/meps12367. Renaut, S, Devred, E, Babin, M. 2018. Northward expan- sion and intensification of phytoplankton growth during the early ice-free season in Arctic. Geophysi- cal Research Letters 45: 10,590–10,598. DOI: http:// dx.doi.org/10.1029/2018GL078995. Serreze, MC, Barry, RG. 2011. Processes and impacts of Arctic amplification: A research synthesis. Global and Planetary Change 77: 85–96. DOI: http://dx. doi.org/10.1016/j.gloplacha.2011.03.004. Ricchiazzi, P, Yang, S, Gautier, C, Sowle, D. 1998. SBDART: A research and teaching software tool for plane-parallel radiative transfer in the Earth’s atmo- sphere. Bulletin of the American Meteorological Soci- ety 79(10): 2101–2114. DOI: http://dx.doi.org/10. 1175/1520-0477(1998)079<2101:SARATS>2.0. CO;2. Sheehan, CE, Petrou, K. 2019. Dimethylated sulfur pro- duction in batch cultures of Southern Ocean phyto- plankton. Biogeochemistry 147(1): 53–69. DOI: http://dx.doi.org/10.1007/s10533-019-00628-8. Sherr, EB, Sherr, BF, Wheeler, PA, Thompson, K. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 9(1) page 30 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ by reduced sulphur fluxes as a result of ocean acid- ification. Nature Climate Change 3(11): 975–978. DOI: http://dx.doi.org/10.1038/nclimate1981. Substrate-controlled succession of marine bacterio- plankton populations induced by a phytoplankton bloom. Science 336(6081): 608–611. DOI: http:// dx.doi.org/10.1126/science.1218344. Slezak, D, Kiene, RP, Toole, DA, Simo´, R, Kieber, DJ. 2007. Effects of solar radiation on the fate of dis- solved DMSP and conversion to DMS in seawater. Aquatic Sciences 69(3): 377–393. DOI: http://dx. doi.org/10.1007/s00027-007-0896-z. Thackeray, CW, Hall, A. 2019. An emergent constraint on future Arctic sea-ice albedo feedback. Nature Cli- mate Change 9: 972–978. DOI: http://dx.doi.org/ 10.1038/s41558-019-0619-1. Smith, DC, Azam, F. 1992. A simple, economical method for measuring bacterial protein synthesis rates in seawater using 3H-leucine. Marine Microbial Food- webs 6(2): 107–114. Theseira, AM, Nielsen, DA, Petrou, K. 2020. Uptake of dimethylsulphoniopropionate (DMSP) reduces free reactive oxygen species (ROS) during late exponen- tial growth in the diatom Thalassiosira weissflogii grown under three salinities. Marine Biology 167(9): 1–7. DOI: http://dx.doi.org/10.1007/ s00227-020-03744-4. Spreen, G, Kaleschke, L, Heygster, G. 2008. Sea ice remote sensing using AMSR-E 89-GHz channels. Journal of Geophysical Research: Oceans 113(C2). DOI: http://dx.doi.org/10.1029/2005JC003384. Toole, DA, Siegel, DA. 2004. Light-driven cycling of di- methylsulfide (DMS) in the Sargasso Sea: Closing the loop. Geophysical Research Letters 31(9): 5–8. DOI: http://dx.doi.org/10.1029/2004GL019581. Stefels, J, Steinke, M, Turner, SM, Malin, G, Belviso, S. 2007. Environmental constraints on the production and removal of the climatically active gas dimethyl- sulphide (DMS) and implications for ecosystem modelling. Biogeochemistry 83: 245–275. DOI: http://dx.doi.org/10.1007/978-1-4020-6214-8_18. Toole, DA, Slezak, D, Kiene, RP, Kieber, DJ, Siegel, DA. 2006. Effects of solar radiation on dimethylsulfide cycling in the western Atlantic Ocean. Deep Sea Res Part I: Oceanographic Research Papers 53(1): 136– 153. DOI: http://dx.doi.org/10.1016/j.dsr.2005.09. 003. Steiner, PA, Sintes, E, Simo´, R, De Corte, D, Pfannku- chen, DM, Ivancˇic´, I, Najdek, M, Herndl, GJ. 2019. Seasonal dynamics of marine snow-associated and free-living demethylating bacterial communities in the coastal northern Adriatic Sea. Environmental Microbiology Reports 11(5): 699–707. DOI: http:// dx.doi.org/10.1111/1758-2229.12783. Trudnowska, E, Lacour, L, Ardyna, M, Rogge, A, Iris- son, JO, Waite, AM, Babin, M, Stemmann, L. 2021. Marine snow morphology illuminates the evo- lution of phytoplankton blooms and determines their subsequent vertical export. Nature Communi- cations 12: 2816. DOI: http://dx.doi.org/10.1038/ s41467-021-22994-4. Stoecker, DK, Hansen, PJ, Caron, DA, Mitra, A. 2017. Mixotrophy in the marine plankton. Annual Review of Marine Science 9: 311–335. DOI: http://dx.doi. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 2003. Temporal and spatial variation in stocks of autotro- phic and heterotrophic microbes in the upper water column of the central Arctic Ocean. Deep-Sea Research Part I: Oceanographic Research 50: 557– 571. DOI: http://dx.doi.org/10.1016/S0967- 0637(03)00031-1. Rowan, KS. 1989. The carotenoids, in Photosynthetic pig- ments of algae. New York, NY: Cambridge University Press: 112–165. Available at https://www. cambridge.org/es/academic/subjects/life-sciences/ plant-science/photosynthetic-pigments-algae? format¼PB. ( ) Simo´, R. 2001. Production of atmospheric sulfur by oce- anic plankton: Biogeochemical, ecological and evo- lutionary links. Trends in Ecology & Evolution 16(6): 287–294. Ruiz-Gonza´lez, C, Galı´, M, Sintes, E, Herndl, GJ, Gasol, JM, Simo´, R. 2012. Sunlight effects on the osmo- trophic uptake of DMSP-sulfur and leucine by polar phytoplankton. PLoS One 7(9): e45545. DOI: http:// dx.doi.org/10.1371/journal.pone.0045545. Simo´, R. 2011. The role of marine microbiota in short- term climate regulation, in Duarte, CM ed., The role of marine biota in the functioning of the biosphere. Bilbao, Spain: Fundacio´n BBVA: 107–130. g j p Saint-Be´at, B, Fath, BD, Aubry, C, Colombet, J, Dinas- quet, J, Fortier, L, Galindo, V, Grondin, PL, Joux, F, Lalande, C, LeBlanc, M. 2020. Contrasting pelagic ecosystem functioning in eastern and west- ern Baffin Bay revealed by trophic network model- ing. Elementa: Science of the Anthropocene 8(1): 1– 24. DOI: http://dx.doi.org/10.1525/elementa.397 Simo´, R, Salo´,V, Almeda, R, Movilla, J, Trepat, I, Saiz, E, Calbet, A. 2018. The quantitative role of microzoo- plankton grazing in dimethylsulfide (DMS) produc- tion in the NW Mediterranean. Biogeochemistry 141: 125–142. DOI: http://dx.doi.org/10.1007/ s10533-018-0506-2. Sanderson, MP, Bronk, DA, Nejstgaard, JC, Verity, PG, Sazhin, AF, Frischer, ME. 2008. Phytoplankton and bacterial uptake of inorganic and organic nitrogen during an induced bloom of Phaeocystis pouchetii. Aquatic Microbial Ecology 51(2): 153–168. DOI: http://dx.doi.org/10.3354/ame01178. Simo-Matchim, A-G, Gosselin, M, Poulin, M, Ardyna, M, Lessard, S. 2017. Summer and fall distribution of phytoplankton in relation to environmental vari- ables in Labrador fjords, northeastern Canada, with special emphasis on Phaeocystis pouchetii. Marine Ecology Progress Series 572: 19–42. DOI: http:// dx.doi.org/10.3354/meps12125. Scarratt, M, Cantin, G, Levasseur, M, Michaud, S. 2000. Particle size-fractionated kinetics of DMS production: Where does DMSP cleavage occur at the microscale. Journal of Sea Research 43: 245–252. DOI: http://dx.doi.org/10.1016/S1385- 1101(00)00019-8. Simon, M, Azam, F. 1989. Protein content and protein synthesis rates of planktonic marine bacteria. Marine Ecology Progress Series 51: 201–213. Six, KD, Kloster, S, Ilyina, T, Archer, SD, Zhang, K, Maier-Reimer, E. 2013. Global warming amplified Art. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 org/10.1146/annurev-marine-010816-060617. Uhlig, C, Damm, E, Peeken, I, Krumpen T, Rabe, B, Korhonen, M, Ludwichowski, K-U. 2019. Sea ice and water mass influence dimethylsulfide concen- trations in the central Arctic Ocean. Frontiers of Earth Science 7: 1–15. DOI: http://dx.doi.org/10. 3389/feart.2019.00179. Strong, C. 2012. Atmospheric influence on Arctic mar- ginal ice zone position and width in the Atlantic sector, February–April 1979–2010. Climate Dynam- ics 39(12): 3091–3102. DOI: http://dx.doi.org/10. 1007/s00382-012-1356-6. Strong, C, Rigor, IG. 2013. Arctic marginal ice zone trend- ing wider in summer and narrower in winter. Geo- physical Research Letters 40(18): 4864–4868. DOI: http://dx.doi.org/10.1002/grl.50928. Vance, T, Davidson, A, Thomson, P, Levasseur, M, Li- zotte, M, Curran, M, Jones, G. 2013. Rapid DMSP production by an Antarctic phytoplankton commu- nity exposed to natural surface irradiances in late spring. Aquatic Microbial Ecology 71(2): 117–129. DOI: http://dx.doi.org/10.3354/ame01670. Sunda, W, Kieber, DJ, Kiene, RP, Huntsman, S. 2002. An antioxidant function for DMSP and DMS in marine algae. Nature 418(6895): 317–320. DOI: http://dx.doi.org/10.1038/nature00851. Varaljay, VA, Robidart, J, Preston, CM, Gifford, SM, Durham, BP, Burns, AS, Ryan, JP, Marin, III R, Kiene, RP, Zehr, JP, Scholin, CA. 2015 Feb 20. Single-taxon field measurements of bacterial gene regulation controlling DMSP fate. ISME Journal 9: 1677–1686. DOI: http://dx.doi.org/10.1038/ismej. 2015.23. Sunda,WG, Hardison, R, Kiene, RP, Bucciarelli, E, Har- ada, H. 2007. The effect of nitrogen limitation on cellular DMSP and DMS release in marine phyto- plankton: Climate feedback implications. Aquatic Sciences 69(3): 341–351. DOI: http://dx.doi.org/ 10.1007/s00027-007-0887-0. Veres, PR, Andrew Neuman, J, Bertram, TH, Assaf, E, Wolfe, GM, Williamson, CJ, Weinzierl, B, Tilmes, S, Thompson, CR, Thames, AB, Schroder, JC, Saiz- Lopez, A, Rollins, AW, Roberts, JM, Price, D, Peischl, J, Nault, BA, Møller, KH, Miller, DO, Mei- nardi, S, Li, Q, Lamarque, J-F, Kupc, A, Kjaer- gaard, HG, Kinnison, D, Jimenez, JL, Jernigan, CM, Hornbrook, RS, Hills, A, Dollner, M, Day, DA, Cuevas, CA, Campuzano-Jost, P, Burkholder, Tang, CCL, Ross, CK, Yao, T, Petrie, B, DeTracey, BM, Dunlap, E. 2004. The circulation, water masses and sea-ice of Baffin Bay. Progress in Oceanography 63(4): 183–228. DOI: http://dx.doi.org/10.1016/j. pocean.2004.09.005. Teeling, H, Fuchs, BM, Becher, D, Klockow, C, Gardeb- recht, A, Bennke, CM, Kassabgy, M, Huang, S, Mann, AJ, Waldmann, J, Weber, M. 2012. Art. 9(1) page 31 of 31 Art. 9(1) page 31 of 31 Galı´ et al: DMS Emissions from the Arctic MIZ 823–840. DOI: http://dx.doi.org/10.1016/j.hal. 2004.12.009. aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 Deming, University of Washington, Seattle, WA, USA Domain Editor-in-Chief: Jody W. Deming, University of Washington, Seattle, WA, USA Associate Editor: Jeff S. Bowman, Scripps Institution of Oceanography, UC San Diego, CA, USA Knowledge Domain: Ocean Science Knowledge Domain: Ocean Science Part of an Elementa Special Feature: Green Edge aded from http://online.ucpress.edu/elementa/article-pdf/9/1/00113/474403/elementa.2020.00113.pdf by guest on 08 April 2022 J, Bui, TP, Brune, WH, Brown, SS, Brock, CA, Bourgeois, I, Blake, DR, Apel, EC, Ryerson, TB. 2020. Global airborne sampling reveals a previously unobserved dimethyl sulfide oxidation mechanism in the marine atmosphere. Proceedings of the National Academy of Sciences U S A 117(9): 4505– 4510. DOI: http://dx.doi.org/10.1073/pnas. 1919344117. Wassmann, P, Reigstad, M. 2011. Future Arctic Ocean seasonal ice zones and implications for pelagic- benthic coupling. Oceanography 24(3): 220–231. DOI: http://dx.doi.org/10.5670/oceanog.2011.74. Webb, AL, van Leeuwe, MA, den Os, D, Venables, H, Stefels, J. 2019. Interannual and interseasonal var- iation in DMS flux from the West Antarctic Penin- sula. Scientific Reports 9: 2233. DOI: http://dx.doi. org/10.1038/s41598-019-38714-4. Verity, PG, Brussaard, CP, Nejstgaard, JC, Van Leeuwe, MA, Lancelot, C, Medlin, LK. 2007. Current under- standing of Phaeocystis ecology and biogeochemis- try, and perspectives for future research. Biogeochemistry 83(1–3): 311–330. DOI: http://dx. doi.org/10.1007/s10533-007-9090-6. Woolf, DK. 1997. Bubbles and their role in gas exchange, in Liss, P, Duce,RAeds.,The sea surfaceandglobal change. Cambridge University Press: 173–206. DOI: http://dx. doi.org/10.1017/CBO9780511525025.007. Vila-Costa, M, del Valle, DA, Gonza´lez, JM, Slezak, D, Kiene, RP, Sa´nchez, O, Simo´, R. 2006. Phyloge- netic identification and metabolism of marine dimethylsulfide-consuming bacteria. Environmental Microbiology 8(12): 2189–2200. DOI: http://dx.doi. org/10.1111/j.1462-2920.2006.01102.x. Woolf, DK. 2005. Parametrization of gas transfer veloci- ties and sea-state-dependent wave breaking. Tellus B 57(2): 87–94. DOI: http://dx.doi.org/10.1111/j. 1600-0889.2005.00139.x. Zapata, M, Jeffrey, SW, Wright, SW, Rodrı´guez, F, Gar- rido, JL, Clementson, L. 2004. Photosynthetic pig- ments in 37 species (65 strains) of Haptophyta: Implications for oceanography and chemotaxon- omy. Marine Ecology Progress Series 270: 83–102. DOI: http://dx.doi.org/10.3354/meps270083. Wang, S, Elliott, S, Maltrud, M, Cameron-Smith, P. 2015. Influence of explicit Phaeocystis parameteriza- tions on the global distribution of marine dimethyl sulfide. Journal Of Geophysical Research Biogeos- ciences 120(11): 2158–2177. DOI: http://dx.doi. org/10.1002/2015JG003017. Zeng,Y-X, Qiao, Z-Y,Yu,Y, Li, H-R, Luo,W. 2016. Diversity of bacterial dimethylsulfoniopropionate degrada- tion genes in surface seawater of Arctic Kongsfjor- den. Scientific Reports 6: 33031. DOI: http://dx.doi. org/10.1038/srep33031. Wassmann, P, Ratkova, T, Reigstad, M. 2005. The con- tribution of single and colonial cells of Phaeocystis pouchetii to spring and summer blooms in the north-eastern North Atlantic. Harmful Algae 4(5): How to cite this article: Galı´, M, Lizotte, M, Kieber, DJ, Randelhoff, A, Hussherr, R, Xue, L, Dinasquet, J, Babin, M, Rehm, E, Levasseur, M. 2021. DMS emissions from the Arctic marginal ice zone. Elementa: Science of the Anthropocene 9(1). DOI: https://doi.org/10.1525/elementa.2020.00113 Domain Editor-in-Chief: Jody W. Published: July 14, 2021 Accepted: May 27, 2021 Submitted: August 7, 2020 Published: July 14, 2021 Accepted: May 27, 2021 Submitted: August 7, 2020 Copyright: © 2021 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. Copyright: © 2021 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. Elem Sci Anth is a peer-reviewed open access j l bli h d b U i i f C lif i P p p journal published by University of California Press. journal published by University of California Press.
https://openalex.org/W4390741048	https://jayapanguspress.penerbit.org/index.php/IJMS/article/download/2976/1304	English	null	Character Education Learning Strategies in Developing Students’ Literacy Skills	International Journal of Multidisciplinary Sciences	2,024	cc-by-sa	3,001	International Journal of Multidisciplinary Sciences Volum ISSN International Journal of Multidisciplinary Sciences Volum ISSN national Journal of sciplinary Sciences Volume 2 Issue 1 (2024) ISSN : 2986-7665 (Media Online) Volume 2 Issue 1 (2024) ISSN : 2986-7665 (Media International Journal of Multidisciplinary Sciences International Journal of Multidisciplinary Sciences Character Education Learning Strategies in Developing Students’ Literacy Skills Maikel Hendrik Domingo Savio Private University (UPDS), Santa Cruz, Bolivia maikelhendrik9@gmail.com Abstract The research aimed to describe character education learning strategies in developing students’ literacy skills. The method used in this research was descriptive qualitative, where most of the data came from literature studies. The lack of interest in reading books (literacy) and critical thinking skills among students is one of the obstacles in Indonesia’s education world nowadays. It is based on the fact that some students who lack interest in reading cannot think critically in processing and obtaining reliable and actionable information. Therefore, efforts, strategies, and methods used in learning are needed to foster students’ ability and interest in reading so that students can have the ability to think critically in processing the information obtained in the hope of creating a golden generation that can compete in this era. Methods The research used descriptive and qualitative methods. This research aimed to describe learning strategies in developing character education literacy. Descriptive qualitative research was intended to explain symptoms or conditions during the research. Qualitative research aims to describe and reduce phenomena, events, social activities, attitudes, beliefs, cognitions, and thoughts of individuals and groups. Qualitative research methods are often called naturalistic research methods because the research is conducted in natural conditions (natural environment) and is said to be an ethnographic method. The data collection method used was a literature study. Introduction In the current globalization era, education plays an important role, is needed in life, and is considered a necessity for everyone. With education, all can learn, gain knowledge, and change a person’s perspective or way of thinking. Currently, education in Indonesia is inferior compared to other countries. There are several causes of low education in Indonesia. One of them is the lack of literacy or reading interest of today’s young generation both among students (elementary, junior high, high school) and among students, which is the cause of the low ability to think critically. Reading (Literacy) is a source of knowledge because a person can learn various things they want or do not know by reading (Literacy). The ability and desire to read affect someone’s knowledge and skills. The more people read, the more they will learn new things and have many abilities. It means that the more things people know, the more they will be able to help them do various things they do not know and have not mastered. Therefore, someone with a high interest in reading will have more quality than someone with a low interest in reading (Benson, 2004; Hornberger, 2009). It should be noted that literacy is an essential skill that every student needs to have in the field of learning. In simple terms, literacy is understood as reading and writing skills. Then, the definition of literacy developed into the abilities and skills individuals possess in language, including reading, writing, listening, and speaking, at a certain level of ability needed in life. Students need literacy skills to understand many subjects. One of the many subjects is character education learning. Character education is the process of planting and learning about the values of life. Character education is an education that serves to shape and develop a strong personality for students. Literacy in character learning emphasizes the skill of receiving information that focuses on several activities, such as reducing, presenting, and verifying information (Faguet & Sanchez, 2008). In this regard, several methods and strategies are used to develop students’ literacy skills in the character education learning process. 43 https://jayapanguspress.penerbit.org/index.php/IJMS Result and Discussions 1. Factors Causing Low Literacy Interest 1. Factors Causing Low Literacy Interest Technology is developing rapidly and has become the needs of every individual today, especially internet technology. The development of internet technology has created many types of social media. It is also supported by the development of mobile phone technology (Smartphones/gadgets), which provides facilities for current social media users. Social media usage makes it easier to communicate with friends and family who cannot meet due to long distances. As it can be observed nowadays, social media is used as a medium to provoke others. In addition, social media also has a negative impact on society. Providing false information or news (Hoax) and others. Social media can be said to be a razor blade that can be used for good things and vice versa; it can be a means of killing and destroying someone. Social media usage can change people’s character and behavior. The emergence and growing development of social media nowadays impact children’s behavior at various levels, such as society being less normal and interacting and communicating with the community because they are too busy with their smartphones and do not care about their surroundings. Social media also usually causes children to become unruly because they do not want to learn and spend their time playing with their devices. The emergence of a ‘top’ technology called the smartphone is one of the causes of low interest in literacy or willingness to read books because the features offered are more interesting, such as Twitter, Friendster, Facebook, Instagram, YouTube, various games, and others that can seize the attention of most people from reading books. The impact of technology and knowledge development is extensive in using teaching materials in schools (Medina-Jerez, Middleton, & Orihuela-Rabaza, 2011). In the learning implementation, most students pay less attention to the material, do not understand it, and only expect grades. 2. Lack of Literacy Impacts 2. Lack of Literacy Impacts a. Lack of ability to understand, master, communicate, and use knowledge and technology to produce high-quality products and develop the country. a. Lack of ability to understand, master, communicate, and use knowledge and technology to produce high-quality products and develop the country. b. Lack of insight and limited knowledge can reduce thinking and easily overwhelm one with negative notions and forces. b. Lack of insight and limited knowledge can reduce thinking and easily overwhelm one with negative notions and forces. c. Lack of literacy skills will make someone’s creativity not develop. Creative thinking can be realized when stakeholders can develop their thinking and react quickly to the environment, which can be trained in literacy activities. The innovative ideas that occur can cause a person to be more productive and beneficial for themselves and those around them. d. Difficulty in improving someone’s quality due to lack of up-to-date information. 44 https://jayapanguspress.penerbit.org/index.php/IJMS e. Ignorance leads to indifference as they are reluctant to expand their knowledge and update themselves with the latest information. It gradually causes the affected person to shut down, become absorbed in their world, and ignore their surroundings. Lack of broad insight tends to cause difficulties in social life, lacking the ability to interact and communicate because they have less input than their friends around them who have broad insight because of the amount of knowledge gained through literacy. g g g g y f. This reluctance of the younger generation is a loss to the nation that has lost the possession of a qualified and productive giver of ideas for the country’s progress. g g g g y f. This reluctance of the younger generation is a loss to the nation that has lost the possession of a qualified and productive giver of ideas for the country’s progress. 3. Implementation of Character Education Literacy p y Character education learning greatly affects values, norms, and personality formation, especially among students, because they are the nation’s next generation (Regalsky & Laurie, 2007). Character education is essential for building a humane personality for the younger generation. Furthermore, the generation is invited to read carefully (literacy) and think of statements that can motivate them to deepen their desire to deepen religious teachings (Lopes Cardozo, 2012). Based on these descriptions, the definition of literacy in character learning emphasizes news comprehension skills that focus on some activities, such as reducing information, displaying information, and verifying information. One way to develop students’ reading and writing skills at school is by implementing or cultivating literacy and through habituation in the classroom learning process in each subject. The implementation of character education literacy can be seen as follows: https://jayapanguspress.penerbit.org/index.php/IJMS a. Implementation The implementation of character education literacy in the learning process, where teachers use the curriculum application with a scientific approach through five stages of activities, which are covered in four core competencies, namely as follows: 1) Observing, in which the process emphasizes the meaning that occurs in learning. 2) Questioning, carried out by the teacher, includes the strength dimension (Strength) (a) inspiring students to improve and develop a strong belief in their potential, (b) having a fighting spirit and not giving up easily when faced with obstacles in the learning process in terms of attitudes, knowledge, and skills. 3) Exploring, the stages of thinking rationally and systematically from observable empirical facts to gain knowledge through reasoning. Associating, this stage is used so that students can use the scientific method and behavior to get answers to problems faced in everyday life. 4) Communicating, networking or collaborative learning defines cooperation as a structure of interactions systematically organized to facilitate the achievement of a common goal. 4) Communicating, networking or collaborative learning defines cooperation as a structure of interactions systematically organized to facilitate the achievement of a common goal. g Methods of Character Education Literacy Implementation g b. Methods of Character Education Literacy Implementation y p 1) Habituation Method The method used in implementing character education literacy at school is the habituation method. By implementing and familiarizing themselves with literacy, students will have broad insight or knowledge that affects their character, especially in thinking, acting according to religious teachings, and doing good. Hornberger & Johnson (2007) states that instilling values or character teachings, such as moral ethics coaching, can overcome student behavior deviations. Thinking, saying, and doing good and right are inseparable and can be used to build human character. 1) Habituation Method The method used in implementing character education literacy at school is the habituation method. By implementing and familiarizing themselves with literacy, students will have broad insight or knowledge that affects their character, especially in thinking, acting according to religious teachings, and doing good. Hornberger & Johnson (2007) states that instilling values or character teachings, such as moral ethics coaching, can overcome student behavior deviations. Thinking, saying, and doing good and right are inseparable and can be used to build human character. 45 https://jayapanguspress.penerbit.org/index.php/IJMS Good behavior can be achieved by controlling behavior with the ability to read and have extensive knowledge related to phenomena, events, and issues owned or obtained through literacy. Students can control themselves in their actions and know what they can and cannot do. With high literacy skills, especially literacy in character, the actions that will be carried out will be based on the teachings of truth 2) Learning Assignment Method The assignment method can encourage students to be more active in learning and understand mysteries and student learning outcomes (Luykx, 1996). Implementing the assignment method to character learning activities encourages students to actively participate and engage in learning activities presented in class and other representative locations. Student assignments are done in various formats, such as a survey on a particular subject, a discussion sequence, or another textbook description. 3) Discussion Method The discussion method is a learning method related to problem-solving learning. The discussion method is a presentation method. In the learning process, students are faced with a problem that will be solved together in the form of a group and can conclude their discussion results. Discussion is an exchange of ideas between several individuals to have a common view of the perceived problem (Klaiber, 2009). g Methods of Character Education Literacy Implementation Discussion is about having several group members exchange ideas intertwined with problematic questions, generate ideas, and review ideas and opinions to solve problems. The ability to have responsive scientific skills (Aalborg, Sullivan, Cortes, Basagoitia, Illanes, & Green, 2016). The discussion method referred to in developing students’ literacy skills is by providing a problem with the aim that students can exchange opinions to get answers, truths, and solutions to the obstacles faced. The discussion method used in learning activities will be able to stimulate students to think systematically, logically, critically, and innovatively in developing their way of thinking in solving a problem. Developing literacy skills in the learning process can develop students’ critical attitude, especially in answering various kinds of phenomena in society related to character. A critical attitude in students is necessary to foster sensitivity regarding everything that happens, especially in the surrounding environment. Thus, students must accept and filter the information obtained through literacy. 4. Learning Strategy Strategy is a plan, step, and sequence in achieving a goal. Strategy is used as a reference in placing the activity process through well-structured and planned steps. Strategy describes the direction of action in achieving a targeted goal (Frisancho, & Reátegui, 2009). b. Discovery Learning Strategy Discovery learning is a learning process where students solve problems to develop their knowledge and skills (Gurven, et al., 2013). The discovery learning model was chosen to foster literacy skills in character education learning because it provides opportunities for students to think, discover, discuss, and collaborate through scientific learning activities to practice and improve critical thinking skills and find solutions, as well as gain knowledge about ideas that affect the improvement of learning outcomes. Implementing a discovery learning strategy requires students to be more active in reading, searching for information, and knowledge to solve problems given by the teacher. It aims for students to have more profound knowledge, memory, vocabulary, and understanding of the material studied longer than students who only get information from the teacher. The learning implementation stage using the discovery learning strategy on character education is when the teacher provides an issue or a problem related to the material to be discussed with the Brainstorming method (collecting ideas or students’ initial knowledge related to the material to be conveyed). Furthermore, students will search and collect data about the material and problems given (exploring) through literacy or reading books. After obtaining data or solutions related to the given issues, students will discuss or present the results obtained (communicating). After the discussion, the teacher will invite students to reflect or conclude together the results of the discussion and the material that has been obtained. a. Inquiry Strategy Inquiry Learning Strategy is a series of learning activities emphasizing a perspective and analytical way to get out of the intended problem (Hornberger, 1990). Inquiry strategy emphasizes students’ activities in searching and finding. Inquiry strategy aims to foster expertise in thinking systematically, logically, and critically in developing students’ thinking skills through implementing reading and writing skills as part of the mental formation process or intellectual skills development. To develop literacy skills through the inquiry strategy, students need to learn how to rely on the material provided and how to use and develop their potential. Developing literacy skills can help students develop their ability to think optimally, solve problems, provide solutions to issues faced, and easily master learning materials. 46 https://jayapanguspress.penerbit.org/index.php/IJMS Conclusions Literacy skills are essential abilities that every student should have. In reality, there was currently a lack of interest in literacy from students. Therefore, character education learning strategies were necessary to foster and improve students’ literacy interests. Several factors caused the low interest in reading. One of them was the rapid development of technology that produced various types of social media and offers various interesting features compared to reading books. Several methods were used to develop students’ literacy skills, such as the habituation method, the learning assignment method, and the discussion method. In addition, the Inquiry Strategy and the Discovery Learning Strategy were used to develop students’ literacy skills. References Aalborg, A., Sullivan, S., Cortes, J., Basagoitia, A., Illanes, D., & Green, M. (2016). Research ethics training of trainers: developing capacity of Bolivian health science and civil society leaders. Acta Bioethica, 22(2). Benson, C. (2004). Do we expect too much of bilingual teachers? Bilingual teaching in developing countries. International Journal of Bilingual Education and Bilingualism, 7(2-3), 204-221. Faguet, J. P., & Sanchez, F. (2008). Decentralization’s effects on educational outcomes in Bolivia and Colombia. World development, 36(7), 1294-1316. Frisancho, S., & Reátegui, F. (2009). Moral education and post‐war societies: the Peruvian case. Journal of Moral Education, 38(4), 421-443. Gurven, M., Von Rueden, C., Massenkoff, M., Kaplan, H., & Lero Vie, M. (2013). How universal is the Big Five? Testing the five-factor model of personality variation among forager–farmers in the Bolivian Amazon. Journal of personality and social psychology, 104(2), 354. 47 https://jayapanguspress.penerbit.org/index.php/IJMS Hornberger, N. H. (1990). Creating successful learning contexts for bilingual literacy. Teachers College Record, 92(2), 212-229. Hornberger, N. H. (2009). Multilingual education policy and practice: Ten certainties (grounded in Indigenous experience). Language Teaching, 42(2), 197-211. Hornberger, N. H., & Johnson, D. C. (2007). Slicing the onion ethnographically: Layers and spaces in multilingual language education policy and practice. Tesol Quarterly, 41(3), 509-532. Klaiber, J. (2009). The Catholic Church, moral education and citizenship in Latin America. Journal of Moral Education, 38(4), 407-420. Lopes Cardozo, M. T. (2012). Transforming pre-service teacher education in Bolivia: from indigenous denial to decolonisation?. Compare: A journal of comparative and international education, 42(5), 751-772. Luykx, A. (1996). From Indios to profesionales: Stereotypes and student resistance in Bolivian teacher training. The cultural production of the educated person: Critical ethnographies of schooling and local practice, 239-272. Medina-Jerez, W., Middleton, K. V., & Orihuela-Rabaza, W. (2011). Using The Dast-C To Explore Colombian And Bolivian Students’images Of Scientists. International Journal of Science and Mathematics Education, 9, 657-690. Regalsky, P., & Laurie, N. (2007). ‘The school, whose place is this’? The deep structures of the hidden curriculum in indigenous education in Bolivia. Comparative Education, 43(2), 231-251. https://jayapanguspress.penerbit.org/index.php/IJMS 48
https://openalex.org/W1548970708	https://journal.unnes.ac.id/nju/index.php/paramita/article/download/2120/2222	Indonesian	null	TINDAKAN NEGARA TERKAIT PERISTIWA G30S: STUDI MAKNA GADAMERIAN PADA PESELAMAT	Paramita/Paramita: Historical Studies Journal	2,012	cc-by	4,088	TINDAKAN NEGARA TERKAIT PERISTIWA G30S: STUDI MAKNA GADAMERIAN PADA PESELAMAT Hamdan Tri Atmaja Jurusan Sejarah, FIS Universitas Negeri Semarang atmaja.hta@gmail.com ABSTRACT Tulisan ini mendeskripsikan secara mendalam tindakan negara terkait peristiwa G30S. Metode penelitian yang digunakan adalah penelitian kualitatif dengan pendekatan hermeneutika yang digagas Gadamer. Hasil penelitian menunjukkan tindakan negara terhadap peselamat adalah mela- kukan penangkapan, pemeriksaan, dan pena- hanan serta memenjarakan mereka ke pulau Buru (untuk peselamat laki-laki) dan Plantun- gan (untuk peselamat perempuan). Peristiwa G30S oleh peselamat dimaknai sebagai kisah pembunuhan para jendral oleh PKI, bentuk kon- spirasi politik memperebutkan kekuasaan Soekarno dengan latar belakang ideologi. Pe- meriksaan dan penangkapan dimaknai peselamat sebagai tindakan tidak beralasan, politik kambing hitam, dan sebagai bentuk kesewenang- wenangan terhadap peselamat. Kehidupan pen- jara dimaknai peselamat sebagai bentuk kerja paksa, hukuman dengan menstigmatisasi dan mengisolasi kaum perempuan. This study aims to gain knowledge of a deep un- derstanding of the state action related to the G30S event. The research method used was a qualitative research approach initiated by Gadamer's hermeneutics. The results showed that state action against survivors were to arrest, investigate, and imprison them to the island of Buru (for men survivors) and Plantungan (for women survivors). The G30S event, by survi- vors, was interpreted as a story of the assassina- tion of the generals by Indonesian Communist Party (PKI), as well as the form of a political conspiracy for Sukarno’s power within ideologi- cal background. Investigation and arrest were interpreted by them as an act of unwarranted, political scapegoat, and a form of abuse against them. While prison life, for survivors, was as a form of forced labor, punishment to stigmatize and isolate women Keywords: State Action, the G30S event, Mean- ing, and Survivor. Kata kunci: Tindakan Negara, Peristiwa G30S, Makna, dan Peselamat. Kata kunci: Tindakan Negara, Peristiwa G30S, Makna, dan Peselamat. Paramita Vol. 22 No. 2 - Juli 2012 [ISSN: 0854-0039] Hlm. 198—206 PENDAHULUAN PKI mengandung monopoli kebenaran yang selalu dipakai Orde Baru sebagai bentuk tafsir tunggal atas kebenaran sejarah pada sekitar tahun 1965 (Notosusanto dan Saleh, 1966; Manafe, 2008). Merujuk pada pemikiran tersebut, maka dalam kajian penelitian ini, istilah yang akan digunakan adalah G30S. Se- mentara kata peselamat yang dimaksud dalam tulisan ini mengacu pada korban hidup (survivor) yaitu mereka yang di- G30S/PKI, adalah istilah yang lazim diwacanakan untuk menggambar- kan peristiwa penculikan para jenderal yang dipimpin Letkol Untung, pada tanggal 1 Oktober 1965 dini hari yang berujung pada tewasnya enam jenderal dan satu orang perwira. Penggunaan istilah tersebut sangat bernuansa politis dan berbau kontroversi. Istilah G30S/ Paramita Vol. 22 No. 2 - Juli 2012 [ISSN: 0854-0039] Hlm. 198—206 : 131—248 Tindakan Negara terkait Peristiwa G 30 S - Hamdan Tri Atmaja tangkap karena anggota PKI dan Or- masnya atau yang dianggap PKI. gara memanfaatkan balik kebisuan se- bagai alat untuk menaklukannya. Di sinilah negara menguasai peselamat, sehingga makna tindakan negara atas peristiwa G30S diproduksi dan dimo- nopoli oleh negara. Makna peristiwa G30S yang diproduksi dan dilestarikan negara, dalam perspektif kekuasaan, memperli- hatkan kekuasaan tidak hanya berada di wilayah lembaga negara, tetapi kekua- san merambah ke wilayah masyarakat. Kekuasaan merembes dalam kehidupan sehari-hari dan memasuki ruang-ruang sosial yang mendefiniskan subjek, objek, dan realitas. Kekuasaan menjadi omni- present, hadir di mana-mana dan mampu mengendalikan wacana, suatu keadaan dimana cara masyarakat me- mandang dan memahami realitas, serta menuturkannya di bawah kendali ke- kuasaan (Supelli, 2000:41; Supelli, 2005). Kekuasaan juga menentukan sistem makna dan kumpulan nilai yang membingkai kemungkinan pemikiran dan ekspresinya, untuk mengeluarkan setiap ketidaksetujuan. Foucoult lebih jauh mengatakan kekuasaan ada di mana-mana, bukan mencakupi segala- galanya, tetapi karena datang dari mana -mana, kekuasaan berfungsi sebagai un- sur yang tak terhitung jumlahnya, ke- kuasaan tidak hanya bersifat represif tetapi juga produktif dan positif (Foucoult, 1997 : 113). Dari serangkaian paparan tentang tindakan negara atas peristiwa G30S seperti tersebut di atas, menunjukkan bahwa kajian tentang G30S dari pers- pektif peselamat, sarat dengan persoa- lan-persoalan kesadaran bersama me- ngenai humanisasi hidup. Kesadaran ber- sama ini akan menghasilkan sebuah ke- sepekatan dialogis, bahwa manusia bisa hancur atau binasa jika negara mem- biarkan kejahatan atas manusia. PENDAHULUAN Dari kesadaran tersebut, perlu secara prinsip menyatakan bahwa ada tindakan- tindakan yang mengingkari martabat manusia, dan tindakan-tindakan itu ti- dak boleh dibiarkan, karena kehadiran bersama manusia memunculkan kesa- daran etis untuk tidak membiarkan munculnya situasi yang akan menye- babkan kebinasaan orang lain, atau situasi yang menyebabkan orang lain menjalani kehidupan tidak manusiawi. Disinilah pentingnya membangun i- ngatan bersama atas masa lampau me- lalui perdebatan-perdebatan jujur se- hingga melahirkan historical accountabil- ity. Ingatan bersama yang bersih dari politisasi, sehingga tidak dijadikan pija- kan untuk mengulangi kekejaman di masa yang akan datang. Situasi semacam ini, membuat para peselamat kehilangan ruang publik untuk menuturkan pengalaman pribadi- nya. Peselamat merasa bahwa penga- laman pribadinya berada di luar realitas yang sedang berjalan. Peselamat tidak mampu mengaktualisasikan pengalam- annya ke dalam wacana yang bukan mengandung pemikirannya. Ketidak- berdayaan ini menjadi keberhasilam ne- gara melakukan perintah untuk bung- kam, Foucoult menyebutnya sebagai arkeologi kebisuan. Kebisuan peselamat dijadikan sebagai garis pertahanan ter- depan oleh negara, yang ampuh untuk menyangkal dan menyingkirkan pese- lamat. Ketika peselamat bertutur, ne- y g g Kajian tentang ingatan bersama dari perspektif peselamat, pernah dila- kukan oleh Urvashi Butalia seorang to- koh feminis dari India. Dengan fokus kajian tentang pembelahan India, Ur- vashi menangkap akibat pembelahan India yang menimbukan kematian, penggusuran, keterenggutan, pemerko- saan, kehilangan rumah, serta kehilang- an keluarga dan sanak saudaranya (Butalia, 2002). Penelitian lain dilaku- kan oleh Jaringan Kerja Budaya (JKB). 199 Paramita Vol. 22, No. 2 - Juli 2012: 131—248 adalah tindakan negara atas peristiwa G30S dari perspektif negara itu sendiri. Dengan menggali intepretasi peselamat tentang tindakan negara atas peristiwa G30S, akan didapatkan keseimbangan intepretasi, sehingga akan muncul makna baru tentang tindakan negara peristiwa G30S, sebagai antitesis atas makna yang diproduksi negara selama ini, sehingga terbuka ruang dialog un- tuk memahami realitas tindakan negara atas peristiwa G30S. Melalui oral histories (sejarah lisan) secara spesifik, tulisan ini mencoba mengungkap tindakan ne- gara terhadap peselamat terkait dengan peristiwa G30S dan bagaimana pese- lamat memaknai tindakan negara terha- dap dirinya terkait dengan peristiwa G30S. Penelitian yang mencoba mengungkap keberagaman ingatan sosial para pese- lamat atas peristiwa G30S (Rossa dkk, 2004). Saskia Elenora Wieringa, dengan mengambil topik sekitar peristiwa G30S dan fokus kajiannya tentang Gerwani. Mencoba mengungkap kebenaran tersembunyi di balik cerita-cerita ten- tang perbuatan seksual yang tidak senonoh dan kebejatan moral yang dila- kukan para gerwani pada saat peristiwa G30S (Wieringa, 1999). PENDAHULUAN g Penelitian lain tentang peselamat yang terkait dengan kajian tentang makna dilakukan oleh Teria Shantall. Penelitian Teria mengungkap penga- laman peselamat dalam sebuah tragedi politik, dengan subjek penelitian adalah para peselamat dalam tragedi Holou- coust di Afrika Selatan (Shantall, 1999). Kajian mendalam tentang makna den- gan menggunakan pendekatan herme- neutika, dilakukan Utan Parlindungan. Sebuah penelitian yang mempertanya- kan mengapa dan bagaimana genjer- genjer dijangkiti ambiguitas makna, yang di satu sisi diletakkan di aras ter- hormat kultural (folksong) sebagai media kritik terhadap penjajahan, dan di sisi lain berakhir secara dramatis sebagai lagu manifesto PKI hingga mengalami pelarangan Orde Baru (Parlindungan, 2007). Kajian tentang penjara Buru per- nah dilakukan oleh Krisnadi (2000). Ia melakukan kajian terhadap tahanan politik di Pulau Buru pada tahun 1969- 1979. Selain itu ada pula penuturan pe- selamat yang menguraikan ceritanya ketika tengah berada pada pengasingan. Demokrano (2006) misalnya mengi- sahkan kisahnya selama 11 tahun dalam penjara Orde Baru. METODE PENELITIAN Jenis penelitian yang digunakan dalam penelitian ini adalah penelitian kualitatif dengan pendekatan interpreta- tif, yang digagas Gadamer yaitu fenomenologi hermeneutika. Dalam fenomenologi hermeneutika, makna suatu tindakan atau fenomena bukanlah sesuatu yang ada pada tindak itu sendiri, tetapi terbuka bagi adanya penafsiran baru sesuai dengan kreativi- tas penafsir (Gadamer, 2004). Sebuah fenomena yang sudah dilempar ke ru- ang publik sepenuhnya menjadi milik penafsir dengan berbagai interpre- tasinya (Raharjo, 2008). Dalam hal ini, interprestasi bukan hanya sebatas mere- produksi makna, tetapi juga mempro- duksi makna. Akhirnya makna yang diproduksi itu sendiri sangat terbuka dengan tafsir makna baru sesuai dengan perkembangan waktu dan zamannya. Dari kajian penelitian tersebut di atas menginspirasi untuk melakukan penelitian, khususnya terkait dengan tindakan negara terkait peristiwa G30S dari perspektif peselamat. Selama ini, yang sudah banyak terpublikasikan Subjek yang perlu didengar, di- mengerti, digali dan terus diungkap dunia realitas di balik yang tampak, agar apa yang implisit menampakkan 200 : 131—248 Tindakan Negara terkait Peristiwa G 30 S - Hamdan Tri Atmaja dirinya menjadi eksplisit adalah pese- lamat yang pernah di penjara di pulau Buru dan Plantungan atau pernah di tahan tetapi tidak sampai Plantungan dan pulau Buru. Berdasarkan pengga- lian data di lapangan, dapat dijaring tujuh peselamat sebagai subjek peneli- tian. Ketujuh peselamat tersebut pada masa lalu tergabung dalam organisasi sosial politik yang berbeda-beda seperti SB Tekstil, IPPI/Lekra, PKI/PR, PGTI dan CGMI, dan dua orang peselamat mengaku tidak pernah ikut dalam or- ganisasi. Disamping itu tujuh peselamat yang terpilih memenuhi kriteria yang unik dan spesifik, seperti memiliki pe- ngalaman khas, memiliki harapan- harapan sendiri, yang pada intinya berusaha menemukan kedalaman (depth), kekayaan (richness) dan kom- pleksitas (complexity) (Sparringa, 2008). dari dalam dirinya sendiri terhadap diri masyarakat tertentu. Secara garis besar, langkah-langkah analisis data dapat dirumuskan sebagai berikut. Pertama, menentukan subjek penelitian. Kedua, mengumpulkan data dengan melaku- kan wawancara mendalam sesuai deng- an panduan wawancara. Ketiga, mela- kukan kategorisasi atau pengelompokan data. Keempat, setelah semua data terkumpul dan diurutkan secara kro- nologis, langkah berikutnya adalah me- reduksi data atau menyusun narasi se- jarah hidup pengalaman peselamat. Ke- lima, melakukan telah dan pengungkap- an keserbamaknaan pengalaman pese- lamat, atau konstruksi peselamat atas pengalamannya terkait dengan tindakan negara atas peristiwa G30S. HASIL DAN PEMBAHASAN Pengumpulan data dalam peneli- tian ini menggunakan wawancara men- dalam. Dalam wawancara mendalam ini, peneliti berupaya menyelam ke dalam dunia psikologis dan sosial sub- jek penelitian. Jenis wawancara men- dalam yang dipilih adalah wawancara sejarah hidup (life history interview), perlu dicatat disini sejarah hidup yang dimaksud bukanlah otobiograsi konven- sional ataupun fiksi, meskipun narasi subjektif penulisnya mirip dengan na- rasi dan pandangan penulis dalam oto- biografi. Sejarah hidup disini lebih memperhatikan pengalaman subjek dan penafsirannya atas dunia dan penga- lamannya. Artinya dengan sejarah hidup berarti menempatkan peran, dan hidup peselamat sebagaimana mereka melihatnya, bukan penafsiran orang lain atau peneliti tetapi penafsirannya sendiri. Peristiwa G30S dalam Perspektif Pese- lamat Peristiwa G30S dalam Perspektif Pese- lamat Peselamat perempuan berpendi- dikan SD dan tidak memiliki latar be- lakang politik apapun, memaknai peristiwa G30S sebagai pembunuhan para jenderal oleh PKI, dengan imbas- nya menghukum orang-orang yang ti- dak bersalah. Pemaknaan ini dibangun tidak didasarkan pada analisis kritis, tetapi hanya sebatas masuknya infor- masi tentang meninggalnya para jenderal oleh PKI. Ketidakmampuan peselamat mencerna informasi yang diperoleh mengenai peristiwa G30S, membenarkan bahwa terbunuhnya para jenderal memang dilakukan oleh PKI. Pembenaran tersebut tidak menunjuk- kan keberpihakan pada salah satu ke- kuatan politik tertentu, tetapi semata- mata karena peselamat merasa tidak tahu apa-apa tentang peristiwa G30S. Teknik analisis yang digunakan dalam penelitian ini adalah teknik anali- sis kualitatif sejarah hidup. Teknik ini mampu untuk melihat reaksi, tangga- pan, intepretasi, pandangan individu Peselamat perempuan, yang memiliki latar belakang komunis, 201 Paramita Vol. 22, No. 2 - Juli 2012: 131—248 seperti aktif dalam ormas IPPI dan Lekra, memaknai peristiwa G30S seba- gai pembunuhan para jenderal oleh PKI. Akan tetapi, tidak semua orang PKI ter- libat dalam peristiwa tersebut. Pe- maknaan ini didasarkan pada apa yang dialami dan dirasakan pada waktu itu. Ketika G30S itu meletus, peristiwa itu benar-benar membingungkan bagi dir- inya. Menurutnya, semua informasi ten- tang peristiwa G30S justru didapatkan dari isu yang berkembang di masyara- kat. Hampir semua orang menyimpul- kan PKI adalah dalang dan pelaku dalam peristiwa G30S dan kesimpulan ini diperoleh secara berulang-ulang mu- lai dari sebelum masuk penjara, ada dalam penjara, sampai dengan keluar dari penjara. Ruang politik seperti ke- luarga yang berhaluan komunis, justru tidak banyak memberi informasi apa- apa tentang peristiwa G30S. meskipun tidak semuanya terlibat dalam dunia politik praktis, tetapi mereka telah menunjukkan ideologi politik ayang dianutnya. Keberagaman tersebut ditunujukan dengan adanya silang pendapat sebagai berikut: G30S sebagai persoalan persaingan antara TNI-AD dan PKI, persoalan G30S meru- pakan konspirasi politik yang menem- patkan PKI dan Amerika sebagai bagian perebutan kekuasaan, dan peristiwa G30S sebagai perebutan kekuasaan Soekarno dengan latar belakang ideologi. Penangkapan dan Pemeriksaan Pese- lamat Pasca-peristiwa G30S, negara se- cara represif melakukan penangkapan terhadap orang-orang yang diduga menjadi anggota PKI atau menjadi ang- gota ormas PKI, tanpa ada sedikit pun celah untuk membela dirinya. Hasil penelitian ini, menunjukkan bahwa penangkapan terhadap mereka dilaku- kan langsung oleh petugas dengan cara mendatangi rumah peselamat. Secara garis besar aktivitas penangkapan dan pemeriksaan terhadap peselamat dapat digambarkan pada tabel 1. Peristiwa G30S dimaknai pula se- bagai perebutan kekuasaan antar-elite politik, hal ini menunjukkan adanya para elite politik yang bertikai mem- perebutkan kekuasaaan. Elite politik mana yang bertikai memperebutkan ke- kuasaan, dalam perspektif peselamat memunculkan berbagai keragaman pen- dapat. Keberagaman pendapat ini menunjukkan adanya keberpihakan pada ideologi politik tertentu pada pese- lamat sesuai dengan keyakinan ideologi yang dianutnya. Ini dapat disadari, mengingat mereka para peselamat Dari tabel tersebut dapat dijelas- kan bahwa tahun penangkapan pada peselamat berbeda-beda. Empat orang peselamat ditangkap beberapa bulan setelah peristiwa G30S, sehingga masih Tabel 1 Penangkapan dan Pemeriksaan Terhadap Peselamat No. Peselamat Tahun Ditangkap Tempat Penangkapan Waktu Tangkap Kekerasan Fisik Pelecehean Seksual 1 2 3 4 5 6 7 Kustini Sunari Sariatun Sumito Suparno Abdullah Wiratmo 1965 1967 1965 1967 1965 1969 1965 Rumah Rumah Perjalanan Rumah Sekolah Rumah Rumah Siang Malam Malam Malam Siang Siang Siang x x x x x x x - - x - - - - Tabel 1 Penangkapan dan Pemeriksaan Terhadap Peselamat 202 : 131—248 Tindakan Negara terkait Peristiwa G 30 S - Hamdan Tri Atmaja kuensi, orang-orang PKI atau mereka yang ada di ormasnya, ditangkap dan diperiksa. Padahal, tidak semua orang PKI atau yang ada di ormasnya menge- tahui soal pembunuhan para jenderal. Para peselamat yang memaknai penang- kapan dan pemeriksaan sebagai politik kambing-hitam, dibangun oleh dua pe- selamat yang berlatar belakang sebagai pegawai. Sebagai pegawai negeri, se- benarnya mereka tidak terlibat dalam dunia politik, hanya secara kebetulan mereka bergabung dalam organisasi profesi seperti PGTI. Bahkan Peselamat lain yang berstatus sebagai pegawai Jawatan Pertanian tdak pernah menjadi anggota PKI ataupun Ormas PKI, hanya secara kebetulan pernah ikut kursus Kader Calon PKI, tetapi mereka tetap ditangkap dan di buang ke pulau Buru. di tahun 1965, dimana suhu politik ma- sih cukup memanas. Tiga orang pese- lamat ditangkap setelah tahun 1965, setelah terjadi pergantian kekuasaan dari Soekarno ke Soeharto. Waktu penangkapan terjadi pada siang dan malam hari, hampir seluruh peselamat mengalami penyiksaan fisik saat pe- meriksaan, dan salah seorang peselamat mengalami pelecehan seksual pada saat pemeriksaan. Penangkapan dan Pemeriksaan Pese- lamat Peselamat perempuan yang tidak merasa menjadi anggota Gerwani, me- maknai penangkapan terhadap dirinya sebagai tindakan tidak beralasan atau mengada-ada. Dengan menggunakan kata “Gerwani” maka semua aktivitas perempuan berhaluan kiri disebut deng- an Gerwani. Kata “Gerwani” dilekatkan pada semua kaum perempuan yang di- tangkap, sehingga semua perempuan yang ditangkap tuduhannya sama seba- gai Gerwani. Padahal, dua peselamat yang dalam pengakuannya hanya men- jadi anggota Serikat Buruh Tekstil dan Lekra tidak pernah merasa menjadi ang- gota Gerwani. Dalam pemeriksaan, pe- selamat terus dipojokkan dengan tudu- han sebagai anggota Gerwani yang per- nah ke Lubang Buaya, membunuh, dan menyiksa para jenderal. Menjawab seju- jurnya dengan mengatakan “bukan Ger- wani”, tidak berarti dibebaskan dari se- mua tuduhan, tetapi justru sebaliknya mendapat penyiksaan. g p g p Penangkapan dan pemeriksaan oleh peselamat dimaknai juga sebagai bentuk kesewenang-wenangan terhadap orang-orang yang memiliki keterkaitan dengan PKI baik secara langsung atau- pun tidak langsung dengan tujuan un- tuk menghancurkan paham komunis di Indonesia. Ketika penangkapan dan pe- meriksaan dimaknai sebagai kesewe- nang-wenangan, maka semua orang yang bersentuhan dengan paham komu- nis, baik dalam bentuk aktivitas politik, budaya, ekonomi, maupun sosial, adalah orang-orang yang dianggap terli- bat dalam peristiwa G30S. Mereka yang pernah bersentuhan dengan PKI dan ormasnya dalam pandangan negara adalah orang-orang yang berbahaya, yang dimungkinkan dapat menum- buhkan benih-benih paham komunis. Di samping itu, negara menganggapnya sebagai orang yang bersalah, yang pan- tas untuk mendapat hukuman. Kesewe- nang-wenangan itu ditunjukkan dengan tidak adanya parameter yang jelas dalam proses penangkapan. Satu- satunya parameter yang digunakan adalah keterkaitan seseorang dengan Penangkapan dan pemeriksaan terhadap orang-orang PKI atau mereka yang ada dalam ormas PKI, dimaknai pula sebagai politik kambing-hitam un- tuk meligitimasi kekuasaan. Kambing- hitam yang dimaksud di sini adalah PKI. Ini menunjukkan pembunuhan para jenderal di Jakarta merupakan perebutan antar elit kekuasaan, ke- mudian digeneralisasikan bahwa semua orang PKI atau mereka yang ada di Or- mas PKI sebagai orang yang terlibat dan bertanggung jawab. Sebagai konse- 203 Paramita Vol. 22, No. 2 - Juli 2012: 131—248 PKI, baik secara langsung ataupun tidak langsung. garis besar keberadaan peselamat dalam penjara dapat digambarkan pada tabel 2. Pemaknaaan penangkapan dan pemeriksaan sebagai bentuk kesewe- nang-wenangan, dibangun oleh pese- lamat yang menjadi anggota PKI. Pese- lamat merasa tidak terlibat dan tidak bersalah atas pembunuhan para jenderal. Memang benar dirinya ang- gota PKI, tetapi walaupun anggota PKI tidak berarti tahu tentang skenario pem- bunuhan para jenderal. Penangkapan dan Pemeriksaan Pese- lamat Peselamat meyakini kalau pembunuhan para jenderal hanya dilakukan oleh elit-elite politik PKI, bukan merupakan kebijakan partai. Maka sudah seharusnya mereka elite politik yang dihukum, bukan se- mua orang PKI dihukum. Pemaknaan ini dibangun pula oleh Peselamat yang merasa tidak pernah menjadi anggota PKI dan tidak pernah terjun dalam politik praktis, bahwa penangkapan dan pemeriksaan yang menimpa dirinya me- rupakan tindakan kesewenang- wenangan. Berdasarkan pada sebaran data tersebut, enam peselamat digolongkan dalam klasifikasi B, dengan penjara tera- khir untuk laki-laki pulau Buru, dan un- tuk wanita Plantungan, serta satu pese- lamat digolongkan dalam klasifikasi C dipenjarakan di LP. Aktivitas inti yang dilakukan untuk tiga peselamat laki-laki bertani, dua peselamat lainnya ditugasi untuk bidang administrasi. Untuk pese- lamat wanita aktivitas utama yang dila- kukan membuat kerajinan, seperti menyulam, menjahit, membuat kristik dan sebagainya disamping aktivitas- aktivitas lainnya. Lamanya waktu di penjara bagi mereka yang dikirim ke Plantungan atau Pulau Buru rata-rata lebih dari sepuluh tahun, empat pese- lamat menjalani hidup di penjara di atas sepuluh tahun, dua peselamat sepuluh tahun, dan satu peselamat yang dipen- jarakan LP, menjalani hidup di penjara selama dua tahun. Kehidupan penjara dimaknai se- bagai kerja paksa. Peselamat diposisikan sebagai sumber daya manusia, yang tenaganya dieksploitasi untuk kepentin- gan kerja tertentu. Ini artinya, penem- patan tahanan politik PKI di Pulau Buru dalam kacamata peselamat bukanlah proyek rehabilitasi mental, tetapi seba- gai proyek membuka Pulau Buru, untuk kepentingan ekonomi. Pemaknaan ini dibangun peselamat yang ditahan Kehidupan Penjara dalam Pandangan Peselamat Setelah dilakukan penangkapan, pemeriksaan, dan penahanan, langkah berikutnya adalah negara memenjara- kan peselamat. Perjalanan menuju pen- jara secara permanen cukup panjang, dan masing-masing peselamat memiliki pengalaman yang berbeda-beda. Secara Tabel 2 Keberadaan Peselamat dalam Penjara Tabel 2 Keberadaan Peselamat dalam Penjara No. Peselamat Gol. Penjara Terakhir Aktivitas Lama di Pen- jara 1 2 3 4 5 6 7 Kustini Sunari Sariatun Sumito Suparno Abdullla Wiratmo B B B B C B B Plantungan Buru Plantungan Buru LP Buru Buru Membuat Kerajinan Bertani Membuat Kerajian Administrasi Adminstrasi Bertani Bertani 13 tahun 11 tahun 14 tahun 10 tahun 2 tahun 10 tahun 13 tahun Tabel 2 Keberadaan Peselamat dalam Penjara 204 : 131—248 Tindakan Negara terkait Peristiwa G 30 S - Hamdan Tri Atmaja karena melindungi orang-orang dari pengejaran petugas dan massa. Pe- maknaan ini tidak jauh berbeda dengan yang dialami peselamat yang ditahan karena dirinya menjadi anggota PKI. Menurutnya, memang benar kehidupan dipenjara seperti kerja paksa. Tahanan diperintahkan untuk kerja, kemudian diawasi secara ketat, dan tahanan merasakan hanya punya waktu longgar saat istirahat siang dan tidur. Pengala- mannya di pulau Buru memperkuat ar- gumentasinya, bahwa peselamat di pu- lau Buru bukanlah melakukan rehabili- tasi mental, tetapi lebih tepat kerja paksa. selamat perempuan. Negara mengang- gap peselamat perempuan sebagai Ger- wani yang berbahaya dan tidak ber- moral, sehingga perlu disingkirkan dari kehidupan keluarga, seperti suami dan anak pada khususnya, serta kehidupun masyarakat pada umumnya. SIMPULAN Berdasarkan temuan dalam penelitian ini, dapat disimpulkan bahwa tindakan yang dilakukan negara terha- dap peselamat adalah melakukan penangkapan, pemeriksaan, dan pena- hanan serta memenjarakan mereka ke pulau Buru (untuk peselamat laki-laki) dan Plantungan (untuk peselamat per- empuan). Semua ini dilakukan negara dengan anggapan, bahwa mereka yang pernah bersentuhan dengan paham ko- munis terlibat dalam peristiwa G30S. Karena itu, negara melakukan penang- kapan terhadap mereka yang pernah menjadi anggota ormas PKI dan mereka yang secara langsung menjadi anggota PKI. Peselamat juga memaknai ke- hidupan penjara sebagai pembuangan peselamat dengan tujuan menying- kirkan dari kehidupan masyarakat. Dalam perspektif ini, negara memposisi- kan orang-orang PKI menjadi orang yang memiliki kemungkinan dapat meracuni dan menyebarkan ideologi komunis di tengah-tengah masyarakat. Pemaknaan ini dikonstruksikan oleh peselamat yang kebetulan berlatar be- lakang pegawai negeri dan pernah ikut Kursus Kader PKI, dan peselamat yang anggota PKI serta peselamat yang per- nah menjadi anggota CGMI. Semua atribut yang menempel pada diri pese- lamat, baik sebagai anggota PKI atau ormas PKI, menjadi keyakinan pada pe- selamat, bahwa dirinya memang harus dibuang ke Pulau Buru, dalam rangka membersihkan PKI sampai ke akar- akarnya. Keberagaman makna tentang peristiwa G30S yang dibangun pese- lamat menunjukkan titik persamaaan yang dirasakan oleh seluruh peselamat, bahwa peselamat merasa tidak menjadi bagian dalam peristiwa tersebut atau dengan kata lain peselamat tidak terli- bat dalam peristiwa G30S. Baik pese- lamat yang menjadi anggota PKI, ang- gota ormas PKI, simpati pada figur Soekarno dan yang tidak berlatar be- lakang ideologi apapun tidak merasa dirinya terlibat dalam peristiwa G30S. Peristiwa G30S benar-benar persoalan para elite politik. Kalau memang benar PKI dianggap terlibat dalam peristiwa tersebut, itu secara resmi bukan kebija- kan partai, tetapi kebijakan para elite politik PKI. Dalam perspektif peselamat per- empuan, kehidupan di penjara di- maknai sebagai bentuk hukuman deng- an menstigmatisasi dan mengisolasi kaum perempuan dari kehidupan masyarakat. Pada konteks ini, perem- puan yang berhaluan komunis dibunuh secara sistematis jatidiri dan karak- ternya. Stigmatisasi Gerwani yang berkonotasi negatif dilekatkan pada pe- Berkaitan dengan keberagaman 205 Paramita Vol. 22, No. 2 - Juli 2012: 131—248 makna tentang penangkapan dan pe- meriksan menunjukkan adanya per- spektif yang sama pada peselamat, bahwa semua tuduhan yang ditujukan pada peselamat tidak benar. Kalau dalam pemeriksaan peselamat harus mengakui semua tuduhan tersebut, hal itu disebabkan faktor keterpaksaan, karena peselamat merasa tidak mampu menahan siksaaan demi siksaan. SIMPULAN Semen- tara itu, keberagaman makna dalam ke- hidupan penjara, menunjukkan adanya pengalaman yang sama dari seluruh peselamat, bahwa kehidupan penjara merupakan kerja paksa dan upaya ne- gara menjauhkan peselamat dari ke- hidupan masyarakat serta stigmatisasi terhadap peselamat sebagai tahanan politik. Demokrano, A. Gumelar. 2006. Dari Kalong sampai Pulau Buru, 11 tahun dalam Sekapan, Penjara, Pembuangan, Dan Kerja Rodi. Yogyakarta: Pusat Sejarah dan Etika Politik. Foucoult, Michel. 1997. Sex dan Kekuasaan: Sejarah Sexualitas. Penerjemah Rahayu S. Hidayat.Jakarta: Gramedia. Foucoult, Michel. 1997. Sex dan Kekuasaan: Sejarah Sexualitas. Penerjemah Rahayu S. Hidayat.Jakarta: Gramedia. Gadamer, Hans-Georg. 2004. Kebenaran dan Metode. Pengantar Filsafat Herme- neutika. Penerjemah Ahmad Sahidah. Yogyakarta: Pustaka Pelajar. Krisnadi, G. 2000. Tahanan Politik Pulau Buru 1969-1979. Jakarta: LP3S. Manafe, Aco. 2008. Teperpu, Mengungkap Pengkhianatan PKI Pada Tahun 1965 Dan Proses Hukum Bagi Para Pelakunya. Jakarta: PT Pustaka Sinar Harapan. Notosusanto, Nugroho dan Saleh, Ismail. 1966. The Coup Attempt of the September 30 Movement in Indonesia. Jakarta: Pembimbing Massa. p Dari seluruh pemaknaan yang di- bangun peselamat, ada konsistensi pe- maknaan, yakni konsistensi bahwa se- luruh tindakan negara terkait dengan peristiwa G30S, sebagai negativitas ke- kuasaan. Negativitas tersebut menyang- kut sikap atau perilaku negara yang dirasakan peselamat, seperti stigma- tisasi, prasangka, trauma, teror atau an- caman, politik kambing hitam, dan hiperpobia. Keberagaman pemaknaan yang dibangun juga menunjukkan penyingkapan “Ada” pada diri pese- lamat yang merasa dirinya sebagai korban, dan penyingkapan “Ada” ne- gara, yang menganggap peselamat seba- gai “Yang lain”. Ketika peselamat di pandang “Yang Lain” oleh negara, maka “Yang Lain” dianggap berbeda dan bukan anggota kelompok. Oleh karena itu “Yang Lain” bisa dianggap sebagai ancaman. Parlindungan, Utan. 2007. Musik dan Politik: Genjer-Genjer, Kuasa dan Kontestasi Makna. Yogyakarta: Laboraorium Ju- rusan Ilmu Pemerintahan Fisipol UGM. Raharjo, Mudjia. 2008. Dasar-Dasar Herm- neutika Antara Intensionalisme dan Gadamerian, Jogyakarta: Ar-Ruzz Me- dia. Rossa, John dkk. 2004. Tahun Yang Tak Per- nah Berakhir: Memahami Pengalamam Korban. Jakarta: Elsam. Shantall, Teria. 1999. “The Experience Mean- ing In Suffering Among Holocoust Survivors. Journal of Humanistic Psy- chology, vol 39 No. 3 Summer 1999. Sparringa, Daniel T. 2000. Kumpulan Bahan Mata Ajaran Metode Penelitian Kualita- tif. Surabaya: FISIP UNAIR. f y Supelli, Karlina. 2001. “Kisah Dialektika Kaum Korban”. Kompas, Februari 2001. --------. 2005. “Mendengarkan Suara Kesuny- ian”. Dalam Mochtar Buchori dan Frans Magnis-Suseno (ed). Etika Politik Dalam Konteks Indonesia. Yogya- karta: Kanisius. DAFTAR PUSTAKA Wieringa, Saskia Eleonora. 1999. Penghan- curan Gerakan Perempuan di Indonesia. Penerjemah Hesri Setiawan. Jakarta: Garba Budaya. Butalia, Urvashi. 2002. Sisi Balik Senyap; Suara-Suara Dari Pemisahan India. Alih Bahasa Landung R Simatupang. Ma- gelang: Indonesia Tera. 206
https://openalex.org/W2554414440	https://boris.unibe.ch/93475/1/2016_Jakob_PubMed%2027839517.pdf	English	null	The impact of extracerebral organ failure on outcome of patients after cardiac arrest: an observational study from the ICON database	Critical care	2,016	cc-by	8,549	© The Author(s). 2016 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. * Correspondence: jlvincent@intensive.org 1Department of Intensive Care, Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium Full list of author information is available at the end of the article The impact of extracerebral organ failure on outcome of patients after cardiac arrest: an observational study from the ICON database Leda Nobile1, Fabio S. Taccone1, Tamas Szakmany2,10, Yasser Sakr3, Stephan M. Jakob4, Tommaso Pellis5, Massimo Antonelli6, Marc Leone7, Xavier Wittebole8, Peter Pickkers9, Jean-Louis Vincent1* and on behalf of the ICON Investigators Open Access RESEARCH Open Access The impact of extracerebral organ failure on outcome of patients after cardiac arrest: an observational study from the ICON database Leda Nobile1, Fabio S. Taccone1, Tamas Szakmany2,10, Yasser Sakr3, Stephan M. Jakob4, Tommaso Pellis5, Massimo Antonelli6, Marc Leone7, Xavier Wittebole8, Peter Pickkers9, Jean-Louis Vincent1* and on behalf of the ICON Investigators The impact of extracerebral organ failure on outcome of patients after cardiac arrest: an observational study from the ICON database Nobile et al. Critical Care (2016) 20:368 DOI 10.1186/s13054-016-1528-6 Nobile et al. Critical Care (2016) 20:368 DOI 10.1186/s13054-016-1528-6 Abstract Background: We used data from a large international database to assess the incidence and impact of extracerebral organ dysfunction on prognosis of patients admitted after cardiac arrest (CA). Methods: This was a sub-analysis of the Intensive Care Over Nations (ICON) database, which contains data from all adult patients admitted to one of 730 participating intensive care units (ICUs) in 84 countries from 8–18 May 2012, except admissions for routine postoperative surveillance. For this analysis, patients admitted after CA (defined as those with “post-anoxic coma” or “cardiac arrest” as the reason for ICU admission) were included. Data were collected daily in the ICU for a maximum of 28 days; patients were followed up for outcome data until death, hospital discharge, or a maximum of 60 days in-hospital. Favorable neurological outcome was defined as alive at hospital discharge with a last available neurological Sequential Organ Failure Assessment (SOFA) subscore of 0–2. Methods: This was a sub-analysis of the Intensive Care Over Nations (ICON) database, which contains data from all adult patients admitted to one of 730 participating intensive care units (ICUs) in 84 countries from 8–18 May 2012, except admissions for routine postoperative surveillance. For this analysis, patients admitted after CA (defined as those with “post-anoxic coma” or “cardiac arrest” as the reason for ICU admission) were included. Data were collected daily in the ICU for a maximum of 28 days; patients were followed up for outcome data until death, hospital discharge, or a maximum of 60 days in-hospital. Favorable neurological outcome was defined as alive at hospital discharge with a last available neurological Sequential Organ Failure Assessment (SOFA) subscore of 0–2. Results: Among the 469 patients admitted after CA, 250 (53 %) had had out-of-hospital CA; 210 (45 %) patients died in the ICU and 357 (76 %) had an unfavorable neurological outcome. Non-survivors had a higher incidence of renal (43 vs. 16 %), cardiovascular (56 vs. 45 %), and respiratory (62 vs. 48 %) failure on admission and during the ICU stay than survivors (all p < 0.05). Similar results were found for patients with unfavorable vs. favorable neurological outcomes. In multivariable analysis, independent predictors of ICU mortality were renal failure on admission, high admission Simplified Acute Physiology Score (SAPS) II, high maximum serum lactate levels within the first 24 h after ICU admission, and development of sepsis. Data collection Demographics and comorbid diseases (including chronic obstructive pulmonary disease, solid or hematologic cancer, liver cirrhosis, heart failure, acquired immunodeficiency syndrome, chronic renal failure, immunosuppression, se- vere malnutrition, and insulin-dependent diabetes mellitus) were collected on admission. Clinical and laboratory data for the Simplified Acute Physiology Acore (SAPS) II [13] and the Acute Physiology and Chronic Health Evaluation (APACHE) II [14] scores were reported as the worst values within the first 24 h after ICU admission. Microbiological and clinical infections were reported daily. A daily assessment of organ function was per- formed using the SOFA score [15]. Abstract Independent predictors of unfavorable neurological outcome were mechanical ventilation on admission, high admission SAPS II score, and neurological dysfunction on admission. Results: Among the 469 patients admitted after CA, 250 (53 %) had had out-of-hospital CA; 210 (45 %) patients died in the ICU and 357 (76 %) had an unfavorable neurological outcome. Non-survivors had a higher incidence of renal (43 vs. 16 %), cardiovascular (56 vs. 45 %), and respiratory (62 vs. 48 %) failure on admission and during the ICU stay than survivors (all p < 0.05). Similar results were found for patients with unfavorable vs. favorable neurological outcomes. In multivariable analysis, independent predictors of ICU mortality were renal failure on admission, high admission Simplified Acute Physiology Score (SAPS) II, high maximum serum lactate levels within the first 24 h after ICU admission, and development of sepsis. Independent predictors of unfavorable neurological outcome were mechanical ventilation on admission, high admission SAPS II score, and neurological dysfunction on admission. ource: https://doi.org/10.7892/boris.93475 Conclusions: In this multicenter cohort, extracerebral organ dysfunction was common in CA patients. Renal failure on admission was the only extracerebral organ dysfunction independently associated with higher ICU mortality. * Correspondence: jlvincent@intensive.org 1Department of Intensive Care, Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium Full list of author information is available at the end of the article Nobile et al. Critical Care (2016) 20:368 Page 2 of 10 Page 2 of 10 Page 2 of 10 Definitions Information on whether extracerebral organ dysfunc- tion influences the outcome of patients after CA could open new lines of research, in particular related to clin- ical management and development of strategies to pre- vent such complications. The objectives of this study were, therefore, to assess the incidence of extracerebral organ failure in patients resuscitated from CA and its impact on prognosis, both in terms of ICU mortality and neurological outcome. For this purpose, we ana- lyzed data from a contemporary, international data- base of ICU patients—the Intensive Care Over Nations (ICON) audit [12]. Most of the definitions have been provided elsewhere [12]. Briefly, infection was defined in accordance with the International Sepsis Forum definitions [16]. Sepsis was defined as the presence of infection with the con- comitant occurrence of at least one organ failure (i.e., a SOFA subscore for the organ in question of >2) [17]. Septic shock was defined as sepsis complicated by car- diovascular failure. For the purposes of this sub-study, patients with “post-anoxic coma” or “cardiac arrest” listed as the reason for ICU admission were considered as having been admitted after CA. CA was considered to have occurred out-of-hospital (OHCA) if patients were admitted through the emergency department and/or by ambulance; all other patients were considered to have had an in-hospital CA (IHCA). Patients in whom the lowest body temperature during the first day after ICU Inclusion criteria Each center prospectively collected data on all adult pa- tients (>16 years of age) who were admitted to their ICU from 8–18 May 2012; patients who remained in the ICU for <24 h (i.e., for routine post-operative surveillance) or who were readmitted were not included. Data were col- lected daily during the ICU stay for a maximum of 28 days, using electronic case report forms through a se- cured web-based platform. Patients were followed up for outcome data until death or hospital discharge, which- ever came first. Decisions regarding withdrawal of care were made according to local practices; patients in whom a decision was made to withdraw care were con- sidered in the final analysis. g There are limited data available on the extent and prognostic value of extracerebral organ dysfunction after CA. In a recent report [8], almost all (96 %) of the 203 patients resuscitated after CA had some degree of organ dysfunction, in particular cardiovascular and respiratory impairments; two-thirds of these patients had dysfunc- tion of at least two extracerebral organ systems. Only al- terations in the cardiovascular and respiratory systems, as assessed by Sequential Organ Failure Assessment (SOFA) subscores, were independently associated with in-hospital mortality. Renal dysfunction has also been re- ported as an independent prognostic factor for mortality among CA survivors [9], although conflicting data have been reported [10]. There are almost no data on coagu- lation dysfunction in this setting, and only one study de- scribed the occurrence of hypoxic hepatitis, but not liver dysfunction, in 11 % of CA survivors, which was also as- sociated with increased intensive care unit (ICU) mortal- ity [11]. Participating centers g Rates of sudden cardiac arrest (CA) vary around the globe, but it is estimated that the incidence is around 55 events per 100,000 person-years [1]. Overall outcomes remain poor, with less than 10 % of patients leaving the hospital alive with good neurological recovery [1]. Brain damage, exacerbated by global ischemia-reperfusion in- jury, is the leading cause of death [2–4]. Therapeutic hypothermia (TH) has been employed following return of spontaneous circulation (ROSC) to limit the extent of brain damage, although concerns remain regarding opti- mal temperature levels, target populations, and duration of cooling [5, 6]. The so-called “post-cardiac arrest syn- drome” links ischemia-reperfusion injury with brain dam- age, myocardial dysfunction, and a systemic inflammatory response that has remarkable similarities to that of sepsis and may result in the development of multi-organ failure (MOF), regardless of whether or not TH is used [7]. Participating centers were recruited by open invitation, through national scientific societies, international meet- ings, and/or individual contacts. Participation was volun- tary with no financial reimbursement. The participating institutions obtained local ethical approval. Methods Full details of methodology have been provided previ- ously [12] and a list of participating ICUs is given in Additional file 1 (Appendix 1). Nobile et al. Critical Care (2016) 20:368 Page 3 of 10 Page 3 of 10 admission was <34 °C were considered to have received TH. At the time of the ICON audit, the targeted temperature management (TTM) study [6], which sug- gested similar effects on outcome for a cooling strategy using 33 °C or 36 °C, had not yet been published and rou- tine practice was to target a temperature of 32–34 °C [18]. Patients were considered as “comatose” (central nervous system (CNS)-SOFA >2) or “non-comatose” (CNS-SOFA 0–2) on admission. ICU mortality rates and overall neuro- logical outcomes were collected: patients who were alive at hospital discharge and in whom the last recorded CNS- SOFA subscore was between 0 and 2 (corresponding to a Glasgow Coma Scale (GCS) score of 10–15) were defined as having a favorable neurological outcome; other patients (non-survivors and survivors with CNS-SOFA of 3–4, i.e., GCS <10) were defined as having an unfavorable neuro- logical outcome. considered as having had a CA. The mean patient age was 66 (52–77) years, and 282 (61 %) patients were male (Table 1). The CA occurred out-of-hospital in 250 (53 %) patients. On admission, the median SAPS II score was 60 (46–75) and the SOFA score was 10 (7–13); 337 (72 %) patients were comatose. A total of 210 (45 %) pa- tients died during the ICU stay and 357 (76 %) had an unfavorable neurological outcome; decisions to limit therapy were made in 170 (36 %) patients. admission was <34 °C were considered to have received TH. At the time of the ICON audit, the targeted temperature management (TTM) study [6], which sug- gested similar effects on outcome for a cooling strategy using 33 °C or 36 °C, had not yet been published and rou- tine practice was to target a temperature of 32–34 °C [18]. g Patients were considered as “comatose” (central nervous system (CNS)-SOFA >2) or “non-comatose” (CNS-SOFA 0–2) on admission. Methods ICU mortality rates and overall neuro- logical outcomes were collected: patients who were alive at hospital discharge and in whom the last recorded CNS- SOFA subscore was between 0 and 2 (corresponding to a Glasgow Coma Scale (GCS) score of 10–15) were defined as having a favorable neurological outcome; other patients (non-survivors and survivors with CNS-SOFA of 3–4, i.e., GCS <10) were defined as having an unfavorable neuro- logical outcome. Mean arterial pressure, pH, and PaO2/FiO2 ratio on the first day were lower, and maximal lactate level higher, in non-survivors than in survivors (Table 1). Non-survivors also had higher SAPS II and SOFA scores on admission, and were more frequently treated with va- sopressors. Similar patterns were found when patients with favorable and unfavorable neurological outcomes were compared (Table 1). Non-survivors had a greater incidence of renal, re- spiratory, and cardiovascular failure than survivors on admission (43 vs. 16 %, 56 vs. 45 %, 62 vs. 48 %, respect- ively; all p < 0.05; Fig. 1) and at any time during the ICU stay (71 % vs. 50 %, 71 vs. 54 %, 75 vs. 57 %, respectively; all p < 0.05; Fig. 1). Similar patterns were found in pa- tients with unfavorable and favorable neurological out- comes (Additional file 1: Figure S1). Patients with OHCA had a greater incidence of hematologic failure than those with IHCA, both on ICU admission (7 % vs. 2 %; p = 0.04) and during the ICU stay (14 % vs. 7 %; p = 0.009; Additional file 1: Figure S2); the occurrence of other organ failures was similar. There was considerable variability in the occurrence of organ failures in different geographical areas: patients from Oceania had the highest incidence of hepatic failure on admission, whereas respira- tory failure was significantly more frequent in patients in Africa and Europe, and hematologic failure occurred more frequently in patients in Africa and Asia (Additional file 1: Figure S3). There were no statistically significant differences in the incidence of organ failure when patients were ana- lyzed according to the GNI (Additional file 1: Figure S4). Statistical analysis y Data are expressed as mean ± SD, median (interquartile range) or count (percentage), as appropriate. For con- tinuous variables, normality assumption checking was performed by inspection of residual and normal plots. Differences between groups were assessed using the ana- lysis of variance (ANOVA), Kruskal-Wallis test, Student’s t test, Mann–Whitney test, χ2 test, or Fisher’s exact test, as appropriate. The occurrence of organ failure was also analyzed according to the location of the arrest (IHCA vs. OHCA), the geographical region (Africa, Europe, Asia, Oceania, and America), and the gross national in- come (GNI) per person (≤US$4035 was defined as low and lower-middle income; $4036–$12,475 as upper- middle income, and >$12,476 as high income) [12]. The time-courses of each SOFA subscore in survivors and non-survivors or in patients with favorable or unfavor- able neurological outcome were analyzed using general- ized estimating equation models. Multivariable logistic regression was used to identify independent predictors of ICU death and of unfavorable neurological outcome. Variables with p < 0.2 in the univariate analysis were considered in the multivariable analyses. Colinearity be- tween variables was excluded prior to modeling. Interac- tions between explanatory variables were also checked. The deviance of the logistic regression model and de- viance and partial residuals were used to check for model adequacy. All reported p values are two-sided and p < 0.05 was considered to indicate statistical signifi- cance. Data were analyzed using IBM® SPSS® Statistics soft- ware, version 22 for Windows (IBM, Armonk, NY, USA). The time-course of several of the SOFA subscores dif- fered in survivors and non-survivors. In particular, the hepatic-SOFA subscore increased over time in the non- survivors but not in the survivors (p = 0.015; Additional file 1: Figure S5), whereas CNS- and cardiovascular- SOFA subscores decreased over time in the survivors but not in the non-survivors (both p < 0.001; Additional file 1: Figures S6 and S7). The renal-SOFA subscore did not change over time in the survivors, but was higher in the non-survivors and decreased progressively over time (p < 0.001, Fig. 2). The time-course of respiratory- and hematologic-SOFA subscores was similar in survivors and non-survivors (Additional file 1: Figures S8 and S9). Renal-SOFA subscores decreased over time in patients with a favorable but not in those with an unfavorable Results Among the 10,069 patients included in the ICON regis- try, 469 had post-anoxic coma (n = 62) or CA (n = 407) as the reported reason for ICU admission and were thus Nobile et al. Results Critical Care (2016) 20:368 Page 4 of 10 Table 1 Characteristics of the study population, according to ICU survival and neurological outcome at hospital discharge All patients (n = 469) Survivors (n = 247) Non-survivors (n = 210) Favorable neurological outcome (n = 97) Unfavorable neuro outcome (n = 357) Age, years 66 (52–77) 65 (52–75) 68 (52–78) 67 (53–75) 66 (52–77) Male, n (%) 282 (61) 149 (61) 126 (61) 60 (62) 222 (60) Weight, kg 75 (65–87) 75 (66–90) 71 (61–83) 75 (65–90) 74 (65–85) OHCA, n (%) 250 (53) 120 (49) 124 (59) 34 (35) 216 (58) IHCA, n (%) 219 (47) 127 (51) 86 (41) 63 (65) 156 (42) No co-morbidities, n (%) 209 (44) 101 (41) 108 (51) 49 (51) 163 (44) COPD/asthma, n (%) 65 (14) 39 (16) 26 (12) 17 (17) 51 (14) Heart failure, n (%) 72 (15) 30 (12) 42 (20) 10 (10) 62 (17) Diabetes, n (%) 45 (10) 20 (8) 25 (12) 12 (12) 33 (9) Cancer, n (%) 38 (8) 18 (7) 20 (10) 10 (10) 28 (8) Chemotherapy, n (%) 9 (2) 4 (2) 5 (2) 1 (1) 8 (2) Liver cirrhosis, n (%) 14 (3) 1 (1) 13 (6) 1 (1) 13 (3) Chronic renal failure, n (%) 50 (11) 21 (9) 29 (14) 10 (10) 40 (11) HIV, n (%) 1 (0) – 1 (1) – 1 (1) Corticosteroids, n (%) 14 (3) 6 (2) 8 (4) 1 (1) 13 (3) Medical admission, n (%) 373 (83) 189 (91) 184 (89) 70 (74) 303 (84) SAPS II score on admission 60 (46–75) 53 (39–64) 71 (57–81)* 38 (29–49) 65(54–79)* SOFA score on admission 10 (7–13) 8 (6–11) 12 (9–14)* 6 (3–8) 11(8–13)* Infections Infection on admission, n (%) 112 (24) 60 (24) 52 (25) 20 (21) 92 (25) Infection any time, n (%) 200 (43) 120 (49) 80 (38) 41 (41) 159 (43) Septic shock on admission, n (%) 75 (16) 40 (16) 35 (17) 11 (11) 64 (17) Septic shock any time, n (%) 133 (28) 74 (30) 59 (28) 20 (21) 113 (30) Vasopressors on admission, n (%) 262 (57) 124 (50) 138 (66)* 42 (43) 220 (62)* Mechanical ventilation on admission, n (%) 408 (87) 205 (83) 203 (97) 67 (69) 341 (92)* Mechanical ventilation any time, n (%) 425 (91) 211 (85) 204 (97) 71 (73) 354 (95)* Hemodialysis/CRRT on admission, n (%) 38 (8) 19 (8) 19 (9) 4 (4) 34 (9) Hemodialysis/CRRT any time, n (%) 103 (23) 59 (24) 44 (21) 17 (17) 86 (23) MAP maximum on first day, mmHg 99 (87–112) 100 (89–113) 97 (85–111) 100 (88–113) 98 (87–112) MAP minimum on first day, mmHg 61 (50–70) 63 (55–71) 56 (45–68) 63 (55–79) 60 (49–70)* pH maximum on first day, mmHg 7.40 (7.30–7.45) 7.41 (7.35–7.46) 7.37 (7.23–7.44)* 7.42 (7.36–7.47) 7.39 (7.28–7.45)* pH minimum on first day, mmHg 7.25 (7.11–7.35) 7.29 (7.19–7.36) 7.18 (7.05–7.29)* 7.32 (7.23–7.38) 7.23 (7.10–7.33)* PaCO2 maximum on first day, mmHg 45 (37–57) 44 (36–55) 46 (37–60) 45 (37–53) 45 (36–60) PaCO2 minimum on first day, mmHg 34 (28–40) 34 (28–40) 33 (27–39) 35 (29–40) 33 (28–40) PaO2 maximum on first day, mmHg 144 (98–229) 135 (94–197) 149 (101–265) 141 (95–190) 145 (99–234) PaO2 minimum on first day, mmHg 77 (61–98) 77 (64–94) 75 (57–105) 78 (66–90) 76 (60–100) PaO2/FiO2 maximum on first day 232 (144–344) 232 (151–337) 233 (132–361) 270 (161–374) 227 (143–343)* PaO2/FiO2 minimum on first day 195 (132–288) 200 (149–294) 186 (120–285)* 218 (165–300) 192 (124–285)* Lactate maximum on first day, mmol/L 3.6 (2.0–7.6) 2.6 (1.5–5.5) 5.8 (2.5–10.5)* 2.1 (1.2–5.5) 4.30 (2.3–8)* ICU stay, days 4 (2–8) 5 (3–10) 3 (1–6)* 4 (2–6) 4 (1–8) Hospital stay days 8 (2–20) 17 (8–31) 3 (1–6)* 15 (8–22) 5 (2–16)* vival and neurological outcome at hospital discharge Page 5 of 10 Nobile et al. Results Critical Care (2016) 20:368 Table 1 Characteristics of the study population, according to ICU survival and neurological outcome at hospital discharge (Continued) Limitation of care, n (%) 170 (36) 40 (16) 126 (61)* 11 (11) 159 (44)* ICU mortality, n (%) 210 (45) – 210 (100)* – 210 (59) Hospital mortality, n (%) 247 (54) 37 (15) 210 (100)* 11 (11) 236 (67)* Values are given as median (interquartile range) unless otherwise stated Missing values: 12 for mortality; 5 for sex; 36 for weight; 23 for source of admission; 15 for vasopressors on admission; 15 for ICU stay, 16 for hospital stay p < 0.05, survivors versus non-survivors; favorable outcome versus unfavorable outcome COPD chronic obstructive pulmonary disease, CRRT continuous renal replacement therapy, FiO2 fraction of inspired oxygen, HIV human immunodeficiency virus, ICU intensive care unit, IHCA in-hospital cardiac arrest, MAP mean arterial pressure, OHCA out-of-hospital cardiac arrest, PaO2 partial pressure of oxygen, PaCO2 partial pressure of carbon dioxide, SAPS Simplified Acute Physiology Score, SOFA Sequential Organ Failure Assessment Table 1 Characteristics of the study population, according to ICU survival and neurological outcome at hospital discharge (Continued) Table 1 Characteristics of the study population, according to ICU survival and neurological outcome at hospital discharge (Continued) Limitation of care n (%) 170 (36) 40 (16) 126 (61) 11 (11) 159 (44)* cs of the study population, according to ICU survival and neurological outcome at hospital discharge p < 0.05, survivors versus non-survivors; favorable outcome versus unfavorable outcome COPD chronic obstructive pulmonary disease, CRRT continuous renal replacement therapy, FiO2 fraction of inspired oxygen, HIV human immunodeficiency virus, ICU intensive care unit, IHCA in-hospital cardiac arrest, MAP mean arterial pressure, OHCA out-of-hospital cardiac arrest, PaO2 partial pressure of oxygen, PaCO2 partial pressure of carbon dioxide, SAPS Simplified Acute Physiology Score, SOFA Sequential Organ Failure Assessment Fig. 1 Occurrence of extracerebral organ failure on intensive care unit (ICU) admission (upper panel) and during the ICU stay (lower panel) in survivors and non-survivors. p < 0.05 Nobile et al. Critical Care (2016) 20:368 Page 6 of 10 Fig. 2 Time-course of renal-SOFA subscore in survivors and non-survivors (upper panel) and in patients with favorable (FO) and unfavorable neurological outcome (UO) (lower panel). SOFA Sequential Organ Failure Assessment Fig. 2 Time-course of renal-SOFA subscore in survivors and non-survivors (upper panel) and in patients with favorable (FO) and unfavorable neurological outcome (UO) (lower panel). Results It is difficult to compare these findings as different confounders, including measured outcomes (e.g., hospital survival vs. neurological outcome), the use of TH, and the proportion of patients with OHCA, may have influenced the results. Moreover, evaluation of the impact of RRT on the prognosis of CA patients may also be influenced by local practice regarding renal support; for example, the use of RRT for correction of electrolyte disturbances and/or fluid overload in pa- tients without overt AKI may be associated with better re- covery and outcome than RRT initiated for severe renal failure. In addition, as a number of patients with extensive brain injury undergo limitation of life-sustaining therapies, regardless of the development of AKI, it is not always pos- sible to assess the impact of AKI per se on outcome. In our study, renal failure was a more significant determinant of outcome than cardiovascular or respiratory dysfunction. Our data differ from those of Roberts et al. [8] who re- ported that extracerebral organ dysfunction was common after CA but found that only cardiovascular dysfunction and altered gas exchange were associated with outcome. However, that study [8] was conducted in a single US aca- demic hospital over several years, whereas in our multi- center international audit, data were collected over a short time period. Moreover, Roberts et al. [8] did not record blood lactate levels and we observed that increased lactate concentrations after CA were significantly associated with ICU mortality. As high lactate concentrations are primar- ily a consequence of prolonged CA and/or severe subse- quent hemodynamic impairment [22–24], lactate can be considered as an extracerebral variable that is predictive of poor outcome in these patients. Abnormalities in tissue perfusion occurring after CA may also potentially contrib- ute to brain hypoperfusion and development of MOF [7]. Thus, monitoring of lactate in this setting may be of more value to assess the severity of tissue hypoxia than just blood pressure or cardiac output. Other studies have also shown the prognostic value of admission blood lactate levels after CA and of changes in lactate levels in the hours after CA [22–25]. In this large database, patients admitted after CA rep- Fig. 3 Occurrence of circulatory, respiratory, and renal failure in patients according to their outcome (alive or dead, with neurological recovery (CNS+) or persistent neurological impairment (CNS–)) were independent predictors of unfavorable neurological outcome (Table 2). Results SOFA Sequential Organ Failure Assessment Fig. 2 Time-course of renal-SOFA subscore in survivors and non-survivors (upper panel) and in patients with favorable (FO) and unfavorable neurological outcome (UO) (lower panel). SOFA Sequential Organ Failure Assessment outcome (p = 0.02; Fig. 2). Similar results were found for the CNS-SOFA subscore (Additional file 1: Figure S10). The analysis of other SOFA subscores showed a similar temporal trend between patients with favorable and un- favorable neurological outcome (Additional file 1: Figures S11–S14). improved neurological function during the ICU stay. These non-survivors with good neurological function more frequently had renal failure than did survivors (p < 0.001 versus survivors with good neurological function and versus survivors with poor neurological function); they also more frequently had cardiovascular failure (p = 0.026) and respiratory failure (p = 0.094) than the ICU survivors with good neurological function (Fig. 3). Among the ICU survivors, 150 (57 %) patients had an unfavorable neurological outcome. Thirty-three (16 %) of the non-survivors had good neurological function (CNS-SOFA score of 0–2) before death (Additional file 1: Figure S15); however, the CNS-SOFA score on admis- sion was also 0–2 in all these patients (0 in 19 patients, 1 in 3 patients and 2 in 11 patients); no patient who was comatose on admission and eventually died had In the multivariable analysis, renal failure on admis- sion, high SAPS II score, high serum lactate levels within the first 24 h after ICU admission, and the development of sepsis were independent predictors of ICU mortality (Table 2). Mechanical ventilation on admission, high SAPS II score, and high CNS-SOFA score on admission Page 7 of 10 Nobile et al. Critical Care (2016) 20:368 Fig. 3 Occurrence of circulatory, respiratory, and renal failure in patients according to their outcome (alive or dead, with neurological recovery (CNS+) or persistent neurological impairment (CNS–)) existing renal dysfunction, older age, longer duration of resuscitation, and the presence of shock [9, 10, 19–21]. Severe acute kidney injury (AKI) was associated with sig- nificantly higher mortality at 30 days after CA in one study [9], but AKI was not an independent predictor of mortality or poor neurological outcome in two others [10, 20]. Interestingly, despite the association of renal failure on admission with increased mortality, the rate of renal replacement therapy (RRT) was not different between any of the groups in our study. Results None of the extracerebral organ fail- ures occurring during the ICU stay were independently associated with ICU mortality or neurological outcome. Discussion This international observational study showed that pa- tients with a poor outcome after CA had a higher inci- dence of renal, cardiovascular, and respiratory failure on admission or during the ICU stay than did patients with good outcomes. Renal failure on admission was an inde- pendent predictor of ICU death, as was the severity of disease, high serum lactate levels, and the development of sepsis. However, extracerebral organ failure occurring later during the ICU stay did not significantly influence ICU mortality or neurological outcome. Renal dysfunction has been reported in nearly 50 % of patients after CA, in particular in patients with pre- Table 2 Multivariable analysis to identify the independent Table 2 Multivariable analysis to identify the independent predictors of ICU mortality and unfavorable outcome Variable p value OR 95 % CI Table 2 Multivariable analysis to identify the independent predictors of ICU mortality and unfavorable outcome Variable p value OR 95 % CI Predictors of ICU mortality SAPS II score on admission <0.001 1.047 1.025–1.069 Lactate max, mEq/L 0.004 1.093 1.029–1.161 Renal failure on admission 0.011 2.413 1.220–4.774 Severe sepsis during ICU stay 0.022 0.537 0.316–0.912 Predictors of unfavorable outcome SAPS II on admission <0.001 1.107 1.077–1.138 MV on admission 0.020 3.787 1.234–11.628 CNS-SOFA on admission <0.001 4.237 3.097–5.796 CI confidence interval, CNS central nervous system, ICU intensive care unit, Lactate max maximal lactate levels within the first 24 h after ICU admission, MV mechanical ventilation, OR odds ratio, SAPS Simplified Acute Physiological Score, SOFA Sequential Organ Failure Assessment In this large database, patients admitted after CA rep- resented around 5 % of the ICU population, similar to values reported in other studies [26]. The SAPS II score was higher in patients with poor neurological outcome or those who died during their ICU stay than in the Nobile et al. Critical Care (2016) 20:368 Page 8 of 10 Page 8 of 10 post-CA disease, thus some key variables (e.g., initial rhythm, time to ROSC, or quality of CPR) were not cap- tured, which limited the degree of adjustment for outcome analyses. Second, our cohort included patients with “post- anoxic coma” and “cardiac arrest” as the reason for ICU admission; it is possible that patients with cerebral injury without a CA (e.g., after hanging) may have been included in the "post-anoxic coma" category. Discussion Moreover, the delay between the occurrence of CA and the presence of neuro- logical impairment was not available. Third, we focused on ICU and hospital mortality, but not on longer-term outcomes. In addition, the use of CNS-SOFA to assess neurological outcome in CA patients has several limita- tions and is not as accurate as other scores, e.g., the Ran- kin scale or the Cerebral Performance Scale (CPC), to quantify the extent of post-anoxic brain injury. Also, we could not evaluate cognitive dysfunction, which may re- flect subtle post-anoxic injury. Fourth, given the substan- tial differences in post-resuscitation care in the countries included in the ICON database, patient management, in- cluding healthcare systems, resources and medical sys- tems, was probably extremely variable among centers. Data about specific treatments (e.g., coronary angiog- raphy) were not available. Decisions regarding withdrawal of care certainly varied among regions. Nevertheless, this study is hypothesis-generating and may help to provide a focus on particular organ dysfunctions that should be ad- equately described in future prospective studies dealing with management and prognostication in such patients. Fifth, data on the duration of hypothermia and potential complications associated with TH were not available. Data on the development of fever were also not available. Sixth, data on “ventilator settings” during mechanical ventilation were not included in the analysis and we cannot comment on the roles of tidal volume, spontaneous breathing, and/ or exposure to high-inspired oxygen fraction on the mea- sured outcomes. Finally, the precise causes of death were not recorded other patients, although others have reported that the SAPS and APACHE scores are not good prognostic tools in patients with CA [27, 28]. Conversely, the highest extracerebral SOFA score at 72 h after CA was inde- pendently associated with in-hospital mortality [8]. Ad- mission factors that have been correlated with an increased SOFA score after CA include non-shockable rhythm, the amount of epinephrine used, use of TH, and elevated stress hormone levels [29]. In particular, the cardiovascular component of the SOFA score has been shown to accurately predict outcomes of patients with CA when combined with neurological examination [30]; our results suggest that the renal component should per- haps also be considered in prediction models. Mechanical ventilation was one of the independent pre- dictors of poor neurological outcome. Sutherasan et al. Conclusions In this multicenter cohort of CA patients, 210 (45 %) pa- tients died in the ICU and 357 (76 %) had an unfavorable neurological outcome. Extracerebral organ dysfunction was common, but renal failure on admission was the only extracerebral organ dysfunction independently associated with higher ICU mortality. Discussion showed that high tidal volume and plateau pressure with lower positive end-expiratory pressure were associated with the occurrence of severe pulmonary complications during the ICU stay in CA patients [31], suggesting a potential role of ventilator settings on outcome. Moreover, patients re- ceiving mechanical ventilation are more exposed to high oxygen levels or abnormal carbon dioxide concentrations, which have been shown to have a significant negative im- pact on brain recovery after post-anoxic injury [32, 33]. Not surprisingly, the severity of brain dysfunction on admission also predicted an unfavorable outcome. In a large cohort of CA patients in Japan, the initial GCS motor score was significantly associated with neuro- logical outcome, as were age, bystander cardiopulmonary resuscitation (CPR), the time from collapse to ROSC, and pupil size [2]. Initial coma was also one of the most important determinants of outcome in another large database of CA patients [30]. Another interesting finding was that 16 % of the patients who eventually died had relatively preserved cerebral func- tion during their ICU stay. In recent years, several studies have focused on identifying markers of poor neurological outcome as if mortality after resuscitated CA were exclu- sively associated with lack of neurological recovery [3, 4, 34]. Neurological recovery was often assessed 3 to 6 months after CA and all non-survivors were considered to have a poor neurological outcome. However, a significant propor- tion of CA patients die from protracted shock and MOF [4], which often occurs early after CA, before any possible neurological outcome assessment. In our study, patients with good neurological function after CA died from other reasons, in particular other organ failures. This finding highlights the importance of repeated neurological assess- ment of CA patients during the ICU stay because a single long-term assessment may underestimate the potential for neurological recovery in some patients. Author details 1 Additional file 1: Appendix 1. Alphabetical list of participating centers by region and country. Figure S1. Occurrence of extra-cerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients with favorable (FO) and unfavorable (UO) neurological outcomes. p < 0.05. Figure S2. Occurrence of extracerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients with in-hospital (IHCA) and out-of-hospital (OHCA) cardiac arrest. p < 0.05. Figure S3. Occurrence of extracerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients according to geographical area. Figure S4. Occurrence of extracerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients according to gross national income. Figure S5. Time course of hepatic-SOFA subscore in survivors and non-survivors. Figure S6. Time course of neurologic (CNS)-SOFA subscore in survivors and non-survivors. Figure S7. Time course of cardiovascular-SOFA subscore in survivors and non-survivors. Figure S8. Time-course of respiratory-SOFA subscore in survivors and non-survivors. Figure S9. Time-course of hematologic-SOFA subscore in survivors and non- survivors. Figure S10. Time-course of neurologic (CNS)-SOFA subscore in patients with favorable (FO) and unfavorable (UO) neurological outcome. Figure S11. Time-course of cardiovascular-SOFA subscore in patients with favorable (FO) and unfavorable (UO) neurological outcome. Figure S12. Time course of hematological-SOFA subscore in patients with favorable (FO) and unfavorable (UO) neurological outcome. Figure S13. Time-course of respiratory-SOFA subscore in patients with favorable (FO) and unfavorable (UO) neurological outcome. Figure S14. Time-course of hepatic-SOFA subscore in patients with favorable (FO) and unfavorable (UO) neurological outcome. Figure S15. Differences in the last available Additional file 1: Appendix 1. Alphabetical list of participating centers by region and country. Figure S1. Occurrence of extra-cerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients with favorable (FO) and unfavorable (UO) neurological outcomes. p < 0.05. Figure S2. Occurrence of extracerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients with in-hospital (IHCA) and out-of-hospital (OHCA) cardiac arrest. 1Department of Intensive Care, Erasme Hospital, Université Libre de Bruxelles, Brussels, Belgium. 2Department of Critical Care, Royal Gwent Hospital, Newport, Wales, UK. 3Department of Anesthesiology and Intensive Care, Uniklinikum Jena, Jena, Germany. 4Department of Intensive Care Medicine, University Hospital Bern, University of Bern, Bern, Switzerland. Ethics approval and consent to participate h b d Each participating institution obtained approval from its local ethics committee for the original ICON audit. Each participating institution obtained approval from its local ethics committee for the original ICON audit. Key messages Analysis of this large database revealed that renal, cardiovascular, and respiratory failure occurring on admission or during the ICU stay was more common in patients with an unfavorable outcome than in those with a favorable outcome. This study has several limitations. First, the ICON database was not designed specifically to record data on Nobile et al. Critical Care (2016) 20:368 Page 9 of 10 Renal failure on admission was the only extracerebral organ dysfunction associated with ICU mortality. Renal failure on admission was the only extracerebral organ dysfunction associated with ICU mortality. helped in the design of the study, and revised the current article for critical content. All authors read and approved the final manuscript. helped in the design of the study, and revised the current article for critical content. All authors read and approved the final manuscript. Extracerebral organ failure occurring later during the ICU stay did not significantly influence ICU mortality or neurological outcome. Author details 1 5Anesthesia and Intensive Care, Santa Maria degli Angeli Hospital, Pordenone, Italy. 6Department of Intensive Care and Anesthesiology, Università Cattolica del Sacro Cuore, Rome, Italy. 7Department of Anesthesia and Intensive Care, Hôpital Nord, AP-HM Aix Marseille Université, Marseille, France. 8Critical Care Department, Cliniques Universitaires St Luc, UCL, Brussels, Belgium. 9Department of Intensive Care, Nijmegen Institute for Infection, Inflammation and Immunity, Radboud University Nijmegen Medical Center, Nijmegen, The Netherlands. 10Department of Anaesthetics, Intensive Care and Pain Medicine, Division of Population Medicine, Cardiff University, Cardiff, UK. p < 0.05. Figure S3. Occurrence of extracerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients according to geographical area. Figure S4. Occurrence of extracerebral organ failure on ICU admission (upper panel) and during the ICU stay (lower panel) in patients according to gross national income. Figure S5. Time course of hepatic-SOFA subscore in survivors and non-survivors. Medicine, Division of Population Medicine, Cardiff University, Cardiff, UK. Medicine, Division of Population Medicine, Cardiff University, Cardiff, UK. Received: 20 May 2016 Accepted: 14 October 2016 References 1. Berdowski J, Berg RA, Tijssen JG, Koster RW. Global incidences of out-of- hospital cardiac arrest and survival rates: systematic review of 67 prospective studies. Resuscitation. 2010;81:1479–87. 1. Berdowski J, Berg RA, Tijssen JG, Koster RW. Global incidences of out-of- hospital cardiac arrest and survival rates: systematic review of 67 prospective studies. Resuscitation. 2010;81:1479–87. prospective studies. Resuscitation. 2010;81:1479–87. 2. Hifumi T, Kuroda Y, Kawakita K, Sawano H, Tahara Y, Hase M, et al. Effect of admission Glasgow Coma Scale motor score on neurological outcome in out-of-hospital cardiac arrest patients receiving therapeutic hypothermia. Circ J. 2015;79:2201–8. 2. Hifumi T, Kuroda Y, Kawakita K, Sawano H, Tahara Y, Hase M, et al. Effect of admission Glasgow Coma Scale motor score on neurological outcome in out-of-hospital cardiac arrest patients receiving therapeutic hypothermia. Circ J. 2015;79:2201–8. 3. Dragancea I, Rundgren M, Englund E, Friberg H, Cronberg T. The influence of induced hypothermia and delayed prognostication on the mode of death after cardiac arrest. Resuscitation. 2013;84:337–42. 3. Dragancea I, Rundgren M, Englund E, Friberg H, Cronberg T. The influence of induced hypothermia and delayed prognostication on the mode of death after cardiac arrest. Resuscitation. 2013;84:337–42. (UO) neurological outcome. Figure S15. Differences in the last available SOFA subscores in survivors (S) and non-survivors (NS). (PDF 1391 kb) 4. Lemiale V, Dumas F, Mongardon N, Giovanetti O, Charpentier J, Chiche JD, et al. Intensive care unit mortality after cardiac arrest: the relative contribution of shock and brain injury in a large cohort. Intensive Care Med. 2013;39:1972–80. Acknowledgments We would like to thank Hassane Njimi PhD for his help with the statistical analyses. Availability of data and materials 8. Roberts BW, Kilgannon JH, Chansky ME, Mittal N, Wooden J, Parrillo JE, et al. Multiple organ dysfunction after return of spontaneous circulation in postcardiac arrest syndrome. Crit Care Med. 2013;41:1492–501. The data that support the findings of this study are available from the corresponding author upon reasonable request. 9. Geri G, Guillemet L, Dumas F, Charpentier J, Antona M, Lemiale V, et al. Acute kidney injury after out-of-hospital cardiac arrest: risk factors and prognosis in a large cohort. Intensive Care Med. 2015;41:1273–80. Abbreviations A Scientific Statement from the International Liaison Committee on Resuscitation; the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiovascular Surgery and Anesthesia; the Council on Cardiopulmonary, Perioperative, and Critical Care; the Council on Clinical Cardiology; the Council on Stroke. Resuscitation. 2008;79:350–79. 7. Nolan JP, Neumar RW, Adrie C, Aibiki M, Berg RA, Bottiger BW, et al. Post- cardiac arrest syndrome: epidemiology, pathophysiology, treatment, and prognostication. A Scientific Statement from the International Liaison Committee on Resuscitation; the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiovascular Surgery and Anesthesia; the Council on Cardiopulmonary, Perioperative, and Critical Care; the Council on Clinical Cardiology; the Council on Stroke. Resuscitation. 2008;79:350–79. Funding There wa g There was no external funding for this study. Competing interests JLV Ed Ch f f JLV is Editor-in-Chief of Critical Care. The authors declare that they have no competing interests. JLV is Editor-in-Chief of Critical Care. The authors declare that they have no competing interests. Ethics approval and consent to participate Each participating institution obtained approval from its local ethics committee for the original ICON audit. Abbreviations AKI: Acute kidney injury; APACHE: Acute Physiology and Chronic Health Evaluation; CA: Cardiac arrest; CNS: Central nervous system; CPR: Cardiopulmonary resuscitation; GCS: Glasgow Coma Scale; GNI: Gross national income; ICON: Intensive Care Over Nations; ICU: Intensive care unit; IHCA: In-hospital cardiac arrest; MOF: Multi-organ failure; OHCA: Out-of- hospital cardiac arrest; ROSC: Return of spontaneous circulation; RRT: Renal replacement therapy; SAPS: Simplified Acute Physiology Score; SOFA: Sequential Organ Failure Assessment; TH: Therapeutic hypothermia 5. Bro-Jeppesen J, Kjaergaard J, Wanscher M, Nielsen N, Friberg H, Bjerre M, et al. The inflammatory response after out-of-hospital cardiac arrest is not modified by targeted temperature management at 33 degrees C or 36 degrees C. Resuscitation. 2014;85:1480–7. 6. Nielsen N, Wetterslev J, Cronberg T, Erlinge D, Gasche Y, Hassager C, et al. Targeted temperature management at 33 degrees C versus 36 degrees C after cardiac arrest. N Engl J Med. 2013;369:2197–206. after cardiac arrest. N Engl J Med. 2013;369:2197 206. 7. Nolan JP, Neumar RW, Adrie C, Aibiki M, Berg RA, Bottiger BW, et al. Post- cardiac arrest syndrome: epidemiology, pathophysiology, treatment, and prognostication. A Scientific Statement from the International Liaison Committee on Resuscitation; the American Heart Association Emergency Cardiovascular Care Committee; the Council on Cardiovascular Surgery and Anesthesia; the Council on Cardiopulmonary, Perioperative, and Critical Care; the Council on Clinical Cardiology; the Council on Stroke. Resuscitation. 2008;79:350–79. 8. Roberts BW, Kilgannon JH, Chansky ME, Mittal N, Wooden J, Parrillo JE, et al. Multiple organ dysfunction after return of spontaneous circulation in postcardiac arrest syndrome. Crit Care Med. 2013;41:1492–501. 9. Geri G, Guillemet L, Dumas F, Charpentier J, Antona M, Lemiale V, et al. Acute kidney injury after out-of-hospital cardiac arrest: risk factors and prognosis in a large cohort. Intensive Care Med. 2015;41:1273–80. 10. Tujjar O, Mineo G, Dell'Anna A, Poyatos-Robles B, Donadello K, Scolletta S, et al. Acute kidney injury after cardiac arrest. Crit Care. 2015;19:169. 11. Champigneulle B, Geri G, Bougouin W, Dumas F, Arnaout M, Zafrani L, Pène F, Charpentier J, Mira JP, Cariou A. Hypoxic hepatitis after out-of-hospital cardiac arrest: Incidence, determinants and prognosis. Resuscitation. 2016;103:60–5. 7. Nolan JP, Neumar RW, Adrie C, Aibiki M, Berg RA, Bottiger BW, et al. Post- cardiac arrest syndrome: epidemiology, pathophysiology, treatment, and prognostication. 33. Roberts BW, Karagiannis P, Coletta M, Kilgannon JH, Chansky ME, Trzeciak S. Effects of PaCO2 derangements on clinical outcomes after cerebral injury: a systematic review. Resuscitation. 2015;91:32–41. Authors’ contributions d d h d FST designed the study, analyzed the data, and drafted the article; LN helped in the design of the study, extracted and analyzed the data, and helped revise the article; TS extracted and analyzed the data, and helped revise the article; YS, SMJ, TP, MA, ML, XW, and PP participated in acquiring data for the original ICON study, and revised this article for critical content; JLV participated in acquiring data for and coordinated the original ICON audit, 10. Tujjar O, Mineo G, Dell'Anna A, Poyatos-Robles B, Donadello K, Scolletta S, et al. Acute kidney injury after cardiac arrest. Crit Care. 2015;19:169. 10. Tujjar O, Mineo G, Dell'Anna A, Poyatos-Robles B, Donadello K, Scolletta S, et al. Acute kidney injury after cardiac arrest. Crit Care. 2015;19:169. 11. Champigneulle B, Geri G, Bougouin W, Dumas F, Arnaout M, Zafrani L, Pène F, Charpentier J, Mira JP, Cariou A. Hypoxic hepatitis after out-of-hospital cardiac arrest: Incidence, determinants and prognosis. Resuscitation. 2016;103:60–5. Page 10 of 10 Page 10 of 10 Nobile et al. Critical Care (2016) 20:368 12. Vincent JL, Marshall JC, Namendys-Silva SA, Francois B, Martin-Loeches I, Lipman J, et al. Assessment of the worldwide burden of critical illness: the intensive care over nations (ICON) audit. Lancet Respir Med. 2014;2:380–6. 13. Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA. 1993;270:2957–63. 14. Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13:818–29. 15. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med. 1996;22:707–10. 16. Calandra T, Cohen J. The international sepsis forum consensus conference on definitions of infection in the intensive care unit. Crit Care Med. 2005;33:1538–48. 17. Levy MM, Fink MP, Marshall JC, Abraham E, Angus D, Cook D, et al. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med. 2003;31:1250–6. 18. Nolan JP, Morley PT, Vanden Hoek TL, Hickey RW, Kloeck WG, Billi J, et al. Therapeutic hypothermia after cardiac arrest: an advisory statement by the advanced life support task force of the International Liaison Committee on Resuscitation. Circulation. 2003;108:118–21. 19. Chua HR, Glassford N, Bellomo R. 34. Taccone F, Cronberg T, Friberg H, Greer D, Horn J, Oddo M, et al. How to assess prognosis after cardiac arrest and therapeutic hypothermia. Crit Care. 2014;18:202. 33. Roberts BW, Karagiannis P, Coletta M, Kilgannon JH, Chansky ME, Trzeciak S. Effects of PaCO2 derangements on clinical outcomes after cerebral injury: a systematic review. Resuscitation. 2015;91:32–41. 34. Taccone F, Cronberg T, Friberg H, Greer D, Horn J, Oddo M, et al. How to assess prognosis after cardiac arrest and therapeutic hypothermia. Crit Care. 2014;18:202. Authors’ contributions d d h d Acute kidney injury after cardiac arrest. Resuscitation. 2012;83:721–7. 19. Chua HR, Glassford N, Bellomo R. Acute kidney injury after cardiac arrest. Resuscitation. 2012;83:721–7. 20. Yanta J, Guyette FX, Doshi AA, Callaway CW, Rittenberger JC. Renal dysfunction is common following resuscitation from out-of-hospital cardiac arrest. Resuscitation. 2013;84:1371–4. 21. Sandroni C, Dell'anna AM, Tujjar O, Geri G, Cariou A, Taccone FS. Acute Kidney Injury (AKI) after cardiac arrest: a systematic review and meta-analysis of clinical studies. Minerva Anestesiol. 2016. Epub ahead of print. 22. Lee DH, Cho IS, Lee SH, Min YI, Min JH, Kim SH, et al. Correlation between initial serum levels of lactate after return of spontaneous circulation and survival and neurological outcomes in patients who undergo therapeutic hypothermia after cardiac arrest. Resuscitation. 2015;88:143–9. 23. Donnino MW, Andersen LW, Giberson T, Gaieski DF, Abella BS, Peberdy MA, et al. Initial lactate and lactate change in post-cardiac arrest: a multicenter validation study. Crit Care Med. 2014;42:1804–11. 24. Lee TR, Kang MJ, Cha WC, Shin TG, Sim MS, Jo IJ, et al. Better lactate clearance associated with good neurologic outcome in survivors who treated with therapeutic hypothermia after out-of-hospital cardiac arrest. Crit Care. 2013;17:R260. 25. Mullner M, Sterz F, Domanovits H, Behringer W, Binder M, Laggner AN. The association between blood lactate concentration on admission, duration of cardiac arrest, and functional neurological recovery in patients resuscitated from ventricular fibrillation. Intensive Care Med. 1997;23:1138–43. 26. Nolan JP, Laver SR, Welch CA, Harrison DA, Gupta V, Rowan K. Outcome following admission to UK intensive care units after cardiac arrest: a secondary analysis of the ICNARC Case Mix Programme Database. Anaesthesia. 2007;62:1207–16. 27. Niskanen M, Kari A, Nikki P, Iisalo E, Kaukinen L, Rauhala V, et al. Acute physiology and chronic health evaluation (APACHE II) and Glasgow coma scores as predictors of outcome from intensive care after cardiac arrest. Crit Care Med. 1991;19:1465–73. 28. Skrifvars MB, Varghese B, Parr MJ. Survival and outcome prediction using the Apache III and the out-of-hospital cardiac arrest (OHCA) score in patients treated in the intensive care unit (ICU) following out-of-hospital, in-hospital or ICU cardiac arrest. Resuscitation. 2012;83:728–33. Submit your next manuscript to BioMed Central and we will help you at every step: Submit your next manuscript to BioMed Central and we will help you at every step: Submit your next manuscript to BioMed Central and we will help you at every step: 29. Ristagno G, Latini R, Plebani M, Zaninotto M, Vaahersalo J, Masson S, et al. Copeptin levels are associated with organ dysfunction and death in the intensive care unit after out-of-hospital cardiac arrest. Crit Care. 2015;19:132. • We accept pre-submission inquiries • Our selector tool helps you to ﬁnd the most relevant journal • We provide round the clock customer support • Convenient online submission • Thorough peer review • Inclusion in PubMed and all major indexing services • Maximum visibility for your research Submit your manuscript at www.biomedcentral.com/submit and we will help you at every step: 30. Rittenberger JC, Tisherman SA, Holm MB, Guyette FX, Callaway CW. An early, novel illness severity score to predict outcome after cardiac arrest. Resuscitation. 2011;82:1399–404. 31. Sutherasan Y, Penuelas O, Muriel A, Vargas M, Frutos-Vivar F, Brunetti I, et al. Management and outcome of mechanically ventilated patients after cardiac arrest. Crit Care. 2015;19:215. 32. Helmerhorst HJ, Roos-Blom MJ, van Westerloo DJ, de Jonge E. Association between arterial hyperoxia and outcome in subsets of critical illness: a systematic review, meta-analysis, and meta-regression of cohort studies. Crit Care Med. 2015;43:1508–19. 32. Helmerhorst HJ, Roos-Blom MJ, van Westerloo DJ, de Jonge E. Association between arterial hyperoxia and outcome in subsets of critical illness: a systematic review, meta-analysis, and meta-regression of cohort studies. Crit Care Med. 2015;43:1508–19.
https://openalex.org/W2152519579	https://eprints.lib.hokudai.ac.jp/dspace/bitstream/2115/60624/1/c5sc03377b.pdf	English	null	Hydrolysis of woody biomass by a biomass-derived reusable heterogeneous catalyst	Chemical science	2,016	cc-by	4,948	Instructions for use Title Hydrolysis of woody biomass by a biomass-derived reusable heterogeneous catalyst Author(s) Kobayashi, Hirokazu; Kaiki, Hiroyuki; Shrotri, Abhijit; Techikawara, Kota; Fukuoka, Atsushi Citation Chemical science, 7(1), 692-696 https://doi.org/10.1039/c5sc03377b Issue Date 2016-01-01 Doc URL http://hdl.handle.net/2115/60624 Rights(URL) https://creativecommons.org/licenses/by/3.0/ Type article File Information c5sc03377b.pdf Instructions for use Title Hydrolysis of woody biomass by a biomass-derived reusable heterogeneous catalyst Author(s) Kobayashi, Hirokazu; Kaiki, Hiroyuki; Shrotri, Abhijit; Techikawara, Kota; Fukuoka, Atsushi Citation Chemical science, 7(1), 692-696 https://doi.org/10.1039/c5sc03377b Issue Date 2016-01-01 Doc URL http://hdl.handle.net/2115/60624 Rights(URL) https://creativecommons.org/licenses/by/3.0/ Type article File Information c5sc03377b.pdf aInstitute for Catalysis, Hokkaido University, Kita 21 Nishi 10, Kita-ku, Sapporo, Hokkaido 001-0021, Japan. E-mail: fukuoka@cat.hokudai.ac.jp bGraduate School of Chemical Sciences and Engineering, Hokkaido University, Kita 13 Nishi 8, Kita-ku, Sapporo, Hokkaido 060-8628, Japan † Electronic supplementary information (ESI) available: Experimental details, raw data of characterisation. See DOI: 10.1039/c5sc03377b EDGE ARTICLE Cite this: Chem. Sci., 2016, 7, 692 Hirokazu Kobayashi,ab Hiroyuki Kaiki,ab Abhijit Shrotri,a Kota Techikawaraab and Atsushi Fukuokaab Hirokazu Kobayashi,ab Hiroyuki Kaiki,ab Abhijit Shrotri,a Kota Techikawaraab and Atsushi Fukuokaab Biomass is the sole carbon-based renewable resource for sustaining the chemical and fuel demands of our future. Lignocellulose, the primary constituent of terrestrial plants, is the most abundant non-food biomass, and its utilisation is a grand challenge in bioreﬁneries. Here we report the ﬁrst reusable and cost-eﬀective heterogeneous catalyst for the depolymerisation of lignocellulose. Air oxidation of woody biomass (Eucalyptus) provides a carbonaceous material bearing an aromatic skeleton with carboxylic groups (2.1 mmol g1) and aliphatic moieties. This catalyst hydrolyses woody biomass (Eucalyptus) to sugars in high yields within 1 h in trace HCl aq. Furthermore, after the reaction, the solid residue composed of the catalyst and insoluble ingredients of woody biomass is easily transformed back to fresh catalyst by the same air oxidation method. This is a self-contained system using woody biomass as both the catalyst source and substrate for realising facile catalyst preparation and recycling. Received 8th September 2015 Accepted 14th October 2015 DOI: 10.1039/c5sc03377b www.rsc.org/chemicalscience Received 8th September 2015 Accepted 14th October 2015 hemicellulose to monomeric sugars (Scheme 1) is the rst step in a biorenery, and this reaction has been studied with both enzymes and chemical catalysts such as soluble mineral acids and solid acids.1,4 However, the practical use of these catalysts is hampered by reuse diﬃculties, high costs of preparation and disposal, and the need for long reaction times.1 Herein, we report the rst reusable and cost-eﬀective heterogeneous cata- lyst to resolve all these issues. Instructions for use Title Hydrolysis of woody biomass by a biomass-derived reusable heterogeneous catalyst Author(s) Kobayashi, Hirokazu; Kaiki, Hiroyuki; Shrotri, Abhijit; Techikawara, Kota; Fukuoka, Atsushi Citation Chemical science, 7(1), 692-696 https://doi.org/10.1039/c5sc03377b Issue Date 2016-01-01 Doc URL http://hdl.handle.net/2115/60624 Rights(URL) https://creativecommons.org/licenses/by/3.0/ Type article File Information c5sc03377b.pdf Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP This journal is © The Royal Society of Chemistry 2016 Open Access Article. Published on 15 October 2015. Downloaded on 02/02/2016 03:58:29. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Hirokazu Kobayashi,ab Hiroyuki Kaiki,ab Abhijit Shrotri,a Kota Techikawaraab and Atsushi Fukuoka*ab Introduction Production of biofuels and bio-chemicals from lignocellulose, the most abundant non-food biomass, is a grand challenge in bioreneries.1–3 Lignocellulose is a composite of cellulose, hemicellulose and lignin, which are insoluble polymers of glucose, related sugars such as xylose, and phenyl propane derivatives, respectively. The hydrolysis of cellulose and The use of heterogeneous catalysts is desired for the eﬃcient depolymerisation of lignocellulose as they are non-corrosive and can be separated from product solution.5–7 It has recently been demonstrated that weakly acidic carbons8–15 and sulpho- nated carbons16–23 are active for the hydrolysis of cellulose as a model substrate. Carbon materials adsorb cellulosic mole- cules through CH–p hydrogen bonds and hydrophobic inter- actions,24,25 and then the adsorbed molecules are hydrolysed by acidic sites.9–13,26 The presence of defect sites on carbon may further improve the activity,10–12 as the adsorption of cellulosic molecules in connement13 changes their conformation to weaken glycosidic bonds.27–29 Regardless of the preferable characteristics of heterogeneous catalysts, the contamination of the catalyst with solid lignin aer the reaction prevents their application in the depolymer- isation of real lignocellulose. Removal of lignin from solid catalysts is oen challenging, essentially rendering the catalyst useless aer the rst reaction. Hence, the lignin fraction must be removed by pretreatment such as the kraprocess before applying lignocellulose to the hydrolysis reaction.9 Another demerit of existing catalysts is the use of large amounts of chemicals such as bases and acids in creating active sites on the catalysts,8–23 leading to high costs and huge quantities of Scheme 1 Hydrolysis of cellulose and hemicellulose fractions in lignocellulose. This journal is © The Royal Society of Chemistry 2016 692 \| Chem. Sci., 2016, 7, 692–696 View Article Online Edge Article Chemical Science Chemical Science neutralisation waste in post-treatments (more than 50 kg of waste per 1 kg of catalyst; see ESI†). organic compounds such as tar. The maximum temperature inside the sample was 575 K during the process, showing no formation of hot spots. Our idea for resolving the issues of conventional heteroge- neous catalysts is to produce a weakly acidic carbon catalyst through simple air oxidation of lignocellulose and lignin residue. Organic materials thermally decompose to form carbonaceous material at an elevated temperature30 and simultaneously gain weakly acidic oxygenated groups (active sites) in the presence of air via oxidation. Thus, we can expect that the carbon material prepared by air oxidation hydrolyses lignocellulose. Whole scheme of our system Fig. 1 represents our system using air oxidation for hydrolysis of lignocellulose. The rst part (Part 1) is the production of a carbon-based catalyst from Eucalyptus by air oxidation, which is necessary only once. The second part (Part 2) is a cyclic process consisting of milling pretreatment, hydrolysis of Euca- lyptus to glucose and xylose in trace HCl, and transformation of the solid residue to fresh catalyst by the same air oxidation. We used Eucalyptus as both a catalyst source and biomass substrate to make this a self-contained system. Eucalyptus is a fast- growing and inexpensive plant [<0.1 pounds (GBP) kg1] that has been cultured as a major feedstock for pulping.31,32 Fig. 2 13C CP/MAS NMR spectra of E-Carbon (red bold solid line), recycled E-Carbon (orange narrow solid line), N2-treated Eucalyptus (green dotted line) and pristine Eucalyptus (blue dashed line). Introduction In this way, the catalyst is readily prepared, and more importantly the used catalyst and residual lignin can be together transformed into fresh catalyst by the same air oxidation method. The chemical structure of E-Carbon was determined with several physicochemical techniques. Solid-state 1H–13C cross polarisation/magic angle spinning nuclear magnetic resonance (CP/MAS NMR) was used to clarify the structural change of Eucalyptus by air oxidation (Fig. 2). The pristine Eucalyptus (blue dashed line) gave major peaks at 110–50 ppm, mainly ascribed to cellulose and hemicellulose.33 Small NMR signals at $110 ppm are derived from lignin (Caromatic–O at 160–140 ppm, Caromatic–C and Caromatic–H at 140–110 ppm).34–36 Aer air oxidation (red solid line), a predominant peak appeared at 125 ppm due to the formation of non-oxygenated aromatic carbons. Smaller broad peaks were observed in the regions for –CO2R (170 ppm), Caromatic–O (150 ppm) and sp3 carbons (<100 ppm).34,35 Infrared (IR) measurement of E-Carbon in a trans- mission mode (Fig. S2†) showed four major peaks ascribed to n(C]O) (1770–1720 cm1), n(C]C, aromatic) (1610 cm1), d(C– H) (1470–1370 cm1) and a mixture of various vibrations such as n(C–O) (1350–1000 cm1),37 indicating the presence of aromatic rings and oxygenated groups. X-ray photoelectron spectroscopy (XPS) in the C 1s region represented a quantitative distribution of functional groups: –CO2R at 288.6 eV (13 1%), C]O at This journal is © The Royal Society of Chemistry 2016 Preparation and characterisation of the catalyst S6†), which are characteristic of amorphous carbon materials with poly- cyclic aromatics.41,42 The amorphous structure was also indi- cated by X-ray diﬀraction measurements (Fig. S7†). Hence, we concluded that E-Carbon consists of an aromatic framework with weakly acidic groups and aliphatic moieties (Fig. 3, the composition is CH0.44O0.42). It was conrmed that the aromatics and acidic sites were not only derived from lignin but also from cellulose fractions; air oxidation of cellulose gave a similar carbon material (named air-oxidised cellulose, see Fig. S2†). 287.2 eV (3 2%), C–O at 286.2 eV (20 5%), and C–C and C]C at 284.6 eV (65 5%) (Fig. S3†).38 The content of carboxylic acid in E-Carbon was 2.1 mmol g1, determined by a titration experiment with NaHCO3.39 Elemental analysis of E-Carbon showed that the amounts of C, H, N and O were 62.5 wt%, 2.2 wt%, <0.3 wt% and 35.0 wt%, respectively, where the oxygen content was estimated by subtracting the weight of C, H and ash from 100 wt%. The ratio corresponds to CH0.43O0.42. E-Carbon adsorbed a large amount of water (1.8 mmol g1 at p/p0 ¼ 0.1; Fig. S4†) at 298 K, which was ca. 160 times greater than the adsorption amount of N2 at 77 K at the same relative pressure (Fig. S5†). This result indicates the condensation of water in E-Carbon due to a high concentration of oxygenated groups.40 The Raman spectrum of E-Carbon contained a broad D-band at 1390 cm1 and a G-band at 1590 cm1 (IG > ID; Fig. S6†), which are characteristic of amorphous carbon materials with poly- cyclic aromatics.41,42 The amorphous structure was also indi- cated by X-ray diﬀraction measurements (Fig. S7†). Hence, we concluded that E-Carbon consists of an aromatic framework with weakly acidic groups and aliphatic moieties (Fig. 3, the composition is CH0.44O0.42). It was conrmed that the aromatics and acidic sites were not only derived from lignin but also from cellulose fractions; air oxidation of cellulose gave a similar carbon material (named air-oxidised cellulose, see Fig. S2†). are encapsulated by lignin in lignocellulose for protecting the carbohydrates from chemical attack, but milling treatment ruptures the composite to expose the sugar polymers.47 More importantly, specic to the mix-milling, good contact between the solid catalyst and solid substrate are created at the same time.48 This treatment required only 2 h for improving the reaction performance in the present study. Open Access Article. Published on 15 October 2015. Downloaded on 02/02/2016 03:58:29. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. The mix-milled solid containing E-Carbon (50 mg) and Eucalyptus (324 mg) was subjected to a hydrolysis reaction in a 120 ppm HCl aqueous solution (pH 2.5) at 488 K (Table 1). This reaction gave glucose in 78% yield based on the carbon content of cellulose in Eucalyptus (Table 1, entry 4). The hemi- cellulose fraction of Eucalyptus was also hydrolysed to xylose in 94% yield. Thus, both cellulose and hemicellulose can be uti- lised in this system. The major by-products were oligosaccha- rides, sugar isomers, 5-hydroxymethylfurfural, levoglucosan and furfural (detailed results are shown in Table S3†). The total amount of the sugars and soluble by-products was almost consistent with the carbohydrate content in the Eucalyptus sample (carbon balance: 96%). In the hydrolysis of Eucalyptus, we assume that E-Carbon and the mild acidic solvent (pH 2.5) synergistically accelerate the formation of monomeric sugars. It has been reported that trace HCl hydrolyses cellulose to produce soluble oligosaccharides, which enables the subsequent hydrolysis of oligosaccharides by solid acid catalysts.49 In contrast, the roles of solid catalyst and HCl are reversed in our system due to the mix-milling.9 Weakly acidic carbons quickly hydrolyse solid cellulose to soluble oligosaccharides owing to the close contact created by mix- milling. As a result, the hydrolysis of oligosaccharides is the rate-determining step, which needs to be accelerated by soluble acid to maximise yields of monomeric sugars. Indeed, the hydrolysis of cellulose by a carbon catalyst in water aer mix- milling almost completely converted cellulose (93%) with soluble oligomers as the main product (70% yield), whereas hydrolysis by HCl gave a low conversion of cellulose (39%) with We also prepared a catalyst by heat-treatment of Eucalyptus under N2 at 573 K as a control. This material had signicantly weaker aromatic peaks in the NMR (green dotted line in Fig. 2) and IR spectra (Fig. S2†). Accordingly, our results show that air oxidation provides more aromatics than N2 treatment at 573 K. This is reasonable as the air oxidation of organic polymers (e.g., polyacrylonitrile) gives aromatic precursors at 473–573 K for the manufacture of carbon bres.43 As for the oxygenated groups in N2-treated Eucalyptus, only weak C]O peaks were observed in the IR spectrum (Fig. S2†) and the area percentage of –CO2R was only 2 1% in the XPS (Fig. S3†). Open Access Article. Published on 15 October 2015. Downloaded on 02/02/2016 03:58:29. This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. The specic amount of carboxylic acid groups determined by titration was 0.11 mmol g1, corresponding to 1/20 of that of E-Carbon. Clearly, air oxidation is essential for introducing a large amount of carboxylic acid groups onto the carbons. Table 1 Hydrolysis of Eucalyptus by Eucalyptus-based catalystsa Entry Solvent Catalyst Product yield/% Glucoseb Xylosec 1 Water No catalyst 3 30 2 Water E-Carbon 31 83 3 HCld No catalyst 32 26 4 HCld E-Carbon (1st cycle) 78 94 5 HCld E-Carbon (2nd cycle) 82 89 6 HCld N2-treated Eucalyptus 28 26 7 HCld Air-oxidised cellulose 77 91 a Reaction conditions: mix-milled sample 374 mg [Eucalyptus 324 mg, catalyst 50 mg (containing 0.11 mmol of carboxylic acid groups)], solvent 40 mL. The reaction temperature was elevated from 298 K to 488 K in 17 min, and then quickly cooled down to 298 K. b Based on carbon content of cellulose. c Based on carbon content of xylan. d 120 ppm HCl (¼ 0.13 mmol). Preparation and characterisation of the catalyst We rst studied the initial air oxidation of Eucalyptus to produce a catalyst. Eucalyptus powder was rst washed with boiling water and dried (see ESI†). The resulting solid was used in all the experiments. The major ingredients of the washed Euca- lyptus powder were cellulose (47 wt%), xylan hemicellulose (10 wt%) and lignin (32 wt%). Ash was present in a small amount (0.12 wt%) and mainly composed of Ca salts (Table S1†). The air oxidation of the Eucalyptus powder was performed at 573 K for 1 h, which were the optimised conditions for preparing an active catalyst (Table S2†). This procedure gave a black material (named E-Carbon, Fig. S1†) in 49% yield based on the carbon content. The remaining part was lost as CO, CO2 and volatile Fig. 2 13C CP/MAS NMR spectra of E-Carbon (red bold solid line), recycled E-Carbon (orange narrow solid line), N2-treated Eucalyptus (green dotted line) and pristine Eucalyptus (blue dashed line). Fig. 1 Schematic representation of the E-Carbon system. HC: xylan hemicellulose. Fig. 1 Schematic representation of the E-Carbon system. HC: xylan hemicellulose. This journal is © The Royal Society of Chemistry 2016 This journal is © The Royal Society of Chemistry 2016 Chem. Sci., 2016, 7, 692–696 \| 693 Edge Article View Article Online Chemical Science Edge Article 287.2 eV (3 2%), C–O at 286.2 eV (20 5%), and C–C and C]C at 284.6 eV (65 5%) (Fig. S3†).38 The content of carboxylic acid in E-Carbon was 2.1 mmol g1, determined by a titration experiment with NaHCO3.39 Elemental analysis of E-Carbon showed that the amounts of C, H, N and O were 62.5 wt%, 2.2 wt%, <0.3 wt% and 35.0 wt%, respectively, where the oxygen content was estimated by subtracting the weight of C, H and ash from 100 wt%. The ratio corresponds to CH0.43O0.42. E-Carbon adsorbed a large amount of water (1.8 mmol g1 at p/p0 ¼ 0.1; Fig. S4†) at 298 K, which was ca. 160 times greater than the adsorption amount of N2 at 77 K at the same relative pressure (Fig. S5†). This result indicates the condensation of water in E-Carbon due to a high concentration of oxygenated groups.40 The Raman spectrum of E-Carbon contained a broad D-band at 1390 cm1 and a G-band at 1590 cm1 (IG > ID; Fig. Preparation and characterisation of the catalyst Although this opti- misation using planetary ball-milling is for laboratory-scale experiments, large-scale milling is applicable to maximise the economic eﬃciency in industry.45,47 Conclusions The air oxidation of biomass feedstock, Eucalyptus, produces a carbon-based catalyst overcoming the limitations of conven- tional catalysts used for the hydrolysis of woody biomass. The catalyst quickly converts lignin-containing Eucalyptus to glucose and xylose in high yields. Lignin remains as a solid together with the catalyst aer the reaction; however, this solid mixture is a source for fresh catalyst and fuel. Therefore, the E-Carbon system drastically reduces the preparation and post-treatment costs of the catalyst. In general, the deactivation or spoiling of catalyst by contaminant is oen a major issue in catalytic reactions. Hence, our idea that converts contaminant to a cata- lyst can be a useful strategy for improving the eﬃciency of catalytic processes. The solid residue recovered aer the reaction with E-Carbon can be converted to a fresh catalyst again by air oxidation as shown in Fig. 1. This is the outstanding characteristic of our catalyst, since all the previous carbon-based catalysts were single use in the hydrolysis of raw biomass.8–23 The treatment at 573 K for 1 h converted the solid residue of 1.63 g [1.12 g derived from Eucalyptus (mainly lignin) and 0.51 g of E-Carbon, obtained in a large-scale experiment] to a black powder of 1.14 g. Accordingly, the catalyst weight increased from 0.51 to 1.14 g aer one cycle in this system. The surplus residue can be used as fuel to power the process, since the solid is derived only from woody biomass and air. The 13C CP/MAS NMR spectrum of the regenerated catalyst contained a strong aromatic carbon peak with small fractions of –CO2R, Caromatic–O and aliphatic groups (Fig. 2, orange solid line). This character is similar to that of original E-Carbon. A portion of the prepared catalyst was again mix-milled with Eucalyptus and subjected to the hydrolysis reaction in the same manner as described above. The reaction produced glucose in 82% yield and xylose in 89% yield (Table 1, entry 5). We also compared the catalytic activity of the rst- and second-cycle E-Carbon at a lower temperature. The experiments at 473 K indicated no decline of catalytic activity by the second air oxidation (Table S4†). It is thus demonstrated that the mixture of lignin and E-Carbon changes to a new active E-Carbon by air oxidation. Since both lignin and E-Carbon are aromatic polymers, their transformation to the new catalyst is easier than the rst synthesis of E-Carbon from lignocellulose. Preparation of E-Carbon Eucalyptus powder was washed with boiling water prior to use for all purposes in this study. 4.00 g of dried Eucalyptus powder was spread with a thickness of 3 mm on a Pyrex dish (ø130) to uniformly prepare the catalyst and avoid hot spots. The sample was calcined under air at atmospheric pressure in an electric furnace with the following program: 298 to 573 K by 5 K min1 and 573 K for 1 h. In the case of reaction residue, the residue of 1.63 g was calcined under the same conditions. The tempera- ture inside the sample was monitored using a thermocouple (ø0.5) equipped with a quartz tube (ca. ø1). Mix-milling Eucalyptus (5.0 g) and catalyst (0.77 g) were milled together in an Al2O3 pot (250 mL) with Al2O3 balls (99.9%, ø15, 210 g) using a Fritsch P-6 planetary ball mill. Milling conditions were 500 rpm for 2 h with a 10 min interval aer every 10 min of milling. Conclusions Our system can leverage the contaminant (lignin) for the preparation of catalyst, which is in sharp contrast to conventional catalytic processes that require removal of the contaminant. This journal is © The Royal Society of Chemistry 2016 Edge Article do not use strong acid but involve weak acids (carboxylic acids). However, they have diﬀerent optimal conditions due to the diﬀerence in their skeletal and active structures. The carbon is composed of an aromatic framework and weak acids, and thereby high temperatures and a wide range of pH are appli- cable in catalytic reactions. On the other hand, cellulase is a protein that requires a conjugated base (carboxylate) in addition to carboxylic acid for the dissociation of glycosidic bonds; thus, low temperature and careful control of the pH with a buﬀer are necessary to keep the enzyme active. Consequently, E-Carbon can work under harsher conditions, which enables the rapid hydrolysis of lignocellulose in trace HCl aq. at high temperature. Moreover, our catalyst is reusable and the price (ca. 0.1 GBP kg1) is two-orders lower than that of cellulase (6.5– 26 GBP kg1).1,51 the formation of glucose as a main product (27% yield).9 Therefore, in the conversion of Eucalyptus, reactions in the absence of HCl or carbon catalyst provide unpractical yields of monomeric sugars (entries 1–3). HCl can be neutralised aer the reaction with very low economic impact, as the acid concentration is less than 1/50 of conventional mineral acid processes.1 Controlled experiments were performed to reveal the important parameters inuencing the catalytic activity of E- Carbon. Since a reaction in aq. HCl without E-Carbon aﬀorded glucose in 32% yield and xylose in 26% yield (entry 3), E-Carbon increases the yield of glucose by 46% and yield of xylose by 68% (subtraction of yields in entry 3 from those in entry 4). The increase corresponds to a turnover number of carboxylic acid of 5.6. The result indicates that E-Carbon acts as a catalyst for the hydrolysis of cellulose and hemicellulose in Eucalyptus. We also found that the air-oxidised carbon prepared from cellulose worked in this reaction similarly to E-Carbon as shown in entry 7 (glucose 77%, xylose 91%). This shows that a cellulose-derived part also constitutes the active catalytic domain in E-Carbon. Contrastingly, the Eucalyptus-based catalyst prepared by N2 treatment was inactive (glucose 28%, xylose 26%; entry 6). Therefore, it is concluded that the air oxidation of woody biomass provides active catalysts for the hydrolysis of lignocellulose. Pretreatment and catalytic reaction Table 1 Hydrolysis of Eucalyptus by Eucalyptus-based catalystsa Entry Solvent Catalyst Product yield/% Glucoseb Xylosec 1 Water No catalyst 3 30 2 Water E-Carbon 31 83 3 HCld No catalyst 32 26 4 HCld E-Carbon (1st cycle) 78 94 5 HCld E-Carbon (2nd cycle) 82 89 6 HCld N2-treated Eucalyptus 28 26 7 HCld Air-oxidised cellulose 77 91 Table 1 Hydrolysis of Eucalyptus by Eucalyptus-based catalystsa E-Carbon and Eucalyptus [1 : 6.48 (wt)] were milled together, named mix-milling,9 by planetary ball-milling44–46 for realising a high-yielding synthesis of sugars. Cellulose and hemicellulose Fig. 3 Proposed structure of E-Carbon. a Reaction conditions: mix-milled sample 374 mg [Eucalyptus 324 mg, catalyst 50 mg (containing 0.11 mmol of carboxylic acid groups)], solvent 40 mL. The reaction temperature was elevated from 298 K to 488 K in 17 min, and then quickly cooled down to 298 K. b Based on carbon content of cellulose. c Based on carbon content of xylan. d 120 ppm HCl (¼ 0.13 mmol). Fig. 3 Proposed structure of E-Carbon. This journal is © The Royal Society of Chemistry 2016 694 \| Chem. Sci., 2016, 7, 692–696 694 \| Chem. Sci., 2016, 7, 692–696 694 \| Chem. Sci., 2016, 7, 692–696 Chemical Science View Article Online Chemical Science View Article Online Edge Article Acknowledgements 25 M. Yabushita, H. Kobayashi, J. Hasegawa, K. Hara and A. Fukuoka, ChemSusChem, 2014, 7, 1443–1450. The authors thank Dr I. Ogino for the water adsorption exper- iment. This work was supported by the Japan Science and Technology Agency (JST) ALCA and a Grant-in-Aid for Young Scientists (KAKENHI, No. 26709060). 26 H. Kobayashi, M. Yabushita, J. Hasegawa and A. Fukuoka, J. Phys. Chem. C, 2015, 119, 20993–20999. 27 O. M. Gazit, A. Charmot and A. Katz, Chem. Commun., 2011, 47, 376–378. Notes and references 28 O. M. Gazit and A. Katz, Langmuir, 2012, 28, 431–437. 29 O. M. Gazit and A. Katz, J. Am. Chem. Soc., 2013, 135, 4398– 4402. 1 J. S. Luterbacher, D. M. Alonso and J. A. Dumesic, Green Chem., 2014, 16, 4816–4838. 30 M. Toda, A. Takagaki, M. Okamura, J. N. Kondo, S. Hayashi, K. Domen and M. Hara, Nature, 2005, 438, 178. 2 P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538–1558. 2 P. Gallezot, Chem. Soc. Rev., 2012, 41, 1538–1558. 3 C. O. Tuck, E. Perez, I. T. Horvath, R. A. Sheldon and M. Poliakoﬀ, Science, 2012, 337, 695–699. 31 R. Rode, H. G. Leite, M. L. D. Silva, C. A. ´A. S. Ribeiro and D. H. B. Binoti, Forest Pol. Econ., 2014, 44, 26–33. D. H. B. Binoti, Forest Pol. Econ., 2014, 44, 26–33 4 H. Kobayashi and A. Fukuoka, Green Chem., 2013, 15, 1740– 1763. 32 F. Ubukata, Journal of Southeast Asian Studies, 2001, 39, 417– 436. 5 A. Fukuoka and P. L. Dhepe, Angew. Chem., Int. Ed., 2006, 45, 5161–5163. 33 E.-L. Hult, P. Larsson and T. Iversen, Cellulose, 2000, 7, 35– 55. 6 M. J. Climent, A. Corma and S. Iborra, Green Chem., 2011, 13, 520–540. 34 K. N. Scott, J. Am. Chem. Soc., 1972, 94, 8564–8568. 35 L. F. Johnson and W. C. Jankowski, Carbon-13 NMR Spectra, John Wiley & Sons, New York, 1972. 7 M. Yabushita, H. Kobayashi and A. Fukuoka, Appl. Catal., B, 2014, 145, 1–9. John Wiley & Sons, New York, 1972. 36 K. M. Holtman, H.-m. Chang, H. Jameel and J. F. Kadla, 36 K. M. Holtman, H.-m. Chang, H. Ja 36 K. M. Holtman, H.-m. Chang, H. Jameel and J. F. Kadla, J. Wood Chem. Technol., 2006, 26, 21–34. , g, J J , J. Wood Chem. Technol., 2006, 26, 21–34. 8 H. Kobayashi, T. Komanoya, K. Hara and A. Fukuoka, ChemSusChem, 2010, 3, 440–443. J. Wood Chem. Technol., 2006, 26, 21–34. 37 P. E. Fanning and M. A. Vannice, Carbon, 1993, 31, 721–730. 9 H. Kobayashi, M. Yabushita, T. Komanoya, K. Hara, I. Fujita and A. Fukuoka, ACS Catal., 2013, 3, 581–587. 38 H. P. Boehm, Carbon, 2002, 40, 145–149. 39 H. P. Boehm, Carbon, 1994, 32, 759–769. 10 P.-W. Chung, A. Charmot, O. A. Olatunji-Ojo, K. A. Durkin and A. Katz, ACS Catal., 2014, 4, 302–310. 40 D. D. Do and H. D. Catalytic reaction The hydrolysis of Eucalyptus was performed in a hastelloy C-22 high-pressure reactor equipped with an agitator operating at 600 rpm and a thermocouple. Mix-milled sample (374 mg) and Comparing E-Carbon with reported catalysts, the active site is slightly similar to those of enzymes (cellulase).50 Both of them This journal is © The Royal Society of Chemistry 2016 695 Chem. Sci., 2016, 7, 692–696 \| 695 View Article Online Chemical Science 21 C. Zhang, H. Wang, F. Liu, L. Wang and H. He, Cellulose, 2012, 20, 127–134. 40 mL of 120 ppm HCl aq. were added into the reactor. The temperature of the reaction mixture was elevated from 298 K to 488 K in ca. 17 min and then quickly lowered to 298 K. Soluble products were analysed by high-performance liquid chroma- tography (Shodex SUGAR SH1011 and Phenomenex Rezex RPM- Monosaccharide Pb++ columns with refractive index detectors). 22 X. Zhao, J. Wang, C. Chen, Y. Huang, A. Wang and T. Zhang, Chem. Commun., 2014, 50, 3439–3442. 23 P. Dornath, H. J. Cho, A. Paulsen, P. Dauenhauer and W. Fan, Green Chem., 2015, 17, 769–775. 24 P.-W. Chung, A. Charmot, O. M. Gazit and A. Katz, Langmuir, 2012, 28, 15222–15232. Notes and references Do, Carbon, 2000, 38, 767–773. 41 A. C. Ferrari and J. Robertson, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 14095–14107. 11 A. Charmot, P.-W. Chung and A. Katz, ACS Sustainable Chem. Eng., 2014, 2, 2866–2872. 42 H. Shinohara, Y. Yamakita and K. Ohno, J. Mol. Struct., 1998, 442, 221–234. 12 G. S. Foo and C. Sievers, ChemSusChem, 2015, 8, 534–543. G. S. Foo and C. Sievers, ChemSusChem, 2015, 8, 53 13 P.-W. Chung, M. Yabushita, A. T. To, Y. Bae, J. Jankolovits, H. Kobayashi, A. Fukuoka and A. Katz, ACS Catal., 2015, 5, 6422–6425. 43 X. Huang, Materials, 2009, 2, 2369. 44 C. F. Burmeister and A. Kwade, Chem. Soc. Rev., 2013, 42, 7660–7667. 14 I. Delidovich and R. Palkovits, Microporous Mesoporous Mater., 2016, 219, 317–321. 45 N. Meine, R. Rinaldi and F. Sch¨uth, ChemSusChem, 2012, 5, 1449–1454. 15 A. T. To, P.-W. Chung and A. Katz, Angew. Chem., Int. Ed., 2015, 54, 11050–11053. 46 A. Shrotri, L. K. Lambert, A. Tanksale and J. Beltramini, Green Chem., 2013, 15, 2761. 16 S. Suganuma, K. Nakajima, M. Kitano, D. Yamaguchi, H. Kato, S. Hayashi and M. Hara, J. Am. Chem. Soc., 2008, 130, 12787–12793. 47 R. G. Blair, in Production of Biofuels and Chemicals with Ultrasound, ed. Z. Fang, R. L. Smith Jr and X. Qi, Springer, Dordrecht, 2014, pp. 269–288. 17 A. Onda, T. Ochi and K. Yanagisawa, Green Chem., 2008, 10, 1033–1037. 48 M. Yabushita, H. Kobayashi, K. Hara and A. Fukuoka, Catal. Sci. Technol., 2014, 4, 2312–2317. 18 R. Rinaldi, N. Meine, J. vom Stein, R. Palkovits and F. Sch¨uth, ChemSusChem, 2010, 3, 266–276. 49 J. Geboers, S. van de Vyver, K. Carpentier, P. Jacobs and B. Sels, Chem. Commun., 2011, 47, 5590–5592. 19 S. van de Vyver, L. Peng, J. Geboers, H. Schepers, F. de Clippel, C. J. Gommes, B. Goderis, P. A. Jacobs and B. F. Sels, Green Chem., 2010, 12, 1560–1563. 50 D. L. Zechel and S. G. Withers, Acc. Chem. Res., 2000, 33, 11– 18. 51 N. A. Bauer and W. R. Gibbons, Int. J. Agric. Biol. Eng., 2012, 5, 48–55. 20 J. Pang, A. Wang, M. Zheng and T. Zhang, Chem. Commun., 2010, 46, 6935–6937. This journal is © The Royal Society of Chemistry 2016 696 \| Chem. Sci., 2016, 7, 692–696
https://openalex.org/W4386028122	http://journal.umg.ac.id/index.php/manajerial/article/download/5968/3372	Indonesian	null	Dorongan Masyarakat Jawa Timur Menjadi Pekerja Migran Indonesia	Jurnal Manajerial	2,023	cc-by	7,465	Abstract Research Limitations - This research was conducted only in certain areas. The northern region is represented by the northern coastal area of Lamongan, the southern region is represented by Trenggalek Regency, the western region is represented by Ngawi Regency, and finally the eastern region is represented by Sumenep Regency and prioritizes the small island in Sumenep, namely Sapudi Island. Research Implications - This research has important implications for the general public, Indonesian immigrant workers as well as for future research. The general public will know what things might happen when they become immigrant workers, and how the immigrants survive in other countries in order to improve their family's economic standard. Immigrant workers can meet their needs in life, while further researchers can dig deeper into other elements that encourage immigrant workers to work abroad. g p g g Research Limitations - This research was conducted only in certain areas. The northern region is represented by the northern coastal area of Lamongan, the southern region is represented by Trenggalek Regency, the western region is represented by Ngawi Regency, and finally the eastern region is represented by Sumenep Regency and prioritizes the small island in Sumenep, namely Sapudi Island. Keyword: Encouragement, Public, East, Java, Immigrant, Workers Dorongan Masyarakat Jawa Timur Menjadi Pekerja Migran Indonesia Maulidyah Amalina Rizqi* Program Studi Manajemen, Fakultas Ekonomi dan Bisnis Universitas Muhammadiyah Gresik, Gresik, Indonesia, maulidyah@umg.ac.id Corresponding author Maulidyah Amalina Rizqi Program Studi Manajemen, Fakultas Ekonomi dan Bisnis Universitas Muhammadiyah Gresik, Gresik, Indonesia, maulidyah@umg.ac.id *Corresponding author P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Abstract Abstract Background - The unemployment rate in East Java province is below the national average and the poverty rate in East Java province has decreased from the previous year. However, the number of Indonesian migrant workers originating from East Java ranks first nationally. Objectives - The aim of this research is to prove that there is a drive from within the community to become migrant workers. Researchers also want to find out what things make the people of East Java to become migrant workers. Design/ Methodology/ Approach - This study uses a Descriptive-Qualitative paradigm approach, with data collection techniques in the form of interviews with active and non-active Indonesian migrant workers (ex immigrants). Diterima : 02 Juli 2023 Direview : 25 Juli 2023 Direvisi : 14 Agustus 2023 Disetujui : 16 Agustus 2023 Background - The unemployment rate in East Java province is below the national average and the poverty rate in East Java province has decreased from the previous year. However, the number of Indonesian migrant workers originating from East Java ranks first nationally. Diterima : 02 Juli 2023 Direview : 25 Juli 2023 Direvisi : 14 Agustus 2023 Disetujui : 16 Agustus 2023 ( g ) Findings - Some of the elements that encourage people to become Indonesian immigrant workers are invitations from family, colleagues or friends, low income, economic improvement, living facilities and low education. Some of these elements are the encouragement or motivation of individuals to meet several needs in personal and social life and each individual wants to fulfill these needs in his life. Conclusion - This research produced several elements that encouraged the people of East Java, especially to become Indonesian immigrant workers, namely invitations from family, colleagues or friends, low income in Indonesia, economic improvement, living facilities and low education. Some of these elements are the encouragement or motivation of individuals to fulfill several needs in personal and social life. f Research Implications - This research has important implications for the general public, Indonesian immigrant workers as well as for future research. The general public will know what things might happen when they become immigrant workers, and how the immigrants survive in other countries in order to improve their family's economic standard. Immigrant workers can meet their needs in life, while further researchers can dig deeper into other elements that encourage immigrant workers to work abroad. py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Temuan - Beberapa unsur yang mendorong masyarakat menjadi pekerja imigran indonesia adalah ajakan dari keluarga, rekan atau teman, pendapatan rendah, peningkatan ekonomi, fasilitas hidup dan pendidikan rendah. Beberapa unsur tersebut merupakan dorongan atau motivasi individu untuk memenuhi beberapa kebutuhan dalam kehidupan pribadi maupun sosial dan setiap individu ingin memenuhi kebutuhan-kebutuhan dalam kehidupannya. y Kesimpulan - Penelitian ini menghasilkan beberapa unsur yang mendorong masyarakat Jawa Timur khususnya untuk menjadi pekerja imigran indonesia, yakni ajakan dari keluarga, rekan atau teman, pendapatan rendah di Indonesia, peningkatan ekonomi, fasilitas hidup dan pendidikan rendah. Beberapa unsur tersebut merupakan dorongan atau motivasi individu untuk memenuhi beberapa kebutuhan dalam kehidupan pribadi maupun sosial. Implikasi Penelitian - Penelitian ini memiliki implikasi penting bagi masyarakat umum, pekerja imigran Indonesia maupun untuk penelitian yang akan datang. Masyarakat umum akan mengetahui hal apa saja yang mungkin terjadi saat menjadi pekerja imigran, dan bagaimana cara para imigran itu untuk bertahan hidup dinegara lain demi meningkatkan taraf ekonomi keluarga. Pekerja imigran dapat memenuhi kebutuhan dalam hidupnya, sedangkan peneliti selanjutnya dapat menggali lebih dalam unsur lain yang mendorong para pekerja imigran untuk bekerja di luar negeri. Batasan Penelitian - Penelitian ini dilakukan hanya pada wilayah-wilayah tertentu. Wilayah utara diwakili area lamongan pesisir utara, wilayah selatan diwakili oleh kabupaten trenggalek, wilayah barat diwakili oleh kabupaten ngawi, dan terakhir wilayah timur diwakili kabupaten sumenep dan diutamakan pada pulau kecil di sumenep yakni pulau sapudi. Kata Kunci: Dorongan, Masyarakat, Jawa, Timur, Pekerja, Imigran Abstrak LatarBelakang - Tingkat pengangguran di provinsi jawa timur dibawah rata-rata nasional dan juga tingkat kemiskinan provinsi jawa timur mengalami penurunan dari tahun sebelumnya. Namun demikian jumlah pekerja migran Indonesia yang berasal dari jawa timur menduduki peringkat pertama secara nasional. Tujuan - Tujuan penelitian ini adalah ingin membuktikan adanya dorongan dari dalam diri masyarakat untuk menjadi pekerja migran. Peneliti juga ingin mencari tahu hal apa saja yang membuat masyarakat jawa timur untuk menjadi pekerja migran. Desain /Metodologi /Pendekatan - Penelitian ini menggunakan pendekatan berparadigma Deskriptif- Kualitatif, dengan teknik pengambilan data berupa wawancara terhadap para pekerja migran indonesia aktif maupun non aktif (ex imigran) 430 Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Tim P-ISSN : 2354-8592 E-ISSN : 2621-5055 Copyright© Creative Commons Attribution 4.0 International License py g ram Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Kata Kunci: Dorongan, Masyarakat, Jawa, Timur, Pekerja, Imigran PENDAHULUAN sebagai pelajar, ibu rumah tangga, pekerja (tenaga Kerja), Wirausaha, purna kerja atau purna tugas (pensiun) bahkan ada juga yang masih menganggur. Indonesia memiliki jumlah penduduk lebih dari 250 juta jiwa. Menurut BPS (Badan Pusat Statistik) jumlah penduduk Indonesia diproyeksikan sebanyak 275,77 juta jiwa pada tahun 2022. Jumlah tersebut naik 1,13% dibandingkan pada tahun lalu 2021 yang sebanyak 272,68 juta jiwa. Ratusan juta penduduk Indonesia tersebut terdiri dari beberapa peran, Berdasarkan data dari badan pusat statistik terdapat perbandingan dan perubahan jumlah tenaga kerja dan pengangguran dari tahun 2021 ke 2022. Berikut dilampirkan data tersebut setiap bulan agustus: Tabel 1. Data Pekerja dan pengangguran Tabel 1. Data Pekerja dan pengangguran Tabel 1. Data Pekerja dan pengangguran Bekerja Pengangguran 2021 2022 2021 2022 Prosentase % 93,51% 94,14% 6,49% 5,86% Jumlah 131.050.520 135.296.710 9.102.050 8.425.930 Sumber: Badan Pusat Statistik, 2022 431 penurunan jumlah pengangguran. Dari 275,77 juta penduduk indonesia per tahun 2022, terdapat 135.296.710 jiwa penduduk indonesia yang memiliki peran sebagai Dari data diatas dapat dilihat bahwa jumlah masyarakat yang bekerja naik sebesar 0,63 % dari tahun sebeumnya, dan hal tersebut juga dibuktikan dengan 431 Copyright© Creative Commons Attribution 4.0 International License Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 tenaga kerja aktif sedangkan jumlah pengangguran per agustus 2022 terdapat 8.425.930 jiwa. Data yang dipaparkan oleh peneliti diatas adalah data di tahun 2022, sedangkan di tahun 2021 jumlah tenaga kerja aktif lebih sedikit daripada tahun 2022. Data diatas menunjukkan bahwa provinsi jawa timur menduduki posisi teratas bahkan selama tahun 2022 jawa timur selalu menduduki jumlah pekerja migran terbanyak. Tingginya jumlah pekerja migran yang berasal dari jawa timur menunjukkan bahwa minat atau keinginan masyarakat Jawa Timur untuk menjadi pekerja migran cukup banyak. Selain data diatas tingkat pengangguran di provinsi jawa timur juga tidak terlalu jauh dari rata-rata nasional. Menurut BPS (Badan Pusat Statistik) data pengangguran nasional per agustus 2022 adalah 5.86% dan jawa timur memiliki tingkat penganguran sebesar 5,49%. Data pengangguran jawa timur hanya selisih 0,35% dari data nasional. Tingkat kemiskinan di provinsi jawa timur mengalami penurunan dari tahun 2021. Kemiskinan provinsi jawa timur sebesar 11,40% di tahun 2021, sedangkan di tahun 2022 menurun menjadi 10,38%. Jumlah pengangguran di Indonesia masih menyentuh angka lebih dari delapan juta jiwa, hal ini menjadi salah satu alasan masyarakat Indonesia mencari lapangan kerja di luar negeri dengan menjadi pekerja migran Indonesia. Terbukti menurut BP2MI (Badan Perlindungan Pekerja Migran Indonesia) jumlah pekerja migran Indonesia hingga November tahun 2022 masih di angka 28.373 jiwa yang tersebar di beberapa negara yakni, Hongkong, Taiwan, Malaysia, Singapura, Saudi arabia, Jepang dan italia. Mayoritas pekerja migran Indonesia bekerja di Malaysia dengan jumlah 9.808 jiwa. Sedangkan menurut Dataindonesia.id dan BP2MI untuk asal daerah pekerja migran Indonesia ini berasal dari jawa timur, jawa tengah, NTB, jawa barat, Sumatera Utara dan Lainnya. Copyright© Creative Commons Attribution 4.0 International License m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Tabel 1. Data Pekerja dan pengangguran Menurut BP2MI jumlah pekerja migran Indonesia berdasarkan asal daerah yakni Jawa Timur sebanyak 6.119 jiwa, Jawa Tengah sebanyak 5.691 jiwa, NTB sebanyak 5.073 Jiwa, jawa Barat sebanyak 4.662 jiwa dan Sumatera Utara sebanyak 2.378 dan Provinsi lainnya sebanyak 4.550 Jiwa. 432 Beberapa data diatas menunjukkan bahwa tingkat pengangguran di provinsi jawa timur dibawah rata-rata nasional dan juga tingkat kemiskinan provinsi jawa timur mengalami penurunan dari tahun sebelumnya. Namun demikian jumlah pekerja migran Indonesia yang berasal dari jawa timur menduduki peringkat pertama secara nasional. Proposisi sementara peneliti adalah banyak dari masyarakat jawa timur ingin menjadi pekerja migran 432 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 ada di sekitarnya, rendah nya Pendidikan, ekonomi keluarga dan lingkungan internal maupun eksternal. Sedangkan menurut Husniawati (2019) terdapat empat faktor yang mempengaruhi motivasi tenaga kerja Indonesia untuk bekerja ke luar negeri di Kabupaten Lombok Timur, yaitu kebutuhan, harapan pribadi, kondisi ekonomi keluarga, dan kondisi lingkungan masyarakat. karena ingin mendapatkan pencapaian pendapatan yang tinggi. Proposisi ini terbentuk juga dikarenakan data Pendidikan akhir masyarakat jawa timur yang cukup rendah. Dikatakan rendah dikarenakan hanya 7,31% penduduk yang mengenyam Pendidikan di perguruan tinggi sedangkan lulusan sekolah dasar menduduki posisi teratas, sebanyak 27,02%. Lebih menyedihkan lagi adalah sebanyak 22,02% penduduk Jawa Timur tidak memiliki Ijazah. ada di sekitarnya, rendah nya Pendidikan, ekonomi keluarga dan lingkungan internal maupun eksternal. Sedangkan menurut Husniawati (2019) terdapat empat faktor yang mempengaruhi motivasi tenaga kerja Indonesia untuk bekerja ke luar negeri di Kabupaten Lombok Timur, yaitu kebutuhan, harapan pribadi, kondisi ekonomi keluarga, dan kondisi lingkungan masyarakat. Berbagai hasil penelitian yang sudah dilakukan sebelumnya menunjukkan hasil yang variatif terhadap motivasi masyarakat yang menjadi tenaga kerja Indonesia (TKI). Maka penelitian ini ingin membuktikan adanya dorongan dari masyarakat untuk menjadi pekerja migran. Peneliti juga ingin mencari tahu hal apa saja yang membuat masyarakat jawa timur untuk menjadi pekerja migran. Kemungkinan besar faktor Pendidikan yang rendah membuat para masyarakat jawa timur memilih menjadi pekerja migran Indonesia. Namun kembali lagi ini hanyalah sebuah proposisi yang perlu dibuktikan. Menurut Sekarimah (2017) bahwa keadaan ekonomi keluarga TKI mengalami ketidakseimbangan antara penghasilan dan kebutuhan yang harus dipenuhi sehingga mengambil keputusan untuk bekerja menjadi TKI di luar negeri yang menjadi alternatif yang sangat menjanjikan. Tabel 1. Data Pekerja dan pengangguran Setelah kembalinya dari bekerja menjadi TKI, kebutuhan dasar keluarga TKI dapat dipenuhi sehingga menjadikan mereka manusia yang lebih baik dalam aspek ekonomi atau non ekonomi lainnya seperti aspek sosial, budaya, keamanan dan psikologi. py g ram Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 kecelakaan kerja, jaminan akan kelangsungan pekerjaannya dan jaminan akan hari tuanya pada saat mereka tidak lagi bekerja. didalam proses produksi yaitu hubungan industrial. Motivasi kerja dapat diartikan sebagai bagian integral dari hubungan/industrial dalam rangka proses pembinaan, pengembangan, dan pengarahan sumber daya manusia dalam suatu perusahaan (Sinungan, 2008:134) c. Kebutuhan sosial (social-need), Jika kebutuhan fisiologis dan rasa aman telah terpuaskan secara minimal, maka akan muncul kebutuhan sosial. Yaitu kebutuhan untuk persahabatan, afiliasi dana interaksi yang lebih erat dengan orang lain. Dalam organisasi akan berkaitan dengan kebutuhan akan adanya kelompok kerja yang kompak, supervise yang baik, rekreasi bersama dan sebagainya. Berikut ini berbagai macam teori motivasi menurut para pakarnya yaitu: Maslow (teori hierarki kebutuhan), McClelland (teori motivasi prestasi), Mc Gregor (teori X dan Y), teori motivasi Hezberg, dan Teori ERG Aldefer. Berikut penjelasannya: py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Teori Motivasi Motivasi adalah keadaan jiwa dan sikap mental manusia yang memberikan energi, mendorong kegiatan atau gerakan dan mengarah atau menyalurkan perilaku ke arah mencapai kebutuhan yang memberikan kepuasan atau mengurangi ketidakseimbangan. Kebutuhan-kebutuhan tersebut timbul akibat dari hubungan antar manusia yang dlama hal ini lebih ditekankan pada hubungan yang terjadi Menurut Rizqi (2018) factor yang menjadikan masyarakat menjadi tenaga kerja Indonesia adalah orang-orang yang 433 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License P St di M j U i it M h di h G ik J Ti I d i Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 mendapatkan penghormatan dan penghargaan masyarakat. potensinya yang meningkat karena orang mengaktualisasikan perilakunya. Seseorang yang didominasi oleh kebutuhan akan aktualisasi diri senang akan tugas- tugas yang menantang kemampuan dan keahliannya. mendapatkan penghormatan dan penghargaan masyarakat. mendapatkan penghormatan dan penghargaan masyarakat. potensinya yang meningkat karena orang mengaktualisasikan perilakunya. Seseorang yang didominasi oleh kebutuhan akan aktualisasi diri senang akan tugas- tugas yang menantang kemampuan dan keahliannya. potensinya yang meningkat karena orang mengaktualisasikan perilakunya. Seseorang yang didominasi oleh kebutuhan akan aktualisasi diri senang akan tugas- tugas yang menantang kemampuan dan keahliannya. potensinya yang meningkat karena orang mengaktualisasikan perilakunya. Seseorang yang didominasi oleh kebutuhan akan aktualisasi diri senang akan tugas- tugas yang menantang kemampuan dan keahliannya. e. Kebutuhan aktualisasi diri, kebutuhan puncak yang paling tinggi, sehingga seseorang ingin mempertahankan prestasinya secara optimal. Teori Motivasi Maslow diatas adalah teori motivasi yang bersifat umum atau general, sedangkan persamaan di dalam dunia kerja dengan teori motivasi Maslow tersebut adalah sebagai berikut: P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Teori Motivasi Maslow Teori hirarki kebutuhan dari Abraham Maslow menurut Sofyandi dan Garniwa (2007:102) terdiri dari: d. Kebutuhan penghargaan (Esteem- need), Kebutuhan ini meliputi kebutuhan keinginan untuk dihormati, dihargai atas prestasi seseorang, pengakuan atas kemampuan dan keahlian seseorang serta efektifitas kerja seseorang. a. Kebutuhan Fisiologis (Physiological- need), Kebutuhan Fisiologis merupakan hirarki kebutuhan manusia yang paling dasar yang merupakan kebutuhan untuk dapat hidup seperti makan, minum, perumahan, oksigen, tidur dan sebagainya. 434 e. Kebutuhan aktualisasi diri (self- actualization need), Aktualisasi diri merupakan hirarki kebutuhan dari Maslow yang paling tinggi. Aktualisasi diri berkaitan dengan proses pengembangan potensi yang sesungguhnya dari seseorang. Kebutuhan untuk menunjukkan kemampuan, keahlian dan potensi yang dimiliki seseorang. Kebutuhan aktualisasi diri ada kecenderungan b. Kebutuhan rasa aman (safety need), Apabila kebutuhan fisiologis relative sudah terpuaskan, maka muncul kebutuhan kedua yaitu kebutuhan akan rasa aman. Kebutuhan akan rasa aman ini meliputi keamanan akan perlindungan dari bahaya 434 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 melebihi, mencapai standar-standar, berusaha keras untuk berhasil. mana mereka bisa menentukan tujuan- tujuan yang cukup menantang. Individu berprestasi tinggi bukanlah penjudi, mereka tidak suka berhasil secara kebetulan. Mereka lebih menyukai tantangan menyelesaikan sebuah masalah dan menerima tanggung jawab pribadi untuk keberhasilan atau kegagalan daripada menyerahkan hasil pada kesempatan atau tindakan individu lain. Yang penting, mereka menghindari apa yang mereka anggap sebagai tugas yang sangat mudah atau sangat sulit. Mereka lebih menyukai tugas-tugas dengan tingkat kesulitan menengah (Robbins, 2011) b. Kebutuhan kekuatan (need for power) : kebutuhan untuk membuat individu lain berperilaku sedemikian rupa sehingga mereka tidak akan berperilaku sebaliknya. c. Kebutuhan hubungan (need for affiliation) : keinginan untuk menjalin suatu hubungan antar personal yang ramah dan akrab. Beberapa individu memiliki dorongan yang kuat untuk berhasil. Mereka lebih berjuang untuk memperoleh pencapaian pribadi daripada memperoleh penghargaan. Mereka memilki keinginan untuk melakukan sesuatu dengan lebih baik atau lebih efisien dibandingkan sebelumnya. Dorongan ini merupakan kebutuhan pencapaian prestasi. Copyright© Creative Commons Attribution 4.0 International License py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Teori Motivasi Prestasi dari Mc. Clelland Teori Motivasi Prestasi dari Mc. Clelland Teori kebutuhan Mc Clelland dalam Robbins (2011:232), (McClelland’s theory of need) dikembangkan oleh David McClelland dan rekan-rekannya. Dikatakan bahwa prestasi (achievement), kekuasaan (power), dan afiliasi (affiliation) adalah motivasi yang kuat pada setiap individu. McClelland mengajukan teori yang berkaitan dengan konsep belajar dimana kebutuhan diperoleh dari budaya dan dipelajari melalui lingkungannya. a. Kebutuhan fisiologis, kebutuhan yang pertama dan utama yang wajib dipenuhi oleh tiap individu. Kebutuhan utama inilah yang mendorong setiap individu untuk melakukan pekerjaan apa saja, karena ia akan memperoleh imbalan, baik berupa uang, ataupun barang yang akan digunakan untuk memenuhi kebutuhan utama ini Karena kebutuhan ini dipelajari, maka perilaku yang diberikan reward cenderung lebih sering muncul. McClelland juga mengungkapkan bahwa terdapat kebutuhan seseorang untuk mencapai tujuannya hal ini juga berkaitan dengan pembentukan perilaku serta pengaruhnya terhadap prestasi akademik, hubungan interpersonal, pemilihan gaya hidup, dan unjuk kerja. Hal-hal tersebut didefinisikan sebagai berikut: b. Kebutuhan keamanan atau perlindungan, tiap individu mendambakan keamanan bagi dirinya, termasuk keluarganya. c. Kebutuhan kebersamaan atau sosial, tiap individu senantiasa perlu pergaulan dengan sesamanya. d. Kebutuhan penghormatan atau penghargaan, tiap individu melakukan pekerjaan atau kegiatan yang memungkinkan ia a. Kebutuhan prestasi (need for achievement) : dorongan untuk 435 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 hukuman dan diarahkan untuk pencapaian tujuan organisasi. py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Teori X dan Y dari Mc. Gregor Teori motivasi yang menggabungkan internal dan eksternal dikembangkan oleh Mc. Gregor. Telah merumuskan dua perbedaan dasar mengenai perilaku manusia. Kedua teori itu disebut teori X dan Y. Teori tradisional terkait kehidupan organisasi banyak diarahkan dan dikendalikan atas dasar teori X. Adapun anggapan yang mendasari teori X menurut Reksohadiprojo dan Handoko (2002). Dari hasil penelitiannya terhadap kebutuhan pencapaian, McClelland menemukan bahwa individu dengan prestasi tinggi membedakan diri mereka dari individu lain menurut keinginan mereka untuk melakukan hal-hal dengan lebih baik. Mereka mencari situasi-situasi dimana bisa mendapatkan tanggung jawab pribadi guna mencari solusi atas berbagai maslah, bisa menerima umpan balik yang cepat tentang kinerja sehingga dapat dengan mudah menentukan apakah mereka berkembang atau tidak, dan di a. Rata-rata pekerja itu malas, tidak suka bekerja dan kalau bisa akan menghindarinya. b. Karena pada dasarnya tidak suka bekerja maka harus dipaksa dan dikendalikan, diperlakukan dengan 436 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Teori Motivasi dari Herzberg 437 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 dalam dorongan apa saja yang membuat masyarakat ingin menjadi pekerja imigran Indonesia. memiliki rata-rata dibawah rata-rata nilai pengangguran secara nasional, hal ini yang membuat peneliti ingin meneliti lebih Gambar 1: Model Alur Fikir Penelitian Status Pekerja Meningkat 0,63% dari tahun 2021 ke 2022 dibuktikan dengan penurunan tingkat pengangguran. Jumlah Pekerja Imigran di jawa timur tertinggi dari provinsi lain sejumlah 6.119 jiwa Jumlah pengangguran Jawa timur dibawah rata- rata nasional. Dorongan Masyarakat Jatim menjadi PMI Wawancara Mendalam Kualitatif Status Pekerja Meningkat 0,63% dari tahun 2021 ke 2022 dibuktikan dengan penurunan tingkat pengangguran. Jumlah Pekerja Imigran di jawa timur tertinggi dari provinsi lain sejumlah 6.119 jiwa Jumlah pengangguran Jawa timur dibawah rata- rata nasional. Wawancara Mendalam Dorongan Masyarakat Jatim menjadi PMI Kualitatif Gambar 1: Model Alur Fikir Penelitian Alasan menggunakan metode ini adalah peneliti akan terlibat langsung terhadap apa yang diteliti dan juga melibatkan diri terhadap obyek yang diteliti, tetapi peneliti tidak perlu ikut campur untuk melakukan perubahan di dalamnya tentang apa yang ada dan terjadi dalam subyek dan konteks penelitian. Dalam penelitian ini menggunakan desain kasus tunggal holistic, desain ini untuk melakukan eksplorasi secara mendalam (spesifik) tentang kejadian tertentu atau beberapa peristiwa dari sebuah fenomena. Penelitian ini hanya berfokus pada sejumlah kecil kejadian yang diselidiki secara mendalam dalam satu rentang waktu, dan dalam penelitian ini hanya Teori Motivasi dari Herzberg Teori motivasi yang dikemukakan oleh Herzberg dan kelompoknya. Teori ini sering disebut dengan M–H atau teori dua faktor, bagaimana manajer dapat mengendalikan faktor-faktor yang dapat menghasilkan kepuasan kerja atau ketidakpuasan kerja. Berdasarkan penelitian telah dikemukakan dua kelompok faktor yang mempengaruhi seseorang dalam organisasi, yaitu ”motivasi”. Disebut bahwa motivasi yang sesungguhnya sebagai faktor sumber kepuasan kerja adalah prestasi, promosi, penghargaan dan tanggung jawab. Teori motivasi yang dikemukakan oleh Herzberg dan kelompoknya. Teori ini sering disebut dengan M–H atau teori dua faktor, bagaimana manajer dapat mengendalikan faktor-faktor yang dapat menghasilkan kepuasan kerja atau ketidakpuasan kerja. c. Rata-rata pekerja lebih senang dibimbing, berusaha menghindari tanggung jawab, mempunyai ambisi kecil, kemamuan dirinya diatas segalanya. Teori ini masih banyak digunakan oleh organisasi karena para manajer bahwa anggapan-anggapan itu benar dan banyak sifat-sifat yang diamati perilaku manusia, sesuai dengan anggapan tersebut teori ini tidak dapat menjawab seluruh pertanyaan yang terjadi pada orgaisasi. Mc. Gregor menjawab dengan teori yang berdasarkan pada kenyataannya. Anggapan dasar teori Y adalah: Berdasarkan penelitian telah dikemukakan dua kelompok faktor yang mempengaruhi seseorang dalam organisasi, yaitu ”motivasi”. Disebut bahwa motivasi yang sesungguhnya sebagai faktor sumber kepuasan kerja adalah prestasi, promosi, penghargaan dan tanggung jawab. Kelompok faktor kedua adalah ”iklim baik” dibuktikan bukan sebagai sumber kepuasan kerja justru sebagai sumber ketidakpuasan kerja. Faktor ini adalah kondisi kerja, hubungan antar pribadi, teknik pengawasan dan gaji. Perbaikan faktor ini akan mengurangi ketidakpuasan kerja, tetapi tidak akan menimbulkan dorongan kerja. Faktor ”iklim baik” tidak akan menimbulkan motivasi, tetapi tidak adanya faktor ini akan menjadikan tidak berfungsinya faktor ”motivasi”. a. Usaha fisik dan mental yang dilakukan oleh manusia sama halnya bermain atau istirahat. b. Rata-rata manusia bersedia belajar dalam kondisi yang layak, tidak hanya menerima tetapi mencari tanggung jawab. c. Ada kemampuan yang besar dalam kecedikan, kualitas dan daya imajinasi untuk memecahkan masalah-masalah organisasi yang secara luas tersebar pada seluruh pegawai. 437 Berikut digambarkan alur fikir dalam penelitian ini, yang berawal dari data pekerja meningkat dan jumlah pengangguran yang menurun dari tahun 2021-2022 namun jumlah pekerja imigran di jawa timur masih menjadi pelopor tertinggi diantara provinsi lainnya bahkan jumlah pengangguran di jawa timur d. Pengendalian dari luar hukuman bukan satu-satunya cara untuk mengarahkan tercapainya tujuan organisasi. Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 terfokus pada hal-hal yang mendorong masyarakat jawa timur untuk menjadi pekerja migran indonesia. data juga perlu diorganisasikan ke dalam satu bentuk tertentu (display data) sehingga terlihat sosok secara lebih utuh, display data tersebut sangat diperlukan untuk memudahkan upaya pemaparan dan penegasan kesimpulan (conclusion)”. Penelitian ini hanya menggunakan teknik pengumpulan data dengan Metode Wawancara secara luring maupun daring. Penelitian ini dilakukan berdasarkan arah wilayah di jawa timur dari utara ke selatan dan dari barat ke timur. Wilayah utara diwakili area lamongan pesisir utara, wilayah selatan diwakili oleh kabupaten trenggalek, wilayah barat diwakili oleh kabupaten ngawi, dan terakhir wilayah timur diwakili kabupaten sumenep dan diutamakan pada pulau-pulau kecil di sumenep. Penentuan informan peneliti menggunakan snowball sistem dari setiap wilayah. Pemilihan ini bukan tanpa alasan, peneliti ingin mencari informasi yang banyak dan juga dalam terhadap seluruh informan. Informan dapat PMI (pekerja migran indonesia) yang masih aktif, sudah purna dan yang ingin berangkat. Dalam penelitian kualitatif rawan sekali data yang diperoleh terjadi bias. Hal tersebut wajar adanya, karena yang dicari adalah kata-kata sehingga tidak mustahil terdapat kata-kata keliru yang tidak sesuai antara yang di katakan dengan keadaan yang sesungguhnya. Guna menjaga keabsahan data dalam penelitian, maka digunakan pemeriksaan keabsahan data, dalam penelitian ini menggunakan Teknik triangulasi sumber sebagai salah satu cara memvalidasi data hasil dari wawancara. Penulis melakukan triangulasi dengan cara melakukan wawancara kembali kepada informan yang berbeda untuk mengklarifikasi jawaban informan lain, apakah mereka merasakan hal yang sama. Jika jawaban mereka sama maka dinyatakan valid. Penelitian ini menggunakan tahapan analisis data Sugiyono (2008) dapat dilakukan dengan beberapa tahapan diantaranya: “Mengumpulkan data dengan analisis data, hasil pengumpulan data tersebut tentu saja perlu direduksi (data reduction), yaitu dengan mengikhtiarkan hasil pengumpulan data selengkap mungkin dan memilahnya ke dalam satuan konsep tertentu, kategori tertentu dan tema tertentu. Seperangkat hasil reduksi py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia METODOLOGI PENELITIAN Penelitian ini menggunakan pendekatan berparadigma Kualitatif, menurut Bogdan dan Taylor dalam kutipan dari Moleong (2006) Metode Penelitian Kualitatif sebagai prosedur penelitian yang menghasilkan data deskriptif berupa kata-kata tertulis atau lisan dari orang dan perilaku yang dapat diamati. Semua yang dikumpulkan berkemungkinan menjadi kunci terhadap apa yang diteliti. Fatchan (2011) penelitian kualitatif berasaskan latar ”alamiah” yang berarti kajian dilakukan bersifat holistik pada suatu gejala individu atau kelompok masyarakat tertentu dalam konteks tertentu. 438 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Hasil penelitian Terdapat beberapa Dorongan yang membuat para Pekerja Migran Indonesia (PMI) untuk mencari penghidupan di negara lain. Adapun beberapa hal tersebut adalah dorongan dari keluarga, dorongan dari teman, pendapatan rendah di indonesia, meningkatkan ekonomi keluarga, ingin mendapatkan fasilitas lebih baik, dan pendidikan yang rendah. Dari hasil penelitian ini peneliti telah melakukan proses reduksi, display dan verifikasi dan peneliti akan menguraikan hasil tersebut satu persatu: Deskripsi Informan Informan dalam penelitian ini berasal dari berbagai kota yang ada di jawa timur. Tidak hanya asal daerah yang berbagai macam, namun dari berbagai profesi pula. Adapun beberapa profesi dari para 439 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 informan dalam penelitian ini adalah sebagai buruh (Asisten Rumah Tangga), Buruh bangunan, Guru dan juga perawat. Negara tempat PMI itu bekerja juga berbagai macam dari Hongkong, Taiwan, dan Malaysia dengan tingkatan kompensasi yang bermacam-macam pula. Penelitian ini tidak menyebutkan nama informan karena para pihak tidak berkenan namanya dipublikasikan. untuk menjadi orang sukses. Sedangkan profesinya hanyalah sebagai perawat, dimana di Indonesia ini gaji perawat tidak seberapa baik, maka yang bersangkutan memilih mencari pekerjaan di luar negeri. Tidak hanya itu, ada beberapa pernyataan PMI bahwa mereka setelah Pendidikan Menengah sudah berangkat ke luar negeri untuk bekerja, karena sudah ada Orang tua dan kakak disana yang sudah sebagai imigran. py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 3. Pendapatan rendah Tingkat pendapatan yang rendah menjadi salah satu pendorong masyarakat untuk mencari pekerjaan yang lebih baik dan memiliki pendapatan yang lebih baik pula. Upah minimum regional maupun kabupaten tidak dapat menjadi tolok ukur bahwa seluruh masyarakat di wilayah tersebut memiliki upah dengan besaran yang sama. Karena realitanya banyak dari masyarakat di wilayah yang memiliki UMR tersbesar juga memilih menjadi PMI dikarenakan UMR tidak mereka dapatkan bahkan dapat dikatakan jauh di bawah rata-rata. Terdapat beberapa informan yang profesi sebelumnya adalah nelayan dan buruh bangunan memiliki pendapatan yang tidak menentu. Jika cuaca mendukung mereka akan mendapatkan penghasilan begitu sebaliknya. Sama dengan informan yang sebagai buruh bangunan, jika ada proyek pembangunan ya kerja dan dapat penghasiln, jika tidak ada proyek maka tidak mendapatkan apapun. Kondisi tidak menentu seperti ini membuat banyak masyarakat dengan profesi tersebut memilih mencari pendapatan di luar kota maupun luar negeri agar pendapatan mereka lebih stabil. 4. Peningkatan ekonomi Hal yang paling mendasari seseorang untuk berpindah atau bermigrasi adalah demi meningkatkan perekonomian keluarga. Informan yang ada dalam penelitian ini dahulunya memiliki pendapatan yang tidak menentu dan akhirnya membuat perekonomian keluarga mereka tidaklah baik. Kondisi yang demikian membuat para informan lebih memilih menjadi PMI di negara lain yang lebih menjanjikan dan dapat meningkatkan ekonomi keluarga mereka. Pendapatan para Imigran diluar negeri cukup besar dan hal tersebut juga tidak menutup kemungkinan Biaya hidup juga besar. Terbukti dari para informan yang menyampaikan bahwa mereka harus hidup hemat di negara lain, jika tidak demikian maka mereka tidak dapat menyisihkan dana untuk keluarga di Indonesia. Jadi, demi meningkatkan perekonomian para imigran harus hidup sederhana di negara lain. 3. Pendapatan rendah Tingkat pendapatan yang rendah menjadi salah satu pendorong masyarakat untuk mencari pekerjaan yang lebih baik dan memiliki pendapatan yang lebih baik pula. Upah minimum regional maupun kabupaten tidak dapat menjadi tolok ukur bahwa seluruh masyarakat di wilayah tersebut memiliki upah dengan besaran yang sama. Karena realitanya banyak dari masyarakat di wilayah yang memiliki UMR tersbesar juga memilih menjadi PMI dikarenakan UMR tidak mereka dapatkan bahkan dapat dikatakan jauh di bawah rata-rata. Terdapat beberapa informan yang profesi sebelumnya adalah nelayan dan buruh bangunan memiliki pendapatan yang tidak menentu. Jika cuaca mendukung mereka akan mendapatkan penghasilan begitu sebaliknya. Sama dengan informan yang sebagai buruh bangunan, jika ada proyek pembangunan ya kerja dan dapat penghasiln, jika tidak ada proyek maka tidak mendapatkan apapun. Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia 2. Rekan Banyak dari para informan terdorong untuk menjadi PMI dikarenakan ajakan dari teman. Para informan sudah melihat kesuksesan para rekan yang ada disana dari sudut pandang ekonomi pribadi maupun ekonomi keluarga. Bahkan ada yang berangkat menjadi PMI benar-benar ajakan rekan untuk bisa mencari pekerjaan disana. Hal ini di sampaikan oleh salah satu pekerja imigran yang berada di negara Taiwan. Mereka menyatakan jika akan lebih baik bekerja di negeri orang daripada kerja di negeri sendiri yang tidak ada apresiasi, yang bersangkutan juga mengatakan jika dia tidak mungkin berangkat menjadi PMI jika tidak ada ajakan dari rekannya tersebut. Selain itu ada PMI yang sengaja menjadi PMI dikarenakan melihat temannya sudah bekerja di luar negeri dengan baik. 1. Keluarga atau Saudara Dari beberapa informan yang menyatakan bahwa mereka menjadi PMI karena dorongan dari keluarga yang sebelumnya sudah menjadi PMI. Tidak hanya itu ada PMI yang berasal dari Trengggalek dengan profesi perawat mengungkapkan bahwa dia menjadi PMI karena dituntun oleh keluarga besarnya 1. Keluarga atau Saudara Dari beberapa informan yang menyatakan bahwa mereka menjadi PMI karena dorongan dari keluarga yang sebelumnya sudah menjadi PMI. Tidak hanya itu ada PMI yang berasal dari Trengggalek dengan profesi perawat mengungkapkan bahwa dia menjadi PMI karena dituntun oleh keluarga besarnya 440 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Kondisi tidak menentu seperti ini membuat banyak masyarakat dengan profesi tersebut memilih mencari pendapatan di luar kota maupun luar negeri agar pendapatan mereka lebih stabil. Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia 5. Fasilitas hidup 5. Fasilitas hidup Fasilitas atau lingkungan hidup menjadi salah satu faktor yang membuat masyarakat rela meninggalkan tanah kelahirannya. Kebanyakan para PMI berasal dari wilayah yang memiliki upah minimum yang rendah, atau bahkan jika mereka tinggal di wilayah perkotaan yang upah minimumnya tinggi tapi 441 5. Fasilitas hidup Fasilitas atau lingkungan hidup menjadi salah satu faktor yang membuat masyarakat rela meninggalkan tanah kelahirannya. Kebanyakan para PMI berasal dari wilayah yang memiliki upah minimum yang rendah, atau bahkan jika mereka tinggal di wilayah perkotaan yang upah minimumnya tinggi tapi 441 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 teman lainnya yang saat itu telah lulus dari bangku perkuliahan (D3). Jadi yang memiliki pendidikan rendah rata-rata hanya akan menjadi asisten rumah tangga ataupun kuli bangunan. Dua pekerjaan tersebut yang mendominasi pekerjaan para PMI dikarenakan pendidikan mereka yang sangat rendah. Pendidikan rendah dikarenakan faktor ekonomi rendah dan keterbatasan lembaga pendidikan yang ada di tempat tinggalnya. mereka tinggal di wilayah perbatasan atau wilayah paling ujung dari kota tersebut. Terdapat salah satu wilayah asal PMI yang menjadi informan sangat terpencil, dimana wilayah tersebut tidak memiliki fasilitas yang memadai. Wilayah tersebut tidak memliki saluran listrik negara (PLN), Air Bersih, bahkan untuk berbelanja mereka harus menyebrang ke pulau lain yang memiliki fasilitas umum lebih memadai. Saat ini jaman teknologi, masyarakat di pulau tersebut tidak dapat menggunakan telepon seluler untuk komunikasi. Untuk berkomunikasi masyarakat harus pergi ke pulau seberang yang memiliki jangkauan sinyal. Pembahasan Dorongan untuk melakukan sesuatu juga disebut dengan motivasi. Beberapa unsur yang mendorong masyarakat menjadi pekerja imigran indonesia adalah ajakan dari keluarga, rekan atau teman, pendapatan rendah, peningkatan ekonomi, fasilitas hidup dan pendidikan rendah. Beberapa unsur ini yang ditemukan peneliti terhadap beberapa informan yang berkenan diwawancarai. Beberapa unsur tersebut merupakan dorongan atau motivasi individu untuk memenuhi beberapa kebutuhan dalam kehidupan pribadi maupun sosial dan setiap individu ingin memenuhi kebutuhan-kebutuhan dalam kehidupannya. py g ram Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia 6. Pendidikan rendah Standart pendidikan merupakan syarat wajib bagi perusahaan atau sebuah instansi untuk memperkerjakan karyawannya. Karena dasar pendidikan dapat memudahkan organisasi atau perusahaan menjalankan bisnisnya dan mencapai tujuan dengan mudah. Kebanyakan para PMI ber imigrasi ke negara lain untuk mencari pekerjaan dikarenakan pendidikan mereka hanya sampai SD dan SMP. Sedangkan ada beberapa yang sudah memiliki diploma 3 dengan profesi seperti perawat ataupun Bidan. Terdapat beberapa informan yang berasal dari kota trenggalek menjadi perawat di jepang dengan beberapa 442 Pada poin sebelumnya sudah dijelaskan beberapa teori motivasi dari berbagai pakar. Salah satunya adalah teori hirarki kebutuhan dari Abraham Maslow 442 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 ingin merubah kehidupannya agar lebih baik dan tidak dipandang sebelah mata oleh lingkungannya dengan cara menjadi pekerja imigran indonesia. ingin merubah kehidupannya agar lebih baik dan tidak dipandang sebelah mata oleh lingkungannya dengan cara menjadi pekerja imigran indonesia. ingin merubah kehidupannya agar lebih baik dan tidak dipandang sebelah mata oleh lingkungannya dengan cara menjadi pekerja imigran indonesia. e. Kebutuhan aktualisasi diri (self- actualization need), Aktualisasi diri merupakan hirarki kebutuhan dari Maslow yang paling tinggi. Aktualisasi diri berkaitan dengan proses pengembangan potensi yang sesungguhnya dari seseorang. Para PMI yang aktif maupun sudah non aktif membutuhkan aktualisasi diri. Kebanyakan mereka yang sudah pernah menjadi PMI akan dapat mengaktualisasikan diri terutama di lingkungan sekitarnya. Sebagai contoh salah satu informan dari wilayah pesisir utara yang pernah menjadi buruh bangunan di negara tetangga. Mereka mendapatkan ruang di lingkungan sekitarnya sebagai pekerja bangunan bahkan hingga saat ini proyek pembangunan yang dipercayakan masyarakat kepadanya tidak pernah berhenti. Hal ini membuktikan bahwa masyarakat sekitar sangat mempercayai keahliannya dalam mendirikan bangunan rumah atau yang lainnya. d. Kebutuhan penghargaan (Esteem- need), Kebutuhan ini meliputi kebutuhan keinginan untuk dihormati, dihargai atas prestasi seseorang, pengakuan atas kemampuan dan keahlian seseorang serta efektifitas kerja seseorang. Setiap manusia ingin mendapatkan penghargaan, tetapi banyak cara yang harus dilakukan untuk memenuhi kebutuhan ini. Seperti ungkapan salah satu informan yang sudah pernah menjadi PMI di negara malaysia sebagai Asisten rumah tangga. Informan tersebut mengungkapkan bahwa sebelum menjadi PMI yang bersangkutan tidak pernah dihargai oleh lingkungannya terutama keluarga besarnya. Namun setelah beberapa tahun menjadi PMI barulah lingkungan sekitar terutama keluarga menghargai keberadaannya. Terutama saat ini setelah pulang menjadi PMI informan tersebut dapat membangun rumah dan meningkatkan perekonomian keluarga maka pemenuhan kebutuhan penghargaan itu telah diperoleh. d. Kebutuhan penghargaan (Esteem- need), Kebutuhan ini meliputi kebutuhan keinginan untuk dihormati, dihargai atas prestasi seseorang, pengakuan atas kemampuan dan keahlian seseorang serta efektifitas kerja seseorang. Setiap manusia ingin mendapatkan penghargaan, tetapi banyak cara yang harus dilakukan untuk memenuhi kebutuhan ini. Seperti ungkapan salah satu informan yang sudah pernah menjadi PMI di negara malaysia sebagai Asisten rumah tangga. Informan tersebut mengungkapkan bahwa sebelum menjadi PMI yang bersangkutan tidak pernah dihargai oleh lingkungannya terutama keluarga besarnya. Namun setelah beberapa tahun menjadi PMI barulah lingkungan sekitar terutama keluarga menghargai keberadaannya. Terutama saat ini setelah pulang menjadi PMI informan tersebut dapat membangun rumah dan meningkatkan perekonomian keluarga maka pemenuhan kebutuhan penghargaan itu telah diperoleh. Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 menurut Sofyandi dan Garniwa (2007:102) terdiri dari: di hari tua nya, dengan cara menabung dan mencari rejeki yang banyak di negara lain. Karena menurut mereka pendapatan dari negara lain lebih banyak daripada hasil mereka bekerja di Indonesia. Saat para PMI berangkat ke negara lain pasti menggunakan proses legalitas, maka dari proses legal tersebut membuat keamanan mereka dalam bekerja cukup terjamin. a. Kebutuhan Fisiologis (Physiological- need), Masyarakat membutuhkan kebutuhan dasar berupa kehidupan yang layak. Kehidupan yang layak dapat berasal dari berbagai unsur, unsur finansial maupun fasilitas kehidupan. Dari hasil penelitian yang sudah dilakukan kebanyakan para Pekerja imigran Indonesia tidak mendapatkan pemenuhan kebutuhan fisiologis yang baik di wilayahnya, karena mereka rata-rata berasal dari wilayah perbatasan atau terpencil sehingga mereka memutuskan untuk menjadi Imigran. Selain posisi wilayah para imigran ini dominan memiliki lapangan pekerjaan yang sempit dikarenakan tingkat pendidikan yang rendah. Tingkat pendidikan yang rendah disebabkan oleh ketidakmampuan mereka secara finansial. c. Kebutuhan sosial (social-need), para pekerja imigran ini sangat membutuhkan pemenuhan kebutuhan sosial. Di Indonesia sendiri unsur sosial tidak dapat terelakkan dari unsur finansial. Karena banyak dari masyarakat yang memiliki karakter hanya mau bersosialisasi dengan masyarakat yang memiliki strata ekonomi yang sama, jadi mereka yang memiliki ekonomi rendah banyak yang tidak dapat bersosial dengan baik. Selain itu masyarakat juga terus memandang rendah kepada orang yang memiliki pendidikan dan ekonomi rendah. Sedangkan semnua manusia membutuhkan persahabatan, afiliasi dana interaksi yang lebih erat dengan orang lain. Maka, fenomena itulah yang membuat banyak masyarakat yang memiliki pendidikan dan ekonomi yang rendah b. Kebutuhan rasa aman (safety need), Kebutuhan akan rasa aman ini meliputi keamanan akan perlindungan dari bahaya kecelakaan kerja, jaminan akan kelangsungan pekerjaannya dan jaminan akan hari tuanya pada saat mereka tidak lagi bekerja. Para PMI terdorong untuk bekerja di luar negeri karena mereka menginginkan kehidupan yang layak b. Kebutuhan rasa aman (safety need), Kebutuhan akan rasa aman ini meliputi keamanan akan perlindungan dari bahaya kecelakaan kerja, jaminan akan kelangsungan pekerjaannya dan jaminan akan hari tuanya pada saat mereka tidak lagi bekerja. Para PMI terdorong untuk bekerja di luar negeri karena mereka menginginkan kehidupan yang layak 443 Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 Copyright© Creative Commons Attribution 4.0 International License ram Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License IMPLIKASI PENELITIAN Penelitian ini memiliki implikasi penting bagi masyarakat umum, pekerja imigran Indonesia maupun untuk penelitian yang akan datang. Masyarakat umum akan KESIMPULAN 444 Penelitian ini menghasilkan beberapa unsur yang mendorong masyarakat Jawa Timur khususnya untuk menjadi pekerja imigran indonesia, yakni ajakan dari keluarga, rekan atau teman, pendapatan rendah di Indonesia, peningkatan ekonomi, Copyright© Creative Commons Attribution 4.0 International License Program Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia P-ISSN : 2354-8592 E-ISSN : 2621-5055 Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 mengetahui hal apa saja yang mungkin terjadi saat menjadi pekerja imigran, dan bagaimana cara para imigran itu untuk bertahan hidup dinegara lain demi meningkatkan taraf ekonomi keluarga. Pekerja imigran dapat memenuhi kebutuhan-kebutuhan dalam hidupnya dengan menjadi pekerja imigran indonesia, sedangkan peneliti selanjutnya dapat menggali lebih dalam unsur lain yang mendorong para pekerja imigran untuk bekerja di luar negeri. fasilitas hidup dan pendidikan rendah. Beberapa unsur ini yang ditemukan peneliti terhadap beberapa informan yang berkenan diwawancarai beberapa waktu yang lalu. Beberapa unsur tersebut merupakan dorongan atau motivasi individu untuk memenuhi beberapa kebutuhan dalam kehidupan pribadi maupun sosial. Beberapa kebutuhan yang didapatkan dan dipenuhi oleh para informan dengan cara menjadi pekerja imigran indonesia adalah Kebutuhan Fisiologis, rasa aman, sosial, penghargaan, dan aktualisasi diri. Hal ini sesuai dengan teori dari Abraham Maslow tentang beberapa kebutuhan manusia. ACKNOWLEDGEMENT Penulis mengucapkan terima kasih yang sebesar-besarnya kepada pihak-pihak yang telah membantu dalam penyelesaian penelitian ini. Terima kasih sebanyaknya kepada para informan yang berkenan memberikan waktu dan infromasi seputar kehidupannya menjadi PMI dan perubahan setelah menjadi PMI. DAFTAR PUSTAKA Abdurrahman Fathoni. 2006. Manajemen Sumber Daya Manuisa. Jakarta: Ri Fahmi, Irham. 2018. Perilaku Organisasi. Bandung: Alfabeta. Fatchan, Ahmad. 2011. Metode Penelitian Kualitatif Beserta Contoh Proposal, Tesis Dan Disertasi. Surabaya. Jenggala Pustaka Utama. Fatchan, Ahmad. 2011. Metode Penelitian Kualitatif Beserta Contoh Proposal, Tesis Dan Disertasi. Surabaya. Jenggala Pustaka Utama. Handoko, T. Hani. 2014. Manajemen Personalia dan Sumber Daya Manusia. Yogyakarta: BPFE. https://www.bp2mi.go.id/ diakses tanggal 3 januari 2023 https://www.bp2mi.go.id/ diakses tanggal 3 januari 2023 https://www.bps.go.id/indicator/6/1953/1/jumlah-dan-persentase-penduduk-bekerja-dan- pengangguran.html diakses tanggal 3 januari 2023 445 Copyright© Creative Commons Attribution 4.0 International License py g m Studi Manajemen Universitas Muhammadiyah Gresik Jawa Timur Indonesia Copyright© Creative Commons Attribution 4.0 International License Jurnal Manajerial, Volume 10 Nomor 03 Tahun 2023 http://dx.doi.org/10.30587/manajerial.v10i03.5968 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 P-ISSN : 2354-8592 E-ISSN : 2621-5055 Husniawati. 2017. Analisis Faktor-Faktor Yang Mempengaruhi Motivasi Tenaga Kerja Indonesia Untuk Bekerja Ke Luar Negeri (Studi Pada Calon Tenaga Kerja Indonesia Di Kabupaten Lombok Timur, Ntb). Jurnal PendidikanEkonomi Undiksha. Volume 9 nomor 2 Jackson, Robert L Mathis & John H. 2002. Manajemen Sumber Daya Manusia. Jakarta: Salemba Empat. Kreitner, Kinicki. 2010. Organizational Behavior. New York: McGraw-Hill Misbach, M, L. 2006. Pengaruh Faktor Sosial Ekonomi Dan Motivasi Kerja Serta Kesejahteraan Keluarga Tenaga Kerja Indonesia Terhadap Minat Kembali Ke Luar Negeri. Surabaya: Universitas Airlangga. Moleong, Lexy J. 2013. Metode Penelitian Kualitatif. Edisi Revisi. Bandung : PT. Remaja Rosdakarya. Muchdarsyah Sinungan. 2008. Produktivitas Apa dan Bagaimana. Edisi Kedua. Jakarta: Bumi Aksara. Nawawi, Hadari. 2006. Evaluasi dan manajemen kinerja di lingkungan perusahaan dan industri. Yogyakarta: Gadjah Mada Univercity Press. Nurtjahjanti., Endah, H., & Suparno. 2017. Persepsi Tenaga Kerja Indonesia Terhadap Pilihan Kerja Di Luar Negeri Studi Diskriptif Calon Tenaga Kerja Indonesia Di Blkln Propinsi Jawa Tengah. Semarang: Fakultas Psikologi Universitas Diponegoro. Raharjo, Susilo & Gunanto, 2013, Pemahaman Individu Teknik Nontes, Jakarta: Kencana Prenada Media. Rizqi, Maulidyah Amalina. 2018. Motivasi Masyarakat Indonesia Menjadi Tenaga Kerja Indonesia (Studi Kasus Pada Masyarakat Kecamatan Panceng Kabupaten Gresik). Jurnal Manajerial. Volume 5 Nomor 2. Robbins SP, dan Judge. 2011. Perilaku Organisasi, Salemba Empat, Jakarta Sekarimah, Iis. 2017. Motivasi kerja menjadi Tenaga Kerja Indonesia (TKI) oleh masyarakat di Kecamatan Maliku Kabupaten Pulang Pisau. Undergraduate thesis, IAIN Palangka Raya. Stephen, Robbins. 2015, Perilaku Organisasi, Jakarta: Penerbit Salemba Empat. Sugiyono. 2008. Methode Penelitian Kuantitatif, Kualitatif, R dan D. Bandung: Alfabeta. Yuliarti, Y., & Intan, D, R. 2016. Migrasi Internasional Dan Dampak Sosial Ekonomi Bagi Keluarga Migran Di Desa Serah, Kabupaten Gresik, Jawa Timur. Malang: Fakultas Pertanian Universitas Brawijaya Malang. 446
https://openalex.org/W2106334845	https://zenodo.org/records/2205582/files/article.pdf	French	null	Observations de planètes	Astronomische Nachrichten	1,906	public-domain	2,257	85 85 4134 86 906.2 5 I .260 .260 .266 .271 . 2 7 I .27 I 274 .288 .288 .332 332 .345 345 ,345 a356 .356 .378 378 a318 -378 .378 .383 .383 -383 .383 .383 a383 .383 .383 .386 .386 .388 .388 .388 .388 .388 a391 .391 .391 .397 .397 ,397 -397 .391 -397 397 .499 I35?0 133.3 134.3 I3 5.1 133.7 '33.0 133.2 133.4 134.3 133.4 135.3 134.4 132.2 132.7 134.8 134.6 134.7 134.4 134.4 134.5 134.1 134.1 134.0 134.7 134.5 134.5 134.5 134.8 '34.4 135.4 133.7. 134.6 134.7 134.9 135.4 136.1 135.1 135.0 134.4 '3 5.4 134.5 136.1 '34.2 '35.5 135.5 I 3 5.5 134.9 134.5 2!'67 13ho 2.68 12.1 2.63 11.0 2.52 7.1 2.61 7.5 2.64 11.5 2.65 8.8 2.68 12.3 2.51 12.8 2.58 15.0 2.51 15.5 2.65 10.2 2.84 10.6 2.79 15.1 2.57 14.0 2.58 14.8 2.66 14.8 2.67 15.2 2.66 16.5 2.70 16.8 17.5 2.62 14.4 2.68 14.7 2.58 15.2 2.70 15.6 2.83 15.8 2.64 16.1 2.69 16.2 2.61 13.8 2.68 14.6 2.53 14.0 11.2 - - 2.78 14.9 2.64 15.4 2.73 15.7 2.75 '5.9 2.67 16.1 13.1 2.56 14.1 2.62 14.6 2.55 15.6 2.67 16.1 2.55 17.2 2.59 17.4 2.67 17.6 2.57 17.8 - 17.8 - 2.60 15.9 + 107 0.0 + 1.0 4 1.8 +0.4 -0.3 -0.1 +O.I -I- 1.1 f 0 . 2 + 2 . 1 f I . 2 - 0.9 - 0.4 +1.7 + 1.5 + 1.6 + 1.3 + 1.3 + 1.4 + 1.0 + 1.0 + 1.0 +1.7 + 1.5 +1.5 + 1.5 + 1.8 + 1.4 +2.4 +O.f +1.6 i- 1.7 + 1.9 + 2.4 43.1 +2.1 + 2.0 + 1.4 + 2.4 + 1.5 +3.1 +2.5 + 2.5 t 2 . 5 + 1.9 +1.7 + 1.2 - 0!08 - 0.0 7 -0.24 -0.14 - -0.12 -0.11 - 0.10 - 0.07 -0.24 -0.17 -0.24 -0.09 + 0.08 + 0.03 -0.19 - 0.1 8 - 0.09 -0.06 -0.14 -0.08 +0.02 -0.18 -0.06 +0.07 -0.07 -0.15 -0.08 -0.23 - 0.03 - 0.09 -0.10 -0.10 - -0.12 -0.12 - 0.0 I - - 0 . 2 0 -0.14 -0.16 - 0.09 -0.17 - 0.09 -0.19 - 0.2 I -0.21 - Kopenhagen, Urania-Sternwarte, 1906 September. 85 Die graphische Ausgleichung der nach Stundenwinkel (6) geordneten Messungen gibt folgende Normalabweichungen : 8 -4 -3 - 5h -2 - 1 0 + I + 2 +3 +4 +5 -1-6 + I Ausgleichung -0?5 -0!'06 -0.4 -0.06 -0.3 -0.06 -0.2 -0.06 +o.o -0.06 +0.3 -0.07 +0.7 -0.11 +r.z -0.12 -1-1.5 -0.13 +1.6 -0.14 +1.7 -0.13 c i . 8 -0.13 +1.8 -0.13 Rechnung -0004 -or06 -0.4 -0.06 -0.4 -0.06 -0.3 -0.06 +o.o -0.07 +0.6 -0.09 +I.3 -0.11 +1.5 -0.12 +1.6 -0.13 +1.6 -0.13 +1.6 -0.13 +1.6 -0.13 +1.6 -0.13 p aPP. I 7502 175.6 173.3 166.7 153.6 134.0 114.4 ro1.3 94.1 92.4 92.8 9 5.2 99.0 Die Gesetze der Abweichungen lassen sich mittels der Werte der scheinbaren Positionswinkel P leicht erkennen. Ftir beide Koordinaten finden wir Co + C,. cos 2 P. Die numerischen Werte der Koeffizienten sind : Distanz Co = -0ol09 C, = +0!'036 Positionswinkel Co = +0?58 C, = - 1P04 Distanz Co = -0ol09 C, = +0!'036 Positionswinkel Co = +0?58 C, = - 1P04 Nach Abzug der systematischen Korrektionen bleiben in Positionswinkel die folgenden Fehler tibrig : Beob. Beob. 1905.36 4 0 0 1 7 1906.37 -005 7 1905.96 -0.1 4 1906.38 -0.2 10 1906.16 - 0 . 2 8 1906.39 t-0.4 8 1906.27 3-0.5 5 1906.40 +0.3 1 1906.33 -0.2 I 1906.50 -0.0 4 Der wahrsch. Fehler einer Messung wird ds = folo53 dp = &0?43, also sdp = folo2o. Die Frage, ob der Campbellsche dunkle Begleiter von & Ursae majoris merkbare Stdrungen hervorruft, wird also am besten durch Messungen der Positionswinkel gelost werden. Meine bisherigen Messun- gen erfordern nicht die Annahme der Ndrlundschen Storung, widerlegen sie aber auch nicht, da eine Trennung der syste- matischen Fehler von der mutmafilichen Storung Beobach- tungen iiber mehrere Perioden erfordern wtirde. H E. Lau. faites A l'observatoire d'Uccle pa r Dr. U; Philippat, astronome-adjoint. faites A l'observatoire d'Uccle pa r Dr. U; Philippat, astronome-adjoint faites A l'observatoire d'Uccle pa r Dr. U; Philippat, as cune des planhtes. Les passages ont C t t observCs A 9 fils par la methode de l'ceil et de l'oreille jusqu'au ier octobre 1905 et A 17 fils au chronographe de Dent, dont la trans- formation a CtC terminte A cette epoque. Si possible, les Ces observations ont CtC faites au cercle mtridien de Repsold (ouverture I 6 cm, foyer I .90 rn, grossissement 80). Les corrections instrumentales ont Ctt? dCterminCes au moyen d'ttoiles fondamentales aussi voisines que possible de cha- cune des planhtes. Les passages ont C t t observCs A 9 fils par la methode de l'ceil et de l'oreille jusqu'au ier octobre 1905 et A 17 fils au chronographe de Dent, dont la trans- formation a CtC terminte A cette epoque. Si possible, les 6' 88 87 4134 deux bords ont CtC observCs. Les dCclinaisons ont t t t obtenues par un point6 au centre du champ pendant qu'un aide lisait les quatre microscopes. 11 servCs. Les dCclinaisons ont t t t obtenues par un point6 au centre du champ pendan copes. 11 deux bords ont CtC observCs. Les dCclinaisons ont t t t obtenues par un point6 au centre du champ pendant qu'un aide lisait les quatre microscopes. 11 1 2 23 10 11 12 13 151 25 171 29 2 3h49m44:75 j +of34 4 57 24.87 +0.02 12 17 11.06 +0.26 1 2 23 16.63 +0.12 12 29 25.87 +0.27 1 2 35 37.40 +0.23 18 10 28.55 +0.45 18 50 21.86 +0.44 2 3 4.56 + O . I ~ 2 54 0.69 +0.24 I 3 2 7 1.70 +0.30 +0?40 +0.09 +0.32 t o . 1 7 +0.32 f 0 . 2 8 +0.48 +0.42 +0.25 -1-0.29 t o . 3 6 re microscopes. +16"57' 15!'2 +1!'9 +1!'6 + 2 1 30 41.8 +3.4 +3.4 + o 16 35.9 3-0.4 +0.4 - o 23 52.9 +0.5 +0.3 - I 5 21.0 +1.3 + I . I - I 47 40.3 -1.8 -1.8 - - - - - - + 8 55 18.6 +0.3 +0.2 +13 36 43.2 -2.6 -2.7 1-15 43 3.8 1 -0.6 -0.7 B0 minus 8 (C. T.) I 8(N. A.) uo minus 'u(C.T.) I u(N.A.) Date 11 uo Merc ur e. +O.OI -0.04 -0.01 1904 Juin Oct. > P 0 1905 Janv. > Mai Juin I) -0.24 -0.19 -0.21 +o.o7 +0.29 +O.IO +0.20 +0.09 + o . faites A l'observatoire d'Uccle pa r Dr. U; Philippat, as I ~ +0.30 I --0.05 1904 Janv. 25 FCvr. 3 > 17 Nov. 23 Dtc. 19 D 2 0 2 1 23 1905 Janv. 13 > I4 Mai 11 1 7 )> 29 Juin 2 Juill. 26 AoQt 23 Sept. 2 2 + 9 1 7 12.4 + 7 18 41.7 +18 53 16.5 + 6 44 19.4 1905 Mai 29 Juin 19 I P 2 1 -0.43 -0.29 -0.26 -0.52 -0.45 1904 Sept. 2 7 Nov. 29 1905 Janv. 13 DCc. 18 -0.12 -17 29 19.1 +O.OI - 1 7 39 17.9 -0.01 -14 8 57.5 -0.01 -17 38 36.4 -0.07 -13 48 7.1 1904 Sept. 21 Oct. I 0 I4 1905 Sept. 15 Oct. 4 17 39 14.12 18 25 58.72 19 39 2.02 18 26 22.82 2 0 41 50.09 2 0 51 38.27 2 1 I 19.92 2 2 36 36.05 2 2 40 51.22 1 45 48.1 I 2 0 46 45.01 1 44 59.80 I 56 13.90 2 3 46.81 5 14. 0.47 7 26 48.90 9 51 45.32 I 4 31 45.80 I4 2 1 52.39 14 2 2 10.12 1 46 5.77 I 18 36.58 I 2 1 52.52 3 45 '3.76 2 1 11 9.79 2 1 8 55.85 2 1 8 45.19 2 2 2 24.30 21 58 22.80 V -0.07 -0.05 -0.15 +0.16 +0.03 - 0.09 +0.06 t o . 1 6 +0.06 +0.05 +0.08 +0.46 +0.3 I - 0 . 2 2 + O . I I -0.11 l U S . - 0.0 I - 0.04 -0.18 +0.23 +0.05 - 0.05 +0.13 t o . 0 5 - 0.03 +0.02 f-0.02 +o.ao -0.10 3-0.20 -0.01 +0.60 +0.41 - 2 1 31 32.7 - 2 2 2 27.8 - 2 0 59 23.5 -25 I 0 2.1 -20 23 54.2 - 2 0 4 22.2 -19 44 12.5 -10 4 21.1 - 9 35 21.5 +13 2 2 47.6 + I I 32 31.0 +IO 28 28.1 +lo 37 28.4 +zo 17 9.8 + 2 0 49 19.6 + I 3 35 44.4 -19 2 30.9 - 1 5 49 22.5 -15 46 51.3 -15 5 1 35.3 - 2.0 - 3.6 - 1.6 +I.7 - 2.6 - 4.5 - 0.9 -4.5 -0.5 - 0.4 - 4.6 - 5.2 -4.2 +0.8 + 0.3 + 1.7 -2.1 - 3.8 - 1.1 f I . La planhte &it plutBt devinee que vue. La planhte &it plutBt devinee que vue. faites A l'observatoire d'Uccle pa r Dr. U; Philippat, as 0 - 0.3 +0.5 + 1.0 - 2.6 - 1.1 +0.6 -3.6 -4.5 - 4.0 - 1.0 -2.5 - 0.8 f 2 . 2 - 2.8 - 4.9 - 1.3 -4.9 - 1.5 - 1.5 - 0.6 - 1.7 -0.9 +0.8 +I.I - I. I +0.3 - 1.4 +0.6 + 2.6 - 1.3 - 0.8 -3.1 +O.I -0.1 + 1.5 - 2.9 - 1.9 -0.2 planhte &it plutBt devinee que vue. 4134 Date Bo minus 8 (C. T.) 1 8 (N. A.) a. minus u (C. T.) 1 u (N. A.) ~ a0 1904 Juill. 7 a 8 > 1 1 AoQt 3 5 1905 Janv. 26 6 2 5 57.72 FCvr. 3 11 6 2 5 11.85 2 1 7 6 24 7.36 17~47~56167 I 7 47 46-16 17 47 17.63 1 7 44 9.20' '7 43 56.34 1904 Juill. 16 30 39-31 16 29 39.99 B 11 16 28 46.69 -23°31'31!'o -23 37 27.5 -23 37 20.3 -23 36 22.0 -23 36 18.0 o!'o +I.O +0.6 -0.3 -0.9 -0J22 -0.23 -0.19 -0.38 -0.16 +oJ13 +0.12 +0.13 -0.06 +0.08 + 2 2 17 43.7 + 2 2 18 26.3 + 2 2 19 33.8 - 0 Y I +0.9 +0.5 - 0.4 - 1.0 -2.1 -2.3 -2.8 +o. I -0.2 -0.6 -0.53 -0.36 -0.26 - 1.9 - I 1.8 -13.1 -0.10 +0.06 +o.17 (3) Junon. 1904 Sept. 7 11 2 0 32 10.91 I - 1 -0.60 11 - 8 48 43.5 I - I -1.9 (4) Vesta. - 2.4 - 2 2 5 31.4 - 2 2 10 1.6 - 2 2 I 4 27.4 - - - Osservazioni di pianeti Osservazioni di pianeti fatte con 1' equatoriale Dembowski (187 mm) d i Padova. 1906 1T.m.Padoval da I dd 1 Cf. i0ss.l a app. 1logp.d 1 6 app. 1ogp.d I Red. ad 1. app. * (68) Leto. - Gen. 1 7 2 0 2 0 23 23 25 26 28 29 31 Febb. I I 0 I 1 +0%2 - 91'6 +0.86 - 9.4 +0.86 - 9.4 +0.90 - 9.2 +0.90 - 9.2 +0.92 - 9.1 +0.92 - 9.0 +0.96 - 8.7 +o.97 - 8.8 +0.98 - 8.4 + I . O I - 7.8 i-0.95 - 8.8 +I.OI - 7.9 Gen. 24 24 28 Febb. faites A l'observatoire d'Uccle pa r Dr. U; Philippat, as I 7 1 7 18 I 2 2 3 3 4 5 6 7 7 8 6 8 1 0 h 2 2 m 9" 1 0 34 3 1 0 59 36 1 0 7 9 10 32 11 1 0 3 54 1 0 49 50 10 2 8 24 9 25 7 9 26 39 8 40 42 8 3' 44 7 58 56 8 55 12.08 9.325, +17 56 9.6 8 55 11.28 9.228, t 1 7 56 19.2 8 51 15.64 8.945, +18 2 0 38.9 8 32 15.69 9.068, + 2 0 11 43.8 8 32 15.45 9.068, + 2 0 11 47.5 8 31 23.56 7.050, + 2 0 16 41.1 + 4" 2 oS63 - 0 23.31 -0 24.31 -0 12.00 -0 13.13 -2 59.21 -I 38.96 -1 47.44 -2 44.69 -I 6.58 + I 11.31 +o 23.08 -0 1 0 . 1 2 0.630 +0.83 - 9.2 0.619 +0.83 - 9.3 0.603 +0.88 - 9.4 0.579 +0.97 - 9.1 0.579 +0.97 - 9.1 0.568 +0.96 - 9.0 -0' 581% -3 58.3 -3 56.4 + 2 24.1 + 2 28.4 - 2 35.2 -0 18.5 - 4 38.6 -2 14.3 -0 4.0 + I 57.1 -1 15.1 -0 41.2 10 42 2 5 11 14 29 9 42 52 9 42 52 10 39 13 I 1 35 19 -0 22.51 -0 32.7 -2 41.88 -0 17.8 -0 58.76 - 7 16.8 -2 4.18 +5 20.2 -0 46.63 -3 38.9 -2 17.94 4-0 18.0 A A F A F F F F F F F F F A A A A A A 8h26m 8S10 8 23 5.34 8 23 4.34 8 2 0 2.56 8 2 0 1.43 8 18 0.60 8 16 57.44 8 I4 57.65 8 14 0.41 8 11 5.48 8 2 57.00 8 2 8.77 8 I 2 1.93 0.437 0.402 0.3 15 0.412 0.382 0.400 0.349 0.42 I 0.406 0.460 0.407 0.446 0.3 5 6 9 10 11 12 '3 14 fatte con 1' equatoriale Dembowski (187 mm) d i Padova.
https://openalex.org/W2778736050	https://europepmc.org/articles/pmc5773297?pdf=render	English	null	Moose selecting for specific nutritional composition of birch places limits on food acceptability	Ecology and evolution	2,017	cc-by	11,149	Hilde K. Wam1 \| Annika M. Felton2 \| Caroline Stolter3 \| Line Nybakken4 \| Olav Hjeljord4 Hilde K. Wam1 \| Annika M. Felton2 \| Caroline Stolter3 \| Line Nybakken4 \| Olav Hjeljord4 1Division of Forestry and Forest Resources, NIBIO, Ås, Norway 2Faculty of Forest Sciences, Swedish University of Agricultural Science, Alnarp, Sweden 3Department of Animal Ecology and Concervation, University of Hamburg, Hamburg, Germany 4Faculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, Norway Correspondence Hilde K. Wam, NIBIO, Ås, Norway. Email: hilde.wam@nibio.no Funding information Research was partly funded by the Research Council of Norway, project #215647/E40 (Intensified harvesting of forests—implications for enterprises related to wild and domestic ungulates) 1Division of Forestry and Forest Resources, NIBIO, Ås, Norway 2Faculty of Forest Sciences, Swedish University of Agricultural Science, Alnarp, Sweden 3Department of Animal Ecology and Concervation, University of Hamburg, Hamburg, Germany 4Faculty of Environmental Sciences and Natural Resource Management, Norwegian University of Life Sciences, Ås, Norway Correspondence Hilde K. Wam, NIBIO, Ås, Norway. Email: hilde.wam@nibio.no Funding information Research was partly funded by the Research Council of Norway, project #215647/E40 (Intensified harvesting of forests—implications for enterprises related to wild and domestic ungulates) Abstract Despite decades of intense research, it remains largely unsolved which nutritional fac- tors underpin food selection by large herbivores in the wild. We measured nutritional composition of birch foliage (Betula pubescens) available to, and used by, moose (Alces alces) in natural settings in two neighboring regions with contrasting animal body mass. This readily available food source is a staple food item in the diet of moose in the high- fitness region, but apparently underutilized by moose in the low-fitness region. Available birch foliage in the two regions had similar concentrations of macronutrients (crude protein [CP], fiber fractions, and water-soluble carbohydrates [WSC]), although a notably lower variation of WSC in the low-fitness region. For minerals, there were several area differences: available birch foliage in the low-fitness region had less Mg (depending on year) and P, but more Ca, Zn, Cu, and Mn. It also had higher concentra- tions of some plant secondary metabolites: chlorogenic acids, quercetins, and espe- cially MeOH-soluble condensed tannins. Despite the area differences in available foliage, we found the same nutritional composition of birch foliage used in the two regions. Hilde K. Wam1 \| Annika M. Felton2 \| Caroline Stolter3 \| Line Nybakken4 \| Olav Hjeljord4 Compared to available birch foliage, moose consistently used birch foliage with more CP, more structural fiber (mainly hemicellulose), less WSC, higher concen- trations of several minerals (Ca, Zn, K, Mn, Cu), and lower concentrations of some secondary metabolites (most importantly, MeOH-soluble condensed tannins). Our study conceptually supports the nutrient-balancing hypothesis for a large herbivore: within a given temporal frame, moose select for plant material that matches a specific nutritional composition. As our data illustrate, different moose populations may select for the same composition even when the nutritional composition available in a given food source varies between their living areas. Such fastidiousness limits the propor- tion of available food that is acceptable to the animal and has bearings on our under- standing and application of the concept of carrying capacity. Ecology and Evolution. 2018;8:1117–1130. \| 1117 www.ecolevol.org This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2017 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd. 7 \| Revised: 26 October 2017 \| Accepted: 15 November 2017 7 \| Revised: 26 October 2017 \| Accepted: 15 November 2017 Received: 21 April 2017 \| Revised: 26 October 2017 \| Accepted: 15 November 2017 DOI: 10.1002/ece3.3715 Moose selecting for specific nutritional composition of birch places limits on food acceptability Hilde K. Wam1 \| Annika M. Felton2 \| Caroline Stolter3 \| Line Nybakken4 \| Olav Hjeljord4 1118 \| 1 \| INTRODUCTION 1118 WAM et al. WAM et al. FIGURE 1 Adult moose (Alces alces) feeding on birches (Betula spp.) in early summer, southern Norway. Photo: Hallgeir B. Skjelstad Foraging decisions are complex trade-offs, particularly for wide- roaming and long-living species like large herbivores (Parker, Barboza, & Gillingham, 2009). A fitting quote is that these animals use “most of the best and least of the worst but some of everything” (Langvatn & Hanley, 1993, p. 168). Accumulated knowledge from the field of nutritional ecology shows with increasingly detail how animal metab- olism and food selection comprise sets of synergetic or antagonistic assimilation and allocation pathways of food constituents (e.g., Boggs, 2009; Felton, Felton, Lindenmayer, & Foley, 2009; Felton et al., 2016; Raubenheimer, Simpson, & Mayntz, 2009; Sperfeld, Martin-Creuzburg, & Wacker, 2012). So far, integrated study approaches which con- sider nutritional components in synchrony are comparatively rare for large herbivores in natural settings (e.g., Beck & Peek, 2005; Shipley, Blomquist, & Danell, 1998; Tixier et al., 1997; Vangilder, Torgerson, & Porath, 1982). FIGURE 1 Adult moose (Alces alces) feeding on birches (Betula spp.) in early summer, southern Norway. Photo: Hallgeir B. Skjelstad typically inferred at the level of food availability (e.g., Herfindal et al., 2013; Wam, Hjeljord, & Solberg, 2010). To better elucidate the po- tentially masked and masking factors in food–fitness relationships, researchers need to address its finer print, that is, the nutritional un- derpinnings driving the animals’ food choices (Parker et al., 2009). Behind all animals’ food selection lies the need to assimilate ad- equate quantities of energy and various nutrients from the environ- ment. The challenge is that these components are only available in sets embedded in a food item (as “food packages”), while each com- ponent has its own functional implications for the animal. Some com- ponents are necessary for maintaining life, while others are dangerous and should be avoided (like toxins used by plants to defer herbivory, Freeland & Janzen, 1974). However, some necessary nutrients can be harmful if ingested in excessive amounts, and some toxins are ben- eficial to the consumer in low quantities (Raubenheimer & Simpson, 2009). To complicate matters further, food components have interac- tive effects (Björndal, 1991). For example, if the food contains high levels of carbohydrates relative to protein, then the animal’s ability to avoid a carbohydrate overdose depends on its capacity to endure a protein shortage. 1118 \| 1 \| INTRODUCTION Likewise, high fiber intake may inhibit mineral ab- sorption (Freeland-Graves, Sanjeevi, & Lee, 2015), while the intake of several macronutrients can influence the effects of toxins (Simpson & Raubenheimer, 2012). Keeping these multiple nutritional factors and complexities in mind, in this study, we measured nutritional composition of foliage from a staple food source (Betula pubescens Ehr.) available to, and used by, moose (Alces alces) (Figure 1) in natural settings of two neigh- boring regions of southern Norway with contrasting animal fitness (here, indexed by body mass, which is found to capture much of the fitness variance among Fennoscandian moose populations, Tiilikainen, Solberg, Nygrén, & Pusenius, 2012). Long-term research focus has not managed to fully explain the contrasts between the populations’ food selection and demographic performance (Hagen, 1983; Hjeljord & Histøl, 1999; Wam, Histøl, Nybakken, Solberg, & Hjeljord, 2016). Although moose in both regions have access to birch in excess per capita (Wam et al., 2010), the low-fitness population utilizes it to a no- ticeably lower extent than does the high-fitness population (Figure 2). The study was initiated to explore whether the nutritional composition of the birch foliage could explain this apparent underutilization. We tested whether contents of food constituents (crude protein [CP], fiber fractions, water-soluble carbohydrates [WSC], minerals, and plant sec- ondary metabolites [PSM]) differed between areas, and between used and available foliage. We then used principal component analyses (PCA) to place the differences in a multidimensional framework, con- sidering constituents in synchrony. Achieving nutritional homeostasis, therefore, involves a complex interplay between variable foods, and multiple and changing needs as the animal goes through different life stages and seasons. While evolution has equipped animals with mechanisms to deal with these complexities (Behmer, 2009), nutritional ecologists are still puzzled to understand them. One major explanation is that the nutritional value and selection of a given food item may show extensive spatiotemporal variation (Morgantini & Hudson, 1989). Designing studies that grasp most of the variance of interest can therefore be difficult without extensive prior knowledge of the study system. For example, Jones, Strickland, et al. (2010) found that the availability of various soil re- sources providing different nutritional planes for deer can explain as much as 78% of the variation in their body mass. McArt et al. Abstract carrying capacity, fitness, mineral, nitrogen, preference, ungulate \| 1117 www.ecolevol.org 1118 \| 1 \| INTRODUCTION (2009) measured protein availability in major browse species for moose and found that the within-species variation between two areas was so large that a similar diet and food intake would yield a substantially different protein balance for the moose. These relationships inferred at the level of nutrient availability appear much stronger than those 2.1 \| Study area The two sample areas used for foliage analyses in this study are situ- ated 100 km apart in southeastern Norway (SandeW at 59°42′N, 10°7′E in the low-fitness region and Rakkestad at 59°30′N, 11°22′E in the high-fitness region). A fjord and densely populated areas practi- cally eliminate exchanges of moose between the low- and high-fitness \| 1119 WAM et al. FIGURE 2 Contrasting moose fitness and utilization of a readily available food source (Betula pubescens Ehr.) in five study areas within two regions of southern Norway (modified from Wam et al., 2010, 2016). Birch density was high in all areas (2,470 ± 252 birches available/ ha in the low-fitness region, and 4,659 ± 311 in the high-fitness region, and <20% of available birches were browsed in both regions). Birch in the diet was estimated from counting browse marks on woody plant species along line transects and corrected for nonwoody diet contents as found by fecal analyses (see Wam & Hjeljord, 2010a). In this paper, one sample area in each region was used to collect and analyze nutritional composition of available and used birch foliage, with the aim to explore why moose in the low-fitness region does not utilize the readily available birch to a larger extent. Presumably, birch could be a remedy if food shortage is their culprit for higher fitness sample area SandeW sample area Rakkestad R² = .5984 R² = .0236 40 45 50 55 60 65 70 75 80 0 10 20 30 40 50 60 70 80 90 Calf body mass (dressed kg) Birch in summer diet (% of all browsed trees) Low-ﬁtness region High-ﬁtness region \| 1119 sample area SandeW sample area Rakkestad R² = .5984 R² = .0236 40 45 50 55 60 65 70 75 80 0 10 20 30 40 50 60 70 80 90 Calf body mass (dressed kg) Birch in summer diet (% of all browsed trees) WAM et al. 1119 Low-ﬁtness region High-ﬁtness region Birch in summer diet (% of all browsed trees) FIGURE 2 Contrasting moose fitness and utilization of a readily available food source (Betula pubescens Ehr.) in five study areas within two regions of southern Norway (modified from Wam et al., 2010, 2016). 2.1 \| Study area Birch density was high in all areas (2,470 ± 252 birches available/ ha in the low-fitness region, and 4,659 ± 311 in the high-fitness region, and <20% of available birches were browsed in both regions). Birch in the diet was estimated from counting browse marks on woody plant species along line transects and corrected for nonwoody diet contents as found by fecal analyses (see Wam & Hjeljord, 2010a). In this paper, one sample area in each region was used to collect and analyze nutritional composition of available and used birch foliage, with the aim to explore why moose in the low-fitness region does not utilize the readily available birch to a larger extent. Presumably, birch could be a remedy if food shortage is their culprit for higher fitness The climate in the study area is continental with cold winters (February norm −4.5°C in SandeW and −5.6°C in Rakkestad) and warm summers (June norm 14.9°C in SandeW and 13.7°C in Rakkestad) (Norwegian Meteorological Institute, 2013). Start of growing sea- son (first day of the year with mean temperature >5°C) is 2 May in SandeW and 25 April in Rakkestad. Normal precipitation during June is 59 mm (SandeW) and 65 mm (Rakkestad). Norms are based on the years 1961–1990. regions to which the sample areas belong (Figure 2). Both regions are part of the boreal forest zone (Moen, 1999), dominated by com- mercially cultivated Norway spruce (Picea abies), with some Scots pine (Pinus sylvestris) on drier sites of poor soil fertility. Younger for- est stages are dominated by birch (Betula spp.), sparsely intermixed with other deciduous species: in SandeW, birch make up 78% of the browse biomass on a typical clearcut (<20 years since logging, inter- mediate soil fertility), compared to 95% in Rakkestad (Wam et al., 2010). In the field layer, bilberry (Vaccinum myrtillus) is the most abun- dant forage plant in the older forest, and grasses in the younger for- est. Comprehensive data on plant abundances are given in Wam and Hjeljord (2010a). are stated per dry weight (mg per DW). Briefly outlined, we ground the samples, conducted four series of cold-methanol extractions and then ran the samples through High Pressure Liquid Chromatography (HPLC, 1100 series, Agilent USA) (for more details, see Nybakken, Hörkkä, & Julkunen-Tiitto, 2012). We quantified phenolic acids and flavonoids at 320 nm. Individual compound concentrations were cal- culated based on available commercial standards. In the result section, we have collected these into major groups, while the individual meas- ures are listed in the appendix (Table A1). We analyzed condensed tannins from the HPLC-extract (MeOH-soluble fraction) and from the dried residue after phenolic extractions (MeOH-insoluble fraction) with the acid butanol assay (Hagerman, 2002). We calculated these concentrations using purified condensed tannins from Betula nana (dwarf birch) leaves. We systematically sampled foliage from N = 9 pristine trees (no obvious signs of herbivory, damage, or disease) per clearcut each year. These samples represent the “available” birch foliage in our study. Sampling was systematically spread out along a fixed cross-sectional pattern of the clearcut, starting with one tree in the center and two each in the four perpendicular directions from center (excluding the last 10 m to avoid edge effects). The aim was to spatially distribute our sampling as much as possible in order to cover the most variance in local growth conditions for birch on the clearcut and to avoid observer bias when selecting trees. We therefore kept a straight path, not yield- ing for hurdles like ditches or piles of logging waste. After walking the set distance, we sampled the nearest available tree. We opted to in- clude only pristine trees when sampling for available foliage, because plant responses to herbivory attacks are so diverse (Kessler & Baldwin, 2002). One can theorize that we thereby omitted attractive foliage, and rather sampled what had been discarded by moose. However, there were thousands of birches on the clearcuts in the study area (4,659 ± 311/ha in Rakkestad, 2,470 ± 252 in SandeW, N = 576 plots), and only 16 ± 2.4% (Rakkestad) and 15 ± 1.9% (SandeW) of the avail- able trees were browsed by moose in summer (H. K. Wam, unpublished data). We therefore consider our samples to be a random selection of the available foliage, and thus a representation of its average nutri- tional composition on the clearcuts. Prior to nutrient analyses, we estimated dry matter concentration by oven-drying subsamples at 103°C. Because dry matter concentra- tions of samples already were so high from air drying (91.1 ± 0.13%), samples analyzed for nutritional contents were not additionally dried in the laboratory. All nutrient measures are stated per dry matter, that is, corrected for remaining moisture in the dried samples. We esti- mated CP using a thermal conductivity detector (Leco FP-528; Leco®, St. Joseph, USA) and the 990.03 calculation (a standard established by the Association of Official Analytical Chemists, AOAC, 2012). We de- termined structural carbohydrates using filter bag techniques (Ankom Technology A200), that is, method 6 for neutral detergent fiber (NDF) (Van Soest, Robertson, & Lewis, 1991), method 5 for acid detergent fiber, and method 9 for acid detergent lignin (Daisy II Incubator, solu- tion as in 973.18, AOAC, 2012). The contents of fiber fractions stated in the text are adjusted for residual starch and protein (i.e., aNDF), but not for residual ash (i.e., not aNDFom). We estimated WSC with a spectrophotometer (Genesys 10S Vis; Thermo Fisher Scientific, Inc., Waltham, USA), following Hall, Hoover, Jennings, and Miller Webster (1999). We analyzed contents of minerals with inductively coupled plasma spectrometry (ICP-AES) (iCAP 6300 Radial; Thermo Fisher Scientific, Inc., Waltham, USA) after microwave digestion (EAM sec. 4.4, FDA, 2013). Trees with signs of current summer browsing by moose (i.e., leaf stripping) were sampled opportunistically throughout the clearcut. Upon visually detecting leaf stripping, we consistently sampled the tree closest to us (sometimes several trees in a cluster were stripped). Samples could not stem from the same cluster of trees. These sam- ples represent the “used” foliage in our study. We assume they reflect the carte blanche choice of moose, that is, their individual nutritional composition had not (yet) been influenced by the browsing, and thus, reflect the nutritional composition that moose were seeking in this fo- liage. This approach is the only option when one wishes not to manip- ulate the foodscape. Three clearcuts in SandeW lacked browsed trees. We defoliated each sample tree mimicking moose browsing along the 20–30 outer cm of the current year’s growth of the leader shoot (including leaves and minor parts of petioles). If the leader shoot on used trees had been browsed by moose, we defoliated the neighboring shoot. 2.4 \| Statistical analyses We tested for differences in nutritional compositions using fac- torial analyses of variance (“lm” in R, version 2.15.3, R Core Team, 2013), with use (available, used), region (high fitness, low fitness) and year (2012, 2013) as categorical predictors (specified as factors). Homogeneity of response variances across each predictor level was checked by graphical inspection of residuals from exploratory linear fits (Zuur, Ieno, & Smith, 2007) and found adequate apart from for sodium. We therefore opted to use observations directly, with no variance-stabilizing transformations. Each response parameter (nu- trient or chemical group of secondary compounds) was tested in a separate model with the explanatory predictors (use, region, year) as fixed effects. Generalized models fitted with logit link function and bi- nomial distribution for proportional data (McCullagh & Nelder, 1989) gave consistently the same outcome as our ordinary linear models. To Summer browsing intensity per tree was very low, with only a few shoots browsed per tree (Wam & Hjeljord, 2010b). Therefore, sampled foliage always stemmed from shoots in the central apex portion of the tree (not from the side branches, which may have a different chemical composi- tion Hjeljord, Høvik, & Pedersen, 1990). Composite samples of available and used foliage, respectively, were combined in sealed plastic bags on site and placed in open paper traces when we returned to the field quar- ters in the afternoon. The foliage was then let to air-dry inside with no exposure to sunlight. After 3–5 days, foliage had reached a constant dry weight concentration of 91.1 ± 0.13%. 2.2 \| Data collection We collected foliage for obtaining nutritional profiles between 19/06/2012 and 06/07/2012 and between 24/06/2013 and 12/07/2013, alternating between the low-fitness and the high- fitness region every 3rd day to avoid bias from sample date. Sites to be sampled were randomly drawn from all available clearcuts of age 5, 10, or 15 (±1 year) years since clearing (8 replicas of each), on intermediate soil fertility (defined as G14 or G17 on the H40 sys- tem) (N = 24 clearcuts each for SandeW and Rakkestad). The same clearcuts were sampled in 2 years in order to account for potential influence of weather, which strongly influences nutrient compo- sition (and moose selection) of browse (Bø & Hjeljord, 1991). June was colder and drier in 2012 than in 2013: mean temperature/pre- cipitation was 12.1°C/83 mm versus 13.8°C/142 mm in SandeW, and 11.8°C/108 mm versus 13.6°C/134 mm in Rakkestad (Norwegian Meteorological Institute, 2013). The soil fertility in the study area is intermediate, and generally a little higher in SandeW, where approximately 60% of area is classified as ≥G14 (in the H40 forest productivity index) compared to 40% in Rakkestad (Wam et al., 2010) (see Tveite, 1977 for details on the H40 index, which indicates tree height when trees are 40 years at breast height = 1.3 m). A typical clearcut on <G14 sites produces only about half as much deciduous browse as do clearcuts on the more fertile sites. Practically, all loggings are performed as clearfelling, and clear- cuts are small (averaging about 1.5 ha) in a global perspective. Tops and branches are traditionally left on site to decompose, and new spruce forest is almost entirely recruited by planting. The use of her- bicides, pesticides, scarification, and fertilizers is generally scarce in the area and had not been applied on the clearcuts used for foliage sampling in this study. 1120 WAM et al. WAM et al. 4 \| DISCUSSION Our study of moose selectivity of nutritional compositions in birch foliage in areas with contrasting fitness produced three key findings. First, the moose showed a clear selection pattern of food constituents, as the nutritional profiles of birch foliage that they used differed sig- nificantly from the birch foliage that was available to them on young clearcuts. Second, our results highlight the complex interconnections between macronutrients and micronutrients and PSM in a staple food source and their potential influences on consumers. Third, the area differences in the availability of foliage with the preferred nutritional composition offer insights into observed contrasting fitness of the dis- tinct moose populations living there. Below, we try to integrate these key findings. It is important not to generalize the specific selection pattern of food constituents in our study (peak of growing season) to other times of the year, because this is likely to differ with season (e.g., Stolter, Ball, & Julkunen-Tiitto, 2013; Tixier et al., 1997). Macronutrient concentrations in the available birch foliage were largely similar in the two regions (Table 1, Figure 3), with only hemi- cellulose and cellulose being slightly lower in the high-fitness region (hemicellulose only in year 2013, Figure 4). Notably, there was also a wider range of available concentrations of WSC in the available foli- age in the high-fitness region, especially in 2013 (thus a significant area × year interaction). Area differences were stronger for minerals: the low-fitness region had available birch foliage with more calcium and zinc (both more so in 2013), as well as more iron, copper, and manganese than the high-fitness region. In contrast, birch foliage in the low-fitness region had less phosphorous and magnesium (Mg only in 2012). According to the PCA, nutrients in the available foliage co- varied in a distinct pattern: calcium, zinc, manganese, and structural carbohydrates formed one cluster, while potassium, copper, phospho- rous, and CP formed another cluster, and magnesium and WSC a third cluster (Figure 5a). It was mainly the first and third of these clusters that separated the two regions in the biplots. 4.1 \| Selection of food constituents: A challenging balancing act Moose showed consistent selection for a specific nutritional com- position across areas, that is, there were practically no significant area × use interactions (Table 1, Figures 3 and A1). Compared to the availability, moose used foliage with more CP, calcium, zinc, manga- nese (particularly in the low-fitness region), potassium, and copper, but less WSC. The selection of structural carbohydrates differed between years (Table 1): Lignin concentration in the used foliage was lower than in the available foliage in 2012, but not so in 2013 (Figure 4). In 2013, the used foliage also had more hemicellulose compared to the avail- able foliage. Irrespectively of region, the moose in our study consistently selected birch foliage with higher concentration of CP and several minerals (most strongly Ca, Zn, and Mn) than in the average birch foliage that was available to them. Because of covariation with other food constit- uents, we must not interpret the positive single statistical coefficients as intended selection for a given food constituent (Table 1), but rather look at them in synchrony. A few previous studies on moose selection of plant material within specific plant species have also shown a corre- lation between food selection and covarying contents of protein and minerals (e.g., Danell, Niemela, Varvikko, & Vuorisalo, 1991; Faber & Lavsund, 1999; Thompson, McQueen, Reichardt, Trenholm, & Curran, 1989). Thompson et al. (1989), for example, concluded that moose selected for stands of balsam fir which had protein contents that met their requirement. Because protein comprises a much larger part of animal diet than do minerals, it is easy to overemphasize by default the role of protein. However, perhaps an animal uses a particular food source to adjust the overall intake of other food constituents. A case facilitate direct interpretation of model output, we prefer not to trans- form data unless necessary. We therefore opted to present the linear models in the paper. Final models were validated by lack of patterns in residuals plotted against fitted values and QQ plots of standard- ized residuals (Zuur et al., 2007). To visualize how nutrient concen- trations covaried, we ran principal component analyses (“prcomp” in R). Because of large differences in concentrations between nutrients (e.g., carbohydrates in the magnitude of 30% vs. trace elements in the magnitude of 3‰), we centered and scaled concentrations for each nutrient prior to the PCA (van den Berg, Hoefsloot, Westerhuis, Smilde, & van der Werf, 2006). MeOH-soluble condensed tannins (and slightly also myricetins) were lower in used than in available foliage. The available foliage had al- most a twice as high concentration of these tannins in the low-fitness region than in the high-fitness region. The low-fitness region also had significantly more chlorogenic acids and quercetins (no difference be- tween available and used). A principal component biplot showed clear area separation of available foliage by differences in the soluble ver- sus insoluble condensed tannins (Figure 5b). Notably, there was also a positive covariation between the soluble condensed tannins and di- gestible carbohydrates (hemicellulose, cellulose, and WSC). Additional biplots indicate that moose used foliage with generally lower concen- trations of PSM than available, and more so in the low-fitness region than in the high-fitness region (Figure A2). 2.3 \| Chemical analyses We measured concentrations of low molecular weight phenolics di- rectly on the air-dried samples from 2013. The phenolic measures WAM et al. 1121 3.2 \| Contents of plant secondary metabolites There were also area differences in the concentrations of plant sec- ondary metabolites (PSM) in the birch foliage, particularly concern- ing MeOH-soluble condensed tannins (Figure 6). The available foliage had less MeOH-insoluble tannins in the low-fitness region than it had in the high-fitness region, but these tannins did not differ be- tween available and used foliages. In contrast, the concentrations of 1122 WAM et al. WAM et al. 1122 \| WAM TABLE 1 Concentrations of nutrients (% of dry matter) and plant secondary compounds (mg per dry weight) in birch foliage available to or useda by moose on boreal forest clearcuts (N = 48)b in two Norwegian regions with contrasting animal fitness (low and high body mass). Foliage was sampled in late June to early July 2012 and 2013 (June 2012 was colder and drier than 2013). Statistical testsc ran as sequential contrasting against the reference level high-fitness region, available foliage, year 2012 (i.e., the intercept). Single coefficients must be interpreted in relation to the reference level and interaction effects. A simplified example of how to read the table: lignin concentration was 7.3 in the available birch foliage for both regions in 2012 (no influence of “area”), while it was 0.6 lower in 2013 and 1.8 lower in used than in available. However, there was a positive year × use interaction which largely outweighs these two negative influences (lignin was actually higher in used than in available foliage in 2013, see also Figure 4). 3.2 \| Contents of plant secondary metabolites We have put coefficients most central to interpreting difference between available and used in bold font Response Coefficients [t, p-value] α (intercept) β1 (area) β2 (year) β3 (use) β1 × β2 β1 × β3 β2 × β3 Crude protein (%) 15.0 n.s n.s 1.2 [2.8, 0.005] n.s n.s — Neutral detergent fiber (NDF) (%)d 30.1 n.s 1.4 [1.5, 0.144] −1.9 [−1.5, 0.138] n.s n.s 4.7 [2.7, 0.007] Acid detergent fiber (ADF) (%) 16.9 n.s −0.7 [−1.8, 0.083] −1.3 [−2.4, 0.016] n.s n.s 2.1 [3.0, 0.004] Lignin (%) 7.3 n.s −0.6 [−2.5, 0.013] −1.8 [−2.4, 0.018] n.s n.s 2.0 [4.6, 0.000] Hemicellulose (NDF – ADF) 13.7 0.3 [0.5, 0.641] 0.1 [0.1, 0.913] −0.6 [−0.7, 0.490] 2.5 [2.4, 0.018] n.s 3.5 [3.1, 0.003] Cellulose (ADF – lignin) 9.3 0.6 [3.0, 0.004] n.s n.s n.s n.s n.s Water-soluble carbohydrates (WSC) (%) 22.3 −2.6 [−2.7, 0.007] −8.1 [−8.6, 0.000] −1.8[−2.2, 0.027] 5.9 [4.5, 0.000] n.s n.s Calcium (Ca) (‰) 5.2 0.4 [1.4, 0.152] −0.3 [−1.3, 0.200] 0.9 [4.2, 0.000] 0.9 [2.6, 0.014] n.s n.s Phosphorous (P) (‰) 2.4 −0.2 [−2.6, 0.010] −0.2 [−2.0, 0.040] n.s n.s n.s n.s Magnesium (Mg) (‰) 3.0 −0.5 [−6.1, 0.000] −0.2 [−2.7, 0.008] n.s 0.4 [3.0, 0.004] n.s n.s Potassium (K) (‰) 7.2 n.s 0.6 [2.4, 0.018] 0.7 [2.1, 0.042] n.s n.s −0.6 [−1.4, 0.150] Sodium (Na) (‰)e 1.0 2.6 [7.4, 0.000] −0.4 [−1.0, 0.312] 0.1 [0.1, 0.910] −2.8 [−5.1, 0.000] −2.8 [−3.5, 0.000] −0.0 [−0.0, 0.985] Iron (Fe) (PPM) 53.8 4.4 [2.2, 0.032] n.s n.s n.s n.s n.s Zinc (Zn) (PPM) 190 98.2 [5.5, 0.000] −7.5 [5.5, 0.000] 69.1 [4.9, 0.000] 132.1 [5.2, 0.000] n.s n.s Copper (Cu) (PPM) 7.3 0.7 [4.2, 0.000] −1.2 [−5.8, 0.000] 0.7 [2.5, 0.013] n.s n.s −0.5 [−1.4, 0.165] Manganese (Mn) (PPM) 1550 990 [7.1, 0.000] n.s 560 [3.4, 0.000] n.s 494 [2.0, 0.051] n.s Molybdenum (Mo) (PPM) 0.2 n.s −0.1 [−3.3, 0.001] n.s n.s n.s n.s MeOH-soluble condensed tannins (mg per DW) 5.6 3.5 [5.1, 0.000] −1.6 [−2.3, 0.025] n.s MeOH-insoluble condensed tannins (mg per DW) 31.9 −15.5 [−10.6, 0.000] n.s n.s Chlorogenic acid and deriva- tives (mg per DW) 1.6 2.1 [7.3, 0.000] n.s n.s Hydroxycinnamic acids (HCAs) (mg per DW) 0.9 n.s n.s n.s Myricetin glycosides (mg per DW) 1.0 n.s −0.2 [−2.0, 0.042] n.s (Continues) (Continues) Coefficients [t, p-value] Our study was conducted at times of peak contents of CP in browse (e.g., Leslie, Starkey, & Vavra, 1984; Marshal, Krausman, & Bleich, 2005). Yet, whether protein contents of food actually contrib- ute to drive food selection depends on its scarcity relative to animal needs (see, e.g., Beck, Flinders, Nelson, & Clyde, 1996; Zweifel-Schielly, Kreuzer, Ewald, & Suter, 2009; Dostaler, Ouellet, Therrien, & Cote, 2011 compared to, e.g., Tixier et al., 1997; Gillingham, Parker, & Hanley, 2001; Zweifel-Schielly et al., 2012). Moose require roughly 6%–8% CP of dry matter intake for maintenance (Schwartz, Regelin, & Franzmann, 1987), but up to 25% for reproduction and growth (review across cer- vids; Dryden, 2011). The birch foliage in our study had about 15% CP, that is, sufficient for maintenance, but possibly deficit for production. Protein is not only a source of amino acids for cell renewal but also energy. This is pertinent to the question of whether a search for protein ever drives the food selection of an animal (Felton et al., 2009): is a selection for food with more protein due to a need for energy or for amino acids? In our study, protein was higher in used than available foliage, while other sources of easily digested energy were not (espe- cially WSC, Figure 3). This indicates that amino acids were of greater physiological importance to moose than was energy when it foraged on birch foliage. To fully disentangle these two attributions, we would have to look at the complete food intake (Felton et al., 2016), taking into account all the potential covariations of importance as indicated by the biplots in our study (Figure 5). n.s The moose in our study actually used birch foliage with contents of WSC being lower than in the available birch foliage. Notably, the within- year variation of WSC in available foliage was lower in the low-fitness region than in the high-fitness region, offering moose less of a choice. WSC have seldom been studied in relation to the diet of large herbivores in natural settings (but see Beck et al., 1996; Faber & Lavsund, 1999; Tixier et al., 1997), and it seems premature to routinely ignore it. WSC are one of the highly fermentable sources of energy that is assumed to increase palatability for most animal species (e.g., Jones & Roberts, 1991). Coefficients [t, p-value] TABLE in two N Statistic the refe “area”), w was actu Respon Crude p Neutral (%)d Acid det Lignin (% Hemice Cellulos Water-s (WSC) Calcium Phospho Magnes Potassiu Sodium Iron (Fe Zinc (Zn Copper Mangan Molybd MeOH- tannins MeOH- tannins Chlorog tives (m Hydroxy (mg pe Myrice DW) WAM et al. 1123 Response α (intercept) β1 (area) β2 (year) β3 (use) β1 × β2 β1 × β3 β2 × β3 Quercetin glycosides (mg per DW) 4.1 1.1 [2.9, 0.005] n.s n.s Kaempferol glycosides (mg per DW) 2.1 0.5 [1.7, 0.099] 0.2 [0.7, 0.471] −0.7 [−1.7, 0.094] Apigenin glycosides (mg per DW) 1.0 −0.4 [−2.3, 0.026] n.s n.s Flavonoids (mg per DW) 6.4 0.9 [1.5, 0.140] n.s n.s Available = foliage from a random sample of undamaged trees that had not (yet) been browsed by moose. Birches were available in very high densities on the clearcuts (mean 3,565 ± 282/ha across study areas), o we consider these samples to represent a cross-section of available birch foliage (not rejected foliage). Used = foliage from trees with recent browsing marks from moose (i.e., leaf stripping). One municipality selected as sampling area in each region. Clearcuts were randomly drawn from all the area’s clearcuts of intermediate site fertility and age 5, 10 or 15 years since clearing (balanced design). Chemical analyses on composite samples per clearcut, made from 9 ± 0.0 (available) and 3 ± 0.2 (used) trees. The same clearcuts were sampled in both years (secondary compounds only measured in 2013). Linear model, no transformations applied. Generalized models with logit link and binomial correction (quasi-binomial, approximated Wald-statistics) gave consistently the same results. Lignin is practically ingestible to moose, so the digestible fractions of food fiber are hemicellulose and cellulose. β1 × β2 × β3 = 2.7 (2.7, 0.009). Available foliage in the low-fitness region in 2012 had a very large variance (and thereby a higher mean) in sodium, causing the 3-way interaction to be significant. in point, the “selection for” higher Ca by moose in our study may actu- ally reflect a need for Zn, traded against the costs of accompanying excess Ca (see below). A selection for food with higher protein contents was expected in our study because large herbivores normally can meet their protein needs only during plant growing season (Mattson, 1980; Parker et al., 2009). Coefficients [t, p-value] However, it could be that there are interactions with other food constit- uents that wildlife research is not yet aware of. One such may be positive covariation with soluble condensed tannins (Figure 5b). In contrast to protein and WSC, the moose in our study appeared to be quite flexible on the fiber structure when selecting birch foliage, with hemicellulose in used foliage being similar as in available foliage in 1 year (2012) and higher in the other year (2013). This is a fine reminder to ecologists that weather conditions produce interannual variation in nutritional compositions (e.g., Vázquez-de-Aldana, García-Ciudad, & García-Criado, 2008), and subsequently in animals’ food selection. Interestingly, the foliage used also had higher lignin concentrations in 2013 than in 2012. Because lignin is practically indigestible to rumi- nants (Van Soest, 1994), it is generally expected that they select for food with lower lignin concentrations. Possibly, the moose had to “accept” 1124 WAM et al. WAM et al. 1124 \| WAM et FIGURE 3 Nutritional profiles of birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, lat June to early July 2012–2013. Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers) for nutrients where used foliage significantly differed from available foliage. See Table 1 for complete nutritional profiles, as well as the influence of year FIGURE 3 Nutritional profiles of birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, late June to early July 2012–2013. Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers) for nutrients where used foliage significantly differed from available foliage. See Table 1 for complete nutritional profiles, as well as the influence of year birch foliage with more lignin in 2013 because the benefits from other food constituents outweigh the reduced digestibility from lignin. Findings in our study that stand out in relation to minerals mainly pertain to Ca, Zn, Cu, and Mn. Although P and Mg were of similar concentrations in used and available birch foliage, it is also worth noting the area difference in their ratios to Ca. The ratio between Ca, P, and Mg is crucial for calcium homeostasis which is one of the most sensitive homeostasis in the body (Arnaud, 1983). Coefficients [t, p-value] The low-fitness region had available birch foliage with more Ca, and simultaneously a lit- tle less P and Mg (ratio Ca:P:Mg = 2.8:1:1.2) than the high-fitness region (ratio 2.2:1:1.3). Particularly, the former ratio is not within recommendations for large ruminants (NRC, 2001), where Ca intake should preferentially stay within 1–1.5 times the P and Mg intake (lower ratios for maintenance than for bone production). The moose in our study did not select birch foliage in line with NRC recom- mendations, however, as the ratios were even higher in used than in available foliage (ratio used 3.3:1:1.2 for low fitness, and 2.4:1:1.2 for high fitness). This may be a result of constraints in selection op- tions rather than overall nutritional preferences, as the minerals (as are all food constituents) are only available to the animals as intri- cate complexes in “food packages.” Possibly the moose could bal- ance these mineral ratios by adjusting the intake of other food items in the diet. consistent across species (Dryden, 2016). The low-fitness region had available birch foliage with more Ca, and simultaneously a lit- tle less P and Mg (ratio Ca:P:Mg = 2.8:1:1.2) than the high-fitness region (ratio 2.2:1:1.3). Particularly, the former ratio is not within recommendations for large ruminants (NRC, 2001), where Ca intake should preferentially stay within 1–1.5 times the P and Mg intake (lower ratios for maintenance than for bone production). The moose in our study did not select birch foliage in line with NRC recom- mendations, however, as the ratios were even higher in used than in available foliage (ratio used 3.3:1:1.2 for low fitness, and 2.4:1:1.2 for high fitness). This may be a result of constraints in selection op- tions rather than overall nutritional preferences, as the minerals (as are all food constituents) are only available to the animals as intri- cate complexes in “food packages.” Possibly the moose could bal- ance these mineral ratios by adjusting the intake of other food items in the diet. protein, as well as P and K. The protein-P-K covariation (in an opposite direction of Ca) has previously been demonstrated in cervid food (e.g., Vangilder et al., 1982). These two major interacting complexes put lim- itations on the moose’ option to compose a nutritionally balanced diet from birch. Coefficients [t, p-value] In large herbivores, Ca in the blood must be maintained within the narrow range of 1.00–1.25 mmol/L (NRC, 2001). The Ca ratios to other min- erals need to be narrowly balanced because minerals affect each other’s absorption in the animal body by forming insoluble com- plexes (Spears, 2003). The Ca:P:Mg mineral ratios are of particular importance for moose during the growing season (i.e., our time of study), which is also the period of intensive bone (in juveniles) and antler (in males) growth. A Ca:P ratio in cervid antlers of 2:1 is highly No other nutrients than fiber had a significant year × use interac- tion in our study, despite the fact that practically all the mineral con- centrations in the available birch foliage varied with year (and area). This indicates high importance of mineral compositions to the animals. Ceacero, Landete-Castillejos, Garcia, Estevez, and Gallego (2010) have documented that cervid individuals are indeed able to adjust their mineral intake according to the nutritional needs. Different minerals are absorbed and function in strong interaction (nutrient stoichiom- etry, Elser et al., 2000), and also in relation to macronutrients. The importance of balancing the intake of various minerals is well-known within livestock research (Reece, Erickson, Goff, & Uemura, 2015) but has received little attention in wildlife literature. \| 1125 WAM et al. FIGURE 4 Fiber structure (hemicellulose and lignin) of birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, late June to early July 2012–2013. Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers). Note the influence of year WAM et al. 1125 FIGURE 4 Fiber structure (hemicellulose and lignin) of birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, late June to early July 2012–2013. Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers). Note the influence of year FIGURE 4 Fiber structure (hemicellulose and lignin) of birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, late June to early July 2012–2013. Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers). Note the influence of year consistent across species (Dryden, 2016). Coefficients [t, p-value] If the moose are in need of protein, P, Cu, or K and use birch to balance their dietary intake, they will have to also accept lower contents of hemicellulose and different concentrations of Ca, Zn, and Mg. Likewise, if the moose are in need of Zn, it may have to accept surplus Ca or Mn, which comprises yet another important trade-off. An excess of Ca intake is known to exacerbate a deficiency of Zn (and a range of other minerals, Spears, 2003). Zn is part of a vast array of enzymes involved in especially amino acid synthesis and cell replica- tion. A deficit therefore typically affects animal tissue growth and re- production (e.g., Enjalbert, Lebreton, & Salat, 2006), and subsequently, causes low body mass. Ohlson and Staaland (2001) found that Zn was one of the minerals that were of higher concentrations in birch than in most other moose foraging plants in southern Norway (and in line with Zn concentrations in our study, i.e., approximately 200 PPM compared to <50 PPM in other plants, see also Suttle, 2010). Birch may therefore be of special interest to moose as a source of Zn. As expected from the mineral-binding properties of fiber (Schwartz, Regelin, Franzmann, & Hubbert, 1987; Whitehead, Goulden, & Hartley, 1985), several of the minerals in our study covaried with hemicellulose (Ca, Zn, and Mn, Figure 5a). Copper on the other hand, covaried with 1126 WAM et al. (a) Crude protein WAM (a) (b) Crude protein Biplots showing covariance in concentrations of (a) nutrients as well as (b) plant secondary metabolites in birch foliage availa Norwegian regions of contrasting animal fitness (high and low), late June to early July 2012–2013. Food constituents on arro r covary the most, and in a differing direction than other such clusters. The longer the arrow, the stronger the variance of a giv ws this clustering pattern. The ellipses around observations are 2/3 confidence intervals. The less overlap between these, the (b) FIGURE 5 Biplots showing covariance in concentrations of (a) nutrients as well as (b) plant secondary metabolites in birch foliage available to moose in two Norwegian regions of contrasting animal fitness (high and low), late June to early July 2012–2013. Food constituents on arrows close together covary the most, and in a differing direction than other such clusters. Coefficients [t, p-value] Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers) for chemical groups where used foliage significantly differed from available foliage FIGURE 6 Concentrations (mg DW-1) of plant secondary metabolites in birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, late June to early July 2012–2013. Shown are median with 1st–3rd quartiles (boxes) and 1.5 cut-off for min and max (whiskers) for chemical groups where used foliage significantly differed from available foliage nutrients in the diet (Villalba & Provenza, 2005). From the above ref- erences on moose and tannins, it seems that for each 1% increase in condensed tannin concentration, the digestibility of CP in shrubs or browse foliage is reduced by 2.5%. Applying these numbers to our study, soluble condensed tannins in the available foliage may reduce protein digestibility by about 22% in the low-fitness region, and by about 12% in the high-fitness region. This falls well in line with Spalinger et al. (2010), who found that the reduction was 38% across a range of natural browse for moose in Alaska. This could be a substantial loss if protein is scarce. Notably, the moose were able to select foliage with 3 times lower concentrations of soluble condensed tannins in the high-fitness region compared to the low-fitness region. The area differences in actual effects of tannins may be even more skewed, as the tannin: protein ratio can determine whether insoluble tannin/protein complexes will form (Hagerman & Robbins, 1987). A follow-up of our study would be to conduct in vivo digestibility trials with the birch foliage. In neither region did the moose show a strong general avoidance of PSM. This is in line with previous studies of moose and PSM avoidance for summer foliage (Stolter et al., 2013), as opposed to for winter twigs (Stolter, 2008). The biplots in the appendix (Figure A2) indicate that moose selected foliage with overall less PSM, but in the constituent- specific models (Table 1), only MeOH-soluble condensed tannins (and to a lesser extent myricetins) had a significantly different concentration of available and used. Very little is known about cervid food selection and myricetins (often considered a beneficial antioxidant to humans, Williamson & Manach, 2005). A previous study on moose showed a positive relationship between food use and myricetins contents in summer (Stolter et al., 2013). Coefficients [t, p-value] The longer the arrow, the stronger the variance of a given nutrient follows this clustering pattern. The ellipses around observations are 2/3 confidence intervals. The less overlap between these, the larger the difference between areas. Ca, calcium; P, phosphorous; K, potassium; Zn, zinc; Mn, manganese; Cu, copper; sol.tannin, MeOH-soluble condensed tannins; ins.tannin, MeOH-insoluble condensed tannins; HCA, hydroxycinnamic acids; ChlAcid, chlorogenic acids (b) (b) FIGURE 5 Biplots showing covariance in concentrations of (a) nutrients as well as (b) plant secondary metabolites in birch foliage available to moose in two Norwegian regions of contrasting animal fitness (high and low), late June to early July 2012–2013. Food constituents on arrows close together covary the most, and in a differing direction than other such clusters. The longer the arrow, the stronger the variance of a given nutrient follows this clustering pattern. The ellipses around observations are 2/3 confidence intervals. The less overlap between these, the larger the difference between areas. Ca, calcium; P, phosphorous; K, potassium; Zn, zinc; Mn, manganese; Cu, copper; sol.tannin, MeOH-soluble condensed tannins; ins.tannin, MeOH-insoluble condensed tannins; HCA, hydroxycinnamic acids; ChlAcid, chlorogenic acids FIGURE 5 Biplots showing covariance in concentrations of (a) nutrients as well as (b) plant secondary metabolites in birch foliage available to moose in two Norwegian regions of contrasting animal fitness (high and low), late June to early July 2012–2013. Food constituents on arrows close together covary the most, and in a differing direction than other such clusters. The longer the arrow, the stronger the variance of a given nutrient follows this clustering pattern. The ellipses around observations are 2/3 confidence intervals. The less overlap between these, the larger the difference between areas. Ca, calcium; P, phosphorous; K, potassium; Zn, zinc; Mn, manganese; Cu, copper; sol.tannin, MeOH-soluble condensed tannins; ins.tannin, MeOH-insoluble condensed tannins; HCA, hydroxycinnamic acids; ChlAcid, chlorogenic acids \| 112 WAM et al. WAM et al. 1127 \| . FIGURE 6 Concentrations (mg DW-1) of plant secondary metabolites in birch foliage available to and used by moose in two Norwegian regions with contrasting animal fitness, late June to early July 2012–2013. Coefficients [t, p-value] Possibly, the negative relationship in our study stems from covariation with the soluble tannins (Figure 5b, A2). The classical studies of Robbins, Hanley, et al. (1987) and Robbins, Mole, Hagerman, and Hanley (1987) found that condensed tannins reduce protein digestibility for ruminants. Subsequent studies have confirmed this (e.g., Hagerman & Robbins, 1993; Jones, Rude, et al., 2010; Spalinger, Collins, Hanley, Casara, & Carnahan, 2010) but also added nuances to the relationship: tannins may actually be beneficial at some concentrations (Clauss et al., 2003; Min, Barry, Attwood, & McNabb, 2003), and its influence on intake may differ with season (Chapman, Bork, Donkor, & Hudson, 2010) or concentrations of other 4.2 \| From food selection to fitness, the next steps Ultimately, research on nutritional ecology is directed to understand higher level ecosystem interactions, typically along the pathway of 1128 WAM et al. WAM et al. animal fitness (DeGabriel et al., 2014). Access to food that better match the preferred nutritional composition is beneficial to wide- roaming animals such as moose for two reasons. One is the improved nutrient absorption discussed throughout the previous section. The other is the reduced energetic costs of obtaining the nutrients in the landscape (locomotion Fryxell, 1991; or predator vigilance Christianson & Creel, 2010). Our study highlights the need to take into account that the realized value of a given food source to an animal may be site-specific. In our study, the most abundant food available to moose had a less optimal nutritional composition in the low-fitness region compared to in the high-fitness region, but still moose in both areas selected for the same nutritional composition of this food source. Such fastidiousness limits the amount of available food that is acceptable to the animal and has bearings on our un- derstanding and application of the concept of carrying capacity. To better find out how, we encourage researchers to conduct food se- lection studies from a multidimensional viewpoint, by assessing food constituents in synchrony. This can clarify the functional roles of different constituents and the animals’ nutritional priorities in times of scarcity. Another key element for future studies that is currently lacking is the bioactivity of specific PSM in the animal body, that is how they impact nutrition and subsequently, animal physiology. With such developments, we would gain a further understanding of the complex trade-offs involved in the foraging decisions made by wide-roaming and long-living herbivores. Behmer, S. T. (2009). Insect herbivore nutrient regulation. Annual Review of Entomology, 54, 165–187. https://doi.org/10.1146/annurev. ento.54.110807.090537 Björndal, K. A. (1991). Diet mixing—Non additive interactions of diet items in an omnivorous fresh-water turtle. Ecology, 72, 1234–1241. https:// doi.org/10.2307/1941097 Bø, S., & Hjeljord, O. (1991). Do continental moose ranges improve during cloudy summers? Canadian Journal of Zoology, 69, 1875–1879. Boggs, C. (2009). Understanding insect life histories and senescence through a resource allocation lens. Functional Ecology, 23, 27–37. https://doi.org/10.1111/j.1365-2435.2009.01527.x Ceacero, F., Landete-Castillejos, T., Garcia, A. J., Estevez, J. A., & Gallego, L. (2010). Can Iberian red deer (Cervus elaphus hispanicus) discriminate among essential minerals in their diet? British Journal of Nutrition, 103, 617–626. https://doi.org/10.1017/S0007114509992091 Chapman, G. A., Bork, E. W., Donkor, N. ACKNOWLEDGMENTS Research was partly funded by the Research Council of Norway, pro- ject #215647/E40 (Intensified harvesting of forests—implications for enterprises related to wild and domestic ungulates). Dostaler, S., Ouellet, J. P., Therrien, J. F., & Cote, S. D. (2011). Are feed- ing preferences of white-tailed deer related to plant constituents? Journal of Wildlife Management, 75, 913–918. https://doi.org/10.1002/ jwmg.118 4.2 \| From food selection to fitness, the next steps T., & Hudson, R. J. (2010). Effects of supplemental dietary tannins on the perfor- mance of white-tailed deer (Odocoileus virginianus). Journal of Animal Physiology and Animal Nutrition, 94, 65–73. https://doi. org/10.1111/j.1439-0396.2008.00883.x Christianson, D., & Creel, S. (2010). A nutritionally mediated risk ef- fect of wolves on elk. Ecology, 91, 1184–1191. https://doi. org/10.1890/09-0221.1 Clauss, M., Lason, K., Gehrke, J., Lechner-Doll, M., Fickel, J., Grune, T., & Streich, W. J. (2003). Captive roe deer (Capreolus capreolus) select for low amounts of tannic acid but not quebracho: Fluctuation of prefer- ences and potential benefits. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 136, 369–382. https://doi. org/10.1016/S1096-4959(03)00244-6 Danell, K., Niemela, P., Varvikko, T., & Vuorisalo, T. (1991). Moose brows- ing on Scots pine along a gradient of plant productivity. Ecology, 72, 1624–1633. https://doi.org/10.2307/1940962 DeGabriel, J. L., Moore, B. D., Felton, A. M., Ganzhorn, J. U., Stolter, C., Wallis, I. R., … Foley, W. J. (2014). Translating nutritional ecology from the laboratory to the field: Milestones in linking plant chemistry to population regulation in mammalian browsers. Oikos, 123, 298–308. https://doi.org/10.1111/j.1600-0706.2013.00727.x CONFLICT OF INTEREST Dryden, G. M. (2011). Quantitative nutrition of deer: Energy, protein and water. Animal Production Science, 51, 292–302. The authors declare they have no conflict of interests regarding this study. Dryden, G. M. (2016). Nutrition of antler growth in deer. Animal Production Science, 56, 962–970. https://doi.org/10.1071/AN15051 Elser, J. J., Sterner, R. W., Gorokhova, E., Fagan, W. F., Markow, T. A., Cotner, J. B., … Weider, L. J. (2000). Biological stoichiometry from genes to ecosystems. Ecology Letters, 3, 540–550. https://doi. org/10.1046/j.1461-0248.2000.00185.x Hilde K. Wam http://orcid.org/0000-0003-1224-7249 Enjalbert, F., Lebreton, P., & Salat, O. (2006). Effects of copper, zinc and sele- nium status on performance and health in commercial dairy and beef herds: Retrospective study. Journal of Animal Physiology and Animal Nutrition, 90, 459–466. https://doi.org/10.1111/j.1439-0396.2006.00627.x ORCID Hilde K. Wam http://orcid.org/0000-0003-1224-7249 L., Barboza, P. S., & Gillingham, M. P. (2009). Nutrition integrates environmental responses of ungulates. Functional Ecology, 23, 57–69. https://doi.org/10.1111/j.1365-2435.2009.01528.x Herfindal, I., Haanes, H., Solberg, E. J., Røed, K. H., Høgda, K. A., & Sæther, B.-E. (2013). Moose body mass variation revisited: Disentangling effects of environmental conditions and genetics. Oecologia, 174, 447–458. R Core Team (2013). R: A language and environment for statistical comput- ing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.R-project.org/ Hjeljord, O., & Histøl, T. (1999). Range-body mass interactions of a northern ungulate—A test of hypothesis. Oecologia, 119, 326–339. https://doi. org/10.1007/s004420050793 Raubenheimer, D., & Simpson, S. J. (2009). Nutritional PharmEcology: Doses, nutrients, toxins, and medicines. Integrative and Comparative Biology, 49, 329–337. https://doi.org/10.1093/icb/icp050 Hjeljord, O., Høvik, N., & Pedersen, H. B. (1990). Choice of feeding sites by moose during summer, the influence of forest structure and plant phenology. Holarctic Ecology, 13, 281–292. Raubenheimer, D., Simpson, S. J., & Mayntz, D. (2009). Nutrition, ecology and nutritional ecology: Toward an integrated framework. Functional Ecology, 23, 4–16. https://doi.org/10.1111/j.1365-2435.2009.01522.x Jones, E. L., & Roberts, E. (1991). A note on the relationship between palat- ability and water-soluble carbohydrates content in perennial ryegrass. Irish Journal of Agricultural Research, 30, 163–167. Reece, W. O., Erickson, H. H., Goff, J. P., & Uemura, E. E. (2015). Dukes’ physiology of domestic animals, 13th ed. Ames, IA: John Wiley & Sons Inc. Jones, P. D., Rude, B., Muir, J. P., Demarais, S., Strickland, B. K., & Edwards, S. L. (2010). Condensed tannins’ effect on white-tailed deer forage di- gestibility in Mississippi. Journal of Wildlife Management, 74, 707–713. https://doi.org/10.2193/2009-116 Robbins, C. T., Hanley, T. A., Hagerman, A. E., Hjeljord, O., Baker, D. L., Schwartz, C. C., & Mautz, W. W. (1987). Role of tannins in defending plants against ruminants: Reduction in protein availability. Ecology, 68, 98–107. https://doi.org/10.2307/1938809 Jones, P. D., Strickland, B. K., Demarais, S., Rude, B. J., Edwards, S. L., & Muir, J. P. (2010). Soils and forage quality as predictors of white-tailed deer Odocoileus virginianus morphometrics. Wildlife Biology, 16, 430– 439. https://doi.org/10.2981/10-041 Robbins, C. T., Mole, S., Hagerman, A. E., & Hanley, T. A. (1987). Role of tannins in defending plants against ruminants: Reduction in dry matter digestion? Ecology, 68, 1606–1615. https://doi.org/10.2307/1939852 Kessler, A., & Baldwin, I. T. (2002). Plant responses to herbivory: The emerg- ing molecular analysis. Annual Review of Plant Biology, 53, 299–328. https://doi.org/10.1146/annurev.arplant.53.100301.135207 Schwartz, C. C., Regelin, W. L., & Franzmann, A. Freeland, W. J., & Janzen, D. H. (1974). Strategies in herbivory by mam- mals—Role of plant secondary compounds. American Naturalist, 108, 269–289. https://doi.org/10.1086/282907 McCullagh, P., & Nelder, J. A. (1989). Generalized linear models. London, UK: Chapman and Hall. https://doi.org/10.1007/978-1-4899-3242-6 Min, B. R., Barry, T. N., Attwood, G. T., & McNabb, W. C. (2003). The ef- fect of condensed tannins on the nutrition and health of ruminants fed fresh temperate forages: A review. Animal Feed Science and Technology, 106, 3–19. https://doi.org/10.1016/S0377-8401(03)00041-5 Freeland-Graves, J. H., Sanjeevi, N., & Lee, J. J. (2015). Global perspectives on trace element requirements. Journal of Trace Elements in Medicine and Biology, 31, 135–141. https://doi.org/10.1016/j.jtemb.2014.04.006 6, 3–19. https://doi.org/10.1016/S0377-8401(03)00041-5 Fryxell, J. M. (1991). Forage quality and aggregation by large herbivores. American Naturalist, 138, 478–498. https://doi.org/10.1086/285227 Moen, A. (1999). National atlas of Norway—Vegetation. Hønefoss, Norway: Norwegian Mapping Authority. Gillingham, M. P., Parker, K. L., & Hanley, T. A. (2001). Habitat use by black- tailed deer in relation to rate of forage intake. Alces, 37, 339–352. Morgantini, L. E., & Hudson, R. J. (1989). Nutritional significance of wa- piti (Cervus elaphus) migrations to alpine ranges in western Alberta, Canada. Arctic and Alpine Research, 21, 288–295. https://doi. org/10.2307/1551568 Hagen, Y. (1983). Elgens vinterbeiting i Norge. Viltrapport 83, Trondheim, Norway: Norwegian Environment Agency (in Norwegian with English summary). Norwegian Meteorological Institute (2013). Weather statistics. Retrieved from www.met.no Hagerman, A. E. (2002). The tannin handbook. Oxford, UK: Miami University. NRC (National Research Council) (2001). Nutrient requirements of dairy cat- tle, 7th ed. Washington, DC: National Academy Press. Hagerman, A. E., & Robbins, C. T. (1987). Implications of soluble tannin- protein complexes for tannin analysis and plant defense mechanisms. Journal of Chemical Ecology, 31, 1243–1259. https://doi.org/10.1007/ BF01020552 Nybakken, L., Hörkkä, R., & Julkunen-Tiitto, R. (2012). Combined enhancements of temperature and UVB influence growth and phenolics in clones of the sexually dimorphic Salix myrsini- folia. Physiologia Plantarum, 145, 551–564. https://doi. org/10.1111/j.1399-3054.2011.01565.x Hagerman, A. E., & Robbins, C. T. (1993). Specificity of tannin-binding sali- vary proteins relative to diet selection by mammals. Canadian Journal of Zoology, 71, 628–633. https://doi.org/10.1139/z93-085 Hall, M. B., Hoover, W. H., Jennings, J. P., & Miller Webster, T. K. (1999). A method for partitioning neutral detergent soluble carbohydrates. Journal of the Science of Food and Agriculture, 79, 2079–2086. https:// doi.org/10.1002/(ISSN)1097-0010 Ohlson, M., & Staaland, H. (2001). Mineral diversity in wild plants: Benefit and bane for moose. Oikos, 94, 442–454. https://doi. org/10.1034/j.1600-0706.2001.940307.x Parker, K. REFERENCES 9–466. https://doi.org/10.1111/j.1439-0396.2006.00627.x Faber, W. E., & Lavsund, S. (1999). Summer foraging on Scots pine Pinus sylvestris by moose Alces alces in Sweden—Patterns and mechanisms. Wildlife Biology, 5, 93–106. AOAC (Association of Official Analytical Chemists) (2012). Official methods of analysis of AOAC International. Arlington, TX: AOAC. Arnaud, C. D. (1983). Hormonal regulation of calcium homeostasis. pp. 1-19 In D. D. Bikle (Ed.), Assay of calcium-regulating hormones. New York, NY: Springer. FDA (US Food and Drug Administration). (2013). Elemental analysis manual. Silver Spring, MD: FDA. Felton, A. M., Felton, A., Lindenmayer, D. B., & Foley, W. J. (2009). Nutritional goals of wild primates. Functional Ecology, 23, 70–78. https://doi.org/10.1111/j.1365-2435.2008.01526.x Beck, J. L., Flinders, J. T., Nelson, D. R., & Clyde, C. L. (1996). Dietary overlap and preference of elk and domestic sheep in Aspen-dominated habi- tats in north-central Utah. Sharing common ground on western range- lands. Proceedings of a Livestock/Big Game Symposium, 343, 81–85. Felton, A. M., Felton, A., Raubenheimer, D., Simpson, S. J., Krizsan, S. J., Hedwall, P. O., & Stolter, C. (2016). The nutritional balancing act of a large herbivore: An experiment with captive moose (Alces alces L.). PLoS ONE, 11(3), e0150870. https://doi.org/10.1371/journal.pone.0150870 Beck, J. L., & Peek, J. M. (2005). Great Basin summer range forage qual- ity: Do plant nutrients meet elk requirements? Western North American Naturalist, 65, 516–527. WAM et al. 1129 Sperfeld, E., Martin-Creuzburg, D., & Wacker, A. (2012). Multiple resource limitation theory applied to herbivorous consumers: Liebig’s minimum rule versus interactive co-limitation. Ecological Letters, 15, 142–150. https://doi.org/10.1111/j.1461-0248.2011.01719.x Wam, H. K., Histøl, T., Nybakken, L., Solberg, E. J., & Hjeljord, O. (2016). Transient nutritional peak in browse foliage after forest clearing advo- cates cohort management of ungulates. Basic and Applied Ecology, 17, 252 261 htt //d i /10 1016/j b 2015 11 006 52–261. https://doi.org/10.1016/j.baae.2015.11.006 Stolter, C. (2008). Intra-individual plant response to moose brows- ing: Feedback loops and impacts on multiple consumers. Ecological Monographs, 78, 167–183. https://doi.org/10.1890/07-0401.1 Wam, H. K., & Hjeljord, O. (2010a). Moose summer and winter diets along a large scale gradient of forage availability in southern Norway. European Journal of Wildlife Research, 56, 745–755. https://doi.org/10.1007/ s10344-010-0370-4 Stolter, C., Ball, J. P., & Julkunen-Tiitto, R. (2013). Seasonal differences in the relative importance of specific phenolics and twig morphology result in contrasting patterns of foraging by a generalist herbivore. Canadian Journal of Zoology, 91, 338–347. https://doi.org/10.1139/cjz-2012-0270 Wam, H. K., & Hjeljord, O. (2010b). Moose summer diet from feces and field-surveys: A comparative study. Rangeland Ecology and Management, 63, 387–395. https://doi.org/10.2111/REM-D-09-00039.1 Suttle, N. F. (2010). Mineral nutrition of livestock, 4th ed. Oxford, UK: CABI. https://doi.org/10.1079/9781845934729.0000 Wam, H. K., Hjeljord, O., & Solberg, E. J. (2010). Differential forage use makes carrying capacity equivocal on ranges of Scandinavian moose (Alces alces). Canadian Journal of Zoology, 88, 1179–1191. https://doi. org/10.1139/Z10-084 Thompson, I. D., McQueen, R. E., Reichardt, P. B., Trenholm, D. G., & Curran, W. J. (1989). Factors influencing choice of balsam fir twigs from thinned and unthinned stands by moose. Oecologia, 81, 506–509. https://doi. org/10.1007/BF00378960 Whitehead, D. C., Goulden, K. M., & Hartley, R. D. (1985). The dis- tribution of nutrient elements in cell wall and other fractions of the herbage of some grasses and legumes. Journal of the Science of Food and Agriculture, 36, 311–318. https://doi.org/10.1002/ (ISSN)1097-0010 Tiilikainen, R., Solberg, E. J., Nygrén, T., & Pusenius, J. (2012). Spatio- temporal relationship between calf body mass and population produc- tivity in Fennoscandian moose Alces alces. Wildlife Biology, 18, 304– 317. https://doi.org/10.2981/10-116 Williamson, G., & Manach, C. (2005). Bioavailability and bioefficacy of poly- phenols in humans. II. Review of 93 intervention studies. The American Journal of Clinical Nutrition, 81, 243S–255S. Tixier, H., Duncan, P., Scehovic, J., Yani, A., Gleizes, M., & Lila, M. (1997). Food selection by European roe deer (Capreolus capreolus): Effects of plant chemistry, and consequences for the nutritional value of their diets. Journal of Zoology, 242, 229–245. Zuur, A. F., Ieno, E. N., & Smith, G. M. (2007). Analysing ecological data. New York, NY: Springer. https://doi.org/10.1007/978-0-387-45972-1 Tveite, B. (1977). Site-index curves for Norway spruce (Picea abies (L.) Karst.). Report Norwegian Forest Research Institute, 33, 1–84. Zweifel-Schielly, B., Kreuzer, M., Ewald, K. C., & Suter, W. (2009). Habitat selection by an Alpine ungulate: The significance of forage characteris- tics varies with scale and season. Ecography, 32, 103–113. https://doi. org/10.1111/j.1600-0587.2008.05178.x van den Berg, R. A., Hoefsloot, H. C., Westerhuis, J. A., Smilde, A. K., & van der Werf, M. J. (2006). Centering, scaling, and transformations: Improving the biological information content of metabolomics data. BMC Genomics, 7, 142. https://doi.org/10.1186/1471-2164-7-142 Zweifel-Schielly, B., Leuenberger, Y., Kreuzer, M., & Suter, W. (2012). A her- bivore’s food landscape: seasonal dynamics and nutritional implications of diet selection by a red deer population in contrasting Alpine habitats. Journal of Zoology, 286, 68–80. Van Soest, P. J. (1994). Nutritional ecology of the ruminant, 2nd ed. Ithaca, NY: Cornell University Press. Van Soest, P. J., Robertson, J. B., & Lewis, B. A. (1991). Methods for dietary fiber, neutral detergent fiber and nonstarch polysaccharides in relation to animal nutrition. Journal of Dairy Science, 74, 3583–3597. https://doi. org/10.3168/jds.S0022-0302(91)78551-2 W. (1987). Protein digestion in moose. The Journal of Wildlife Management, 51, 352–357. https://doi. org/10.2307/3801015 Langvatn, R., & Hanley, T. A. (1993). Feeding-patch choice by red deer in relation to foraging efficiency. Oecologia, 95, 164–170. https://doi. org/10.1007/BF00323486 Schwartz, C. C., Regelin, W. L., Franzmann, A. W., & Hubbert, M. (1987). Nutritional energetics of moose. Swedish Wildlife Research Supplement, 1(part 1), 265–280. Leslie, D. M., Starkey, E. E., & Vavra, M. (1984). Elk and deer diets in old- growth forests in Western Washington. Journal of Wildlife Management, 48, 762–775. https://doi.org/10.2307/3801423 Shipley, L. A., Blomquist, S., & Danell, K. (1998). Diet choices made by free- ranging moose in northern Sweden in relation to plant distribution, chemistry, and morphology. Canadian Journal of Zoology, 76, 1722– 1733. https://doi.org/10.1139/z98-110 Marshal, J. P., Krausman, P. R., & Bleich, V. C. (2005). Rainfall, tempera- ture, and forage dynamics affect nutritional quality of desert mule deer forage. Rangeland Ecology & Management, 58, 360–365. https://doi. org/10.2111/1551-5028(2005)058[0360:RTAFDA]2.0.CO;2 Simpson, R. J., & Raubenheimer, D. (2012). The nature of nutrition: A unify- ing framework from animal adaptation to human obesity. Princeton, NJ: Princeton University Press. https://doi.org/10.1515/9781400842803 Mattson, W. J. (1980). Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics, 11, 119–161. https://doi. org/10.1146/annurev.es.11.110180.001003 Spalinger, D. E., Collins, W. B., Hanley, T. A., Casara, N. E., & Carnahan, A. M. (2010). The impact of tannins on protein, dry matter, and energy digestion in moose (Alces alces gigas). Canadian Journal of Zoology, 88, 977–987. https://doi.org/10.1139/Z10-064 McArt, S. H., Spalinger, D. E., Collins, W. B., Schoen, E. R., Stevenson, T., & Bucho, M. (2009). Summer dietary nitrogen availability as a potential bottom-up constraint on moose in south-central Alaska. Ecology, 90, 1400–1411. https://doi.org/10.1890/08-1435.1 977–987. https://doi.org/10.1139/Z10-064 Spears, J. W. (2003). Trace mineral bioavailability in ruminants. Journal of Nutrition, 133, 1506S–1509S. 1130 WAM et al. WAM et al. SUPPORTING INFORMATION Additional Supporting Information may be found online in the sup- porting information tab for this article. Vangilder, L. D., Torgerson, O., & Porath, W. R. (1982). Factors influencing diet selection by white-tailed deer. Journal of Wildlife Management, 46, 711–718. https://doi.org/10.2307/3808563 Vázquez-de-Aldana, B. R., García-Ciudad, A., & García-Criado, B. (2008). Interannual variations of above-ground biomass and nutritional quality of Mediterranean grasslands in Western Spain over a 20-year period. Australian Journal of Agricultural Research, 59, 769–779. https://doi. org/10.1071/AR07359 How to cite this article: Wam HK, Felton AM, Stolter C, Nybakken L, Hjeljord O. Moose selecting for specific nutritional composition of birch places limits on food acceptability. Ecol Evol. 2018;8:1117–1130. https://doi.org/10.1002/ece3.3715 How to cite this article: Wam HK, Felton AM, Stolter C, Nybakken L, Hjeljord O. Moose selecting for specific nutritional composition of birch places limits on food acceptability. Ecol Evol. 2018;8:1117–1130. https://doi.org/10.1002/ece3.3715 Villalba, J. J., & Provenza, F. D. (2005). Foraging in chemically diverse envi- ronments: Energy, protein, and alternative foods influence ingestion of plant secondary metabolites by lambs. Journal of Chemical Ecology, 31, 123–138. https://doi.org/10.1007/s10886-005-0979-z
https://openalex.org/W2608379538	https://europepmc.org/articles/pmc5402977?pdf=render	Breton	null	Correction: Virtual Hand Feedback Reduces Reaction Time in an Interactive Finger Reaching Task	PloS one	2,017	cc-by	611	CORRECTION Correction: Virtual Hand Feedback Reduces Reaction Time in an Interactive Finger Reaching Task Johannes Brand, Marco Piccirelli, Marie-Claude Hepp-Reymond, Manfred Morari, Lars Michels, Kynan Eng Johannes Brand, Marco Piccirelli, Marie-Claude Hepp-Reymond, Manfred Morari, Lars Michels, Kynan Eng The standard deviations of the oculomotor parameters in Table 2 are incorrect. Please see the corrected Table 2 here. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Brand J, Piccirelli M, Hepp-Reymond M- C, Morari M, Michels L, Eng K (2017) Correction: Virtual Hand Feedback Reduces Reaction Time in an Interactive Finger Reaching Task. PLoS ONE 12 (4): e0176655. https://doi.org/10.1371/journal. pone.0176655 Published: April 24, 2017 Copyright: © 2017 Brand et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. 1 / 2 PLOS ONE \| https://doi.org/10.1371/journal.pone.0176655 April 24, 2017 light: point light; c. hand: cartoon hand; v. hand: virtual hand. p-values were Bonferroni corrected. PLOS ONE \| https://doi.org/10.1371/journal.pone.0176655 April 24, 2017 Table 2. Analyses of oculomotor parameters. Table 2. Analyses of oculomotor parameters. Parameter Action Act.–Obs. Observation cursor p. light c. hand v. hand p cursor p. light c. hand v. hand Sac. (2s-1) 3.7 ± 0.4 3.7 ± 0.3 3.6 ± 0.4 3.7 ± 0.5 0.017* 4.2 ± 0.4 3.9 ± 0.3 4.1 ± 0.4 4.2 ± 0.5 Fix. (2s-1) 3.0 ± 0.4 3.0 ± 0.3 2.9 ± 0.4 3.1 ± 0.5 0.004* 3.5 ± 0.4 3.3 ± 0.3 3.5 ± 0.4 3.6 ± 0.5 Bli. (2s-1) 0.3 ± 0.1 0.4 ± 0.1 0.3 ± 0.1 0.3 ± 0.1 <0.001* 0.5 ± 0.1 0.5 ± 0.1 0.6 ± 0.1 0.5 ± 0.1 ΔSac.(ms) 45.9 ± 3.9 49.7 ± 4.6 44.7 ± 4.0 48.5 ± 6.9 0.017* 56.3 ± 5.9 52.5 ± 4.6 57.2 ± 6.0 56.9 ± 6.2 ΔFix.(ms) 407.4 ± 36.7 459.8 ± 30.6 396.5 ± 32.1 399.4 ± 37.9 0.017* 346.7 ± 25.5 336.1 ± 23.9 355.3 ± 26.4 347.3 ± 32.2 ΔBli.(ms) 51.6 ± 10.2 51.6 ± 10.8 43.1 ± 10.1 42.8 ± 12.2 <0.001* 67.3 ± 10.8 62.1 ± 10.3 68.0 ± 11.1 71.3 ± 12.5 Δx(px) 111.5 ± 7.9 95.6 ± 6.7 98.8 ± 6.4 106.7 ± 6.3 1.000 108.1 ± 6.0 94.0 ± 6.6 94.2 ± 5.6 102.9 ± 5.9 Δy(px) 182.4 ± 4.6 176.8 ± 2.1 181.3 ± 5.1 181.0 ± 4.0 1.000 183.1 ± 4.5 174.2 ± 4.2 173.0 ± 3.3 180.8 ± 3.5 Dvx px s 48.3 ± 1.9 43.2 ± 1.2 43.9 ± 1.6 47.8 ± 2.0 1.000 53.1 ± 4.6 44.1 ± 2.2 45.6 ± 1.4 43.6 ± 1.5 Dvy px s 63.0 ± 2.0 61.4 ± 1.3 63.4 ± 1.9 66.9 ± 1.9 1.000 66.3 ± 2.3 61.3 ± 2.2 65.4 ± 1.8 64.3 ± 2.1 Mean ± standard error of oculomotor parameters. Sac.: Saccade; Fix.: Fixation; Bli.: Blink; ΔSac.: Saccade duration; ΔFix.: Fixation duration; ΔBli.: Blink duration; Δx: amplitude x; Δy: amplitude y; Δvx: velocity x; Δvy: velocity y; px: pixel; Act.–Obs.: factor Action–Observation; p. light: point light; c. hand: cartoon hand; v. hand: virtual hand. p-values were Bonferroni corrected. Mean ± standard error of oculomotor parameters. Sac.: Saccade; Fix.: Fixation; Bli.: Blink; ΔSac.: Saccade duration; ΔFix.: Fixation duration; ΔBli.: Blink duration; Δx: amplitude x; Δy: amplitude y; Δvx: velocity x; Δvy: velocity y; px: pixel; Act.–Obs.: factor Action–Observation; p. https://doi.org/10.1371/journal.pone.0176655.t001 1. Brand J, Piccirelli M, Hepp-Reymond M-C, Morari M, Michels L, Eng K (2016) Virtual Hand Feedback Reduces Reaction Time in an Interactive Finger Reaching Task. PLoS ONE 11(5): e0154807. doi:10. 1371/journal.pone.0154807 PMID: 27144927 Reference 2 / 2
https://openalex.org/W4387018446	https://jneuroengrehab.biomedcentral.com/counter/pdf/10.1186/s12984-023-01254-0	English	null	Virtual reality—enhanced walking in people post-stroke: effect of optic flow speed and level of immersion on the gait biomechanics	Journal of neuroengineering and rehabilitation	2,023	cc-by	13,578	© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecom‑ mons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Journal of NeuroEngineering and Rehabilitation Journal of NeuroEngineering and Rehabilitation De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 https://doi.org/10.1186/s12984-023-01254-0 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 https://doi.org/10.1186/s12984-023-01254-0 Journal of NeuroEngineering and Rehabilitation Abstract Background Optic flow—the apparent visual motion experienced while moving—is absent during treadmill walk‑ ing. With virtual reality (VR), optic flow can be controlled to mediate alterations in human walking. The aim of this study was to investigate (1) the effects of fully immersive VR and optic flow speed manipulation on gait biomechanics, simulator sickness, and enjoyment in people post-stroke and healthy people, and (2) the effects of the level of immer‑ sion on optic flow speed and sense of presence. Methods Sixteen people post-stroke and 16 healthy controls performed two VR-enhanced treadmill walking sessions: the semi-immersive GRAIL session and fully immersive head-mounted display (HMD) session. Both con‑ sisted of five walking trials. After two habituation trials (without and with VR), participants walked three more trials under the following conditions: matched, slow, and fast optic flow. Primary outcome measures were spatiotemporal parameters and lower limb kinematics. Secondary outcomes (simulator sickness, enjoyment, and sense of presence) were assessed with the Simulator Sickness Questionnaire, Visual Analogue Scales, and Igroup Presence Questionnaire. Results When walking with the immersive HMD, the stroke group walked with a significantly slower cadence (-3.69strides/min, p = 0.006), longer stride time (+ 0.10 s, p = 0.017) and stance time for the unaffected leg (+ 1.47%, p = 0.001) and reduced swing time for the unaffected leg (− 1.47%, p = 0.001). Both groups responded to the optic flow speed manipulation such that people accelerated with a slow optic flow and decelerated with a fast optic flow. Compared to the semi-immersive GRAIL session, manipulating the optic flow speed with the fully immersive HMD had a greater effect on gait biomechanics whilst also eliciting a higher sense of presence. Conclusion Adding fully immersive VR while walking on a self-paced treadmill led to a more cautious gait pattern in people post-stroke. However, walking with the HMD was well tolerated and enjoyable. People post-stroke altered their gait parameters when optic flow speed was manipulated and showed greater alterations with the fully-immer‑ sive HMD. Further work is needed to determine the most effective type of optic flow speed manipulation as well as which other principles need to be implemented to positively influence the gait pattern of people post-stroke. Trial registration number: The study was pre-registered at ClinicalTrials.gov (NCT04521829). Virtual reality—enhanced walking in people post‑stroke: effect of optic flow speed and level of immersion on the gait biomechanics Emma De Keersmaecker1,2,3,4, Anke Van Bladel4,5, Silvia Zaccardi1,3,6, Nina Lefeber1,7, Carlos Rodriguez‑Guerrero8, Eric Kerckhofs1, Bart Jansen3,6,9 and Eva Swinnen1,2,3,4 Virtual reality—enhanced walking in people post‑stroke: effect of optic flow speed and level of immersion on the gait biomechanics Emma De Keersmaecker1,2,3,4, Anke Van Bladel4,5, Silvia Zaccardi1,3,6, Nina Lefeber1,7, Carlos Rodriguez‑Guerrero8, Eric Kerckhofs1, Bart Jansen3,6,9 and Eva Swinnen1,2,3,4 RESEARCH Open Access Abstract Keywords Virtual reality, Optic flow, Gait biomechanics, Stroke Correspondence: Emma De Keersmaecker emma.de.keersmaecker@vub.be Full list of author information is available at the end of the article Correspondence: Emma De Keersmaecker emma.de.keersmaecker@vub.be Full list of author information is available at the end of the article Background Two-thirds of all stroke survivors suffer from walking impairments, causing them to experience a decrease in activities of daily living, level of participation and qual- ity of life [4–6]. People post-stroke often have an asym- metric gait pattern characterized by a shorter stance time and longer swing time of the affected limb and a longer stance time and shorter swing time of the unaffected limb [7]. This asymmetry leads to alterations in step length and a reduced walking speed and cadence [7]. In order to improve these impairments, people post-stroke often receive treadmill training, a repetitive and task-spe- cific gait training that has the potential to enhance neural plasticity—the ability to create permanent structural and functional changes of the brain and spinal cord—which is vital to trigger the learning process of the sensorimotor system [8, 9]. Two key aspects of VR are immersion and sense of presence. Based on the level of immersion, VR devices and systems can be classified into two categories: (1) Semi-immersive or non-immersive VR systems, who let the user perceive both real world and a part of the vir- tual environment (e.g. TV-screens, projection screens), and (2) Fully immersive VR systems, who fully integrate the user into the virtual environment, by blocking out perception of the real world (e.g. head-mounted dis- plays (HMD)) [24]. The level of immersion has an impact on the user’s VR experience by influencing the sense of presence (i.e. the feeling of being physically present in the virtual world), with stronger feelings of ‘being physi- cally present’ during exposure in more immersive virtual environments [25, 26]. With semi-immersive VR systems participants are still perceiving the real environment and thus also the real optic flow while walking. Therefore, it is expected that the effect of optic flow speed manipulations on gait will be more limited in a non – or semi-immersive virtual environment, compared to a fully immersive vir- tual environment. However, to the best of our knowledge, no research has been performed so far on the effect of immersion on optic flow speed manipulations. Controlling our locomotion is a complex, multisensory process and involves the integration of visual, vestibu- lar, and proprioceptive information [10]. An important source of visual information used to guide locomotion is optic flow. Optic flow refers to the pattern of visual motion experienced while moving around and is being projected onto the retina of the eye. Background populations [10, 13]. In general, it seems that healthy people will increase their walking speed with a slower optic flow and decrease their speed with a faster optic flow [14–16, 18]. This strategy can be altered in patients with neurological diseases due to damage in brain areas involved in the perception and use of optic flow [10]. For example, Schubert and colleagues (2005) found that due to the overreliance on visual information in Parkin- son’s disease patients, optic flow speed manipulations led to exaggerated walking speed responses compared to healthy people [20]. On the other hand, the study by Lim et al. reported that cerebral palsy children used an oppo- site strategy and increased their walking speed with a fast optic flow speed and vice versa [23]. It is assumed that people post-stroke still have the ability to use optic flow information during walking, but alterations are possible and responses can be heterogeneous between individu- als, depending on the location of the brain lesion [10]. With the use of VR, the selective manipulation of optic flow could be used to induce desired locomotor changes, such as an increase in walking speed, and therefore has the potential to advance the field of post-stroke gait reha- bilitation. However, studies about the effect of optic flow speed on locomotion in people post-stroke are still scarce [13, 22]. Given the potential of optic flow speed manip- ulation, further exploration is necessary to determine how such manipulation could be useful for rehabilitation purposes. A more in-depth analysis of how optic flow speed influences the gait pattern in people post-stroke is needed. In recent years, virtual reality (VR) has been on the rise in the field of healthcare. Over the last 20 years, the pop- ularity and use of VR for physical rehabilitation alone increased remarkably, with increasing evidence sup- porting its use [1, 2]. However, VR that is used today for physical rehabilitation are often video gaming consoles and were initially designed for entertainment purposes instead of rehabilitation [3]. Consequently, they do not incorporate rehabilitation and motor learning principles to optimally enhance motor rehabilitation. Hence, VR games specifically built for different rehabilitation pur- poses are required to achieve optimal rehabilitation [3]. One of these rehabilitation purposes that could benefit from VR is post-stroke gait rehabilitation. Post-stroke gait rehabilitation remains a major clinical challenge. © The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecom‑ mons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 2 of 18 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (20 Background It provides us with information about the direction and speed of locomotion [10, 11]. During normal walking, the optic flow and pro- prioceptive information are congruent. However, with the use of VR, the speed of optic flow can be manipulated in such way that there is a mismatch between the optic flow and the proprioceptive information of the lower limbs [12]. As a result, people will adjust their gait pat- tern in order to diminish this incongruity [13].ll Optic flow speed and its influence on locomotion has been examined in the healthy population [14–18] and more specific in older adults [19], but also in several clin- ical populations, such as neurological patients [13, 20– 22]. It is suggested that optic flow can exert an influence on locomotion, but there are conflicting results between Page 3 of 18 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 For these reasons, the aim of this study was three-fold: (1) to investigate the effect of adding fully immersive VR while walking on a self-paced treadmill on the gait bio- mechanics, simulator sickness and enjoyment, in peo- ple post-stroke and healthy people, (2) to investigate the effect of optic flow speed manipulation (two times faster and two times slower than their comfortable walking speed) on the gait biomechanics, in people post-stroke and healthy people and (3) to investigate the effect of the level of immersion (semi-immersive vs. fully immer- sive) during walking with different optic flow speeds on the gait biomechanics and level of presence, in people post-stroke and healthy people. We hypothesized that: (1) adding fully immersive VR while walking on a self- paced treadmill will alter the gait biomechanics in both groups, (2) both healthy people and people post-stroke will alter their gait pattern in response to the optic flow speed manipulation and (3) the effect of optic flow speed manipulation and the level of presence will be larger with the fully immersive VR. pattern (Functional Ambulation Categories (FAC) score 2, 3 or 4), (5) ability to walk on a treadmill for 4 times 8 min without bodyweight support, (6) to signal pain, fear and discomfort and (7) to give informed consent. People post-stroke were excluded if they had (1) other neurological deficits leading to impaired gait (e.g. Par- kinson’s disease, multiple sclerosis), (2) comorbidities (e.g. Background COPD, severe osteoporosis, cardiovascular insta- bility), (3) visual and/or vestibular disorders that can interfere with the VR (e.g. Meniere’s disease), (4) severe spasticity of the lower limbs (Modified Ashworth Scale > 2), (5) acute medical illness, (6) the inability to understand and carry out instructions and (7) severe unilateral spatial neglect. For the healthy participants, the following inclusion criteria were used: (1) normal or corrected-to-normal vision with glasses or contact lenses and (2) no locomo- tion impairments. Participants were excluded if they (1) have had significant lower extremity injuries during the last two years that might affect their gait and (2) had any type of vestibular/visual deficiency. i Based on a sample size calculation (GPower 3.1.9.4) (F-tests, repeated measures ANOVA, within-between subjects) with Cohen’s f of 0.25 (moderate effect size), type 1 error probability of 0.05, power of 0.80 for 2 groups and 4 conditions, a minimum of 24 participants, divided equally in 2 groups, had to be recruited. Apparatus Participants walked on the treadmill of the GRAIL sys- tem, an integrative motion capture system consisting of 10 optical motion cameras (Vicon Inc., UK), a dual belt treadmill with integrated force sensors, a 180-degree cylindrical projection screen and D-Flow software (Motekforce Link, Netherlands). The treadmill of the GRAIL system has two modes: fixed walking speed or self-paced. For this study, the treadmill was self-paced, meaning that the participants had control over the speed of the treadmill and could start, stop and change speed at will. For safety reasons, participants wore a safety harness and the maximum walking speed was set at 2 m/s. The 180-degree cylindrical projection screen of the GRAIL system provided the semi-immersive VR, while the HMD VR system ‘Oculus Rift’ (Oculus, LLS, US) ensured the fully immersive VR (Figs. 1 and 2). The vir- tual environment used in this study was a standard envi- ronment provided by Motek and represented a city street in the Italian Alps. For this study, the game elements of this environment (i.e. collecting ingredients on the street) were removed and participants only had to walk forward. The same virtual environment was shown in both VR devices. Material and methods Study design An experimental, 2-group, repeated measures single- center trial was conducted in which people post-stroke and healthy people performed two VR-enhanced tread- mill walking sessions. Both sessions were identical and carried out on two separate points in time within 10 days, only the VR system used to manipulate the optic flow speed differed: the semi-immersive Gait Real-time Interactive Lab (GRAIL) system and the fully immersive head-mounted display (HMD). The study took place at the Smart Space lab of the University Hospital in Ghent, Belgium. The study was approved by the Ethics Com- mittee of the University of Brussels and the University Hospital of Ghent (B1432020000120) and pre-registered at ClinicalTrials.gov (NCT04521829). The results of this study has been reported in two different papers. The first paper reports on the effect of adding and manipulating optic flow speed in a semi-immersive virtual environ- ment (GRAIL session, paper submitted). The current paper reports on the effect of adding and manipulating optic flow speed in a fully immersive virtual environment (HMD session) and compares the GRAIL and HMD ses- sion to investigate the effect of the level of immersion. Participants Chronic, ambulatory stroke patients and age – and sex matched healthy adults were included. The follow- ing inclusion criteria were used for the stroke popula- tion: (1) diagnosed with stroke (as defined by the World Health Organization), (2) adult (≥ 18 years), (3) stroke onset ≥ 3 months, (4) ambulatory with an impaired gait De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 4 of 18 Fig. 1 The GRAIL system with the semi-immersive projection screen (left) and with the fully immersive HMD ‘Oculus Rift’ (right) Fig. 1 The GRAIL system with the semi-immersive projection screen (left) and with the fully immersive HMD ‘Oculus Rift’ (right) g. 1 The GRAIL system with the semi-immersive projection scr Fig. 2 A picture of a participant walking on the GRAIL system with the reflective markers while wearing the HMD Statistical analysis IBM SPSS Statistic IBM SPSS Statistics version 28, custom-made scripts in Python 3.7. and Matlab (R2022a) were used for statistical analysis. Level of significance was set at α = 0.05. Baseline characteristics between groups were compared using an independent sample t-test and Mann–Whitney U test for respectively normally and not-normally distributed con- tinuous variables and a Chi-squared test for categorical variables. Our secondary outcome measures were simulator sick- ness, sense of presence and level of enjoyment. Simulator sickness was assessed with the Simulator Sickness Ques- tionnaire (SSQ) [28]. The SSQ is a widely used question- naire to evaluate simulator sickness when using VR and consists of 16 symptoms. Before and after the walking trial, participants had to indicate on a four-point Likert scale ranging from 0 (none) to 3 (severe) how much each symptom was affecting them at that moment. The overall score is measured by adding the scores of the 16 items and multiplying the achieved sum by 3.74 [28]. The total score can serve as an indicator of the severity of the simu- lator sickness and ranges between 0 – 179.52 with higher scores indicating higher levels of simulator sickness.h To investigate the effect of fully immersive VR on the gait biomechanics, the averages during the last 30 s of the trial without VR were compared to those obtained during the last 30 s of the habituation trial. For spati- otemporal data, linear mixed-effect models (LMM) were used. LMM were conducted with condition (no VR, with VR) and group (post-stroke, healthy) as fixed factors, accounting for the within subject correlations and a ran- dom intercept of participants. Multiple models were built in SPSS. With the use of the Akaike’s Information Crite- ria (AIC) value, the best fitted model was chosen (with smaller AIC values indicating a better model). The within subject covariance was unstructured. For kinematic data, statistical parametric mapping (SPM) was used [30]. A SPM two-way Analysis of Variance (ANOVA) was per- formed to examine the effect of condition (no VR, with VR) and group (post-stroke, healthy): the F-statistic (SPM{F}) was calculated at each point of the time-series. Where SPM{F} crossed a threshold equivalent to α = 0.05, post-hoc Bonferroni analyses were performed using SPM paired t-tests. For post-hoc comparisons, the SPM{t} sta- tistic was calculated for each comparison. Experimental procedure In each session, both groups underwent five walking tri- als (see Fig. 3 for the protocol timeline). The first trial consisted of 8 min where participants walked without VR to get familiarized with self-paced walking [27]. For the second trial, the habituation trial, the VR was added and participants walked for 5 min with either the large pro- jection screen of the GRAIL system (GRAIL session) or with the VR glasses (HMD session) to get used to walking with VR. Thereafter, participants underwent three more walking trials of 8 min during which the optic flow speed was being manipulated: 2 times slower than, equal to and 2 times faster than their comfortable walking speed. The duration of 8 min was chosen to investigate if and how long the changes that are expected immediately after the manipulation of optic flow maintain. A longer duration was not feasible and/or would evoke fatigue. The com- fortable walking speed of the participants was defined as the average walking speed during the 5-min habitu- ation trial. The optic flow speed manipulation occurred after one minute and lasted for the remaining 7 min. The order of the two sessions (GRAIL, HMD) and the optic flow speed manipulation within the session (matched, slow, fast) was randomized through block randomiza- tion in Microsoft Excel®. Participants were not informed about this manipulation. In-between walking rounds, Fig. 2 A picture of a participant walking on the GRAIL system with the reflective markers while wearing the HMD Fig. 3 A protocol timeline for a possible GAIL or HMD session Fig. 3 A protocol timeline for a possible GAIL or HMD session Fig. 3 A protocol timeline for a possible GAIL or HMD session De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 5 of 18 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation participants were given a 5-min rest period to complete a few short questionnaires. agreeing with the statement). The maximus score of the IPQ in total is 84. The maximum score for the subscales are 6 for the general item, 30 for spatial presence, 24 for involvement and 24 for experienced realism. At the end of each session, participants had to fill in the IPQ. Outcomes and pre‑processing Our primary outcome measures were lower limb kin- ematics (i.e. hip, knee and ankle in the sagittal plane) and spatiotemporal gait parameters (i.e. walking speed, cadence, stride time, step length, swing – and stance time, step width). i Lastly, the level of enjoyment was assessed with two Visual Analogue Scales (VAS). After walking without the VR and with the VR, participants were asked to answer following two questions: VAS1 – Indicate on the line below how much you enjoyed walking on the treadmill under these conditions, VAS2 – Indicate on the line below whether you would like to do this type of gait training during your rehabilitation (stroke group only). Participants had to answer these questions by drawing a line on a 10 cm horizontal line. At both ends of the line, opposite answers were provided. Using a ruler, the score can be determined by measuring the distance (cm) on the 10-cm line between the beginning of the line (left side) and the participant’s mark, providing a range of scores from 0–10. Kinematic data were recorded with the use of a 10-camera VICON Vero 1.3 system at 100 Hz using the full body Plug-in-Gait model provided by Vicon. In this study we only used marker data from the lower limbs. Sagittal kinematic marker data of the hip, knee and ankle were processed using Vicon Nexus software. Gait cycle segmentation of kinematic data and calculation of the spatiotemporal gait parameters (i.e. cadence, stride time, step length, swing – and stance time, step width) were performed in Python 3.7. (Anaconda Inc., USA) with custom-made scripts. Walking speed was measured con- tinuously and was derived directly from the treadmill sys- tem. Data were resampled to 100 Hz with custom-made scripts in Python 3.7. Statistical analysis IBM SPSS Statistic The critical threshold was set equivalent to α = 0.0253 to account for The sense of presence experienced in a virtual envi- ronment was assessed with the Igroup Presence Ques- tionnaire (IPQ) [29]. The IPQ consists of 14 questions divided in three subscales (spatial presence, involvement, experienced realism) and one additional general item not belonging to a subscale. “Spatial presence” measures the sense of being physically present (e.g. “Somehow I felt that the virtual world surrounded me”), “involvement” measures the attention devoted to the virtual environ- ment and involvement experienced (e.g. “How aware were you of the real world surrounding while navigating in the virtual world?”) and “Experienced realism” meas- ures the subjective experience of realism (e.g. “How real did the virtual world seem to you?”). The IPQ is scored on a seven-point Likert scale ranging from 0 (when totally disagreeing with the statement) to 6 (when totally De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 6 of 18 Table 1 Subjects’ demographic and clinical characteristics BDI Beck Depression Inventory. FAC Functional Ambulation Categories. Values are expressed in number (percentage) or mean ± standard deviation Characteristic Stroke (n = 16) Healthy (n = 16) p-value Age (years) 53.88 ± 11.43 53.75 ± 11.61 0.976 Sex Male (n, %) Female (n, %) 10 (62.5) 6 (37.5) 10 (62.5) 6 (37.5) 1.000 Height (cm) 172.00 ± 8.52 173.50 ± 6.82 0.587 Weight (kg) 74.18 ± 14.35 73.41 ± 11.13 0.867 BDI score 10.88 ± 8.68 1.81 ± 2.46 < 0.001 Time since stroke (months) 44.24 (49.20) - Paretic side Left (n, %) Right (n, %) 9 (56.25) 7 (43.75) - - FAC score 4 - Fugl-Meyer lower limb (/34) 22.69 (6.87) - Table 1 Subjects’ demographic and clinical characteristics multiple comparisons. The t-statistic (SPM{t}) was calcu- lated at each point of the time-series and where SPM{t} crossed the threshold, significant differences were found.fl , gif To investigate the effect of optic flow speed on the gait biomechanics, four time points were compared to exam- ine both the short-term and long-term effect: the aver- ages during the 30 s before the manipulation, compared to those obtained during the 30 s immediately after the manipulation, the middle 30 s and the last 30 s of the 8-min trial. Statistical analysis IBM SPSS Statistic LMM were conducted for spatiotemporal parameters, with optic flow condition (matched, fast, slow), time (pre manipulation, post manipulation, mid- dle and end of the trial) and group (post-stroke, healthy) as fixed factors, accounting for the within subject corre- lations and a random intercept of participants. Multiple models were again built in SPSS and with the use of the AIC the best fitted model was chosen. The within sub- ject covariance was unstructured. For the kinematic data, a SPM two-way repeated measures ANOVA was per- formed to examine the effect of time (pre manipulation, post manipulation, middle and end of the trial) and group (post-stroke, healthy) in each optic flow condition: the SPM{F} was calculated at each point of the time-series. Where SPM{F} crossed a threshold equivalent to α = 0.05, post-hoc Bonferroni analyses were performed using SPM paired t-tests. For post-hoc comparisons, the SPM{t} statistic was calculated for each comparison. The criti- cal threshold was set equivalent to α = 0.017 to account for multiple comparisons. Significant differences were recorded where the SPM{t} crossed this threshold.f BDI Beck Depression Inventory. FAC Functional Ambulation Categories. Values are expressed in number (percentage) or mean ± standard deviation Effect of fully immersive VR on the gait biomechanics Spatiotemporal gait parametershf Effect of fully immersive VR on the gait biomechanics Spatiotemporal gait parametershf The resulting LMM focusing on the effect of condi- tion and group suggested that no significant interac- tion effect between condition and group was found for all spatiotemporal gait parameters (see Additional file 1: Table S1 for all models). For cadence, stride time, stance time (unaffected leg post-stroke), swing time (unaffected leg post-stroke) and step width a main effect of condition was found (Table 2). When walk- ing with the immersive VR, people post-stroke walked with a significantly slower cadence (MD -3.69strides/ min [− 6.22; − 1.15], p = 0.006), a longer stride time (MD 0.10 s [0.02;0.18], p = 0.017) and stance time of the unaffected leg (MD 1.47% [0.61;2.32], p = 0.001) and a shorter swing time of the unaffected leg (MD -1.47% [− 2.32; − 0.61], p = 0.001). The healthy controls signifi- cantly reduced their step width when walking with the VR (MD -1.93 cm [− 3.09; − 0.77], p = 0.002). To investigate the effect of the level of immersion dur- ing walking with different optic flow speeds on the gait biomechanics, the mean differences (MD) of three com- parisons (pre–post manipulation, pre–mid trial, and pre–end trial) of the GRAIL session were compared to those obtained from the HMD session in both groups with a paired sample t-test. For the sense of presence, LMM were conducted with condition (GRAIL, HMD) and group (post-stroke, healthy) as fixed factors, account- ing for the within subject correlations. Models were build using the AIC in SPSS. The within subject covariance was unstructured. Subjects characteristics SPM two-way ANOVA analyses were performed on 15 subjects in each group, due to missing data of one per- son in the stroke group (missing data was due to one or more Vicon markers that fell off while walking). To maintain equal group sizes, the healthy matched par- ticipant was also removed. The SPM two-way ANOVA revealed no significant interaction effect, nor a main effect of condition (Additional file 1: Figs. S1 and S2). Sixteen people post-stroke and 16 age—and sex matched healthy controls participated in this study. There were no significant differences in baseline characteristics observed between groups, with exception for the score on the Beck Depression Inventory (BDI). People post- stroke scored significantly higher on this questionnaire (higher total scores indicate more severe depressive symptoms) (Table 1). De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 7 of 18 Table 2 Effect of fully immersive virtual reality on the spatiotemporal gait parameters Values are reported in mean with 95% confidence interval and MD (mean difference) with 95% confidence interval. VR virtual reality, %GC: percentage of gait cycle. ♦for the healthy participants the average of the left and right side is used. The asterisk indicates a significant difference No VR with VR MD No VR vs. Subjects characteristics VR p value Walking speed (m/s) Stroke 0.88 [0.72;1.04] 0.79 [0.66;0.93] − 0.09 [− 0.19; 0.01] 0.085 Healthy 1.39 [1.24;1.55] 1.37 [1.23;1.50] − 0.02 [− 0.13; 0.08] 0.625 Cadence (strides/min) Stroke 48.07 [44.04;52.09] 44.38 [40.63;48.14] − 3.69 [− 6.22; − 1.15] 0.006 Healthy 56.51 [52.61;60.41] 55.65 [52.01;59.28] − 0.86 [− 3.32; 1.59] 0.477 Stride time (sec) Stroke 1.32 [1.20;1.44] 1.42 [1.30;1.54] 0.10 [0.02; 0.18] 0.017* Healthy 1.07 [0.95;1.18] 1.08 [0.97;1.20] 0.02 [− 0.06; 0.10] 0.641 Step length affected♦ (cm) Stroke 50.97 [46.11;55.82] 51.83 [47.37;56.29] 0.86 [− 2.84; 4.57] 0.637 Healthy 66.79 [62.09;71.49] 67.85 [63.53;72.16] 1.06 [− 2.53; 4.65] 0.551 Step length unaffected♦ (cm) Stroke 49.82 [44.90;54.73] 49.29 [44.60;53.98] − 0.53 [− 3.98; 2.93] 0.757 Healthy 66.79 [62.03;71.55] 67.85 [63.31;72.39] 1.06 [− 2.28; 4.41] 0.523 Stance time affected♦ (%GC) Stroke 67.14 [65.62;68.67] 67.98 [66.77;69.18] 0.83 [− 0.26; 1.93] 0.132 Healthy 65.18 [63.71;66.66] 65.77 [64.61;66.94] 0.59 [− 0.48; 1.65] 0.268 Stance time unaffected♦ (%GC) Stroke 69.85 [68.36;71.35] 71.32 [69.71;72.93] 1.47 [0.61; 2.32] 0.001* Healthy 65.18 [63.73;66.63] 65.77 [64.21;67.33] 0.59 [− 0.24; 1.41] 0.156 Swing time affected♦ (%GC) Stroke 32.86 [31.33;34.38] 32.02 [30.82;33.23] − 0.83 [− 1.93; 0.26] 0.123 Healthy 34.82 [33.34;36.29] 34.23 [33.06;35.39] − 0.59 [− 1.65; 0.48] 0.268 Swing time unaffected♦ (%GC) Stroke 30.15 [28.65;31.64] 28.68 [27.07;30.29] − 1.47 [− 2.32; − 0.61] 0.001* Healthy 34.82 [33.37;36.27] 34.23 [32.67;35.79] − 0.59 [− 1.41; 0.24] 0.156 Step width affected♦ (cm) Stroke 18.08 [16.02;20.14] 17.11 [15.07;19.14] − 0.97 [− 2.17; 0.23] 0.109 Healthy 14.03 [12.03;16.03] 12.10 [10.13;14.07] − 1.93 [− 3.09; − 0.77] 0.002* Step width unaffected♦ (cm) Stroke 18.07 [15.98;20.16] 16.98 [14.93;19.03] − 1.09 [− 2.31; 0.14] 0.080 Healthy 14.03 [12.01;16.05] 12.10 [10.11;14.09] − 1.93 [− 3.12; − 0.74] 0.002* Table 2 Effect of fully immersive virtual reality on the spatiotemporal gait parameters Values are reported in mean with 95% confidence interval and MD (mean difference) with 95% confidence interval. VR virtual reality, %GC: percentage of gait cycle. ♦for the healthy participants the average of the left and right side is used. The asterisk indicates a significant difference Values are reported in mean with 95% confidence interval and MD (mean difference) with 95% confidence interval. VR vi ♦for the healthy participants the average of the left and right side is used. The asterisk indicates a significant difference Enjoyment Both groups indicated that they enjoyed walking on the treadmill with VR more compared to walking on the treadmill without VR as indicated by a significant increase in VAS1 (Table 3). The difference between groups was not significant. The stroke group would also like to implement VR in their gait training as indicated by a higher score on the VAS2 with VR compared to without VR. Spatiotemporal gait parametersh The resulting LMM focusing on optic flow condi- tion (matched, fast, slow), time (pre manipulation, post manipulation, middle and end of the trial) and group (post-stroke, healthy) suggested interactions between condition and time with a main effect of group for step length, stance and swing time (affected leg post-stroke) and step width (see Additional file 1: Table S2 for all models). A three-way interaction between optic flow condition, time and group was suggested for walking speed, cadence, stride time, stance and swing time (unaf- fected leg post-stroke).l Table 4 shows the MD between optic flow condi- tion, group and time for all spatiotemporal gait param- eters. Significant interaction effects revealed that in both groups, the slow and fast optic flow speed manipulation led to significant changes in several spatiotemporal gait parameters. Immediately after the fast optic flow manip- ulation, both groups significantly decreased their walk- ing speed (stroke: MD -0.10 m/s [− 0.16;-0.04], p < 0.001; healthy: MD − 0.12 m/s [− 0.18; − 0.06], p < 0.001). This decrease in walking speed was maintained over time in the stroke group only (till mid trial). In the slow optic flow condition, immediately after the manipulation both groups significantly increased their walking speed (stroke: MD 0.06 m/s [0.02;0.10], p < 0.001; healthy: MD 0.07 m/s [0.03;0.10], p < 0.001). This increase in walk- ing speed was only maintained over time in the healthy group.h Immediately after the slow OF manipulation, people post-stroke had an increase in plantar flexion of 1.99° at 66% of the gait cycle, an increase in knee flexion of 1.07° at 18% of the gait cycle, and an increase in hip flexion of 1.39° at 20% of the gait cycle (Fig. 4). At mid trial, healthy people had an increase in dorsiflexion with a maximum of 1.02° between 15 and 25% of the gait cycle, an increase in knee flexion with a maximum of 2.18° between 6 and 21% and 65–69% of the gait cycle, and an increase in hip flexion with a maximum of 2.48° between 0 and 25% and 87–100% of the gait cycle (Fig. 5). Spatiotemporal gait parametersh At the end of the trial, healthy people had an increase in dorsiflexion with a maximum of 3.62° between 17 and 24% and at 64% of the gait cycle, an increase in knee flexion with a maximum of 2.61° between 56 – 70% of the gait cycle, and an increase in hip flexion with a maximum of 3.03° between 0 and 24% and 92–100% of the gait cycle (Fig. 6). The decrease in walking speed of the stroke group in the fast optic flow condition was accompanied by a slower cadence, longer stride time, decreased step length of the affected and unaffected leg and an increased stance time and decreased swing time of the unaffected leg. The decrease in walking speed of the healthy group was accompanied by a slower cadence and decreased step length. The increase in walking speed of the stroke group in the slow condition was accompanied by a faster cadence, shorter stride time, increased step length of the affected and unaffected leg and a decreased stance time and increased swing time of the unaffected leg. The increase in walking speed of the healthy group was accompanied by a faster cadence and increased step length. Effect of level of immersion during walking with different optic flow speeds The effect of the level of immersion on the gait biome- chanics was only investigated in the spatiotemporal gait parameters since optic flow speed had no or only a very limited effect on the lower limb kinematics. In both ses- sions, the kinematic changes were very small and did not reach the minimal clinical important difference and are therefore not considered clinically relevant [31, 32]. Table 5 shows the MD between the fast and slow optic flow condition, group and time for both sessions for the most relevant spatiotemporal gait parameters. Both in people post-stroke and healthy people, manipulating the optic flow speed in the fully immersive HMD had a greater effect on spatiotemporal gait parameters com- pared to the semi-immersive GRAIL system. The most Simulator sickness Only the stroke group had a significant increase in the SSQ after walking with the VR, from 4.68(± 7.03) points to 11.69(± 11.97) points (MD 7.01(± 6.53) points, p = 0.003). The healthy group had a non-significant increase from 1.17(± 2.25) points to 2.10(± 3.04) points (MD 0.94(± 2.55) points, p = 0.157). The difference between groups was significant (p = 0.003). Table 3 Results of the two Visual Analogue Scales Values are reported in mean with SD and MD (mean difference) with SD. VAS1: How much did you enjoy walking on the treadmill under these conditions? VAS2: Would you like to do this type of gait training during your rehabilitation (people post-stroke only)? The asterisk indicates a significant difference No VR With VR MD Within group p-value Between group p-value VAS1 (/10) Stroke 4.99(± 2.01) 6.76(± 2.73) 1.77(± 2.32) 0.010* 0.395 Healthy 6.11(± 2.23) 7.11(± 2.14) 1.01(± 1.57) 0.005* VAS2 (/10) Stroke 6.47(± 2.49) 7.90(± 1.90) 1.43(± 2.04) 0.015* – Table 3 Results of the two Visual Analogue Scales Values are reported in mean with SD and MD (mean difference) with SD. VAS1: How much did you enjoy walking on the treadmill under these conditions? VAS2: Would you like to do this type of gait training during your rehabilitation (people post-stroke only)? The asterisk indicates a significant difference Page 8 of 18 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Effect of optic flow speed manipulation on the gait biomechanics participants were removed from the analyses. The SPM analyses revealed a significant interaction effect in the matched condition for the unaffected ankle and knee joint, in the fast condition for the ankle (both affected and unaffected) and affected hip joint, and in the slow condition for the ankle (both affected and unaffected) and unaffected knee joint. A significant main effect of time was found for the ankle and hip (both affected and unaffected side post-stroke) in the matched and fast condition and for the ankle, knee and hip (both affected and unaffected side post-stroke) in the slow condition (Additional file 1: Fig. S3 through 8). In the post-hoc SPM t-tests, the critical thresholds were only exceeded in the slow condition for the ankle, knee and hip joint in both groups (unaffected side post-stroke) when comparing pre and post manipulation (people post-stroke), pre manipu- lation and mid trial (healthy group), and pre manipula- tion and end trial (healthy group). Kinematicsh The SPM two-way repeated measures ANOVA was per- formed on 11 subjects in each group due to missing data (4 healthy participant, 1 stroke patient, missing data was due to one or more Vicon markers that fell off while walking). To maintain equal group sizes, the matched De Keersmaecker et al. Kinematicsh Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 9 of 18 Table 4 Effect of optic flow speed on the spatiotemporal gait parameters Condition Group Time point Walking speed (m/s) Cadence (stride/min) Stride time (s) Step length (cm) MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value Affected leg Unaffected leg Matched optic flow Stroke Pre Post 0.03 [0.00;0.06] .081 0.21 [-0.44;0.87] 1.000 -0.01 [-0.03;0.01] 1.000 1.33 [-0.32;2.97] .185 1.61 [-0.29;3.51] .138 Mid 0.04 [− 0.06;0.14] 1.000 0.56 [− 0.96;2.09] 1.000 − 0.02 [− 0.06;0.02] 1.000 2.02 [− 1.44;5.48] .662 2.34 [− 1.29;5.97] .478 End − 0.05 [− 0.14;0.04] .812 − 1.28 [− 2.61;0.05] .066 0.05 [0.01;0.09] .011* 0.08 [− 2.64;2.81] 1.000 0.46 [− 2.80;3.73] 1.000 Healthy Pre Post 0.02 [− 0.01;0.05] .270 0.06 [− 0.60;0.71] 1.000 0.00 [− 0.02;0.02] 1.000 1.16 [− 0.49;2.81] .335 1.16 [− 0.73;3.06] .566 Mid 0.09 [− 0.01;0.19] .091 1.32 [− 0.21;2.84] .125 − 0.03 [− 0.07;0.02] .557 2.67 [− 0.79;6.13] .227 2.67 [− 0.96;6.30] .282 End 0.06 [− 0.03;0.15] .458 1.01 [− 0.33;2.35] .247 − 0.02 [− 0.06;0.02] .916 2.19 [− 0.53;4.92] .184 2.13 [− 1.14;5.40] .460 Fast optic flow Stroke Pre Post − 0.10 [− 0.16;− 0.04] < .001* − 2.41 [− 3.58;− 1.23] < .001* 0.10 [0.04;0.16] < .001* − 4.58 [− 7.29;− 1.87] < .001* − 5.01 [− 7.99;− 2.02] < .001* Mid − 0.08 [− 0.16;0.00] .037* − 2.12 [− 3.96;− 0.29] .016* 0.08 [0.01;0.14] .011* − 2.13 [− 4.66;0.39] .142 − 2.50 [− 5.19;0.18] .079 End − 0.07 [− 0.19;0.04] .512 − 2.57 [− 4.63;− 0.51] .008* 0.13 [− 0.02;0.29] .127 − 2.41 [− 6.21;1.38] .498 − 2.47 [− 6.82;1.88] .715 Healthy Pre Post − 0.12 [− 0.18;− 0.06] < .001* − 1.41 [− 2.58;− 0.24[ .012* 0.04 [− 0.02;0.10] .574 − 4.37 [− 7.08;− 1.66] < .001* − 4.37 [− 7.35;− 1.39] .001* Mid − 0.01 [− 0.09;0.07] 1.000 − 0.32 [− 2.18;1.53] 1.000 0.01 [− 0.05;0.07] 1.000 − 0.16 [− 2.73;2.40] 1.000 − 0.11 [− 2.83;2.61] 1.000 End − 0.10 [− 0.21;0.02] .155 − 0.89 [− 2.98;1.21] 1.000 0.02 [− 0.13;0.17] 1.000 − 1.30 [− 5.18;2.58] 1.000 − 1.30 [− 5.72;3.11] 1.000 Slow optic flow Stroke Pre Post 0.06 [0.02;0.10] < .001* 1.44 [0.66;2.21] < .001* − 0.05 [− 0.08;− 0.02] < .001* 2.27 [0.80;3.74] < .001* 2.40 [0.61;4.19] .004* Mid 0.08 [0.00;0.15] .053 1.35 [− 0.40;3.10] .219 − 0.03 [− 0.11;0.05] 1.000 1.78 [− 0.76;4.33] .344 1.67 [− 1.66;5.00] 1.000 End 0.08 [− 0.02;0.18] .204 1.05 [− 1.26;3.35] 1.000 − 0.03 [− 0.10;0.05] 1.000 1.73 [− 1.70;5.17] .952 1.93 [− 1.53;5.39] .734 Healthy Pre Post 0.07 [0.03;0.10] < .001* 1.15 [0.37;1.93] .001* − 0.02 [− 0.05;0.01] .310 1.69 [0.22;3.16] .017* 1.69 [− 0.10;3.48] .074 Mid 0.15 [0.07;0.23] < .001* 2.42 [0.70;4.14] .003* − 0.05 [− 0.13;0.03] .621 3.99 [1.49;6.50] < .001* 3.99 [0.71;7.28] .011* De Keersmaecker et al. Kinematicsh Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 10 of 18 Table 4 (continued) Condition Group Time point Walking speed (m/s) Cadence (stride/min) Stride time (s) Step length (cm) MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value Affected leg Unaffected leg End 0.16 [0.05;0.26] .001* 2.92 [0.61;5.23] .007* − 0.06 [− 0.13,0.02] .262 4.69 [1.26;8.11] .001* 4.59 [1.13;8.05] .005* Condition Group Time point Stance time (% gait cycle) Swing time (% gait cycle) Step width (cm) MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value Affected leg Unaffected leg Affected leg Unaffected leg Affected leg Unaffected leg Matched optic flow Stroke Pre Post − 0.28 [− 0.62;0.07] .189 − 0.19 [− 0.51;0.13] .603 0.28 [− 0.07;0.62] .189 0.19 [− 0.13;0.51] .603 0.03 [− 0.57;0.64] 1.000 0.00 [− 0.60;0.60] 1.000 Mid − 0.43 [− 1.05;0.18] .336 − 0.40 [− 1.09;0.28] .652 0.43 [− 0.18;1.05] .336 0.40 [− 0.28;1.09] .652 0.33 [− 0.28;0.94] .812 0.34 [− 0.24;0.92] .651 End 0.21 [− 0.52;0.94] 1.000 0.21 [− 0.36;0.77] 1.000 − 0.21 [− 0.94;0.52] 1.000 − 0.21 [− 0.77;0.36] 1.000 0.43 [− 0.27;1.14] .554 0.36 [− 0.38;1.10] 1.000 Healthy Pre Post − 0.15 [− 0.50;0.19] 1.000 − 0.15 [− 0.47;0.17] 1.000 0.15 [− 0.19;0.50] 1.000 0.15 [− 0.17;0.47] 1.000 0.06 [− 0.55;0.66] 1.000 0.06 [− 0.54;0.66] 1.000 Mid − 0.34 [− 0.96;0.27] .747 − 0.34 [− 1.03;0.35] 1.000 0.34 [− 0.27;0.96] .747 0.34 [− 0.35;1.03] 1.000 0.15 [− 0.46;0.76] 1.000 0.15 [− 0.44;0.73] 1.000 End − 0.27 [− 1.00;0.46] 1.000 − 0.27 [− 0.84;0.30] 1.000 0.27 [− 0.46;1.00] 1.000 0.27 [− 0.30;0.84] 1.000 0.31 [− 0.39;1.02] 1.000 0.32 [− 0.42;1.06] 1.000 Fast optic flow Stroke Pre Post 0.53 [− 0.67;1.72] 1.000 1.46 [0.57;2.35] < .001* − 0.53 [− 1.72;0.67] 1.000 − 1.46 [− 2.35;− 0.57] < .001* − 0.59 [− 1.28;0.10] .132 − 0.56 [− 1.32;0.20] .285 Mid 0.83 [0.09;1.56] .020* 0.83 [0.28;1.37] < .001* − 0.83 [− 1.56,− 0.09] .020* − 0.83 [− 1.37;− 0.28] < .001* − 0.29 [− 0.94;0.36] 1.000 − 0.18 [− 0.86;0.50] 1.000 End 0.95 [− 0.16;2.05] .133 1.11 [0.21;2.01] .010* − 0.95 [− 2.05;0.16] .133 − 1.11 [− 2.01;0.21] .010* 0.06 [− 0.82;0.95] 1.000 0.06 [− 0.80;0.92] 1.000 Healthy Pre Post 0.82 [− 0.38;2.01] .383 0.82 [− 0.07;1.71] .088 − 0.82 [− 2.01;0.38] .383 − 0.82 [− 1.71;0.07] .088 0.46 [− 0.23;1.15] .430 0.46 [− 0.30;1.22] .606 Mid 0.16 [− 0.59;0.90] 1.000 0.16 [− 0.39;0.71] 1.000 − 0.16 [− 0.90;0.59] 1.000 − 0.16 [− 0.71;0.39] 1.000 0.38 [− 0.27;1.04] .650 0.39 [− 0.30;1.07] .746 End 0.37 [− 0.75;1.49] 1.000 0.39 [− 0.53;1.31] 1.000 -0.37 [− 1.49;0.75] 1.000 − 0.39 [− 1.31;0.53] 1.000 0.60 [− 0.32;1.51] .452 0.58 [− 0.30.1.74] .443 De Keersmaecker et al. Kinematicsh Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 11 of 18 Table 4 (continued) Condition Group Time point Stance time (% gait cycle) Swing time (% gait cycle) Step width (cm) MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value MD [95% CI] p-value Affected leg Unaffected leg Affected leg Unaffected leg Affected leg Unaffected leg Slow optic flow Stroke Pre Post − 0.36 [− 0.90;0.18] .421 − 0.84 [− 1.30;− 0.38] < .001* 0.36 [− 0.18;0.90] .421 0.84 [0.38;1.30] < .001* 0.44 [− 0.26;1.14] .523 0.40 [− 0.28;1.09] .647 Mid − 0.32 [− 1.67;1.03] 1.000 − 0.63 [− 1.44;0.19] .224 0.32 [− 1.03;1.67] 1.000 0.63 [− 0.19;1.44] .224 0.21 [− 0.64;1.06] 1.000 0.14 [− 0.68;0.97] 1.000 End 0.05 [− 1.26;1.37] 1.000 − 0.74 [− 1.53;0.04] .073 − 0.05 [− 1.37;1.26] 1.000 0.74 [− 0.04;1.53] .073 0.57 [− 0.02;1.17] .067 0.61 [0.02;1.21] .042* Healthy Pre Post − 0.33 [− 0.88;0.21] .552 − 0.33 [− 0.79;0.12] .294 0.33 [− 0.21;0.88] .552 0.33 [− 0.12;0.79] .294 0.09 [− 0.60;0.79] 1.000 0.09 [− 0.59;0.78] 1.000 Mid − 0.70 [− 2.04;0.64] .860 − 0.70 [− 1.51;0.11] .122 0.70 [− 0.64;2.04] .860 0.70 [− 0.11;1.51] .122 0.29 [− 0.55;1.13] 1.000 0.29 [− 0.53;1.10] 1.000 End − 0.82 [− 2.13;0.49] .508 − 0.82 [− 1.60;− 0.03] .037* 0.82 [− 0.49;2.13] .508 0.82 [0.03;1.60] .037* 0.43 [− 0.17;1.02] .313 0.42 [− 0.18;1.01] .337 Values are reported in MD (mean difference) and 95%CI (95% confidence interval). The asterisk indicates a significant difference Page 12 of 18 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 12 of 18 Fig. 4 Results of the post-hoc analyses, paired sample t-test for the slow condition (pre vs. post manipulation) in the stroke group (unaffected side). Horizontal axis is percentage gait cycle. First row is mean joint angles ± 1 standard deviation for people post-stroke pre manipulation (green) and post manipulation (blue). Second row shows SPM(t) value throughout the gait cycle. The dashed red line is equivalent to α = 0.02. Third row shows mean difference with 95% confidence interval between pre and post manipulation Fig. 4 Results of the post-hoc analyses, paired sample t-test for the slow condition (pre vs. post manipulation) in the stroke group (unaffected side). Horizontal axis is percentage gait cycle. Walking with immersive VR Results of this study demonstrated that immersive VR- enhanced treadmill walking with the use of a HMD was accepted by people post-stroke and healthy people. All participants were able to complete all walking trials with- out having any signs of severe simulator sickness, as indi- cated by the low total scores on the SSQ in both groups. People post-stroke also reported that they liked walking with the VR more than without and would like to imple- ment VR-enhanced treadmill walking in their gait reha- bilitation. These results are in line with recent studies examining the potential of immersive VR for the rehabili- tation of neurological patients [33, 34]. Discussion prominent differences were found for walking speed and step length (both affected and unaffected leg post-stroke). With the slow optic flow speed, both groups increased their walking speed and step length more with the fully immersive HMD compared to the semi-immersive GRAIL system. The opposite was seen with the fast optic flow speed, where both groups decreased there walking speed and step length more when the optic flow speed was manipulated with the fully immersive HMD. Results for all optic flow conditions and all spatiotemporal gait parameters can be found in Additional file 1: Table S3. Kinematicsh First row is mean joint angles ± 1 standard deviation for people post-stroke pre manipulation (green) and post manipulation (blue). Second row shows SPM(t) value throughout the gait cycle. The dashed red line is equivalent to α = 0.02. Third row shows mean difference with 95% confidence interval between pre and post manipulation Sense of presence T bl h h Table 6 shows the mean and SD of the IPQ subscales for each group and condition. Additional file 1: Table S4 shows the results of the LMM. The LMM for the general item only revealed a main effect of condition and sug- gested that in both groups, the fully immersive HMD resulted in a significantly higher feeling of being there than the semi-immersive GRAIL screen (MD 0.92 points, p = 0.004). The LMM for the subscales spatial presence and involvement also revealed a significant main effect of condition and suggested that in both groups, the HMD resulted in a significantly higher spatial presence (MD 0.58 points, p < 0.001) and higher involvement (MD 0.98 points, p < 0.001). The LMM for the subscale experienced realism revealed no main effects. The recent study of Winter et al. (2021) reported an increase in walking speed when walking with fully immersive VR compared to walking with no VR in healthy individuals, individuals post-stroke, and individ- uals with multiple sclerosis [33]. Participants walked in a virtual environment that was designed to increase their motivation during training and consisted of an engag- ing storyline with the implementation of gamification elements (i.e. rebuilding a virtual world by walking on a path). This is in contrast with the virtual environment used in our study, where participants walked forward in an endless city street with no gaming elements. We found that when people post-stroke walked with the HMD, they Page 13 of 18 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 13 of 18 Fig. 5 Results of the post-hoc analyses, paired sample t-test for the slow condition (pre manipulation vs. mid trial) in the healthy group. Horizontal axis is percentage gait cycle. First row is mean joint angles ± 1 standard deviation for heathy people pre manipulation (green) and mid trial (blue). Second row shows SPM(t) value throughout the gait cycle. The dashed red line is equivalent to α = 0.02. Third row shows mean difference with 95% confidence interval between pre manipulation and mid trial Fig. 5 Results of the post-hoc analyses, paired sample t-test for the slow condition (pre manipulation vs. mid trial) in the healthy group. Horizontal axis is percentage gait cycle. Sense of presence T bl h h First row is mean joint angles ± 1 standard deviation for heathy people pre manipulation (green) and mid trial (blue). Second row shows SPM(t) value throughout the gait cycle. The dashed red line is equivalent to α = 0.02. Third row shows mean difference with 95% confidence interval between pre manipulation and mid trial Fig. 6 Results of the post-hoc analyses, paired sample t-test for the slow condition (pre manipulation vs. end trial) in the healthy group. Horizontal axis is percentage gait cycle. First row is mean joint angles ± 1 standard deviation for heathy people pre manipulation (green) and end trial (blue). Second row shows SPM(t) value throughout the gait cycle. The dashed red line is equivalent to α = 0.02. Third row shows mean difference with 95% confidence interval between pre manipulation and end trial Fig. 6 Results of the post-hoc analyses, paired sample t-test for the slow condition (pre manipulation vs. end trial) in the healthy group. Horizontal axis is percentage gait cycle. First row is mean joint angles ± 1 standard deviation for heathy people pre manipulation (green) and end trial (blue). Second row shows SPM(t) value throughout the gait cycle. The dashed red line is equivalent to α = 0.02. Third row shows mean difference with 95% confidence interval between pre manipulation and end trial De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 14 of 18 Table 5 Effect of the level of immersion on the spatiotemporal gait parameters Gait parameter Group Condition Time point N GRAIL MD (SD) HMD MD (SD) GRAIL vs. Optic flow speed manipulation This might be explained by the difference in the type of virtual environment (engaging storyline with gaming elements versus an endless city street without gaming elements), but perhaps also by the difference in the type of treadmill system. In the study of Winter et al. the speed was changed manually by the participants with the use of buttons on the handles. This difference in regulation (top-down versus bottom-up) may result in different responses. Lastly, it must also be noted that prior to this study, patients were not familiar with fully immersive VR-enhanced treadmill walking and changes in spatiotemporal gait parameters may be attributed to a more cautious gait pattern. It is very likely that this more cautious gait pattern will diminish when patients are more acquainted with the HMD. Nonetheless, these results highlight the need to incorporate valuable prin- ciples (such as performance feedback, gaming elements or competition) in the virtual environment to influence a person’s gait pattern. More research about implementing such valuable principles in the virtual environment that walked with a slower cadence, a longer stride time, and a longer stance and shorter swing time of the unaffected leg compared to walking without VR. This contrasts with the increased walking speed reported in the study of Winter et al. (2021). This might be explained by the difference in the type of virtual environment (engaging storyline with gaming elements versus an endless city street without gaming elements), but perhaps also by the difference in the type of treadmill system. In the study of Winter et al. the speed was changed manually by the participants with the use of buttons on the handles. This difference in regulation (top-down versus bottom-up) may result in different responses. Lastly, it must also be noted that prior to this study, patients were not familiar with fully immersive VR-enhanced treadmill walking and changes in spatiotemporal gait parameters may be attributed to a more cautious gait pattern. It is very likely that this more cautious gait pattern will diminish when patients are more acquainted with the HMD. Nonetheless, these results highlight the need to incorporate valuable prin- ciples (such as performance feedback, gaming elements or competition) in the virtual environment to influence a person’s gait pattern. More research about implementing such valuable principles in the virtual environment that f The choice for a constant one-time speed manipulation was based on existing literature [14, 15, 17, 18]. Sense of presence T bl h h Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 15 of 18 Table 5 (continued) Gait parameter Group Condition Time point N GRAIL MD (SD) HMD MD (SD) GRAIL vs. HMD p-value Mid 15 0.36 (2.98) − 0.16 (2.75) 0.552 End 14 0.69 (2.91) − 1.74 (4.54) 0.039* Slow Pre Post 15 1.40 (1.52) 1.74 (2.35) 0.616 Mid 15 1.84 (2.12) 4.09 (2.40) 0.010* End 14 1.57 (1.77) 5.14 (2.72) < 0.001* Values are reported in MD (mean difference) and SD (standard deviation). GRAIL Gait Real-time Interactive Lab, HMD head-mounted display. The asterisk indicates a significant difference could positively influence the gait pattern of people post- stroke and could be used during gait training is needed. Table 6 Results of the igroup presence questionnaire Values are reported in mean and SD (standard deviation). IPQ: Igroup Presence Questionnaire, GRAIL: Gait Real-time Interactive Lab, HMD: head-mounted display IPQ subscale Condition Stroke Healthy GRAIL vs. HMD Mean (SD) Mean (SD) p-value General item GRAIL 3.27 (1.39) 2.94 (1.77) 0.004 HMD 3.94 (1.48) 4.06 (1.34) Spatial pres‑ ence GRAIL 3.16 (0.62) 3.09 (0.80) < 0.001 HMD 4.00 (0.54) 3.46 (0.75) Involvement GRAIL 2.70 (1.62) 2.50 (1.43) < 0.001 HMD 3.78 (1.39) 3.36 (1.18) Realness GRAIL 2.35 (1.34) 1.86 (1.01) 0.053 HMD 2.56 (1.44) 2.26 (1.15) Table 6 Results of the igroup presence questionnaire Sense of presence T bl h h HM p-value Walking speed (m/s) Stroke Fast Pre Post 16 − 0.01 (0.08) − 0.10 (0.07) 0.001* Mid 16 − 0.01 (0.11) − 0.08 (0.14) 0.058 End 16 − 0.01 (0.12) − 0.07 (0.12) 0.018* Slow Pre Post 16 0.00 (0.05) 0.06 (0.05) 0.005* Mid 16 0.00 (0.06) 0.08 (0.13) 0.087 End 16 0.01 (0.04) 0.08 (0.15) 0.088 Healthy Fast Pre Post 16 − 0.04 (0.03) − 0.12 (0.10) 0.013* Mid 16 0.01 (0.07) − 0.01 (0.09) 0.410 End 16 0.01 (0.08) − 0.10 (0.21) 0.043* Slow Pre Post 16 − 0.01 (0.03) 0.07 (0.06) < 0.001* Mid 16 0.01 (0.04) 0.15 (0.09) < 0.001* End 16 0.01 (0.03) 0.18 (0.11) < 0.001* Cadence (stride/min) Stroke Fast Pre Post 16 − 1.02 (1.59) − 2.41 (1.79) 0.841 Mid 16 − 0.88 (2.55) − 2.12 (3.32) 0.051 End 15 − 1.07 (2.61) − 2.66 (3.74) 0.001* Slow Pre Post 16 1.18 (1.27) 1.44 (1.40) 0.377 Mid 15 1.10 (2.81) 1.13 (3.00) 0.393 End 14 1.23 (2.58) 0.66 (4.50) 0.077 Healthy Fast Pre Post 16 − 0.38 (0.44) − 1.41 (1.85) 0.030* Mid 15 − 0.03 (0.95) − 0.41 (1.85) 0.302 End 13 − 0.04 (1.08) − 0.76 (2.29) 0.143 Slow Pre Post 15 0.65 (0.63) 1.16 (0.83) 0.068 Mid 15 0.88 (0.93) 2.35 (1.80) 0.008* End 14 1.13 (1.61) 2.93 (2.26) 0.017* Stride time (s) Stroke Fast Pre Post 16 0.03 (0.05) 0.10 (0.11) 0.030* Mid 16 0.03 (0.09) 0.08 (0.12) 0.083 End 15 0.03 (0.10) 0.14 (0.32) 0.127 Slow Pre Post 16 − 0.04 (0.12) − 0.05 (0.06) 0.581 Mid 15 − 0.04 (0.12) − 0.01 (0.15) 0.642 End 14 − 0.05 (0.11) − 0.02 (0.16) 0.446 Healthy Fast Pre Post 16 0.01 (0.01) 0.04 (0.05) 0.050 Mid 15 0.00 (0.02) 0.01 (0.04) 0.219 End 13 0.00 (0.02) 0.02 (0.05) 0.107 Slow Pre Post 15 − 0.01 (0.01) − 0.02 (0.02) 0.030* Mid 15 − 0.02 (0.02) − 0.05 (0.04) 0.009* End 14 − 0.02 (0.03) − 0.06 (0.05) 0.020* Step length (cm) Affected leg Stroke Fast Pre Post 16 0.43 (3.55) − 4.58 (4.30) < 0.001* Mid 16 1.20 (5.17) − 2.13 (4.46) 0.030* End 15 2.36 (5.54) − 2.60 (6.55) 0.006* Slow Pre Post 16 2.09 (1.71) 2.27 (2.03) 0.813 Mid 15 2.68 (5.21) 1.93 (4.71) 0.729 End 13 2.51 (3.77) 2.89 (5.74) 0.827 Step length (cm) Unaffected leg Stroke Fast Pre Post 16 0.42 (3.84) − 5.01 (5.02) 0.001* Mid 16 − 0.02 (4.99) − 2.50 (4.84) 0.099 End 15 0.57 (7.47) − 2.71 (8.01) 0.141 Slow Pre Post 16 1.79 (3.57) 2.40 (2.94) 0.646 Mid 15 3.10 (9.15) 1.95 (6.43) 0.726 End 13 3.23 (8.39) 3.12 (5.21) 0.967 Healthy Fast Pre Post 16 − 1.55 (1.35) − 4.37 (3.63) 0.011* Table 5 Effect of the level of immersion on the spatiotemporal gait parameters De Keersmaecker et al. Optic flow speed manipulation Both groups responded to the optic flow speed manipu- lation by adjusting their spatiotemporal gait parameters. However, relatively small changes in spatiotemporal gait parameters were reported. The changes in the lower limb joint kinematics were too small to be of any clini- cal value [31, 32]. Both people post-stroke and healthy controls increased their walking speed with a slow optic flow speed and decreased their walking speed with a fast optic flow speed. However, only the decrease in walking speed with the fast optic flow speed reached the minimal clinically important difference of 0.10 m/s [35]. Improv- ing patients’ walking speed is an important therapeutic outcome and is often a goal of post-stroke rehabilita- tion [6]. The fact that also the stroke group responded to the optic flow speed manipulation and showed altera- tions in their gait pattern, provides a rationale to incor- porate such manipulations in a VR-enhanced training to promote faster walking speeds. The increased walk- ing speed in people post-stroke was accompanied by a faster cadence and longer step length of both the affected and unaffected leg. However, it must be noted that the increase in walking speed was only maintained over time by the healthy group and not by the people post-stroke. This may indicate that in people post-stroke, the effect of a single manipulation is rather short lasting and may not be sufficient to influence their locomotion. Therefore, it is advisable to further explore the effect of different types of optic flow speed manipulations, such as multiple inter- mittent manipulations of a constant optic flow speed over a longer period. Before optic flow speed manipulations can be implemented in such a training, further work is needed to determine the most optimal type of optic flow speed manipulation as well as to investigate the carry- over effects to overground walking.h Values are reported in mean and SD (standard deviation). IPQ: Igroup Presence Questionnaire, GRAIL: Gait Real-time Interactive Lab, HMD: head-mounted display Values are reported in mean and SD (standard deviation). IPQ: Igroup Presence Questionnaire, GRAIL: Gait Real-time Interactive Lab, HMD: head-mounted display walked with a slower cadence, a longer stride time, and a longer stance and shorter swing time of the unaffected leg compared to walking without VR. This contrasts with the increased walking speed reported in the study of Winter et al. (2021). Optic flow speed manipulation Based on De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 16 of 18 Page 16 of 18 the results of the study of Lamontagne et al. (2007), it is suggested that for people post-stroke constant optic flow speeds are easier to perceive and to integrate than con- tinuously changing optic flow speeds (e.g. sinusoidally patterns of optic flow speed) and could therefore elicit a greater effect on patient’s gait pattern [13]. The one-time manipulation in our study lasted for 7 min, which was longer than most previous studies investigating constant optic flow speed manipulations [15, 17, 18]. Our results stipulated that in people post-stroke, there were mainly changes in spatiotemporal gait parameters immediately after the manipulation, by mid or end trial (respectively 3 or 6 min after the manipulation) these changes were no longer visible. It is therefore suggested that the effect of a one-time optic flow speed manipulation is rather short lasting. More research about different types of manipu- lations, such as multiple intermittent manipulations of a constant optic flow speed over a longer period of time, is therefore needed. semi-immersive GRAIL projection screen, participants were still aware of their real environment and thus also of the real optic flow. Furthermore, during the GRAIL session, people also had more visual information because they could look down while walking. It was only dur- ing the HMD session that participants were completely immersed in the virtual environment. It is likely that the optic flow speed manipulations were much more notice- able for the participants when walking with the HMD, which could explain its greater effect on locomotion. These results are promising and support the use of more immersive VR devices for rehabilitation.h This study is an initial step to establish fully immersive VR-enhanced treadmill training for people post-stroke. Fully immersive VR devices are still not widely used for rehabilitation today, despite their advantages over less immersive VR systems. Compared to the GRAIL system, the HMD has some important assets: the HMD is a much more affordable system and requires much less space when combined with treadmill walking, making the HMD more suitable to implement in rehabilitation – or even home settings. To date, limited studies investigated long-term fully immersive VR interventions. Semi‑immersive vs. fully immersive VRhi This is the first study to examine the effect of immersion on sense of presence and during walking with different optic flow speeds by providing a direct comparison of manipulating the optic flow speed in a semi-immersive (GRAIL) and fully immersive (HMD) virtual environ- ment. As hypothesized, both groups reported a higher sense of presence when the virtual environment was pre- sented via the HMD, compared to the semi-immersive GRAIL projection screen. Manipulating the optic flow speed in the fully immersive virtual environment also had a larger effect on the spatiotemporal gait param- eters compared to the semi-immersive virtual environ- ment. While walking on the self-paced treadmill with the Optic flow speed manipulation Therefore, in the future it will be important to also investigate the long-term effect of fully immersive VR interventions for stroke rehabilitation.hhi Other factors that could have influenced the current results and should be investigated in future studies are stroke severity, stroke onset and stroke location. People post-stroke included in this study were all chronic, ambu- latory stroke patients who could walk independently, but still experienced some difficulties with stairs or uneven surfaces. The average time post-stroke was 44.24 months but ranged from 3.4 months to 202.5 months (16.8 years). An important limitation of our study is that we did not include the stroke location as a patient characteristic. There is a complex cortical network that is responsible for the perception and use of optic flow during locomo- tion and involves several visual, multisensory and vestib- ular areas [36]. When the stroke is located in one of these brain areas, the perception and use of optic flow can be affected and patients could react differently on the optic flow speed manipulations [10]. It is therefore advisable for future research to include specific information about the stroke location as a patient characteristic. This study also included some limitations. The first limitation is the limited inventoried patients’ character- istics included in the study. An important characteristic that was not included as a baseline characteristic is par- ticipants’ degree of visual dependency. Controlling our locomotion is a complex task and involves the integra- tion of visual, vestibular, and proprioceptive information. People do not always rely equally on these three sources of information to control their locomotion. For example, it is possible that people have changed their weighting of vision and become more or less visual dependent which could influence their response to optic flow speed manip- ulations. A second limitation is related to the specific inclusion criteria. Therefore, results apply only to the population studied and are not generalizable to all people post-stroke. Lastly, to investigate the effect of patients’ characteristics that could influence the perception of the optic flow (such as visual dependency), a larger sample size is needed. Funding The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: E.D.K. is a Strategic Basic Research fellow funded by the Research Foundation – Flanders (FWO) [1S58419N]. Availability of data and materials The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request. Conclusionh This study demonstrated that adding fully immersive VR while walking on a self-paced treadmill influenced the gait pattern of people post-stroke and led to a slightly more cautious gait pattern. However, walking with the HMD was well tolerated and enjoyable. Manipulat- ing the optic flow speed in a fully immersive virtual De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 17 of 18 environment mainly influenced the spatiotemporal gait parameters of people post-stroke and healthy people. A negative relationship between optic flow speed and walk- ing speed was observed in both groups, meaning that people walked faster with a slower optic flow speed and slower with a faster optic flow speed. Manipulating the optic flow speed in a fully immersive virtual environment had a greater effect on spatiotemporal gait parameters compared to the semi-immersive virtual environment and elicited a greater sense of presence. Further work is needed to determine the most optimal type of optic flow speed manipulation as well as which other principles need to be implemented to positively influence the gait pattern of people post-stroke. Acknowledgements h h ld l k The authors would like to thank the participants for taking part in this trial, Smart Space lab for letting us use the GRAIL system and the Support for Quantitative and Qualitative Research (SQUARE) of the Vrije Universiteit Brussel for their support with the data analysis. 6. Dobkin BH. Clinical practice. Rehabilitation after stroke (in eng). N Engl J Med. 2005;352(16):1677–84. https://doi.org/10.1056/NEJMcp043511. 7. Beyaert C, Vasa R, Frykberg GE. Gait post-stroke: Pathophysiology and rehabilitation strategies (in eng). Neurophysiol Clin. 2015;45(4–5):335–55. https://doi.org/10.1016/j.neucli.2015.09.005. Author details 2-way repeated measures ANOVA SPM analyses matched condition (affected side post stroke). Figure S4. 2-way repeated measures ANOVA SPM analyses fast condition (affected side post stroke). Figure S5. 2-way repeated measures ANOVA SPM analyses slow condition (affected side post stroke). Figure S6: 2-way repeated measures ANOVA SPM analyses matched condition (unaffected side post stroke). Figure S7. 2-way repeated measures ANOVA SPM analyses fast condition (unaffected side post stroke). Figure S8. 2-way repeated measures ANOVA SPM analyses slow condition (unaffected side post stroke) Competing interests The authors declare that they have no competing interests. Competing interests The authors declare that they have no competing interests. References 1. Keshner EA, Fung J. The quest to apply VR technology to rehabilitation: tribulations and treasures. J Vestib Res Equilib Orientat. 2017;27(1):1–5. https://doi.org/10.3233/VES-170610. 1. Keshner EA, Fung J. The quest to apply VR technology to rehabilitation: tribulations and treasures. J Vestib Res Equilib Orientat. 2017;27(1):1–5. https://doi.org/10.3233/VES-170610. 2. Levac DE, et al. Promoting therapists’ use of motor learning strategies within virtual reality-based stroke rehabilitation (in eng). PLoS ONE. 2016;11(12):e0168311. https://doi.org/10.1371/journal.pone.0168311. 3. Gatica-Rojas V, Mendez-Rebolledo G. Virtual reality interface devices in the reorganization of neural networks in the brain of patients with neuro‑ logical diseases (in eng). Neural Regen Res. 2014;9(8):888–96. https://doi. org/10.4103/1673-5374.131612. (unaffected side post stroke). Figure S8. 2-way repeated measures ANOVA SPM analyses slow condition (unaffected side post stroke) 4. Benjamin EJ, et al. Heart disease and stroke statistics-2017 update: a report from the American Heart Association (in eng). Circulation. 2017;135(10):e146–603. https://doi.org/10.1161/cir.0000000000000485. 5. Jørgensen HS, Nakayama H, Raaschou HO, Olsen TS. Recovery of walking function in stroke patients: the Copenhagen Stroke Study (in eng). Arch Phys Med Rehabil. 1995;76(1):27–32. https://doi.org/10.1016/s0003- 9993(95)80038-7. Ethics approval and consent to participate Ethical approval for the experiments was given by the Ethics Committee of the University of Brussels and the University Hospital of Ghent (B1432020000120). All subjects gave written informed consent before participating in the study. Consent for publication Consent for publication Consent for the publication of subjects’ data has been obtained for this study. Abbreviations Abbreviations VR Virtual reality HMD Head-mounted display GRAIL Gait Real-time Interactive Lab SSQ Simulator Sickness Questionnaire IPQ Igroup Presence Questionnaire VAS Visual Analogue Scale LMM Linear mixed-effect models AIC Akaike’s Information Criteria SPM Statistical parametric mapping MD Mean difference BDI Beck Depression Inventory FAC Functional Ambulation Categories Author details 1 Rehabilitation Research, Department of Physiotherapy, Human Physiology and Anatomy, Vrije Universiteit Brussel, Brussels, Belgium. 2 Center for Neuro‑ sciences (C4N), Vrije Universiteit Brussel, Brussels, Belgium. 3 Brussels Human Robotic Research Center (BruBotics), Vrije Universiteit Brussel, Brussels, Belgium. 4 Alliance research group REBI (Rehabilitation technology for people with a brain injury), Vrije Universiteit Brussel & Ghent University, Brussels, Belgium. 5 Faculty of Medicine and Health Sciences, Department Rehabilita‑ tion Sciences, Campus UZ Gent, Ghent, Belgium. 6 Department of Electronics and Informatics, Engineering Sciences, Vrije Universiteit Brussel, Brussels, Belgium. 7 Movement and Nutrition for Health and Performance, Vrije Univer‑ siteit Brussel, Brussels, Belgium. 8 Department of Mechanical Engineering, KU Leuven, Heverlee, Leuven, Belgium. 9 Imec, Leuven, Belgium. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s12984-023-01254-0. The online version contains supplementary material available at https://doi. org/10.1186/s12984-023-01254-0. Received: 25 October 2022 Accepted: 18 September 2023 Additional file 1: Table S1. Linear mixed models for all spatiotempo‑ ral gait parameters – effect VR. Table S2. Linear mixed models for all spatiotemporal gait parameters – effect optic flow speed. Table S3. Effect of level of immersion on the spatiotemporal gait parameters. Table S4. Linear mixed models for the Igroup Presence Questionnaire. Figure S1. 2-way ANOVA SPM analyses for the joint angles (affected side post-stroke). Figure S2. 2-way ANOVA SPM analyses for the joint angles (unaffected side post-stroke). Figure S3. 2-way repeated measures ANOVA SPM analyses matched condition (affected side post stroke). Figure S4. 2-way repeated measures ANOVA SPM analyses fast condition (affected side post stroke). Figure S5. 2-way repeated measures ANOVA SPM analyses slow condition (affected side post stroke). Figure S6: 2-way repeated measures ANOVA SPM analyses matched condition (unaffected side post stroke). Figure S7. 2-way repeated measures ANOVA SPM analyses fast condition (unaffected side post stroke). Figure S8. 2-way repeated measures ANOVA SPM analyses slow condition (unaffected side post stroke) Additional file 1: Table S1. Linear mixed models for all spatiotempo‑ ral gait parameters – effect VR. Table S2. Linear mixed models for all spatiotemporal gait parameters – effect optic flow speed. Table S3. Effect of level of immersion on the spatiotemporal gait parameters. Table S4. Linear mixed models for the Igroup Presence Questionnaire. Figure S1. 2-way ANOVA SPM analyses for the joint angles (affected side post-stroke). Figure S2. 2-way ANOVA SPM analyses for the joint angles (unaffected side post-stroke). Figure S3. Author contributions BJ adapted the virtual environment from the GRAIL system to work with the HMD; EDK and AVB recruited the subjects and performed the experiments. All authors were involved in designing the experiments, analyzing and interpret‑ ing the data. EDK wrote the original document draft. All authors provided substantial feedback on the manuscript. All authors read and approved the final manuscript. 8. Cano Porras D, Sharon H, Inzelberg R, Ziv-Ner Y, Zeilig G, Plotnik M. Advanced virtual reality-based rehabilitation of balance and gait in clini‑ cal practice. Ther Adv Chronic Dis. 2019. https://doi.org/10.1177/20406 22319868379. De Keersmaecker et al. Journal of NeuroEngineering and Rehabilitation (2023) 20:124 Page 18 of 18 Page 18 of 18 9. Mehrholz J, Thomas S, Elsner B. Treadmill training and body weight sup‑ port for walking after stroke (in eng). Cochrane Database Syst Rev. 2017. https://doi.org/10.1002/14651858.CD002840.pub4. 32. Guzik A, Drużbicki M, Wolan-Nieroda A, Turolla A, Kiper P. Estimating mini‑ mal clinically important differences for knee range of motion after stroke (in eng). J Clin Med. 2020. https://doi.org/10.3390/jcm9103305. 10. Keshner EA, Lamontagne A. The untapped potential of virtual reality in rehabilitation of balance and gait in neurological disorders (in eng). Front Virtual Real. 2021. https://doi.org/10.3389/frvir.2021.641650. 33. Winter C, Kern F, Gall D, Latoschik ME, Pauli P, Käthner I. Immersive virtual reality during gait rehabilitation increases walking speed and motiva‑ tion: a usability evaluation with healthy participants and patients with multiple sclerosis and stroke (in eng). J Neuroeng Rehabil. 2021;18(1):68. https://doi.org/10.1186/s12984-021-00848-w. 11. Prokop T, Schubert M, Berger W. Visual influence on human locomotion— modulation to changes in optic flow. Exp Brain Res. 1997;114(1):63–70. https://doi.org/10.1007/PL00005624. 34. Moan ME, Vonstad EK, Su X, Vereijken B, Solbjør M, Skjæret-Maroni N. Experiences of stroke survivors and clinicians with a fully immersive vir‑ tual reality treadmill exergame for stroke rehabilitation: a qualitative pilot study (in eng). Front Aging Neurosci. 2021;13:735251. https://doi.org/10. 3389/fnagi.2021.735251. 12. Adamovich SV, Fluet GG, Tunik E, Merians AS. Sensorimotor training in virtual reality: a review. NeuroRehabilitation. 2009;25(1):29–44. https://doi. org/10.3233/NRE-2009-0497. 13. Lamontagne A, Fung J, McFadyen BJ, Faubert J. Modulation of walking speed by changing optic flow in persons with stroke (in eng). J Neuroeng Rehabil. 2007;4:22. https://doi.org/10.1186/1743-0003-4-22. 35. Bohannon RW, Glenney SS. Minimal clinically important difference for change in comfortable gait speed of adults with pathology: a systematic review (in eng). J Eval Clin Pract. 2014;20(4):295–300. https://doi.org/10. 1111/jep.12158.l 14. Katsavelis D, Mukherjee M, Decker L, Stergiou N. Author contributions The effect of virtual reality on gait variability (in eng). Nonlinear Dynamics Psychol Life Sci. 2010;14(3):239–56. 36. Uesaki M, Ashida H. Optic-flow selective cortical sensory regions associated with self-reported states of vection (in eng). Front Psychol. 2015;6:775. https://doi.org/10.3389/fpsyg.2015.00775. 15. Mohler BJ, Thompson WB, Creem-Regehr SH, Pick HL Jr, Warren WH Jr. Visual flow influences gait transition speed and preferred walking speed (in eng). Exp Brain Res. 2007;181(2):221–8. https://doi.org/10.1007/ s00221-007-0917-0. 19. !!! INVALID CITATION !!! [19–21]. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: 28. Kennedy RS, Lane NE, Berbaum KS, Lilienthal MG. Simulator Sickness Questionnaire: An Enhanced Method of Quantifying Simulator Sickness. Int J Aviat Psychol. 1993;3(3):203–20. 28. Kennedy RS, Lane NE, Berbaum KS, Lilienthal MG. Simulator Sickness Questionnaire: An Enhanced Method of Quantifying Simulator Sickness. Int J Aviat Psychol. 1993;3(3):203–20. 29. Thomas WS. The sense of presence in virtual environments: a three- component scale measuring spatial presence, involvement, and realness. Zf ur Medienpsychologie. 2003;15:69–71. 29. Thomas WS. The sense of presence in virtual environments: a three- component scale measuring spatial presence, involvement, and realness. Zf ur Medienpsychologie. 2003;15:69–71. 30. Pataky TC. One-dimensional statistical parametric mapping in Python (in eng). Comput Methods Biomech Biomed Engin. 2012;15(3):295–301. https://doi.org/10.1080/10255842.2010.527837. 30. Pataky TC. One-dimensional statistical parametric mapping in Python (in eng). Comput Methods Biomech Biomed Engin. 2012;15(3):295–301. https://doi.org/10.1080/10255842.2010.527837. 31. 19. !!! INVALID CITATION !!! [19–21]. 20. Schubert M, Prokop T, Brocke F, Berger W. Visual kinesthesia and locomo‑ tion in Parkinson’s disease (in eng). Mov Disord. 2005;20(2):141–50. https://doi.org/10.1002/mds.20281. 21. Lim H. Effect of the modulation of optic flow speed on gait parameters in children with hemiplegic cerebral palsy. J Phys Ther Sci. 2014;26(1):145–8. https://doi.org/10.1589/jpts.26.145. 21. Lim H. Effect of the modulation of optic flow speed on gait parameters in children with hemiplegic cerebral palsy. J Phys Ther Sci. 2014;26(1):145–8. https://doi.org/10.1589/jpts.26.145. 22. Kang HK, Kim Y, Chung Y, Hwang S. Effects of treadmill training with optic flow on balance and gait in individuals following stroke: randomized controlled trials (in eng). Clin Rehabil. 2012;26(3):246–55. https://doi.org/ 10.1177/0269215511419383. 22. Kang HK, Kim Y, Chung Y, Hwang S. Effects of treadmill training with optic flow on balance and gait in individuals following stroke: randomized controlled trials (in eng). Clin Rehabil. 2012;26(3):246–55. https://doi.org/ 10.1177/0269215511419383. 23. Lim H. Effect of the modulation of optic flow speed on gait parameters in children with hemiplegic cerebral palsy (in eng). J Phys Ther Sci. 2014;26(1):145–8. https://doi.org/10.1589/jpts.26.145. 24. Vinas-Diz S, Sobrido-Prieto M. Virtual reality for therapeutic purposes in stroke: a systematic review. Neurologia. 2016;31(4):255–77. https://doi. org/10.1016/j.nrl.2015.06.012. 24. Vinas-Diz S, Sobrido-Prieto M. Virtual reality for therapeutic purposes in stroke: a systematic review. Neurologia. 2016;31(4):255–77. https://doi. org/10.1016/j.nrl.2015.06.012. g j 25. Rose T, Nam CS, Chen KB. Immersion of virtual reality for rehabilitation— review (in eng). Appl Ergon. 2018;69:153–61. https://doi.org/10.1016/j. apergo.2018.01.009. g j 25. Rose T, Nam CS, Chen KB. Immersion of virtual reality for rehabilitation— review (in eng). Appl Ergon. 2018;69:153–61. https://doi.org/10.1016/j. apergo.2018.01.009. g 26. Tieri G, Morone G, Paolucci S, Iosa M. Virtual reality in cognitive and motor rehabilitation: facts, fiction and fallacies (in eng). Expert Rev Med Devices. 2018. https://doi.org/10.1080/17434440.2018.1425613. 26. Tieri G, Morone G, Paolucci S, Iosa M. Virtual reality in cognitive and motor rehabilitation: facts, fiction and fallacies (in eng). Expert Rev Med Devices. 2018. https://doi.org/10.1080/17434440.2018.1425613. p g 27. Meyer C, et al. Familiarization with treadmill walking: How much is enough? (in eng). Sci Rep. 2019;9(1):5232. https://doi.org/10.1038/ s41598-019-41721-0. 27. Meyer C, et al. Familiarization with treadmill walking: How much is enough? (in eng). Sci Rep. 2019;9(1):5232. https://doi.org/10.1038/ s41598-019-41721-0. Publisher’s Note 16. O’Connor SM, Donelan JM. Fast visual prediction and slow optimization of preferred walking speed. J Neurophysiol. 2012;107(9):2549–59. https:// doi.org/10.1152/jn.00866.2011. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. 17. Powell WA, Hand S, Stevens B, Simmonds M. Optic flow in a virtual environment: sustained influence on speed of locomotion. Cyberpsychol Behav. 2006;9(6):710–710. 18. Salinas MM, Wilken JM, Dingwell JB. How humans use visual optic flow to regulate stepping during walking (in eng). Gait Posture. 2017;57:15–20. https://doi.org/10.1016/j.gaitpost.2017.05.002. p g j g p 19. !!! INVALID CITATION !!! [19–21]. 19. !!! INVALID CITATION !!! [19–21]. Guzik A, Drużbicki M, Perenc L, Wolan-Nieroda A, Turolla A, Kiper P. Establishing the minimal clinically important differences for sagittal hip range of motion in chronic stroke patients (in eng). Front Neurol. 2021;12:700190. https://doi.org/10.3389/fneur.2021.700190. 31. Guzik A, Drużbicki M, Perenc L, Wolan-Nieroda A, Turolla A, Kiper P. Establishing the minimal clinically important differences for sagittal hip range of motion in chronic stroke patients (in eng). Front Neurol. 2021;12:700190. https://doi.org/10.3389/fneur.2021.700190.
https://openalex.org/W4287545551	https://zenodo.org/records/6478689/files/5.%20Estrategias.pdf	es		Estrategias que utilizan los docentes para el desarrollo del pensamiento lógico-matemático en los estudiantes universitarios	Zenodo (CERN European Organization for Nuclear Research)	2,020	cc-by	5,225	ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES ESTRATEGIAS QUE UTILIZAN LOS DOCENTES PARA EL DESARROLLO del pensamiento lógico-matemático en los estudiantes universitarios Strategies used by teachers for the development of logical-mathematical thinking in university students Eduin A. Arriaga Valderrama1 RECIBIDO ACEPTADO PUBLICADO eduinarriaga@gmail.com. ORCID: https://orcid.org/0000-0002-7068-8395 [04/10/2020] [13/11/2020] [30/12/2020] Pág. 52-64 Edwin J. Arriaga Rincón2 ejosue_22@hotmail.com. ORCID: https://orcid.org/0000-0001-6457-7851 Saray M. Arriaga Rincón3 smaria_95@hotmail.com. ORCID: https://orcid.org/0000-0003-2718-6047 1 Docente de Matemática del programa de Ingeniería y Tecnología de la Unermb. 2 Docente del MPPE. 3 Licenciada en Administración con mención en Aduana. RESUMEN tuvo como objetivo formular estrategias para el desarrollo del pensamiento lógico-matemático en los estudiantes de Ingeniería y Tecnología de la Unermb. La investigación fue de tipo descriptiva, con enfoque cuantitativo y diseño no experimental. La población estuvo constituida por 1130 estudiantes del programa Ingeniería y Tecnología (PIT) de la Unermb, sede Puertos de Altagracia, y la muestra estuvo El pensamiento lógico permite a los estudiantes solucionar problemas, tanto a nivel educativo como laboral y social, por lo cual, resulta esencial para el desarrollo personal. En ese sentido, los docentes se han enfocado en ejecutar estrategias que permitan desarrollar este tipo de pensamiento. Sobre la base de ello, esta investigación 52 Vol. 2, N.º 1, septiembre-diciembre 2020 https://bit.ly/38T7Bdo Palabras clave Estrategias, Pensamiento Lógico-Matemático, Docentes, Estudiantes, Pedagogía Keywords ABSTRACT Strategies, Logical-mathematical thinking, Teachers, Students, Pedagogy Logical thinking allows students to solve problems, both at the educational, work and social levels, and is therefore essential for personal development. In that sense, teachers have focused on implementing strategies that allow them to develop this type of thinking. Based on this, this research had the objective of formulating strategies for the development of logical-mathematical Introducción En el proceso de enseñanza-aprendizaje es común que los docentes busquen métodos y estrategias adecuadas para dar sentido a las clases y construir 53 ARTÍCULOS DE INVESTIGACIÓN thinking in students of Engineering and Technology at Unermb. The research was of a descriptive type, with a quantitative approach and non-experimental design. The population was constituted by 1130 students of the Engineering and Technology program (PIT) of Unermb, headquarters of Puertos de Altagracia, and the sample was conformed by 137 students that conform the project of Gas Engineering. Regarding the technique and instruments of data collection, the survey and the questionnaire were used, respectively, with response categories in the Likert scale. It was evidenced that the students present apathy, little sense of participation and transformation to the DPLM, due to the scarce strategies applied by the teachers, where the teaching methods are mechanistic and traditional, this from the fact that the teachers are not involved in the human part of the students, because they are dedicated to dictate and develop exercises without taking into account the student. Finally, it was concluded that the alternative solution is to establish teaching strategies for DPLM in students, in order to train professionals, through a constructivist pedagogical current supported by liberating and critical pedagogy. conformada por 137 estudiantes que conformen el proyecto de Ingeniería de Gas. Respecto a la técnica e instrumentos de recolección de datos, se utilizó la encuesta y el cuestionario, respectivamente, con categorías de respuesta en la escala de Likert. Se evidenció que los estudiantes presentan apatía, poco sentido de participación y transformación al DPLM, debido a las escasas estrategias aplicadas por los docentes, donde los métodos de enseñanza son mecanicistas y tradicionales, esto a partir de que los docentes se involucran poco en la parte humana de los educandos, pues se dedican a dictar y desarrollar ejercicios sin tomar en cuenta al estudiante. Por último, se concluyó que la alternativa de solución es establecer estrategias docentes para el DPLM en los estudiantes, con el fin de formar profesionales, mediante una corriente pedagógica constructivista apoyada en la pedagogía liberadora y crítica. ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES soluciones significativas para los estudiantes universitarios, con el fin de mejorar sus habilidades y destrezas en el pensamiento lógico-matemático y, con ello, brindar herramientas que faciliten la solución de problemas. la formación integral del ser humano (Velásquez et al. 2008). Según Fernández (2003), el pensamiento lógico hace referencia al desarrollo sensomotriz que se realiza, fundamentalmente, a través de los sentidos. En este marco, el conjunto de experiencias individuales realizado de forma consciente, con base en la percepción sensorial, en relación con los demás y los objetos del entorno, transmiten a la mente, hechos sobre los cuales se elabora una serie de ideas, que sirven para relacionarse con el exterior. Por consiguiente, el desarrollo del pensamiento lógico es un proceso de adquisición de nuevos códigos, que abren las puertas del lenguaje y permiten la comunicación con el entorno, es decir, constituye la base indispensable para el desarrollo de los conocimientos en todas las áreas académicas y es un instrumento a través del cual se asegura la interacción humana. De allí la importancia del progreso de capacidades de pensamiento lógico esenciales para la formación integral del ser humano. Asimismo, el pensamiento lógico propicia la adquisición de nuevos conocimientos, los cuales son codificados o decodificados para una comprensión eficaz, efectiva y eficiente, de modo tal que se desarrollen competencias donde el estudiante consiga su propio estilo de aprendizaje y aprenda a su propio ritmo. En efecto, el basamento consiste en una concepción pedagógica asentada sobre el desarrollo integral del educando universitario y en sus características, intereses y necesidades, así como en una pedagogía orientadora y flexible, que no se convierta en una prescripción de tareas, y que se destaque por fomentar la comunicación y el desarrollo moral en la formación integral de los actores sociales. Entonces, el pensamiento o razonamiento lógico-matemático es importante, ya que permite a los estudiantes elaborar y diseñar una serie de herramientas que faciliten la solución de problemas o situaciones existentes, donde los mismos adquieren habilidades y destrezas para la interacción eficaz con su entorno. Por tanto, se hace necesarioelempleodeestrategiasefectivas, por parte del docente, que incentiven al estudiante a desarrollar las actividades académicas, debido a que se deben transformar los procesos de enseñanza y aprendizaje, a fin de lograr captar la atención de los mismos; en ese sentido, se debe contextualizar los contenidos, con la finalidad de que el estudiante conozca la importancia de aplicar el razonamiento lógico-matemático. El desarrollo del pensamiento lógicomatemático es un proceso de adquisición de nuevos códigos, que abren las puertas del lenguaje y permiten la comunicación con el entorno, por lo cual, constituye la base indispensable para la adquisición de los conocimientos de todas las áreas académicas y, además, es un instrumento a través del cual se asegura la interacción humana. De allí la importancia del desarrollo de competencias del pensamiento lógico-matemático esenciales para Además de lo indicado, existen debilidades en la formación matemática 54 Vol. 2, N.º 1, septiembre-diciembre 2020 https://bit.ly/38T7Bdo A partir de lo expuesto, esta investigación tuvo como objetivo principal formular estrategias para el desarrollo del pensamiento lógico-matemático en los estudiantes de Ingeniería y Tecnología de la Unermb. Cabe destacar que el logro de este objetivo contribuirá a la promoción del aprendizaje exitoso, incrementando las teorías existentes respecto a la enseñanza de la matemática, además de contribuir en la formación de los estudiantes universitarios mediante una corriente pedagógica constructivista, apoyada en la pedagogía liberadora, crítica y práctica, aplicadas en la educación, nueva o activa, lo que permite consolidar las características de cada individuo, potenciando el desarrollo de la nación. Al tomar en cuenta las consideraciones planteadas, se evidencia la necesidad de determinar estrategias que faciliten o conlleven al estudiantado a pensar y reflexionar más respecto a la importancia de la matemática como medio para el desarrollo del pensamiento lógico-matemático, es decir, que los objetos de estudio creen nuevos métodos de aprendizaje. Cabe destacar que la mayoría de los docentes explican ejercicios matemáticos sin darle a los estudiantes, las herramientas que les permitan comprender la realidad de los mismos; de igual forma, se ha observado la deserción de estos por la poca creatividad e innovación de los docentes en su praxis educativa. Estrategias docentes El concepto de estrategia se ha incorporado recientemente a la psicología del aprendizaje y la educación, como una forma de resaltar el carácter procedimental que tiene todo aprendizaje. Al respecto, Solé (2006) indica que las estrategias son procedimientos que implican la planificación de acciones. Por su parte, Goodman (1996) afirma que una estrategia “es un amplio esquema para obtener, evaluar y utilizar información” (p. 62). Por otro lado, Nerici, citado por Parra (2002), considera que la estrategia es un procedimiento lógico, estructurado a nivel psicológico, formulado con el objetivo de dirigir el aprendizaje de los estudiantes, a fin de alcanzar los objetivos de la enseñanza. En ese marco, para fines de esta investigación, se revisó el estudio de Villavicencio (2018), quien señala que la normalización de estrategias, respecto a los tres momentos de enseñanza (inicio, ejecución y final), conlleva a que estas se efectúen con éxito, potenciando y enriqueciendo las capacidades del estudiante. Adicionalmente, se tomó en cuenta la investigación de Herrada y Baños (2018), donde se resalta que las estrategias enfocadas en el aprendizaje cooperativo estimulan la adquisición de competencias, especialmente en la enseñanza de matemáticas, permitiendo, a su vez, el desarrollo individual, social y laboral del estudiante. Por otra parte, es preciso resaltar que el concepto de estrategia se relaciona con la psicología cognitiva y 55 ARTÍCULOS DE INVESTIGACIÓN de los estudiantes, tales como la carencia de estrategias innovadoras que incentiven a los mismos a potenciar habilidades y destrezas en la solución de problemas matemáticos, en los cuales se profundiza su proceso lógico de razonar y desarrollar con mayor rapidez, situaciones presentadas durante su vida cotidiana. ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES la teoría pedagógica constructivista, ya que la estrategia representa un instrumento esencial para potenciar la capacidad organizadora y activa del estudiante, quien, a partir de este elemento pedagógico, puede adquirir capacidades y potenciar sus habilidades (Morales et al., 2018). su propio aprendizaje, ya que este va a construir conocimientos a partir de sus experiencias. Para ello, el docente debe generar un ambiente dinámico y proactivo que conduzca al estudiante a crear, participar y buscar nuevas alternativas de solución a los diferentes problemas que se le presenten. En este orden de ideas, los docentes asumen estrategias que sirven para generar en sus estudiantes el deseo de aprender, de allí que sea necesario plantear estrategias docentes de enseñanza y aprendizaje. En esta línea, Rajadell (2002) señala que la estrategia docente es una “actuación secuenciada potencialmente consciente del profesional en educación, guiada por uno o más principios de la didáctica, y encaminada hacia la optimización del proceso de enseñanza-aprendizaje” (p. 65). A partir de esta definición, se observa que el concepto de estrategia docente acoge una doble vertiente, aunque íntimamente complementaria: estrategias de enseñanza y estrategias de aprendizaje. Díaz y Hernández (2006) señalan que el nuevo paradigma exige al docente la utilización de estrategias que le permitan al estudiante utilizar más su cerebro y equilibrar las facultades de los hemisferios cerebrales (izquierdo para la lógica, derecho para la creatividad), de tal forma que se conjugue lo emocional con lo cognitivo, lo radical con lo creativo, así como lo lógico con lo imaginativo; de esta forma, se podrá acelerar los procesos de aprendizaje, ya que la activación de las múltiples inteligencias implica pensar, sentir y hacer, lo cual debe ser estimulado en la acción docente, a través de estrategias innovadoras y transformadoras. Adicionalmente, Gómez (1993) indica que todo proceso instructivo se determina a partir de la estrategia pedagógica establecida, la cual deriva de una reflexión didáctica respecto a qué, cómo o por qué enseñar. Esta reflexión debe concretarse a partir de los objetos de conocimiento, como las habilidades, capacidades y competencias a adquirir, así como el método, técnicas y recursos docentes más adecuados para la comprensión, desarrollo y adquisición de conocimientos. Asimismo, la estrategia docente debe contar con un mecanismo de feedback, que permita controlar su ejecución y corregir los fallos o debilidades observados. En el mismo orden de ideas, las estrategias de enseñanza son los métodos, técnicas, procedimientos y recursos que se planifican, de acuerdo con las necesidades de la población a la cual va dirigida y que tiene por objeto hacer más efectivo el proceso de enseñanza-aprendizaje. Cabe destacar que, en la actualidad, las estrategias deben fomentarse mediante un enfoque constructivista, capaz de transformar al estudiante en un sujeto activo y protagonista de Cabe resaltar que, para el logro de los objetivos pedagógicos, en cuanto al aprendizaje lógico-matemático, el docente puede tomar en cuenta elementos como la motivación, 56 Vol. 2, N.º 1, septiembre-diciembre 2020 https://bit.ly/38T7Bdo Proporciona una organización adecuada de los datos del material (conexiones internas) que se ha de aprender al representarla en forma gráfica o escrita, mejorando su significatividad lógica y, por ende, el aprendizaje significativo. Este grupo de estrategias se pueden emplear en los distintos momentos de la enseñanza. Entre las principales estrategias que destacan en esta organización se encuentran las representaciones viso-espaciales, como los mapas o redes semánticas, y las representaciones lingüísticas, como los resúmenes o cuadros sinópticos. Por su parte, Díaz y Hernández (2006) afirman que los procesos cognitivos, para promover el aprendizaje, pueden organizarse con base en los siguientes tipos de estrategias: a. Estrategias para activar o crear conocimientos previos y para establecer expectativas adecuadas en los estudiantes d. Estrategias para promover el enlace entre los conocimientos previos y la nueva información que se ha de aprender, es decir, las conexiones externas La activación del conocimiento previo cumple una doble función, ya que, por un lado, permite conocer lo que saben sus estudiantes y, por otro, permite utilizar dicho conocimiento como base para promover nuevos aprendizajes. Cabe resaltar que este proceso de integración asegura una mayor significatividad de los aprendizajes logrados y se recomienda usarlas antes o durante la instrucción. Este tipo de estrategias se encuentran constituidas por aquellas de inspiración ausbeliana, como son los organizadores previos (comparativos y expositivos) y las analogías. b. Estrategias para orientar la atención de los estudiantes Sirven para focalizar y mantener la atención de los aprendices durante un discurso o texto. Estas estrategias se deben emplear de manera continua durante el desarrollo de la clase, instruyendo a los estudiantes sobre qué puntos, conceptos o ideas deben centrar sus procesos de atención, codificación y aprendizaje. Los mecanismos que deben incluirse en este tipo de estrategias son las preguntas insertadas, el uso de pistas o claves para explorar la estructura del discurso y el uso de ilustraciones. Entonces, es de suma importancia que los docentes rompan con la tradicional y comprobada ineficacia de la forma de enseñanza que, desde una perspectiva conductual, se maneja hasta la actualidad, y que conozcan y empleen las mencionadas estrategias de enseñanza que, concebidas desde el modelo cognitivo, facilitan la adquisición de aprendizajes significativos y el desarrollo 57 ARTÍCULOS DE INVESTIGACIÓN c. Estrategias para organizar la información que se ha de aprender intereses reales de los estudiantes, ambiente motivante, adecuado proceso enseñanza-aprendizaje, posibilidad del educando para modificar o reforzar su comportamiento, utilizar los recursos naturales del medioambiente, adecuados a la realidad de las situaciones de aprendizaje, entre otros (Parra, 2014). ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES de la capacidad de los estudiantes a través del desarrollo de las inteligencias múltiples. Al respecto, las estrategias de enseñanza se definen como procedimientos o recursos usados por el agente de enseñanza, con la finalidad de promover el aprendizaje en los estudiantes, es decir, estas estrategias permiten que el docente amplíe y facilite el aprendizaje significativo en el alumnado. Las estrategias de enseñanza se clasifican en preinstruccionales, coinstruccionales y posinstruccionales (Good y Brophy, 2000). que hayan adquirido, definiendo las intenciones que pretende lograr al término de las actividades. Al respecto, Díaz y Hernández (2006) explican que, por la utilidad de este tipo de estrategias, los agentes educadores tienen la posibilidad de plantear los objetos que se ha propuesto que obtengan los alumnos, esto a partir de la explicación de lo que van a realizar en clase y deben aprender. Consecuentemente, esta metodología permite a los estudiantes, prepararse previamente para entender los aspectos de trabajo; así, por sí mismo, podrá establecer su nivel de rendimiento. Estrategias preinstruccionales Algunas de las estrategias preinstruccionales son los organizadores previos, ilustraciones, así como plantear los objetivos que se pretenden para la clase, la unidad o el proyecto, además de realizar resúmenes, con los cuales el docente puede elaborar y darse cuenta acerca de cuánto saben los estudiantes acerca de determinado contenido, ya que, si a este le interesa, surgirán otras inquietudes. Se utilizan en el momento de iniciar las actividades, y tienen como propósito estimular a los estudiantes, con el propósito de captar su atención e interés. Para alcanzar este objetivo, el agente de enseñanza desarrolla actividades que resulten atractivas y despierten la curiosidad. Este tipo de estrategias preparan y alertan al estudiante respecto a qué y cómo va a instruirse (adquirir conocimientos y experiencias previas adecuadas), a fin de que esto le permita situarse en un contexto de aprendizaje pertinente. Entonces, la activación de conocimientos previos, ayuda al docente en dos aspectos fundamentales: conocer lo que conocen y saben los estudiantes, y utilizar estos conocimientos como base para promover nuevos conocimientos y experiencias (Díaz y Hernández, 2006). Estrategias coinstruccionales Las estrategias coinstruccionales son aquellas que potencian el contenido curricular durante el proceso de enseñanza, ya que permiten detectar la información esencial de determinado texto, conceptualizar esta nueva información, delimitar la estructura e interrelación de este material de aprendizaje, y mantener la atención y motivación del estudiante. En este marco, con estas estrategias se procura que el estudiante se mantenga motivado a aprender y adquirir nuevos conocimientos, mediante el proceso de enseñanza y el hacer, fundamentalmente en el nivel básico, momento en el cual debe enfatizarse en la realización de actividades durante la clase (Díaz y Hernández, 2006). Las estrategias pre-instruccionales cumplen con la función de activar (o generar) conocimientos previos de los estudiantes para generar lo que aún no existe en su marco de aprendizaje. Además, sirven para esclarecer ideas 58 Vol. 2, N.º 1, septiembre-diciembre 2020 https://bit.ly/38T7Bdo Se presentan después del contenido que se ha de aprender y permiten al estudiante formar una visión sintética, integradora y crítica del contenido enseñado, posibilitando valorar su propio aprendizaje. Durante la aplicación de este tipo de estrategias, el docente puede utilizar preguntas intercaladas, resúmenes, mapas conceptuales y redes semánticas. Además, es propicio brindar al estudiante la información que las estrategias posinstruccionales van a permitir concluir respecto al tema o actividad realizada, a fin de estimularlo a cerrar el tema. Cabe resaltar que esta acción facilita la oportunidad del docente para realizar retroalimentación con los estudiantes, con el objetivo de verificar si se logró a cabalidad el proceso de aprendizaje, siendo este el verdadero sentido de la actividad docente. Por lo expuesto, se puede determinar que los docentes, mediante la utilización de estrategias coinstruccionales, tienen la posibilidad de utilizar señalizaciones, mapas mentales, analogías, entre otros, a fin de estimular al estudiante durante la clase y convertir el proceso de aprendizaje en un momento didáctico. Esto último se afirma a raíz de que el docente, junto a los alumnos, analiza y descompone la información, lo que conduce a la obtención del conocimiento del tema tratado en clase y conlleva al logro del aprendizaje significativo. Entonces, las estrategias posinstruccionales promueven la interrelación de los conocimientos previos con la nueva información que se va a aprender; sin embargo, el uso de estos nuevos conocimientos depende de las características del proceso de aprendizaje planteado por el agente de enseñanza, como tareas, actividades didácticas efectuadas, entre otros. No obstante, es necesario resaltar que, durante la aplicación de estrategias coinstruccionales, el agente de enseñanza no debe desligarse ni del inicio ni del cierre de la actividad. Por ello, se recomienda al docente ser paciente, amable, comprensivo, utilizar un tono de voz adecuado, hacer la actividad dinámica, considerar el uso de recursos audiovisuales, entre otros aspectos relevantes que propicien un ambiente de aprendizaje significativo para los estudiantes. Estrategias motivacionales para la enseñanza de la matemática El educador debe acudir a estrategias motivacionales, que permitan al estudiante incrementar sus potencialidades, ayudarlo a incentivar su deseo de aprender y enfrentarlo a situaciones en las que tenga que utilizar su capacidad de discernimiento, para conseguir solucionar determinadas 59 ARTÍCULOS DE INVESTIGACIÓN Estrategias posinstruccionales Por su parte, Shuell (1996) afirma lo anterior y expone que las estrategias coinstruccionales permiten que el estudiante mejore su capacidad de atención y, con base en ello, detecte la información principal, la decodifique y la conceptualice, además de estructurarla e interrelacionarla con otros conocimientos, es decir, este tipo de estrategias se relaciona con el logro del proceso de aprendizaje, con base en la comprensión por parte del estudiante. A partir de ello, esta categoría estratégica integra estrategias como mapas conceptuales, analogías, entre otras. ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES de la Unermb, sede Puertos de Altagracia, mientras que, para la selección de la muestra, se aplicó la fórmula de Sierra (1994), con un margen de 8 %, obteniendo un total de 137 estudiantes pertenecientes al proyecto de Ingeniería de Gas. cuestiones. En ese marco, esta investigación define las estrategias motivacionales como las técnicas y recursos que debe utilizar el docente, para hacer más efectivo el aprendizaje de la matemática, manteniendo las expectativas del estudiante. Por otro lado, respecto a las técnicas de recolección de datos, se aplicó la encuesta, y el instrumento utilizado fue el cuestionario, el cual fue diseñado mediante categorías de respuesta a escala de Likert. Estos datos fueron tabulados a través de herramientas infoestadísticas, y mediante métodos estadísticos se probaron las hipótesis establecidas en el desarrollo de la investigación, además de establecer con exactitud, patrones de comportamiento en el proceso de enseñanza-aprendizaje. Desde este punto de vista, es importante que el docente haga una revisión de las prácticas pedagógicas que emplea en el aula de clase y reflexione sobre la manera cómo, hasta ahora, ha impartido conocimientos a sus estudiantes, para que, de esta manera, pueda conducir su enseñanza con técnicas y recursos adecuados, que le permitan al educando construir de manera significativa el conocimiento y alcanzar el aprendizaje de una forma efectiva. En este sentido, Chiavenato (1999) señala que la motivación se relaciona con el impulso individual, generado a partir de un estímulo externo (ambiente) o interno (proceso mental) para actuar de determinada forma o conducir a un comportamiento específico. Por ello, la motivación, como estrategia didáctica, ayuda al estudiante a valorar el aprendizaje, ya que puede evidenciar cómo el docente, a través de la aplicación de estrategias motivadoras, logran el aprendizaje efectivo del alumno. Para el análisis de los datos obtenidos, se utilizó la estadística descriptiva, la cual consiste en el cálculo de las distribuciones de frecuencias relativas (fr), orientadas a la consecución de los resultados que darán solución al planteamiento del problema. El mismo se calculó a partir de las puntuaciones una vez obtenidas en la muestra. A partir de lo indicado, se estimaron medidas de tendencia central, con el objetivo de obtener el puntaje que represente las estrategias y pensamiento lógico-matemático en cada una de las dimensiones analizadas. Los resultados recogidos se consignaron en tablas, para describir el comportamiento de cada uno de los indicadores con un baremo preestablecido para cada una de las dimensiones intervinientes en el estudio. METODOLOGÍA Esta investigación, de tipo descriptivo, se realizó con base en el paradigma positivista, por lo cual, el enfoque fue de tipo cuantitativo, mientras que, el diseño de investigación fue no experimental, ya que no se manipularon los datos primarios obtenidos. Respecto a la población, esta estuvo constituida por 1130 estudiantes del programa Ingeniería y Tecnología (PIT) 60 Vol. 2, N.º 1, septiembre-diciembre 2020 https://bit.ly/38T7Bdo Figura 1. ¿Los docentes desarrollan estrategias motivacionales que contribuyan al proceso de aprendizaje? 9% 14 % 24 % 25 % Nunca Casi nunca Casi siempre Siempre Como se observa en la Figura 1, el 28 % de los encuestados respondió que, a veces, los docentes desarrollan estrategias que contribuyen a promover un ambiente de confianza y autoconfianza, estimulando así a los estudiantes a aplicar técnicas y métodos relacionados con su nivel de aprendizaje y, con ello, adaptarlos a sus conocimientos. El 25 % indicó que siempre los docentes realizan lo indicado; un 24 % expresó que casi siempre; el 14 %, casi nunca, y solo el 9 % resaltó que nunca evidencian la aplicación de 28 % A veces estrategia educativas. Con base en los resultados obtenidos, se determinó que los docentes, medianamente, deciden promover estrategias didácticas y motivacionales para desarrollar el pensamiento lógico-matemático. Análisis de la varianza Con el propósito de proponer las estrategias para el desarrollo del pensamiento lógico-matemático, se aplicó el análisis de la varianza (Anova de un factor). Tabla 1. Análisis de varianza Suma de cuadrados gl Media cuadrática F Sig Intergrupos 2,504,673 11 227,698 9,100,739 0,000 Intragrupos 40,832 1632 2.50E-02 2,545,505 1643 Total Como se evidencia en la Tabla 1, se obtuvo un nivel de significación observado (Sig.) de 0,000, asociado a un valor del estadístico F = 9100,739, lo cual indica que se rechaza la hipótesis nula, es decir, todos los componentes asociados a los ítems de la dimensión estrategias no son iguales, ya que el Sig. es menor 61 ARTÍCULOS DE INVESTIGACIÓN RESULTADOS ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES que 0,05. En este sentido, de acuerdo con Villavicencio (2018), una estrategia planificada, es decir, que incluya los tres momentos de enseñanza, potencia las capacidades de los alumnos. Con base en ello, el docente debe realizar sus clases tomando en cuenta los tres momentos de enseñanza: inicio, desarrollo y cierre, lo cual permitirá promover la comprensión, el análisis y aplicación efectiva de los procesos numéricos. • El docente estimula el aprendizaje dentro y fuera del aula. • El docente promueve en los estudiantes el entusiasmo sobre el aprendizaje. • El profesor realiza actividades que permitan al estudiante crear conciencia. • El profesor estimula actividades para humanizar la conducta en los estudiantes. • El docente anticipa el logro del aprendizaje, haciéndole ver al estudiante la facilidad del aprendizaje del pensamiento lógico-matemático. En efecto, el docente debe hacerle saber al estudiante, tanto los logros adquiridos como sus debilidades, a fin de que pueda mejorarlas con la ayuda de su docente tutor y compañeros de clase, como forma de aprendizaje cooperativo, esto haciendo conocimiento del propósito de aprender y aplicar diversos ejercicios lógicos matemáticos con su profesión o vida cotidiana. Respecto a esto último, Herrada y Baños (2018) indican que el aprendizaje cooperativo promueve la adquisición de competencias, además de optimizar el rendimiento académico de los estudiantes, por lo cual, resulta un método de enseñanza-aprendizaje adecuado para el desarrollo del pensamiento lógico-matemático. • El docente enseña el propósito de realizar diversos ejercicios. • El profesor realiza las dinámicas de reflexión al inicio, desarrollo o al cierre de la clase. • El profesor motiva, al final de la clase, a que el estudiante diga si la entendió. • El profesor termina la clase felicitando a los estudiantes por haber entendido la clase. CONCLUSIONES A partir de los resultados obtenidos, se concluyó que la mayoría de educandos demuestra apatía por los contenidos desarrollados en asignaturas de orden numérico, debido a la complejidad que estas implican en la resolución de problemas, los cuales se ausentan de la cotidianidad, es decir, que son pocos contextualizados. De igual forma, se develó que la gran mayoría de los estudiantes de ingeniería de la Unermb, núcleo de los Puertos de Altagracia, dispone escasamente de las habilidades y destrezas numéricas necesarias para la comprensión de los contenidos propios de la carrera de Ingeniería. Por lo tanto, se interpreta que existen diferencias significativas, que repercuten en el aprendizaje significativo, entre los siguientes ítems: • El docente emplea actividades que promuevan un ambiente armónico dentro del aula. • Se promueve la confianza estudiante/ docente dentro del aula. • Las condiciones ambientales son agradables para incentivar el estímulo a los estudios. 62 Vol. 2, N.º 1, septiembre-diciembre 2020 https://bit.ly/38T7Bdo espacios de aprendizajes o ponerse a hacer otra cosa en sus cuadernos. Adicionalmente, al momento de determinar las estrategias que utiliza el docente en el desarrollo del pensamiento lógico-matemático, se determinó que estas son escasas, dado a que pocas veces emplean procedimientos que incentiven a los estudiantes a la construcción de propuestas para desarrollar su propio pensamiento, siendo a su vez crítico y reflexivo sobre el cómo resolver problemas relacionados con su área objeto de estudio y, al mismo tiempo, trasladarlos a la realidad con fenómenos estudiados en áreas numéricas. REFERENCIAS Chiavenato, L. (1999). Teorías generales de la administración educativa. Mc Graw Hill. Escritos-Una-Perspectiva-TransaccionalSociopsicolinguistica Díaz, F y Hernández G, (2006). Estrategias docentes para un aprendizaje significativo. Una interpretación constructivista. McGraw-Hill. Herrada, R. y Baños, R. (2018). Experiencias de aprendizaje cooperativo en matemáticas. Espiral Cuadernos del Profesorado, 11(23), 99-108. https://core.ac.uk/download/pdf/161848308.pdf Fernández, N. (2003). Capacitación en el uso de herramientas básicas para el desarrollo de Educación Continua. RIED Revista Iberoamericana de Educación a Distancia, 6(1), 107-121. DOI: https://doi.org/10.5944/ried.6.1.1104 Morales, L., García, O., Torres, A. y Lebrija, A. (2018). Habilidades cognitivas a través de la estrategia de aprendizaje cooperativo y perfeccionamiento epistemológico en matemática de estudiantes de primer año de universidad. Formación Universitaria, 11(2), 45-56. https:// scielo.conicyt.cl/scielo.php?pid=S0718-500620 18000200045&script=sci_arttext&tlng=e Gómez,M.(1993).Sociologíadeldisciplinamiento escolar. Centro Editor de América Latina. Good, L. y Brophy, K. (2000). Psicología Educativa Contemporánea. McGraw-Hill. Nérici, G. (2000). Hacia una didáctica general dinámica. Kapelusz SA. Goodman, K (1996). La lectura, la escritura y los textos escritos una perspectiva transaccional sociopsicolingüística. Textos en contexto, 2, 78-89. https://es.scribd.com/ document/418443065/Goodman-Kenneth1996-La-Lectura-La-Escritura-y-Los-Textos- Parra, K. (2014). El docente y el uso de la medicación en los procesos de enseñanza y aprendizaje. Revista de Investigación, 38(83), 155-180. https://www.redalyc.org/ pdf/3761/376140398009.pdf 63 ARTÍCULOS DE INVESTIGACIÓN Seguidamente, al indagar las estrategias motivacionales que utiliza el docente, en el desarrollo del pensamiento lógico-matemático, se encontró que estas carecen de aplicación, debido a que los docentes, generalmente, mantienen una relación armónica con sus estudiantes, es decir, pocas veces se involucran en la parte humana de estos, por lo que solamente se dedican a dictar clases y desarrollar ejercicios, sin tomar en cuenta si los estudiantes prestan atención o captan cada la explicación dada. Con base en ello, los estudiantes mantienen una conducta inadecuada en el aula, como salir del aula, conversar con otros compañeros, entrar y salir repetidas veces de los ARTÍCULOS DE INVESTIGACIÓN REVISTA DE CIENCIAS SOCIALES Rajadell, N. (2002). La importancia de las estrategias docentes para la resolución de conflictos en el aula. Revista Electrónica Interuniversitaria de Formación del Profesorado, 5(3). https://dialnet.unirioja.es/ servlet/articulo?codigo=1034539 Rivas, F. (2008). Un acercamiento al estado del conocimiento del campo profesores de educación básica y normal en el estado de Chihuahua [tesis de maestría, Universidad Pedagógica Nacional Hidalgo de Parral]. Repositorio Universidad Pedagógica Nacional Hidalgo de Parral. Shuell, T.J. (1996). Teaching and learning in a classroom context. In D. C. Berliner y R. C. Calfee (Eds.), Handbook of Educational Psychology (pp. 726-764). Simon & Schuster Macmillan. Sierra, R. (1994). Técnicas de Investigación Social. Teoría y ejercicios. Paraninfo S.A. Solé, I. (2006). Estrategias de lectura. Aique. Velásquez, J., Flórez, G., Ruiz, F. y Tamayo, O. (2008). Modelización de procesos de enseñanza en profesores de ciencias de la ciudad de Manizales (Colombia) desde el concepto contenido pedagógico del conocimiento. Enseñanza de las Ciencias. https://ddd. uab.cat/pub/edlc/edlc_a2009nEXTRA/edlc_ a2009nExtrap2714.pdf Villavicencio, J. (2018). Normalización de la estrategia de enseñanza-aprendizaje Giras de Campo, para planificación y ejecución de sus tres momentos de ejecución: el antes, durante y después. Revista Torreón Universitario, 7(19), 62-73. DOI: https://doi. org/10.5377/torreon.v7i19.7911 64
https://openalex.org/W2413841588	https://www.nature.com/articles/srep16590.pdf	English	null	Catestatin attenuates endoplasmic reticulum induced cell apoptosis by activation type 2 muscarinic acetylcholine receptor in cardiac ischemia/reperfusion	Scientific reports	2,015	cc-by	7,806	Catestatin attenuates endoplasmic reticulum induced cell apoptosis by activation type 2 muscarinic acetylcholine receptor in cardiac ischemia/reperfusion ived: 05 June 2015 d: 16 October 2015 16 November 2015 OPEN received: 05 June 2015 accepted: 16 October 2015 Published: 16 November 2015 Feng Liao1,, Yang Zheng1,, Junyan Cai1, Jinghui Fan1, Jing Wang1, Jichun Yang1,2, Qinghua Cui1,2, Guoheng Xu1, Chaoshu Tang1 & Bin Geng1,2 Catestatin (CST) is a catecholamine secretion inhibiting peptide as non-competitive inhibitor of nicotinic acetylcholine receptor. CST play a protective role in cardiac ischemia/reperfusion (I/R) but the molecular mechanism remains unclear. Cardiomyocytes endogenously produced CST and its expression was reduced after I/R. CST pretreatment decreased apoptosis especially endoplasmic reticulum (ER) stress response during I/R. The protection of CST was confirmed in H9c2 cardiomyoblasts under Anoxia/reoxygenation (A/R). In contrast, siRNA-mediated knockdown of CST exaggerated ER stress induced apoptosis. The protective effects of CST were blocked by extracellular signal-regulated kinases 1/2 (ERK1/2) inhibitor PD90895 and phosphoinositide 3-kinase (PI3 K) inhibitor wortmannin. CST also increased ERK1/2 and protein kinase B (Akt) phosphorylation and which was blocked by atropine and selective type 2 muscarinic acetylcholine (M2) receptor, but not type 1 muscarinic acetylcholine (M1) receptor antagonist. Receptor binding assay revealed that CST competitively bound to the M2 receptor with a 50% inhibitory concentration of 25.7 nM. Accordingly, CST inhibited cellular cAMP stimulated by isoproterenol or forskolin, and which was blocked by selective M2 receptor antagonist. Our findings revealed that CST binds to M2 receptor, then activates ERK1/2 and PI3 K/Akt pathway to inhibit ER stress-induced cell apoptosis resulting in attenuation cardiac I/R injury. Catestatin (CST) is a 21 amino acid-residue, hydrophobic neuroendocrine peptide derived from chro- mogranin A (ChgA)1. It is co-stored in the secretory granule and co-released with catecholamine in adrenal chromaffin cells and adrenergic neurons as an endogenous noncompetitive antagonist of nico- tine acetylcholine (nAch) receptor and inhibition catecholamine secretion in mammals2. CST stimulated histamine release from mast cells3,4. CST also regulated rostral ventrolateral medulla neuron activity and caused reduction sympathetic barosensitivity and parapheral chemoreflex5,6. Clearly, CST has a signifi- cant anti-hypertensive effect7. ypf CST is also expressed and generated in the heart8. CST reduces isoproterenol and endothelin-1– induced cardiac contractility9,10 through the PI3 K-Akt-endothelial nitric oxide synthase (eNOS) path- way11. Many clinical trials have revealed an association between plasma CST level and cardiac diseases including acute myocardial infarction, heart failure and cardiac remodeling12–15. These studies suggest 1Department of Physiology and Pathophysiology, School of Basic Medical Science, Peking University, P.R. China. 2Center for Noncoding RNA Medicine, Peking University Health Science Center Beijing 100191, China. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports There are amendments to this paper received: 05 June 2015 accepted: 16 October 2015 Published: 16 November 2015 Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 Results CST CST pretreatment improved left ventricular (LV) function and decreased I/R injury. Ischemia reperfusion procedure and CST pretreatment (CP) (CST supplementation for 15 min before I/R) were performed (Fig. 1a). After 30 min of ischemia and 1 h of reperfusion, cardiac endogenous ChgA mRNA (see supplementary Fig. S1 online) and protein expression were no changes, but CST fragments (21 kD and 7.5 kD) reduced comparison to control (Fig. 1b). Different dose of CST perfusion for 1 h did not induce lactate dehydrogenase (LDH) leakage (supplementary Fig. S2 online). CST (25 nM, 50 nM and 100 nM) pretreatment decreased LDH leakage and cardiac troponin I (cTNI) level in perfusate at dif- ferent reperfusion time point (supplementary Fig. S3 online), which indicated that CST pretreatment lowered the ischemia reperfusion (I/R) injury, especially CST at 100 nM exhibited well protection. So we selected 100 nM CST in following experiments. As Fig. 1c showed, I/R caused global cardiac infarc- tion, and CP significantly reduced infarct size as compared with I/R alone (9.8 ± 4.2% vs 27.5 ± 8.2%, P = 0.0001). We also assessed LV diastolic pressure (LVEDP), developed LV pressure (LV systolic pressure minus LVEDP, dLVP) and maximal rate of LV pressure development (LV ± dp/dtmax) for LV function. The I/R rat heart showed greatly increased LVEDP and impaired dilation; CP lowered the LVEDP during each time point of reperfusion (All P < 0.05, Fig. 1d). CP also improved the dLVP recovery (All P < 0.05, Fig. 1e) and LV ± dp/dtmax (All P < 0.05 Fig. 1f). Thus, CST pretreating improved cardiac function during I/R. CST pretreatment inhibited I/R–induced cell apoptosis. We used cleaved caspase-3 (a marker of cell apoptosis) antibody immunofluorescence staining to assess cell apoptosis. I/R increased the number of apoptotic cardiomyocytes, and CP significantly reduced the cell apoptosis (Fig. 2a,b). Consistently, I/R increased apoptosis markers (caspase-9, 7 and -3 cleavage, and poly (ADP-ribose) polymerase (PARP) cleaved into 89- and 24-kDa segments). CP lowered cleaved caspase-9, -7, -3 and PARP (Fig. 2c,e) for a role in reduced apoptosis. p p I/R induced ER-stress response by assessing phosphorylated PERK and chaperone protein Grp78 expression. ER stress-initiated apoptosis signals evaluated by caspase-12 cleavage, Chop protein expres- sion and phosphorylated JNK. CP also lowered the protein expression of ER stress markers and ER stress-initiated apoptosis signals in cardiac I/R (Fig. Catestatin attenuates endoplasmic reticulum induced cell apoptosis by activation type 2 muscarinic acetylcholine receptor in cardiac ischemia/reperfusion ived: 05 June 2015 d: 16 October 2015 16 November 2015 OPEN These authors contributed equally to this work. Correspondence and requests for materials should be addressed to B.G. (email: bingeng@hsc.pku.edu.cn) 1 Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 1 www.nature.com/scientificreports/ that CST may play an essential role in the pathogenesis of ischemic heart diseases. More recently, Penna et al. found that CST post-conditioning or post-treatment decreased infarct area and diastolic left ven- tricular (LV) pressure, and increased cardiomyocyte viability during cardiac ischemic-reperfusion (I/R)16. PI3 K and protein kinase Cε inhibitors, mitochondrial KATP channel blocker and reactive oxygen spe- cies (ROS) scavenger abolished the protective effects of CST in cardiac I/R17. Clearly, to investigate the protective mechanism of endogenous CST in cardiac diseases will be very useful and necessary for its potential clinical use. p It had been reported that activation of nAch (α 7nAchR)18 or type 2 muscarinic Ach receptor (M2 receptor)19 reduced cardiac I/R injury. Because CST is a noncompetitive inhibitor of nAch receptor1, it is likely that nAch receptor is not involved in protective effect of CST. Interestingly, Gi/o protein partici- pated in the cardiac negative inotropism of CST10,20. Gi/o protein is also an important intracellular target of M2 and M4 receptors21. Based on these previous findings, it is reasonably to speculate that M receptor may play essential roles in CST’s protective effects during cardiac I/R injury. y p y pf g j y In the present study, we presented novel evidence that CST directly bound to and activated M2 recep- tor to attenuate ER-stress linked apoptosis via activation of ERK1/2 and Akt pathways in cardiac I/R. Results CST 2d,e).i p p g g To confirm the CST regulatory role in ER stress-induced apoptosis, we used siRNA knockdown of the CST precursor ChgA in H9c2 cardiomyoblasts. Firstly, we measured the cell viability by CCK8 kit and found that different CST treatment or knockdown did not affect H9c2 cell viability (see supplementary Fig. S4a,S4b online). siRNA treatment for 48 h lowered CST precursor-ChgA protein expression in H9c2 cells (Fig. 3a). In H9c2 cell with A/R, siRNA mediated knockdown of ChgA increased the number of apoptotic cells as compared with scramble siRNA (Fig. 3b). ChgA siRNA also upregulated ER stress response markers (phosphorylated PERK and JNK, Grp78 and Chop expression, and cleaved caspase-12) and an apoptosis marker (cleaved caspase-3) (Fig. 3c). Calcium overload is an important injury mech- anism in I/R and an inducer of ER stress. Thapsigargin (calcium pump inhibitor induce intracellular calcium overload) is a typical ER stress inducer. Knockdown of CST also increased thapsigargin-induced cell apoptosis (Fig. 3d) and especially levels of the ER stress markers and cell apoptosis marker (Fig. 3e and supplementary Fig. S5 online). Above results suggest that endogenous CST may reduce ER stress in part with I/R injury. ERK and PI3 K pathway is involved in the protective effect of CST. To investigate the possible signals of CST protection, we treated H9c2 cells with specific inhibitors for 30 min to assess the blocking effects. Intriguingly, PD98059 and wortmannin reversed the CST protection on cell apoptosis (Fig. 4a,b) and ER-stress response (Fig. 4c,d). Knockdown CST also lowered the phosphorylation of ERK1/2 and Akt during A/R (see supplementary Fig. S6 online). These results indicate that ERK1/2 and PI3 K pathway involved in the protection of CST. In isolated heart, we also found CP significantly increased ERK1/2, Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 2 www.nature.com/scientificreports/ ntificreports/ Figure 1. Catestatin (CST) pretreatment (CP) reduced infarct size and maintained cardiac function after cardiac ischemia-reperfusion (I/R) injury in rats. (a) The procedure of global cardiac I/R (n = 18) and CP (n = 18) and original recording of left ventricular (LV) pressure. (b) The ChgA protein and its cleavage fragments expression were measured by western blot. (c) Heart tissue was stained with triphenyltetrazolium chloride and infarct size (white color) was measured by volume (n = 6 for each group). Results CST Evaluation of cardiac function including LV end diastolic pressure (LVEDP)(b), LV developed pressure (dLVP, LV systolic pressure–diastolic pressure) (d), and LV ±dp/dtmax (e) during I/R and CP (total 36 rats were recruited). Data are mean ± SD. P < 0.05, *P < 0.01 vs I/R group. #P < 0.05 vs parameters before ischemia. Figure 1. Catestatin (CST) pretreatment (CP) reduced infarct size and maintained cardiac function after cardiac ischemia-reperfusion (I/R) injury in rats. (a) The procedure of global cardiac I/R (n = 18) and CP (n = 18) and original recording of left ventricular (LV) pressure. (b) The ChgA protein and its cleavage fragments expression were measured by western blot. (c) Heart tissue was stained with triphenyltetrazolium chloride and infarct size (white color) was measured by volume (n = 6 for each group). Evaluation of cardiac function including LV end diastolic pressure (LVEDP)(b), LV developed pressure (dLVP, LV systolic pressure–diastolic pressure) (d), and LV ±dp/dtmax (e) during I/R and CP (total 36 rats were recruited). Data are mean ± SD. P < 0.05, *P < 0.01 vs I/R group. #P < 0.05 vs parameters before ischemia. igure 1. Catestatin (CST) pretreatment (CP) reduced infarct size and maintained cardiac function after gu e . Catestat (CS ) p et eat e t (C ) educed a ct s e a d a ta ed ca d ac u ct o afte cardiac ischemia-reperfusion (I/R) injury in rats. (a) The procedure of global cardiac I/R (n = 18) and CP (n = 18) and original recording of left ventricular (LV) pressure. (b) The ChgA protein and its cleavage fragments expression were measured by western blot. (c) Heart tissue was stained with triphenyltetrazolium chloride and infarct size (white color) was measured by volume (n = 6 for each group). Evaluation of cardiac function including LV end diastolic pressure (LVEDP)(b), LV developed pressure (dLVP, LV systolic pressure–diastolic pressure) (d), and LV ±dp/dtmax (e) during I/R and CP (total 36 rats were recruited). Data are mean ± SD. P < 0.05, P < 0.01 vs I/R group. #P < 0.05 vs parameters before ischemia. Akt and eNOS phosphorylation (Fig. Results CST 4e,f) compared with I/R, which confirmed ERK1/2 and PI3 K/Akt were partly activated by CST to reduce the ER-stress and increase cell survival.i To confirm these pathways involved in ER stress-induced cellular apoptosis by CST, we measured the effect of these inhibitors on DTT-, tunicamycin- and thapsigargin-induced cellular apoptosis. In situ staining of apoptotic cells with cleaved caspase-3 antibody revealed that CST reduced the number of apoptotic cells induced by DTT, tunicamycin and thapsigargin, and which also were blocked by PD98059 and wortmannin (see supplementary Fig. S7–S9 online). Consistently, the major apoptosis pathway of ER stress—Chop expression, caspase-12 cleavage and JNK phosphorylation, ER stress response markers including phosphorylated PERK, Grp78 protein expression—were also lowered by CP while inducing by DTT (Fig. 5a, and supplementary Fig. S10 online), tunicamycin (Fig. 5b and supplementary Fig. S11 online) and thapsigargin (Fig. 5c and supplementary Fig. S12 online), which were also reversed by two inhibitors. Therefore, CST inhibited apoptotic pathways of the unfolded protein response in part via ERK and PI3 K signaling pathways. Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 3 www.nature.com/scientificreports/ Figure 2. CST pretreatment reduced ER-stress and decreased apoptosis. (a) Cleaved caspase-3 immunofluorescence staining and nuclear staining; apoptotic cells were stained in pink Scale bar = 100 μ M. (b) Apoptosis cell numbers were counted in five different field of vision. (c) Western blot analysis of effect of I/R and CP on apoptosis and caspase family members and PARP and (d) proteins related to the ER stress pathway. Relative protein expression was analyzed by band gray degree. β -actin is a normalization control (e). Six independent experiments were performed for above studies. Figure 2. CST pretreatment reduced ER-stress and decreased apoptosis. (a) Cleaved caspase-3 l Figure 2. CST pretreatment reduced ER stress and decreased apoptosis. (a) Cleaved caspase 3 immunofluorescence staining and nuclear staining; apoptotic cells were stained in pink Scale bar = 100 μ M. (b) Apoptosis cell numbers were counted in five different field of vision. (c) Western blot analysis of effect of I/R and CP on apoptosis and caspase family members and PARP and (d) proteins related to the ER stress pathway. Relative protein expression was analyzed by band gray degree. β -actin is a normalization control (e). Six independent experiments were performed for above studies. CST acts as an M2 receptor agonist. CST might activate Gi/o protein20. Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 Results CST In 5 subtypes (M1 to M5) muscarinic acetylcholine receptors, only M2 and M4 receptors bind with Gi/o protein21. Thus, muscarinic acetylcholine receptor signaling might be involved in the cardioprotective effect of CST. To verify this hypothesis, the nonspecific M receptor antagonist atropine, selective M1 receptor antagonist pirenzepine, and selective M2 receptor antagonist AF-DX116 were used. We found that carbamoylcholine lowered Grp78 and Chop protein expression and increased phosphorylated ERK1/2 and Akt levels effect which was similar to the CST (Fig. 6a). Atropine and AF-DX116 but not pirenzepine blocked the CST-altered level of these proteins (Fig. 6a). Therefore, M2 but not M1 receptor may mediate the CST action. To confirm this finding, we used another selective M2 receptor antagonist-methoctramine, and the selective M2/M4 receptor antagonist himbacine, which also blocked the effect of CST on the ER stress response and ERK1/2 and Akt phosphorylation during A/R (Fig. 6b). In isolated neonatal rat cardiomyocytes, CP also reduced cell apoptosis (PARP and caspase-3 cleavage), ER-stress response (Grp78 and Chop protein expression) by I/R, and increased phosphorylated ERK1/2 and Akt. The protection were also blocked with selective M2 receptor inhibitors (Fig. 6c,d). In isolated heart, AF-DX116 also blocked the protective effects of CP on infarct size, LDH leakage and cTNI level (supplementary Fig. S13a online). These data suggested that M2 receptor may be a target protein interacting with CST. gg p y g p g To investigate the interaction of CST with M2 receptor and as an endogenous ligand, we used radi- oactive isotope [125-I]-labeled CST (0.4 μ M) as a ligand, and N-methylscopolamine (NMS:0.045 nM to 0.3 μ M) as a “cold” ligand were co-incubated with membrane protein (200 μ g) in H9c2 cardiomy- blasts. Radioactivity of binding CST was assessed and the competitive inhibition curve is in Fig. 7a. The half-maximal inhibition (IC50) of binding occurred with 25.7 nM NMS (Fig. 7a). Scatchard analysis revealed a single binding site with KD 18.6 nM (Fig. 7b). To confirm the specific binding of CST to the Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 4 www.nature.com/scientificreports/ Figure 3. siRNA knockdown of CST increased ER stress-induced apoptosis in H9c2 cardiomyoblasts during anoxia/reperfusion (A/R). (a) siRNA knockdown effect by CST precursor (ChgA) protein expression. Immunofluorescence staining by cleaved caspase-3 for cell apoptosis (b) and western blot analysis of ER stress-marker expression (c). Results CST (e) Western blot analyzed the changes of phosphorylated ERK1/2, Akt and eNOS during CP in isolated heart; and expressions analysis (f) h d Th i d d t i t d f it Figure 4. The role of ERK1/2 and PI3 K/Akt pathway in CST’s protective effects during A/R in cardiomyoblasts. H9c2 cardiomyoblasts were pretreated with ERK inhibitor PD98059 (20 μ M), PI3 K/Akt inhibitor wortmannin (10 nM) for 30 min, followed by treatment with CST (100 nM) for 30 min, then cells underwent A/R. (a) Apoptosis was assessed by cleaved caspase-3–positive cells (a) and counted (b). Scale bars = 50 μ M. Western blot analysis of ER stress-markers expression(c), and relative protein expression was analyzed (d). Six independent experiments were performed for above studies. (e) Western blot analyzed the changes of phosphorylated ERK1/2, Akt and eNOS during CP in isolated heart; and expressions analysis (f) was showed. Three independent experiments were done for it. Figure 4. The role of ERK1/2 and PI3 K/Akt pathway in CST’s protective effects during A/R in Figure 4. The role of ERK1/2 and PI3 K/Akt pathway in CST’s protective effects during A/R in cardiomyoblasts. H9c2 cardiomyoblasts were pretreated with ERK inhibitor PD98059 (20 μ M), PI3 K/Akt inhibitor wortmannin (10 nM) for 30 min, followed by treatment with CST (100 nM) for 30 min, then cells underwent A/R. (a) Apoptosis was assessed by cleaved caspase-3–positive cells (a) and counted (b). Scale bars = 50 μ M. Western blot analysis of ER stress-markers expression(c), and relative protein expression was analyzed (d). Six independent experiments were performed for above studies. (e) Western blot analyzed the changes of phosphorylated ERK1/2, Akt and eNOS during CP in isolated heart; and expressions analysis (f) was showed. Three independent experiments were done for it. Figure 4. The role of ERK1/2 and PI3 K/Akt pathway in CST’s protective effects during A/R in cardiomyoblasts. H9c2 cardiomyoblasts were pretreated with ERK inhibitor PD98059 (20 μ M), PI3 K/Akt inhibitor wortmannin (10 nM) for 30 min, followed by treatment with CST (100 nM) for 30 min, then cells underwent A/R. (a) Apoptosis was assessed by cleaved caspase-3–positive cells (a) and counted (b). Scale bars = 50 μ M. Western blot analysis of ER stress-markers expression(c), and relative protein expression was analyzed (d). Six independent experiments were performed for above studies. Results CST Cells with CST siRNA knockdown were treated with thapsigargin; immunohistochemistry of apoptosis and (d) western blot analysis of ER stress-markers expression (e). Five independent experiments were performed for these studies. Scale bar = 100 μ M. Figure 3. siRNA knockdown of CST increased ER stress-induced apoptosis in H9c2 cardiomyoblasts during anoxia/reperfusion (A/R). (a) siRNA knockdown effect by CST precursor (ChgA) protein expression. Immunofluorescence staining by cleaved caspase-3 for cell apoptosis (b) and western blot analysis of ER stress-marker expression (c). Cells with CST siRNA knockdown were treated with thapsigargin; immunohistochemistry of apoptosis and (d) western blot analysis of ER stress-markers expression (e). Five independent experiments were performed for these studies. Scale bar = 100 μ M. M2 receptor, we added AF-DX116 (final concentration: 50 nM, IC50 50 nM for M2 receptors), and the IC50 of binding was significantly increased to 36.0 nM (Fig. 7a); scatchard analysis revealed a KD of 45.8 nM (Fig. 7b). Therefore, CST specifically binds to the M2 receptor. gh pi y p Activation M2 receptor can decrease intracellular cAMP level under catecholamine stimulation. To identify the biological function of CST binding to the M2 receptor, we measured cAMP level with carbamoylcholine as a positive control. CST and carbamoylcholine lowered cAMP level induced by isoproterenol in H9c2 cells (Fig. 7c). AF-DX116 and himbacine did not affect increased cAMP level with isoproterenol, but blocked the effect of CST. To distinguish the β -receptor inhibition, we used the adenylyl-cyclase activator forskolin to directly increase intracellular cAMP level. CST also lowered cAMP Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 5 www.nature.com/scientificreports/ Figure 4. The role of ERK1/2 and PI3 K/Akt pathway in CST’s protective effects during A/R in Figure 4. The role of ERK1/2 and PI3 K/Akt pathway in CST’s protective effects during A/R in cardiomyoblasts. H9c2 cardiomyoblasts were pretreated with ERK inhibitor PD98059 (20 μ M), PI3 K/Akt inhibitor wortmannin (10 nM) for 30 min, followed by treatment with CST (100 nM) for 30 min, then cells underwent A/R. (a) Apoptosis was assessed by cleaved caspase-3–positive cells (a) and counted (b). Scale bars = 50 μ M. Western blot analysis of ER stress-markers expression(c), and relative protein expression was analyzed (d). Six independent experiments were performed for above studies. Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 Results CST (e) Western blot analyzed the changes of phosphorylated ERK1/2, Akt and eNOS during CP in isolated heart; and expressions analysis (f) was showed. Three independent experiments were done for it. Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 6 www.nature.com/scientificreports/ Figure 5. CST protected against ER-stress induced apoptosis via ERK1/2 and PI3 K pathway in H9c2 cardiomyoblast. Western blot analysis of protein expression of ER stress-marker proteins and apoptosis cell number with treatment with stress inducers dithiothreitol (DTT) (a), tunicamycin (b), and thapsigargin (c) after pretreatment with PD98059 or wortmannin. P < 0.01 vs single ER-stress inducers, #P < 0.05 vs single catestatin treatment. Eight independent experiments were performed for above studies. Figure 5. CST protected against ER-stress induced apoptosis via ERK1/2 and PI3 K pathway in H9c2 cardiomyoblast. Western blot analysis of protein expression of ER stress-marker proteins and apoptosis cell number with treatment with stress inducers dithiothreitol (DTT) (a), tunicamycin (b), and thapsigargin (c) after pretreatment with PD98059 or wortmannin. P < 0.01 vs single ER-stress inducers, #P < 0.05 vs single catestatin treatment. Eight independent experiments were performed for above studies. Figure 5. CST protected against ER-stress induced apoptosis via ERK1/2 and PI3 K pathway in H9c2 Figure 5. CST protected against ER-stress induced apoptosis via ERK1/2 and PI3 K pathway in H9c2 cardiomyoblast. Western blot analysis of protein expression of ER stress-marker proteins and apoptosis cell number with treatment with stress inducers dithiothreitol (DTT) (a), tunicamycin (b), and thapsigargin (c) after pretreatment with PD98059 or wortmannin. P < 0.01 vs single ER-stress inducers, #P < 0.05 vs single catestatin treatment. Eight independent experiments were performed for above studies. level, which was blocked by AF-DX116 and himbacine but not pirenzepin (Fig. 7d). Thus, CST functions as an agonist of the M2 receptor. Discussion I h In the present study, we identify the CST is an endogenous ligand of the M2 receptor. CST pretreatment activates M2 receptor, thereby activates PI3 K and ERK1/2 pathway results in reducing ER stress-induced cell apoptosis and increasing cell survival (Fig. 8), thus lowers infarct size and recovers cardiac pump function under I/R.it Firstly we identified the endogenous CST reducing after I/R, but its precursor ChgA expression were no changes. That means increased enzymes (which catalyzed ChgA to generate CST) activity during I/R, although the enzymes were not identified until now. CST pretreatment compensated to endogenous CST decrease, thus lowered the infarct size and improved cardiac function in rats under I/R, which agreed with studies of CST post-conditioning16,17. However, human wild type CST (100 nM) treatment for 2 h during reperfusion phase increased, but Gly364Ser variant peptide lowered infarct size; because of lower inhibitory activity on nicotinic cholinergic receptor22. In our present study or CST post-conditioning, short-times supplementation reduced the inhibitory effects on nicotinic cholinergic receptor, which might partly explain the paradoxical effects of CST on cardiac I/R. g p y p pf Apoptosis occurs in cardiomyocytes during I/R. Genetic or pharmacological inhibition of cell apoptosis reduced infarct size and improved cardiac function23. CST pretreatment similar with CST post-conditioning24, decreased apoptosis during I/R. In contrast, knockdown CST increased apoptosis. 7 www.nature.com/scientificreports/ Figure 6. The role of the muscarinic receptors in CST’s protective effect on apoptosis. (a) Western blot analysis of effect of non-selective M receptor antagonist atropine, selective M1 receptor antagonis pirenzepine and selective M2 receptor antagonist AF-DX116 on ER-stress markers expression and phosphorylation of ERK1/2 and Akt and (b) M2 receptor-specific blocking confirmed with methoctra and himbacine. Quantitative analysis of protein expression was shown in the bottom of figure. P < 0 vs A/R treatment; #P < 0.01 vs A/R + CP group. Six independent experiments were performed for abo studies. In neonatal rat cardiomyocytes, we measured the apoptosis markers-PARP and caspase-3 clea ER-stress markers-Grp78 and Chop expression, ERK1/2 and Akt phosphorylation; and the role of M2 receptor using selective M2 receptor antagonists(c). Relative protein expression was performed (d). T Figure 6. The role of the muscarinic receptors in CST’s protective effect on apoptosis. (a) Western blot analysis of effect of non selective M receptor antagonist atropine selective M1 receptor antagonis Figure 6. The role of the muscarinic receptors in CST’s protective effect on apoptosis. Discussion I h (a) Western blot analysis of effect of non-selective M receptor antagonist atropine, selective M1 receptor antagonist pirenzepine and selective M2 receptor antagonist AF-DX116 on ER-stress markers expression and phosphorylation of ERK1/2 and Akt and (b) M2 receptor-specific blocking confirmed with methoctramine and himbacine. Quantitative analysis of protein expression was shown in the bottom of figure. P < 0.05 vs A/R treatment; #P < 0.01 vs A/R + CP group. Six independent experiments were performed for above studies. In neonatal rat cardiomyocytes, we measured the apoptosis markers-PARP and caspase-3 cleavage, ER-stress markers-Grp78 and Chop expression, ERK1/2 and Akt phosphorylation; and the role of M2 receptor using selective M2 receptor antagonists(c). Relative protein expression was performed (d). Three independent experiments were performed for it. ER stress induces apoptosis involved in regulation during I/R25. CST pretreatment lowered ER stress induced apoptosis during I/R, even directly stimulated by ER-stress inducers; oppositely, knockdown CST aggravated ER-stress induced apoptosis, suggesting endogenous CST stabilized ER-stress response to anti-apoptosis. p p Next, we assessed the possible signal involved in the CST regulation. A PI3 K inhibitor blocked the Ca2+ transient inhibition of CST after β -adrenergic stimulation11 and cardiac function recovery with Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 8 www.nature.com/scientificreports/ Figure 7. CST bound to M2 receptor and reduced cellular cAMP levels. (a) Competitive inhibitory receptor binding curve involved [125-I]-labeled CST (0.4 μ M) as a ligand, N-methylscopolamine (NMS, 0.045 nM to 0.3 μ M) as “cold” ligand, and AF-DX116 as selective inhibitor. (b) Equilibrium dissociation constant (KD) values were counted by scatchard analysis. Selective M-receptor antagonist effect with CST on intracellular cAMP level after stimulation with isoproterenol (c) or forskolin (d). Data are mean ± SD. (a) P < 0.05 vs untreated group; (b) P < 0.01 vs single treatment with isoproterenol (ISO) or drugs; (c) P < 0.01 vs CST. Six independent experiments were performed for these studies. Figure 7. CST bound to M2 receptor and reduced cellular cAMP levels. (a) Competitive inhibitory Figure 7. CST bound to M2 receptor and reduced cellular cAMP levels. (a) Competitive inhibitory receptor binding curve involved [125-I]-labeled CST (0.4 μ M) as a ligand, N-methylscopolamine (NMS, 0.045 nM to 0.3 μ M) as “cold” ligand, and AF-DX116 as selective inhibitor. (b) Equilibrium dissociation constant (KD) values were counted by scatchard analysis. Selective M-receptor antagonist effect with CST on intracellular cAMP level after stimulation with isoproterenol (c) or forskolin (d). Discussion I h Data are mean ± SD. (a) P < 0.05 vs untreated group; (b) P < 0.01 vs single treatment with isoproterenol (ISO) or drugs; (c) P < 0.01 vs CST. Six independent experiments were performed for these studies. Figure 7. CST bound to M2 receptor and reduced cellular cAMP levels. (a) Competitive inhibitory receptor binding curve involved [125-I]-labeled CST (0.4 μ M) as a ligand, N-methylscopolamine (NMS, 0.045 nM to 0.3 μ M) as “cold” ligand, and AF-DX116 as selective inhibitor. (b) Equilibrium dissociation constant (KD) values were counted by scatchard analysis. Selective M-receptor antagonist effect with CST on intracellular cAMP level after stimulation with isoproterenol (c) or forskolin (d). Data are mean ± SD. (a) P < 0.05 vs untreated group; (b) P < 0.01 vs single treatment with isoproterenol (ISO) or drugs; (c) P < 0.01 vs CST. Six independent experiments were performed for these studies. Figure 8. Proposed mechanism of CST’s protective effects in cardiac I/R injury. CST acts as an endogenous M2 receptor agonist, activates M2 receptor and causes inhibition of adenylyl cyclase (AC) activity via the α subunit (α i) of Gi, thereby decreasing cAMP level while stimulating by isoproterenol or forskolin. During cardiac I/R, CST binds to M2 receptor and activates ERK1/2 and PI3 K/Akt signaling pathway to reduce ER-stress response and cell apoptosis. Figure 8. Proposed mechanism of CST’s protective effects in cardiac I/R injury. CST acts as an endogenous M2 receptor agonist, activates M2 receptor and causes inhibition of adenylyl cyclase (AC) activity via the α subunit (α i) of Gi, thereby decreasing cAMP level while stimulating by isoproterenol or forskolin. During cardiac I/R, CST binds to M2 receptor and activates ERK1/2 and PI3 K/Akt signaling pathway to reduce ER-stress response and cell apoptosis. Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 9 www.nature.com/scientificreports/ CST post-conditioning17. We found that wortmannin blocked the effect of CST on ER stress-linked apoptosis; and knockdown CST lowered phosphorylated Akt. Therefore, the PI3 K-Akt pathway is an important signal cascade of CST in cardiac diseases. The canonical mitogen-activated protein kinases (MAPKs) JNK, ERK1/2 and p38 are activated in response to ER stress, each with different roles. JNK is considered a downstream target of IRE1-dependent ER stress26. I/R–induced ROS leads to sustained calmodulin-dependent protein kinase II activation, which promotes JNK signaling27, thereby promoting apoptosis26. Discussion I h Here we found that CST inhibited JNK phosphorylation, and siRNA knockdown of CST acti- vated JNK. Thus, the JNK signal is involved in inhibiting ER stress-induced apoptosis with CST. In gen- eral, ERK1/2 activation is considered to promote cell survival. ER stress-promoted activation of ERK1/2 is associated with increased cell proliferation and decreased apoptosis28. Pharmacological or genetic inhi- bition of ERK1/2 promoted ER stress-induced apoptosis29. We found that PD98059-inhibited MAPK kinase reduced ERK1/2 activation and reversed the protection of CST on ER stress, so CST increase ERK1/2 phosphorylation to promote cell survival. These findings indicate that PI3 K and ERK1/2 path- way involved in the protection of endogenous CST during I/R. y p g g We asked how CST activates PI3 K and ERK1/2 signal and which the receptor of CST is. Here we showed CST was similar to carbamoylcholine, reduced ER-stress response and activated ERK1/2 and PI3 K signals. This effect was blocked by a selective M2 but not M1 receptor antagonist. Receptor bind- ing assay showed that CST is a ligand of the M2 receptor. CST bound to M2 receptor activation directly activating Gi protein, thus reducing cAMP production while stimulated by isoproterenol30 or forskolin; and the effect was also blocked by M2 selective antagonism, suggesting CST functions as an M2 receptor agonist. M2 receptor activation increases the NOS/NO cGMP pathway31, which also explains in part the CST anti-adrenergic effect by NOS/NO cGMP11. All these findings demonstrate that endogenous CST as an agonist of M2 receptor to participate in pathogenesis of the cardiac I/R. g p p p p g Muscarinic receptor expression is abundant in ventricular muscle. Acetylcholine has protective role in cardiac I/R in vitro, but it is non-selective ligand and activates all cholinergic receptor subtype, which limited its application due to enormous side effects. Here we identified that CST is a selective ligand of M2 receptor with high affinity and its intracellular signal (cAMP, PI3 K and ERK1/2) is consistent with M2 receptor agonist. Berberine also has potential binding ability to M2 receptor (KD values 5.4 μ M) and has protective effects in cardiac diseases32. These data indicate that M2 receptor might be a new target for cardiac I/R, but it should be confirmed in cardiomyocytes specific M2 receptor knockout mice in the future. Different amino acid variant in CST affected CST inhibitory activity on nicotinic cholinergic receptor33. Discussion I h Thus, we also increasing CST binding to M2 receptor by modification replacement amino acid, if the sites and model of CST binding to M2 receptor are clear. A synthesized small-molecule chem- ical according to the pharmacophore model of CST lowered blood pressure in hypertensive animals34. According to the model, more selective and effective M2 agonist might be designed and synthesized for treating cardiac ischemic diseases. Methods Animals and materials. All animal procedures complied with the Animal Management Rule of the Ministry of Health, People’s Republic of China (document No. 55, 2001) and the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, updated 2011). The care and use of the laboratory animals were approved by Laboratory Animal Ethics Committee of Peking University. Total 36 male Sprague–Dawley (SD) rats (180–200 g, 8 week) were sup- plied by the Animal Center, Health Science Center, Peking University. H9c2 rat cardiomyoblasts were pur- chased from the Stem Cell Bank, Chinese Academy of Sciences. The human wild-type CST, hChgA352–372 (SSMKLSFRARGYGFRGPGPQL), was purchased from Phoenix Pharmaceuticals (Belmont, CA, USA). The extracellular signal-regulated kinase (ERK) inhibitor PD98059, PI3 K/Akt inhibitor wortmannin, selective M1 receptor antagonist pirenzepine, selective M2 receptor antagonists AF-DX116 (Otenzepad) and methoctramine, selective M2/M4 receptor antagonist himbacine, and non-specific M receptor antag- onist atropine, carbamoylcholine, tunicamycin, dithiothreitol (DTT), and thapsigargin were from Sigma- Aldrich (St. Louis, MO, USA). ChgA siRNA (sense: GAU GAU GAU GGU CAG UCG GdTdT, antisense: CCG ACU GAC CAU CAU CAU CdTdT) and scramble siRNA were synthesized by Sigma. Langendorff perfusion procedure and experimental protocol. Rats fed a standard diet were heparinized (200 IU) and anesthetized with sodium pentobarbital (50 mg/kg, intraperitoneally). Hearts were excised and rapidly moved to a Langendorff perfusion system as described35. After 30-min stabiliza- tion, hearts underwent 30-min normoxia, zero-flow global ischemia for 30 min, followed by 60-min rep- erfusion (I/R group). For the CP group, CST (100 nM) was infused for 15 min before ischemia (Fig. 1A). Cardiac function including left ventricular end diastolic pressure (LVEDP), LV systolic pressure (LVSP), and maximal and minimal LV pressure development (± dp/dtmax) were recorded. p p p max At the end of reperfusion, hearts were frozen at − 20 °C, then cut into 2-mm thin slices, which were stained with 1% triphenyltetrazolium chloride (TTC) in phosphate buffer (pH 7.4) at 37 °C for 10 min, then fixed in 4% paraformaldehyde solution to enhance the contrast of the stain. Images in each slice were analyzed by image pro plus software. The volume of infarcted tissue (white) and the risk zone (total) was then calculated by each area plus slice thickness and summing the results. The infarct size was expressed as the percentage of the risk zone36. Total 12 rats were sacrificed for the infarct size assay. Methods Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 10 www.nature.com/scientificreports/ Other rats, while finished I/R procedure, the apex of heart (about 100 mg) was quickly removed for western blot assay. The residual heart were cut about 5 mm high (from the site of apex), embedded with OCT then the frozen slice were prepared for apoptosis staining (Methods of western blot and apoptosis assay are available in the Supplementary methods online). Anoxia/reoxygenation (A/R) and induced ER stress in H9c2 and neonatal cardiomyo- cytes. Ten two-day-old Sprague–Dawley rats were anesthetized with sodium pentobarbital (50 mg/kg, intraperitoneally), and then decapitated by heavy scissors. Their ventricles were excised and dissociated using trypsin digestion as described previously37. The isolated cells were pre-plated for 0.5 h in minimal essential medium with 10% fetal calf serum. Nonattached cells were re-plated in another plate in the same medium with 0.1 mM bromodeoxyuridine (BrDU) for 4 days and then were removed. The neonatal cardiomyocytes were used for experiments after a further 2 or 3 days. y y pt y H9c2 rat cardiomyoblasts or neonatal cardiomyocytes were cultured in DMEM with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 °C, 5% CO2. Medium was changed to fresh medium (serum-free), and cells were cultured under anoxia with 95% N2 and 5% CO2 for 12 h at 37 °C, then incubated in a standard incubator for 4 h (A/R)38. Cells were pretreated with PD98059 (20 μ M), wortmannin (10 nM), atropine (10 nM), pirenzepine (1 μ M), AF-DX116 (100 nM), methoctramine (10 nM), or himbacine (1 μ M) for 30 min, then CST (100 nM) or carbamoylcholine (100 μ M) for 30 min, then underwent A/R or were treated with the ER stress inducers tunicamycin (5 μ g/mL, 8 h), dithiothreitol (5 mM, 4 h), or thapsigargin (5 μ M, 8 h).l y g g g Cells at 40% to 50% confluence were transfected with siRNA or scramble siRNA (50 pM) by use of Lipofectamine 2000 (Invitrogen), then incubated at 37 °C in a CO2 incubator for 48 h before A/R. Competitive receptor binding assay. Membranes were prepared from H9c2 cells as described39. Membrane protein was resuspended in HEPES buffer (20 mM HEPES, 1 mM MgCl2, pH 7.4). An amount of 200 μ g membrane protein was incubated in 1 mL HEPES buffer containing 0.4 μ M [125I]-labeled CST (200000 cpm) for 60 min at 37 °C. References 1. Mahata, S. K. et al. Novel autocrine feedback control of catecholamine release. A discrete chromogranin a fragment is a noncompetitive nicotinic cholinergic antagonist. J Clin Invest 100, 1623–1633 (1997). p g g 2. Mahapatra, N. R. Catestatin is a novel endogenous peptide that regulates cardiac function and blood pressure. Cardiovasc Re 80, 330–338 (2008). 3. Kruger, P. G., Mahata, S. K. & Helle, K. B. Catestatin (CgA344-364) stimulates rat mast cell release of histamine in a manner comparable to mastoparan and other cationic charged neuropeptides. Regul Pept 114, 29–35 (2003). 4. Kennedy, B. P., Mahata, S. K., O’Connor, D. T. & Ziegler, M. G. Mechanism of cardiovascular actions of the chromogranin A fragment catestatin in vivo. Peptides 19, 1241–1248 (1998). fragment catestatin in vivo. Peptides 19, 1241–1248 (1998). g p ( ) 5. Gaede, A. H. & Pilowsky, P. M. Catestatin, a chromogranin A-derived peptide, is sympathoinhibitory and attenuates sympathetic 5. Gaede, A. H. & Pilowsky, P. M. Catestatin, a chromogranin A 5. Gaede, A. H. & Pilowsky, P. M. Catestatin, a chromogranin A-derived peptide, is sympathoinhibitory and attenuate barosensitivity and the chemoreflex in rat CVLM. Am J Physiol Regul Integr Comp Physiol 302, R365–372 (2012).f , y, , g p p , y p y y p barosensitivity and the chemoreflex in rat CVLM. Am J Physiol Regul Integr Comp Physiol 302, R365–372 (2012).f yl y g g p y 6. Gaede, A. H., Inglott, M. A., Farnham, M. M. & Pilowsky, P. M. Catestatin has an unexpected effect on the intrathecal actions of PACAP dramatically reducing blood pressure. Am J Physiol Regul Integr Comp Physiol 303, R719–726 (2012). y g p y g g p y 7. Biswas, N. et al. Novel peptide isomer strategy for stable inhibition of catecholamine release: application to hypertension. Hypertension 60, 1552–1559 (2012). yp 8. Biswas, N., Curello, E., O’Connor, D. T. & Mahata, S. K. Chromogranin/secretogranin proteins in murine heart: myocardia production of chromogranin A fragment catestatin (Chga(364-384)). Cell Tissue Res 342, 353–361 (2010). 9. Mazza, R. et al. Catestatin (chromogranin A344-364) is a novel cardiosuppressive agent: inhibition of isoproterenol and endothelin signaling in the frog heart. Am J Physiol Heart Circ Physiol 295, H113–122 (2008).h g g g y y ( ) 10. Angelone, T. et al. The antihypertensive chromogranin a peptide catestatin acts as a novel endocrine/paracrine modulator of cardiac inotropism and lusitropism. Endocrinology 149, 4780–4793 (2008). gy 1. Methods [125I]-labeled CST dissociation was initiated by the addition of N-methylscopolamine (NMS, 0.045 nM to 0.3 μ M) with and without AF-DX116 (50 nM). A control with nothing added was included in each experiment. Binding was terminated by rapid vacuum filtration onto Whatman GF/B filters. In all experiments, non-specific binding was determined with 1 μ M NMS. Filter-bound radioligand was measured by use of a scintillation counter (PerkinElmer, MA, USA). Data were analyzed by non-linear regression with Prism. Statistical analysis. Statistical analysis was performed by SPSS Statistics version 15.0 for windows. Data are presented as mean ± SD. Differences among groups were analyzed by one-way ANOVA, then Student-Newman-Keuls test. Differences between 2 groups were analyzed by two-tail nonparametric t test. Acknowledgementsh g Financial disclosure: This work was supported by the Major State Basic Research Development Program of the People’s Republic of China (nos 2012CB517806 and 2012CB517504) and the National Natural Science Foundation of the People’s Republic China (nos. 81170235, 81470552, 91339106 and 81422006). g inancial disclosure: This work was supported by the Major State Basic Research Development Program f the People’s Republic of China (nos 2012CB517806 and 2012CB517504) and the National Natura www.nature.com/scientificreports/ doi: 10.1007/978-3-642-23274-9_13 1. Yamamoto, S., Kawana, S., Miyamoto, A., Ohshika, H. & Namiki, A. Propofol-induced depression of cultured rat ventricula myocytes is related to the M2-acetylcholine receptor-NO-cGMP signaling pathway. Anesthesiology 91, 1712–1719 (1999). 32. Salehi, S. & Filtz, T. M. Berberine possesses muscarinic agonist-like properties in cultured rodent cardiomyocytes. Pharmaco 63, 335–340 (2011).h 33. Mahata, S. K. et al. The catecholamine release-inhibitory “catestatin” fragment of chromogranin a: naturally occurring human variants with different potencies for multiple chromaffin cell nicotinic cholinergic responses. Mol Pharmacol 66, 1180–1191 (2004). ( ) 34. Tsigelny, I. F., Kouznetsova, V. L., Biswas, N., Mahata, S. K. & O’Connor, D. T. Development of a pharmacophore model for the catecholamine release-inhibitory peptide catestatin: virtual screening and functional testing identify novel small molecule therapeutics of hypertension. Bioorg Med Chem 21, 5855–5869 (2013).f 35. Geng, B. et al. H2S generated by heart in rat and its effects on cardiac function. Biochem Biophys Res Commun 313, 362–368 (2004). 36. Minners, J. et al. Dinitrophenol, cyclosporin A, and trimetazidine modulate preconditioning in the isolated rat heart: suppo a mitochondrial role in cardioprotection. Cardiovasc Res 47, 68–73 (2000).f p 7. Simpson, P. & Savion, S. Differentiation of rat myocytes in single cell cultures with and without proliferating nonmyocardial cells Cross-striations, ultrastructure, and chronotropic response to isoproterenol. Circ Res 50, 101–116 (1982).fl 38. Liu, S. X. et al. Upregulation of heme oxygenase-1 expression by hydroxysafflor yellow A conferring protection from anoxia/ reoxygenation-induced apoptosis in H9c2 cardiomyocytes. Int J Cardiol 160, 95–101 (2012).f yg p p y y 9. Jakubik, J., Randakova, A., El-Fakahany, E. E. & Dolezal, V. Divergence of allosteric effects of rapacuronium on binding and function of muscarinic receptors. BMC Pharmacol 9, 15 (2009). Author Contributions F.L. and Y.Z. performed experiments, analyzed data and wrote the manuscript. J.C., J.F., J.W., Q.C. and J.Y. analyzed data and wrote the manuscript. G.X., C.T. and B.G. contributed to the study design, interpretation of data and reviewed and edited the manuscript. www.nature.com/scientificreports/ www.nature.com/scientificreports/ 7. Perrelli, M. G. et al. Catestatin reduces myocardial ischaemia/reperfusion injury: involvement of PI3 K/Akt, PKCs, mitochondria KATP channels and ROS signalling. Pflugers Arch 465, 1031–1040 (2013).l g g fl g 8. Xiong, J. et al. Postconditioning with alpha7nAChR agonist attenuates systemic inflammatory response to myocardial ischemia— reperfusion injury in rats. Inflammation 35, 1357–1364 (2012). reperfusion injury in rats. Inflammation 35, 1357–1364 (2012). p j y fl 9. Miao, Y. et al. Acetylcholine attenuates hypoxia/ reoxygenation-induced mitochondrial and cytosolic ROS formation in H9c2 cells via M2 acetylcholine receptor. Cell Physiol Biochem 31, 189–198 (2013).h y p y 20. Imbrogno, S., Garofalo, F., Cerra, M. C., Mahata, S. K. & Tota, B. The catecholamine release-inhibitory peptide catestatin (chromogranin A344-363) modulates myocardial function in fish. J Exp Biol 213, 3636–3643 (2010). g yi p 21. Haga, T. Molecular properties of muscarinic acetylcholine receptors. Proc Jpn Acad Ser B Phys Biol Sci 89, 226–256 (2013) l d d ff ll l f h h f d h l 22. Brar, B. K. et al. Human catestatin peptides differentially regulate infarct size in the ischemic-reperfused rat heart. Regul 165, 63–70 (2010).i 3. Whelan, R. S., Kaplinskiy, V. & Kitsis, R. N. Cell death in the pathogenesis of heart disease: mechanisms and significance. Annu Rev Physiol 72, 19–44 (2010). 24. Penna, C. et al. Catestatin Increases the Expression of Anti-Apoptotic and Pro-Angiogenetic Factors in the Post-Ischemic Hypertrophied Heart of SHR. PLoS One 9, e102536 (2014). yp p ( ) 25. Minamino, T. & Kitakaze, M. ER stress in cardiovascular disease. J Mol Cell Cardiol 48, 1105–1110 (2010). y . Minamino, T. & Kitakaze, M. ER stress in cardiovascular diseas 26. Nishitoh, H. et al. ASK1 is essential for endoplasmic reticulum stress-induced neuronal cell death triggered by expa polyglutamine repeats. Genes Dev 16, 1345–1355 (2002). 27. Timmins, J. M. et al. Calcium/calmodulin-dependent protein kinase II links ER stress with Fas and mitochondrial apoptosis pathways. J Clin Invest 119, 2925–2941 (2009). 8. Nguyen, D. T. et al. Nck-dependent activation of extracellular signal-regulated kinase-1 and regulation of cell survival during endoplasmic reticulum stress. Mol Biol Cell 15, 4248–4260 (2004). 9. Zhang, L. J. et al. Inhibition of MEK blocks GRP78 up-regulation and enhances apoptosis induced by ER stress in gastric cancer cells. Cancer Lett 274, 40–46 (2009).f 30. Harvey, R. D. Muscarinic receptor agonists and antagonists: effects on cardiovascular function. Handb Exp Pharmacol, 299–316 (2012). References Bassino, E. et al. A novel catestatin-induced antiadrenergic mechanism triggered by the endothelial PI3 K-eNOS pathway in the myocardium. Cardiovasc Res 91, 617–624 (2011). y 12. Meng, L. et al. Plasma catestatin level in patients with acute myocardial infarction and its correlation with ventricular remodelling. Postgrad Med J 89, 193–196 (2013). g 13. Liu, L. et al. Plasma levels and diagnostic value of catestatin in patients with heart failure. Peptides 46, 20–25 (2013). Zh f l d h h f l 14. Zhu, D., Wang, F., Yu, H., Mi, L. & Gao, W. Catestatin is useful in detecting patients with stage B heart failure. Biomarkers 16, 691–697 (2011). 15. Wang, X. et al. Dramatic changes in catestatin are associated with hemodynamics in acute myocardial infarction. Biomarkers 16, 372–377 (2011). 16. Penna, C. et al. Catestatin improves post-ischemic left ventricular function and decreases ischemia/reperfusion injury in heart. Cell Mol Neurobiol 30, 1171–1179 (2010). Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 11 Additional Information Supplementary information accompanies this paper at http://www.nature.com/srepihi Competing financial interests: The authors declare no competing financial interests. How to cite this article: Liao, F. et al. Catestatin attenuates endoplasmic reticulum induced cell apoptosis by activation type 2 muscarinic acetylcholine receptor in cardiac ischemia/reperfusion. Sci. Rep. 5, 16590; doi: 10.1038/srep16590 (2015). How to cite this article: Liao, F. et al. Catestatin attenuates endoplasmic reticulum induced cell apoptosis by activation type 2 muscarinic acetylcholine receptor in cardiac ischemia/reperfusion. Sci. Rep. 5, 16590; doi: 10.1038/srep16590 (2015). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Com- mons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports \| 5:16590 \| DOI: 10.1038/srep16590 12
https://openalex.org/W3212879835	https://zenodo.org/records/5726059/files/BDJ_article_75362.pdf	English	null	The fauna of aquatic invertebrates in the river impacted by wastewaters from the pulp and paper industry (Komi Republic)	Biodiversity data jurnal	2,021	cc-by	8,069	Maria Baturina , Olga Kononova , Elena Fefilova , Olga Loskutova ‡ ‡ ‡ ‡ ‡ Institute of Biology of Komi Science Centre of the Ural Branch of the Russian Academy of Sciences, Syktyvkar, Russia Corresponding author: Maria Baturina (baturina@ib.komisc.ru), Olga Kononova (kon@ib.komisc.ru) Academic editor: Florian Leese Received: 16 Sep 2021 \| Accepted: 13 Nov 2021 \| Published: 17 Nov 2021 Citation: Baturina M, Kononova O, Fefilova E, Loskutova O (2021) The fauna of aquatic invertebrates in the river impacted by wastewaters from the pulp and paper industry (Komi Republic). Biodiversity Data Journal 9: e75362. https://doi.org/10.3897/BDJ.9.e75362 Corresponding author: Maria Baturina (baturina@ib.komisc.ru), Olga Kononova (kon@ib.komisc.ru) Academic editor: Florian Leese Received: 16 Sep 2021 \| Accepted: 13 Nov 2021 \| Published: 17 Nov 2021 Citation: Baturina M, Kononova O, Fefilova E, Loskutova O (2021) The fauna of aquatic invertebrates in the river impacted by wastewaters from the pulp and paper industry (Komi Republic). Biodiversity Data Journal 9: e75362. https://doi.org/10.3897/BDJ.9.e75362 Biodiversity Data Journal 9: e75362 doi: 10.3897/BDJ.9.e75362 Data Paper Biodiversity Data Journal 9: e75362 doi: 10.3897/BDJ.9.e75362 Data Paper Data Paper © Baturina M et al. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. The fauna of aquatic invertebrates in the river impacted by wastewaters from the pulp and paper industry (Komi Republic) Maria Baturina , Olga Kononova , Elena Fefilova , Olga Loskutova ‡ ‡ ‡ ‡ Background Invertebrates are important elements of aquatic ecosystems and play a crucial role in the transformation of matter and energy in continental water bodies. Communities of aquatic invertebrates are characterised by high sensitivity to pollution by nutrients and toxic substances and acidification of water bodies; they serve as good bioindicators of the quality of the aquatic environment and impacts on hydroecosystems. All hydrobionts participate in the processes of self-purification of water bodies. The presented dataset provides information on the aquatic invertebrate community of a large northern river. During 2018-2020, we collected data on changes in the quantitative indicators of the development of benthic and planktonic communities, as well as the species diversity of their fauna. The dataset combines information about the occurrence and abundance of benthic and planktonic invertebrates and summarises data of aquatic invertebrate species found in the Vychegda River in the zone of influence from the pulp and paper mill. Baturina M et al Baturina M et al 2 Keywords sampling event, benthos and plankton invertebrates, Oligochaeta, Cladocera, Copepoda, Rotifera, Ephemeroptera, Plecoptera, Trichoptera, wastewaters, pulp and paper industry, Vychegda River, Komi Republic New information The presented dataset is part of a monitoring programme of the river ecosystems in the production area of Mondi Syktyvkar JSC (the European North-East of Russia, Komi Republic). The dataset describes the structure of benthic invertebrate and plankton communities in the Northern Dvina River Basin. The data on the finding and abundance of large taxa of aquatic invertebrates and species of some groups: Oligochaeta, Cladocera, Copepoda, Rotifera, Ephemeroptera, Plecoptera and Trichoptera are presented. In total, the resource includes 8720 findings of invertebrates, of which 6041 are for zoobenthos organisms and 2679 for zooplankton organisms. Introduction Freshwater ecosystems are complex, dynamic and diverse. However, they are more vulnerable than marine and terrestrial environments and their biodiversity is declining much faster (Darwall 2018, Schmidt-Kloiber et al. 2019). Currently, the main factor causingglobal environmental problems, such as a reduction in biological diversity, destruction of biotopes, climate change and rapid spread of alien species, is the anthropogenic impact (Alimov et al. 2009). At the same time, there is an inverse correlation between water quality and biodiversity: to maintain a satisfactory water quality, it is necessary to maintain functionally active biodiversity of aquatic ecosystems (Ostroumov 2002). For freshwater environmental monitoring and biodiversity simulation, data on the structure of communities are collected. Pooled species distribution data are now becoming an important (and sometimes primary) source of information in biodiversity research. However, they are often likely to be incomplete. Data gaps for a number of geographic areas and taxonomic groups, including aquatic invertebrates, exist and complicate access to solve the problems associated with global ecological and biogeographic analyses (Balian et al. 2008, Gaiji et al. 2009). Integration of data on findings of taxa is an important objective to biodiversity studies and improving prediction of future environmental changes (Shashkov et al. 2017). It is known (Culp et al. 2000, Baturina and Kononova 2021) that wastewater from pulp and paper mills can negatively affect aquatic organisms and communities, changing their composition and structure, as well as the biology of certain species. The Vychegda (Northern Dvina River Basin) is one of thelargest rivers in the European North-East of Russia; by the Basin area (the catchment size is 121,000 km ), it is the second largest river in the Komi Republic. Taking into account its crucial role in the region, monitoring the 2 The fauna of aquatic invertebrates in the river impacted by wastewaters ... 3 state of the aquatic organisms inhabiting it is very important. After a prolonged large-scale modernisation of the manufacturing facilities of the Mondi Syktyvkar JSC and a reconstruction of the sewage treatment facilities, observations of the state of the river communities acquired the monitoring character – they were carried out at permanent posts from 2018 to 2020 as part of a joint project of the Mondi Syktyvkar JSC and the Institute of Biology of Komi Scientific Center of the Ural Branch of the Russian Academy of Sciences. The presented dataset (Baturina et al. General description Purpose: The main purpose of the work was to prepare a dataset on sample occurrences of zoobenthos and zooplankton invertebrates in the zone impacted by wastewaters from the pulp and paper industry (Komi Republic, Vychegda River). For datasets on sample occurrences, data from 2018 to 2020 were included and, during the period of monitor observations, a basis was prepared for assessing and analysing changes in: (1) quantitative indicators of the development of zoobenthos and zooplankton communities and (2) species diversity of aquatic invertebrate fauna, under the impact of wastewaters from the pulp and paper industry (Komi Republic, Vychegda River). Introduction 2021b) provides information on the current state of diversity and species richness of the Vychegda River in the zone affected by large industrial production and can serve as a basis for water quality monitoring. It is associated with scientific articles (Baturina et al. 2021, Kononova 2021, Baturina et al. 2021a, Kononova 2021) and is an important study on the taxonomic diversity and abundance dynamics of aquatic organisms in the zone of active manufacturing, using the example of a pulp and paper mill. There is an opinion (Fefilova et al. 2021) that the fauna of aquatic invertebrates in the Russian part of the Arctic is poorly studied. The information presented in this work enriches the fundamental knowledge about the composition of zooplankton and zoobenthos in northern rivers. This information is important both for assessing regional resources in order to carry out nature conservation measures and for studying the biogeography of freshwater invertebrates, for a detailed and in-depth analysis of the distribution of certain species or groups and for determining their ranges and ecology. Title: "Russia 2021" Distribution, systematics and spatial organisation of the aquatic invertebrate fauna in a northern river impacted by wastewaters from the pulp and paper industry (Komi Republic, Vychegda River). Assessment of long-term impact of paper mill industry on the biological diversity in the production area. Baturina M et al 4 4 Study area description: All material was collected in the European part of Russia, Komi Republic: Vychegda River Vychegda is the largest right tributary of the Northern Dvina River. It flows through the territory of the Komi Republic and the Arkhangelsk Oblast (). The length of the river is 1,130 km, the catchment area is 121,000 km . Vychegda is a typical lowland river. The average density of the river network in the Vychegda River basin is 0.62 km/km . Vychegda belongs to rivers with incomplete meandering. Its discharge ranges from 162 to 601 m /s. In the river bed of the Vychegda River, the top (346 km long), middle (489 km long) and bottom (296 km long) parts are distinguished. 2 2 3 Our study area is located in the middle reaches of the river. The Middle Vychegda Basin occupies a vast valley; the floodplain is wide, usually bilateral, boggy, with numerous channels. Low-level, sloping shores alternate with high steep ones (3–18 m). The riverbed has a width of 100 to 680 m. There are many shallows and low alluvial sandy islands; the bottom of the river consists of sand, clay and pebbles. The current velocity is from 0.3–0.6 m/s (summer low water) to 1.5–1.8 m/s (high water) (Taskaev 1997). The ecosystem of the Vychegda River is experiencing a strong anthropogenic load as a result of municipal, agricultural and industrial effluents. In the waters of the river, a number of chemical elements and compounds exceed the Maximum Permissible Concentrations (MPC) (Kondratenok 2020). In some sections of the river, there is an increased temperature background (Elsakov and Shchanov 2016), which is the result of the inflow of wastewater from the sewage treatment facilities of the Mondi Syktyvkar JSC and the municipal wastewater of the City of Syktyvkar into the channel. The temperature increase over the background values is observed in the area over a distance up to 15 km from the source(Fig. 1). The research was carried out in zone impacted by wastewaters of Mondi Syktyvkar JSC, in the middle reaches of the Vychegda River. Title: "Russia 2021" On the section of the river (55 km long), seven sampling sites were selected: IB – background zone, above the zone of direct influence of wastewater from the pulp and paper industry (near Sedkyrkeshch Village), I – zone of direct influence of wastewater from the pulp and paper industry, II – 22.8 km lower than point I (near Kochchojyag Village), III – 6.4 km lower than point II (near Sluda Village), IV – zone of direct influence of wastewater from the pulp and paper industry (near Gavrilovka Village), V – 11.8 km lower than point IV (near Sotchem-vyv Village) and VI – 5.5 km lower than point V (near Ust'-Pozheg Village). Throughout the period of our studies, weather conditions changed from year to year. For instance, in 2018 and 2020, the temperature indicators were similar: the average monthly air temperature in July reached +19.5°С and 20.0°С, respectively. In July 2019, the lowest average monthly temperature over the period of our work was registered: +15.4°C (with a deviation from the norm by -2.1°C). The water temperature in the studied section of the river varied in accordance with the air temperature. The minimum water temperature values were registered in July 2019: from 16.3°C to 17.1°C. In warmer 2018 and 2020, the July water temperature ranged from 21.8°C to 25.8°C and from 23.8°C to 24.7°C, respectively. Water levels at the time of sampling in 2018 and 2019 were similar. However, The fauna of aquatic invertebrates in the river impacted by wastewaters ... 5 in 2018, sampling was conducted during the period of an intensive decrease in the water level, and, in 2019 – during the period of an intensive rise in the water level due to significant precipitation. In 2020, during the sampling period, the water level was significantly lower than in the previous two years (Patova et al. 2021). Our studies in 2018-2020 showed improvement of the quality of the river water in places of discharge of wastewater and changes in the environmental situation in connection with the modernisation of the treatment facilities of the Mondi Syktyvkar JSC (in 2014-2019). In 2020, concentrations of some ions in wastewater discharge sites: Cl-, SO42-, PO43-, NH4+, Stat, K, Na, phenol, as well as COD, permanganate value, turbidity and, electrical conductivity increased relative to background points (Patova et al. 2021). Design description: The Dataset (Baturina et al. Title: "Russia 2021" 2021b) provides current data on aquatic invertebrates fauna (zooplankton and zoobenthos) of the Vychegda River (the Northern Dvina Basin) in the zone impacted by wastewaters from the pulp and paper industry (Mondi Syktyvkar JSC). The goal of the project is to prepare a basis for assessing and analysing changes in benthic invertebrate and plankton communities and species diversity of aquatic invertebrate, under the influence of treated wastewater from the pulp and paper production. The collection of hydrobiological samples was carried out mainly during the summer dry season. Standard hydrobiological methods were used. The data mainly concern the description of the species composition of each hydrobiological sample. Data include 23 taxonomic groups of zoobenthos and 152 lower-rank taxa for some groups: Oligochaeta, Mollusca, Cladocera, Copepoda, Coleoptera, Ephemeroptera, Plecoptera, Figure 1. The map of the study area. Figure 1. The map of the study area. Design description: The Dataset (Baturina et al. 2021b) provides current data on aquatic invertebrates fauna (zooplankton and zoobenthos) of the Vychegda River (the Northern Dvina Basin) in the zone impacted by wastewaters from the pulp and paper industry (Mondi Syktyvkar JSC). The goal of the project is to prepare a basis for assessing and analysing changes in benthic invertebrate and plankton communities and species diversity of aquatic invertebrate, under the influence of treated wastewater from the pulp and paper production. The collection of hydrobiological samples was carried out mainly during the summer dry season. Standard hydrobiological methods were used. The data mainly concern the description of the species composition of each hydrobiological sample. Data include 23 taxonomic groups of zoobenthos and 152 lower-rank taxa for some groups: Oligochaeta, Mollusca, Cladocera, Copepoda, Coleoptera, Ephemeroptera, Plecoptera, 6 Baturina M et al 6 Trichoptera and 120 taxa of zooplankton (Rotifera, Cladocera, Copepoda). A total of 8720 findings are included in the resource. Trichoptera and 120 taxa of zooplankton (Rotifera, Cladocera, Copepoda). A total of 8720 findings are included in the resource. Funding: The Ministry of Education and Science of the Russian Federation. Project No АААА-А17-117112850235-2 "Distribution, systematics and spatial organization of fauna and animals population in taiga and tundra landscapes and ecosystems at the Northeast European Russia"; agreement with Mondi Syktyvkar JSC № 45-2018/180405 "Assessment of long-term impact of Mondi Syktyvkar JSC on the biological diversity in the production area". Sampling methods Study extent: (Fig. 2) The data paper is based on one dataset (8720 occurrences). Figure 2. Figure 2. The map of the research area (from Patova et al. 2021). Figure 2. Figure 2. The map of the research area (from Patova et al. 2021). The dataset provides information on the number of individuals of aquatic invertebrates in zoobenthos samples and the number of individuals per cubic metre in zooplankton samples. Hydrobiological samples were collected on the part of the river located in the zone influenced by the pulp and paper industry (Mondi Syktyvkar JSC). Sampling description: The research was carried out in zone impacted by wastewaters of Mondi Syktyvkar JSC, in the middle reaches of the Vychegda River. The material includes data collected during the period of modernisation of the enterprise. The fauna of aquatic invertebrates in the river impacted by wastewaters ... The fauna of aquatic invertebrates in the river impacted by wastewaters ... 7 7 On the section of the river (55 km long), seven points were selected (Fig. 2), which are located at different distances from the wastewater inflow locations of wastewaters discharge of the enterprise (Mondi Syktyvkar JSC). At each point, 2-3 samples of zoobenthos and 2-3 samples of zooplankton were taken, from the right, left bank and middle of the river (if possible). The studies were carried out in July 2018-2020. On the section of the river (55 km long), seven points were selected (Fig. 2), which are located at different distances from the wastewater inflow locations of wastewaters discharge of the enterprise (Mondi Syktyvkar JSC). At each point, 2-3 samples of zoobenthos and 2-3 samples of zooplankton were taken, from the right, left bank and middle of the river (if possible). The studies were carried out in July 2018-2020. Zooplankton samples were collected using plankton nets with subsequent filtration through mesh nylon nets with 82-100 µm mesh size (Morduhaj-Boltovskoj 1975). Quantitative zoobenthic samples were taken from different depths (from 0.5 to 3.5 m). For each point, a general description of the studied river was done and pH, mineralisation and temperature were measured. The identification of organisms was carried out according to the keys to freshwater invertebrates. More than 40 % of taxa (10 of model groups from 24 groups of zooplankton and zoobenthos) were determined up to the species level or genus. Geographic coverage Description: The studied area is located in the middle part of the Vychegda River (Еuropean North-East of Russia, Komi Republic) (Fig. 1). The Vychegda is the largest right tributary of the Northern Dvina River. The study was carried out in the areas impacted by wastewaters from the Mondi Syktyvkar JSC, which is the largest pulp and paper industry enterprise in the European part of Russia. On the section of the river (55 km long), seven points were selected (Fig. 2): IB – background zone, above the zone directly influenced bywastewaters from the pulp and paper industry, I – zone directly influenced by wastewater from the pulp and paper industry, II – 22.8 km lower than point I, III – 6.4 km lower than point II, IV – zone directly influenced by wastewater from the pulp and paper industry, V – 11.8 km lower than point IV, VI – 5.5 km lower than point V. The dataset includes data on aquatic invertebrates from 40 zoobenthos samples and 23 zooplankton samples. Baturina M et al 8 Baturina M et al 8 (Zinchenko 2014) for zoobenthosand a net with mesh nylon nets (82–100 μm) for zooplankton (Morduhaj-Boltovskoj 1975). Samples were preserved in 4% formaldehyde (in the field) and examined under light microscopes in the laboratory. Further identification of aquatic invertebrates was carried out in laboratory conditions. The identification of species of invertebrates was carried out with the preparation of temporary or permanent specimens, under a microscope using keys to identify each taxonomic group. For each sample, the following were described: collection date, locality (with geographic coordinates), device description, sampling depth, water temperature, habitat (with substrate type, aquatic vegetation type, distance to the zone impacted by wastewaters from the Mondi Syktyvkar JSC), collector name, identificator name. For some species, permanent preparations have been made. Design of sampling was based on the regular arrangement. The material includes data collected during the period of modernisation and reduction of emissions of the enterprise. (Zinchenko 2014) for zoobenthosand a net with mesh nylon nets (82–100 μm) for zooplankton (Morduhaj-Boltovskoj 1975). Samples were preserved in 4% formaldehyde (in the field) and examined under light microscopes in the laboratory. Further identification of aquatic invertebrates was carried out in laboratory conditions. The identification of species of invertebrates was carried out with the preparation of temporary or permanent specimens, under a microscope using keys to identify each taxonomic group. For each sample, the following were described: collection date, locality (with geographic coordinates), device description, sampling depth, water temperature, habitat (with substrate type, aquatic vegetation type, distance to the zone impacted by wastewaters from the Mondi Syktyvkar JSC), collector name, identificator name. For some species, permanent preparations have been made. Design of sampling was based on the regular arrangement. The material includes data collected during the period of modernisation and reduction of emissions of the enterprise. Sampling methods Each sample was provided with a description that includes: collection date, locality (with geographic coordinates), device description, depth of sampling, water temperature, habitat (with substrate type, aquatic vegetation type, distance to the zone impacted by wastewaters from the Mondi Syktyvkar JSC), collector name, determined by (identification). Quality control: The data were collected and identified by specialists from the Institute of Biology of Komi Scientific Center of the Ural Branch of the Russian Academy of Sciences. Morphological analysis of specimens were performed using compound microscopes BIMAM R13-1 and Leica DM 4000B. Most of the Rotifera, Cladocera and Copepoda organisms from zooplankton samples and Oligochaeta, Cladocera, Copepoda, Coleoptera, Ephemeroptera, Plecoptera and Trichoptera from benthos samples were identified to the species level. The rest of the organisms from benthos samples were identified only as higher level taxa. For identification of species and higher-level taxa, we used both standards keys and data reported in modern studies specifically addressing the taxonomy of these groups (Rylov 1948, Chekanovskaja 1962, Lepneva 1964, Manujlova 1964, Smirnov 1971, Smirnov 1976, Kutikova 1977, Kutikova and Starobogatov 1977, Dumont and Pensaert 1983, Elliott et al. 1988, Lillehammer 1988, Koste and Shiel 1989, Boruckij et al. 1991, Shiel and Koste 1992, Tsalolikhin 1994, Nogrady et al. 1995, Segers 1995, Tsalolikhin 1995, De Smet 1996, De Smet 1997, Tsalolikhin 1997, Narchuk and Tumanov 2000, Tsalolikhin 2001, Nogrady and Segers 2002, Ueda and Reid 2003, Korovchinsky 2004, Starobogatov et al. 2004, Kotov and Stifter 2006, Teslenko and Zhiltzova 2009, Timm 2009, Alekseev 2010, Chertoprud and Chertoprud 2010, Bekker et al. 2012, Krivosheina 2012, Klimovsky and Kotov 2015, Tsalolikhin 2016, Wilke et al. 2019). Step description: The dataset included our own species list of plankton and benthos fauna of the Vychegda River, based on 63 zooplankton and zoobenthos samples collected in 2018-2020. Samples were taken by standard hydrobiological methods using the Petersen dredge (sampling area 0.025 m ), while those from shallow depths and on rocky bottoms (i.e. gravel) were collected by a hydrobiological scraper (0.09 m ) with mesh size ≤158 μm 2 2 Taxonomic coverage Description: The dataset contains information obtained from sampling for aquatic invertebrates (zooplankton and zoobenthos) (Table 1). Some invertebrates were identified from type to species or taxa (subspecies) of a lower rank. The dataset includes detailed information: 1. taxa (family, genus, species, subspecies) of some (model) groups of zoobenthos: Oligochaeta (1720 occurrences), Mollusca (161), Cladocera (680), Copepoda (680), Coleoptera (160) , Ephemeroptera (1120), Plecoptera (240), Trichoptera (480) and Diptera (360); 2. taxa of the highest rank of zoobenthos groups: Hydrozoa, Nematoda, Hirudinea, Ostracoda, Tardigrada, Hydrachnidia, Araneae, Collembola, Hemiptera, Megaloptera and Odonata (440 occurrences); 3. taxa (species, subspecies) of zooplankton: Rotifera (1638), Cladocera (762) and Copepoda (279). A total of 8720 The fauna of aquatic invertebrates in the river impacted by wastewaters ... The fauna of aquatic invertebrates in the river impacted by wastewaters ... 9 occurrences are included in the resource. In the dataset, full taxonomic affiliation is given for each record, including: Type, Class, Subclass, Order, Family, Subfamily, Genus, Subgenus, Species, lifestage, individuals per sample (for zoobenthos) and individuals per m (for zooplankton). 3 ( p ) Rank Scientific name Common name phylum Cnidaria class Hydrozoa hydras phylum Nematoda nematodes Phylum Rotifera rotifers class Eurotatoria subclass Bdelloidea order Bdelloida subclass Monogononta order Ploima order Flosculariacea phylum Tardigrada tardigrades phylum Annelida class Clitellata subclass Oligichaeta oligochaetes order Tubificida order Lumbriculida order Enchytraeida subclass Hirudinea leeches phylum Mollusca shellfish class Gastropoda phylum Arthropoda class Branchiopoda subclass Diplostraca order Anomopoda order Ctenopoda order Onychopoda order Haplopoda Table 1. List of taxa. 10 Baturina M et al Baturina M et al Rank Scientific name Common name class Hexanauplia subclass Copepoda copepods order Cyclopoida order Harpacticoida class Ostracoda ostracods class Hydrachnidia water mites class Arachnida order Araneae spiders class Collembola springtail class Insecta insects order Hemiptera bedbugs subclass Pterygota order Ephemeroptera mayflies order Plecoptera stoneflies order Megaloptera alderflies order Coleoptera beetles order Trichoptera caddisflies order Odonata dragonflies order Diptera flies Our study showed how wastewater from a pulp and paper mill can affect aquatic invertebrates. For this purpose, we chose six sampling sites in the middle reaches of the Vychegda River, in the zone affected by the treated wastewater of the Mondi Syktyvkar JSC at different distances from the wastewater discharge points. According to the state of invertebrate communities in general, the river waters in the studied area can be classified as "oligotrophic" (in terms of zooplankton) or "satisfactory" (in terms of zoobenthos). Taxonomic coverage The exception is the sites located downstream of the wastewater discharge points that are classified as "eutrophic" (in terms of zooplankton) and the treated wastewater discharge points which are classified as "unsatisfactory" (in terms of zoobenthos). However, given the high indicators of the quantitative development of aquatic invertebrate communities and their high species diversity, we can say that there is no toxic effect of wastewater in the studied area, but there are processes of anthropogenic eutrophication of the river. Analysis of the species composition of model groups – microcrustaceans, oligochaetes, non-chironomid amphibiotic insects – revealed a sufficiently high level of community diversity. The diversity in the groups varied depending on the sampling site, but the differences in values for most of the sites were insignificant. The greatest diversity for all groups was observed at points III and V. The fauna of aquatic invertebrates in the river impacted by wastewaters ... The fauna of aquatic invertebrates in the river impacted by wastewaters ... 11 11 Amongst all the diversity of species of aquatic invertebrates found in the investigated section of the Vychegda River, we noted new species for the regional fauna, Elaphoidella bidens (Schmeil, 1983) and Moina macrocopa (Straus, 1820), as well as a rare species, Brachycercus harisella (Curtis, 1834), listed in the Red Book of the Komi Republic. Amongst all the diversity of species of aquatic invertebrates found in the investigated section of the Vychegda River, we noted new species for the regional fauna, Elaphoidella bidens (Schmeil, 1983) and Moina macrocopa (Straus, 1820), as well as a rare species, Brachycercus harisella (Curtis, 1834), listed in the Red Book of the Komi Republic. Temporal coverage Notes: Data sources provided the dates when the species was detected for the first time in the zone impacted by wastewaters from the pulp and paper enterprise of the 8720 occurrences included in the dataset. The earliest first record dates back to 2018 and the most recent event occurred in 2020. Field studies of zoobenthos were carried out on 23, 24 July 2018, 13-17 July 2019 and 20, 21 July 2020. At each point, from two to five samples were taken. In points I and IV was carried out annually, in item III - in 2018 and 2019, the rest were collected once. Usage licence Usage licence: Creative Commons Public Domain Waiver (CC-Zero) IP rights notes: IP rights notes: This work is licensed under a Creative Commons Attribution (CC-BY) 4.0 License IP rights notes: IP rights notes: This work is licensed under a Creative Commons Attribution (CC-BY) 4.0 License. Data resources Data package title: The fauna of aquatic invertebrates in the river impacted by wastewaters from the pulp and paper industry (Komi Republic) Number of data sets: 1 Download URL: 5a6c4b09-bd03-4a8a-b673-4a5d5430eea2, http://ib.komisc.ru:8088/ ipt/resource?r=cop Data format: Darwin Core Archive format 12 Baturina M et al scription: The dataset includes two tables related by the eventID field – Events and sociated occurrences Associated occurrences Column label Column description eventID (Event Core) An identifier for the event (layer). eventDate (Event Core) The date-time or interval during which an Event occurred. year (Event Core) The year in which the Event occurred, according to the Common Era Calendar. month (Event Core) The month in which the Event occurred. day (Event Core) The day in which the Event occurred. habitat (Event Core) A category or description of the habitat in which the Event occurred. verbatimDepth (Event Core) The original description of the depth below the local surface. eventRemarks (Event Core) Comments or notes about the Event. samplingProtocol (Event Core) The descriptions of the methods used during an Event. sampleSizeValue (Event Core) A numeric value for a measurement of the size of a sample in a sampling event. sampleSizeUnit (Event Core) The unit of measurement of the size (time duration, length, area or volume) of a sample in a sampling event. locationID (Event Core) An identifier for the set of location information. continent (Event Core) The name of the continent in which the Location occurs. country (Event Core) The name of the country in which the Location occurs. countryCode (Event Core) The standard code for the country in which the Location occurs. stateProvince (Event Core) The name of the next smaller administrative region than country (republic) in which the Location occurs. locality (Event Core) The specific description of the place. locationRemarks (Event Core) Comments or notes about the Location. decimalLatitude (Event Core) The geographic latitude. decimalLongitude (Event Core) The geographic longitude geodeticDatum (Event Core) The ellipsoid, geodetic datum or spatial reference system (SRS) upon which the geographic coordinates given in decimalLatitude and decimalLongitude are based. georeferencedBy (Event Core) A list of names of people who determined the georeference (spatial representation) for the Location. The fauna of aquatic invertebrates in the river impacted by wastewaters ... 13 coordinateUncertaintyInMetres (Event Core) The horizontal distance (in metres) from the given decimalLatitude and decimalLongitude describing the smallest circle containing the whole of the Location. occurrenceID (Occurrence Extension) An identifier for the Occurrence. basisOfRecord (Occurrence Extension) The nature of the related resource. Phylum (Occurrence Extension) The full scientific name of the phylum or division in which the taxon is classified. Acknowledgements We are grateful to employees of Institute of Biology (Syktyvkar, Russia) for their help in field material collecting, to Ivan Chadin (Syktyvkar, Russia) for participation in hosting the datasets on GBIF. The research was conducted within the framework of the state assignment topic "Distribution, systematics and spatial organization of fauna and animals population in taiga and tundra landscapes and ecosystems at the Northeastern European Russia" (no. AAAA-A17-117112850235-2), with the financial support of the project "Assessment of long-term impact of Mondi Syktyvkar JSC on the biological diversity in the production area" (Аgreement no. 45-2018/180405). Baturina M et al 14 Baturina M et al A list of publication of literature associated with the Occurrence. Author contributions Maria Baturina took part in fieldwork, processing of the zoobenthic samples, identified benthic invertebrates and Oligochaeta species and wrote the metadata description and manuscript. Olga Kononova took part in processing of the zooplankton samples, identified plankton species (Copepoda, Cladocera, Rotifera), prepared the dataset and wrote the manuscript. Elena Fefilova took part in fieldwork, identified species meiobenthic Crustacea (Cladocera, Cyclopoida, Harpacticoida) prepared the dataset and wrote the manuscript. Olga Loskutova took part in identified species amphibiotic Insecta (Ephemeroptera, Plecoptera, Trichoptera) prepared the dataset. Olga Loskutova took part in identified species amphibiotic Insecta (Ephemeroptera, Plecoptera, Trichoptera) prepared the dataset. Number of data sets: 1 Class (Occurrence Extension) The full scientific name of the class in which the taxon is classified. order (Occurrence Extension) The full scientific name of the order in which the taxon is classified. family (Occurrence Extension) The full scientific name of the family in which the taxon is classified. genus (Occurrence Extension) The full scientific name of the genus in which the taxon is classified. specificEpithet (Occurrence Extension) The name of the first or species epithet of the scientificName. infraspecificEpithet (Occurrence Extension) The name of the lowest or terminal infraspecific epithet of the scientificName, excluding any rank designation. identificationQualifier (Occurrence Extension) A brief phrase or a standard term ("cf.", "aff.") to express the determiner's doubts about the Identification. scientificNameAuthorship (Occurrence Extension) The authorship information for the scientificName formatted according to the conventions of the applicable nomenclaturalCode. scientificName (Occurrence Extension) The full scientific name, with authorship and date information, if known. organismRemarks (Occurrence Extension) Comments or notes about the Organism instance. taxonRank (Occurrence Extension) The taxonomic rank of the most specific name in the scientificName. Lifestage (Occurrence Extension) The age class or life stage of the Organism(s) at the time the Occurrence was recorded. organismQuantityType (Occurrence Extension) The type of quantification system used for the quantity of organisms. individualCount (Occurrence Extension) The number of individuals present at the time of the Occurrence. occurrenceStatus (Occurrence Extension) A statement about the presence or absence of a Taxon at a Location. recordedBy (Occurrence Extension) A list of names of people responsible for recording the original Occurrence. identifiedBy (Occurrence Extension) A list of names of people who assigned the Taxon to the subject. The fauna of aquatic invertebrates in the river impacted by wastewaters ... 15 15 • Baturina M, Kononova O, Fefilova E, Loskutova O (2021b) The fauna of aquatic invertebrates in the river impacted wastewaters of pulp and paper industry (Komi Republic). Institute of Biology of Komi Scientific Centre of the Ural Branch of the Russian Academy of Sciences. GBIF. Sampling event dataset. https://doi.org/ 10.15468/9dpp55 • Baturina MA, Kononova ON (2021) Impact of waste waters from the pulp and paper industry on aquatic zoocenoses: a review of the literature. Contemporary Problems of Ecologу 28 (6). • Baturina MA, Fefilova EB, Loskutova OA (2021) State of benthic communities of Vychegda River under influence of treated wastewater from pulp and paper industry. Contemporary Problems of Ecologу 28 (6). • Bekker E, Kotov AA, Taylor DJ (2012) A revision of the subgenus Eurycercus (Eurycercus) Baird, 1843 emend. nov.(Cladocera: Eurycercidae) in the Holarctic with the description of a new species from Alaska. Zootaxa 3206: 1‑40. https://doi.org/ 10.11646/zootaxa.3206.1.1 Boruckij EV, Stepanova LA, Kos MS (1991) Keys to Calanoida of fresh waters of the USSR. Nauka, Leningrad, 504 pp. [In Russian]. • Boruckij EV, Stepanova LA, Kos MS (1991) Keys to Calanoida of fresh waters of USSR. Nauka, Leningrad, 504 pp. [In Russian]. Chekanovskaja OV (1962) Aquatic oligochaetes of the fauna of the USSR. Nauka, Moscow-Leningrad, 412 pp. [In Russian]. • Chertoprud MV, Chertoprud ES (2010) A short guide to freshwater invertebrates in the center of European Russia. KMK Scientific Press, Moscow, 184 pp. [In Russian]. • Chertoprud MV, Chertoprud ES (2010) A short guide to freshwater invertebrates in the center of European Russia. KMK Scientific Press, Moscow, 184 pp. [In Russian]. Culp JM, Cash KJ, Wrona FJ (2000) Cumulative effects assessment for the Northern • Culp JM, Cash KJ, Wrona FJ (2000) Cumulative effects assessment for the Northern River Basins Study. Journal of Aquatic Ecosystem Stress and Recovery 8: 87‑94. https://doi.org/10.1023/A:1011404209392 • Culp JM, Cash KJ, Wrona FJ (2000) Cumulative effects assessment for the Northern River Basins Study. Journal of Aquatic Ecosystem Stress and Recovery 8: 87‑94. // / / • Darwall W, et al. (2018) The alliance for freshwater life: a global call to unite efforts for freshwater biodiversity science and conservation. Aquatic Conservation: Marine and Freshwater Ecosystems 28: 1015‑1022. https://doi.org/10.1002/aqc.2958 Darwall W, et al. (2018) The alliance for freshwater life: a global call to unite efforts for freshwater biodiversity science and conservation. References • Alekseev VR (Ed.) (2010) Key to identification of zooplankton and zoobenthos of fresh water of European Russia. Zooplankton. 1. KMK Scientific Press, Moscow, 495 pp. [In Russian]. [ISBN 978-5-87317-684-7] • Alekseev VR (Ed.) (2010) Key to identification of zooplankton and zoobenthos of fresh water of European Russia. Zooplankton. 1. KMK Scientific Press, Moscow, 495 pp. [In Russian]. [ISBN 978-5-87317-684-7] • Alimov AF, Buljon VV, Golubkov SM (2009) Resource potential of species, communities and ecosystems of continental reservoirs. Uspekhi Sovremennoj Biologii 129 (6): 516‑527. [In Russian]. • Balian EV, Segers HK, Lévêque C, Martens C (2008) The freshwater animal diversity assessment: an overview of the results. Hydrobiologia 595 (1): 627‑637. https://doi.org/ 10 1007/ 10750 007 9246 3 • Balian EV, Segers HK, Lévêque C, Martens C (2008) The freshwater animal diversity assessment: an overview of the results. Hydrobiologia 595 (1): 627‑637. https://doi.org/ 10.1007/s10750-007-9246-3 10.1007/s10750-007-9246-3 • Baturina M, Fefilova E, Loskutova O (2021a) The list of zoobenthos taxa of the Vychegda River (in the areas impacted wastewaters of pulp and paper industry). Mendeley Data. V1. https://doi.org/10.17632/86zs8ts2fv.1. The fauna of aquatic invertebrates in the river impacted by wastewaters ... The fauna of aquatic invertebrates in the river impacted by wastewaters ... Aquatic Conservation: Marine and • De Smet WH (1996) Rotifera, volume 4: The Proalidae (Monogononta). SPB Academic Publishing, Amsterdam, 102 pp. • De Smet WH (1997) Rotifera. Part 5: The Dicranophoridae and the Ituridae (Monogononta). In: Dumont HJ (Ed.) Guides to the identification of the microinvertebrates of the continental waters of the world. 12. SPB Academic Publishing, Amsterdam, 1-325, 341-344 pp. [ISBN 9789051031355]. • Dumont HJ, Pensaert J (1983) A revision of the Scapholeberinae (Crustacea: Cladocera). Hydrobiologia 100: 3‑45. https://doi.org/10.1007/BF00027420 • Elliott JM, Humpesch UH, Macan TT (1988) Larvae of the British Ephemeroptera: a key with ecological notes. Freshwater Biological Association, 145 pp. [ISBN 9780900386473] Elliott JM, Humpesch UH, Macan TT (1988) Larvae of the British Ephemeroptera: a key with ecological notes. Freshwater Biological Association, 145 pp. • Elsakov VV, Shchanov VM (2016) Satellite methods in the analysis of changes in ecosystems of the Vychegda River basin. Modern Problems of Remote Sensing of the Earth from Space 13 (4): 135‑145. [In Russian]. • Fefilova E, Dubovskaya O, Kononova O, Frolova L, Abramova E, Nigamatzyanova G (2021) Data on taxa composition of freshwater zooplankton and meiobenthos across Arctic regions of Russia. Data in Brief 36: 107‑112. https://doi.org/10.1016/j.dib. 2021.107112 • Gaiji S, King N, Andelman S (2009) The Global Biodiversity Information Facility (GBIF): Infrastructure, standards, and access to forecast agricultural production in the face of 16 Baturina M et al • Kondratenok BM, et al. (2020) State report "On the state of the environment of the Komi Republic in 2019 / Ministry of Natural Resources and Environmental Protection of the Komi Republic. Territorial Information Fund of the Komi Republic, Syktyvkar, 162 pp. • Kononova O (2021) The list of zooplankton taxa of the Vychegda river (in the areas impacted wastewaters of pulp and paper industry). Mendeley Data. V1. https://doi.org/ 10 17632/2yrvb4s7kw 1 • Kondratenok BM, et al. (2020) State report "On the state of the environment of the Komi Republic in 2019 / Ministry of Natural Resources and Environmental Protection of the Komi Republic. Territorial Information Fund of the Komi Republic, Syktyvkar, 162 pp. • Kononova O (2021) The list of zooplankton taxa of the Vychegda river (in the areas impacted wastewaters of pulp and paper industry). Mendeley Data. V1. https://doi.org/ 10 17632/2yrvb4s7kw 1 • Kononova ON (2021) Zooplankton of the Vychegda River under the conditions of treated wastewater from pulp and paper industry. Contemporary Problems of Ecologу 28 (6). The fauna of aquatic invertebrates in the river impacted by wastewaters ... • Korovchinsky NM (2004) Cladocerans of the Order Ctenopoda in the World Fauna: Morphology, Systematic, Ecology, Zoogeography. KMK Scientific Press, Moscow, 410 pp. [In Russian]. [ISBN 5-87317-188-2] • Koste W, Shiel RJ (1989) Rotifera from Australian inland waters. III. Euchlanidae, Mytilinidae and Trichotriidae (Rotifera: Monogononta). Transactions of the Royal Society of South Australia 113: 85‑114. • Kotov AA, Stifter P (2006) Ilyocryptidae of the world. Guides to the identification of the microivertebrates of the Continental Waters of the world. 22. Kenobi Productions, Ghent & Backhuys Publishers, Leiden, 172 pp. • Krivosheina MG (2012) Keys to the flora and fauna of Russia. Key to the families and genera of Palaearctic Diptera insects of the suborder Nematocera by larvae. KMK Scientific Press, Moscow, 244 pp. [In Russian]. • Krivosheina MG (2012) Keys to the flora and fauna of Russia. Key to the families and genera of Palaearctic Diptera insects of the suborder Nematocera by larvae. KMK Scientific Press, Moscow, 244 pp. [In Russian]. • Kutikova L, Starobogatov Y (Eds) (1977) Keys to freshwater invertebrates of European part of Russia (plankton and benthos). Gidrometeoizdat, Leningrad, 511 pp. [In Russian]. • Kutikova L, Starobogatov Y (Eds) (1977) Keys to freshwater invertebrates of European part of Russia (plankton and benthos). Gidrometeoizdat, Leningrad, 511 pp. [In Russian]. • Kutikova LA (1977) Rotifers of the fauna of the USSR (Rotatoria). Nauka, Leningrad, 744 pp. [In Russian]. Kutikova LA (1977) Rotifers of the fauna of the USSR (Rotatoria). Nauka, Leningrad, 744 pp. [In Russian]. Lepneva SG (1964) Fauna of the USSR. Caddisflies. Zoological Institute of the USSR • Lepneva SG (1964) Fauna of the USSR. Caddisflies. Zoological Institute of the USSR Academy of Sciences, Moscow, Leningrad, 562 pp. [In Russian]. Academy of Sciences, Moscow, Leningrad, 562 pp. [In Russian]. • Lillehammer A (1988) Stoneflies (Plecoptera) of Fennoscandian and Denmark. Fauna Entomologica Scandinavica. 21. Brill, Leiden, 169 pp. [ISBN 978-9004086951] Lillehammer A (1988) Stoneflies (Plecoptera) of Fennoscandian and Denmark. Fauna lehammer A (1988) Stoneflies (Plecoptera) of Fennoscandian and Denmark. Fauna ehammer A (1988) Stoneflies (Plecoptera) of Fennoscandian and Denmark. Fauna tomologica Scandinavica. 21. Brill, Leiden, 169 pp. [ISBN 978-9004086951] Entomologica Scandinavica. 21. Brill, Leiden, 169 pp. [ISBN 978-9004086951] Entomologica Scandinavica. 21. Brill, Leiden, 169 pp. [ISBN 978-9004086951 • Manujlova EF (1964) Cladocera of the fauna of the world. Nauka, Leningrad, 328 pp. [In Russian]. Manujlova EF (1964) Cladocera of the fauna of the world. Nauka, Leningrad, 328 pp. The fauna of aquatic invertebrates in the river impacted by wastewaters ... • Starobogatov YI, Prozorova LA, Bogatov VV, Sayenko EM (2004) Keys to freshwater invertebrates of Russia and adjacent lands. Mollusces, Polychaetes, Nemerteans. 6. Nauka, St. Petersburg, 526 pp. [In Russian]. • Starobogatov YI, Prozorova LA, Bogatov VV, Sayenko EM (2004) Keys to freshwater invertebrates of Russia and adjacent lands. Mollusces, Polychaetes, Nemerteans. 6. Nauka, St. Petersburg, 526 pp. [In Russian]. Taskaev AI (Ed.) (1997) Atlas on climate and hydrology of the Komi Republic. Drofa, DiK, Moscow, 116 pp. [In Russian]. • Taskaev AI (Ed.) (1997) Atlas on climate and hydrology of the Komi Republic. Drofa, DiK, Moscow, 116 pp. [In Russian]. • Teslenko VA, Zhiltzova LA (2009) Key to the stoneflies (Insecta, Plecoptera) of Russia and adjacent countries. Imagines and nymphs. Dalnauka, Vladivostok, 382 pp. [In Russian]. • Teslenko VA, Zhiltzova LA (2009) Key to the stoneflies (Insecta, Plecoptera) of Russia and adjacent countries. Imagines and nymphs. Dalnauka, Vladivostok, 382 pp. [In Russian]. • Timm T (2009) A guide to the freshwater Oligochaeta and Polychaeta of northern and central Europe. 66. Mauch, Lauterbornia, 235 pp. • Timm T (2009) A guide to the freshwater Oligochaeta and Polychaeta of northern and central Europe. 66. Mauch, Lauterbornia, 235 pp. • Tsalolikhin SY (Ed.) (1994) Keys to freshwater invertebrates of Russia and adjacent lands. 1. Zoological Institute of the Russian Academy of Sciences, St. Petersburg, 395 pp. [In Russian]. • Tsalolikhin SY (Ed.) (1994) Keys to freshwater invertebrates of Russia and adjacent lands. 1. Zoological Institute of the Russian Academy of Sciences, St. Petersburg, 395 pp. [In Russian]. • Tsalolikhin SY (Ed.) (1995) Keys to freshwater invertebrates of Russia and adjacent lands. 2. Zoological Institute of the Russian Academy of Sciences, St. Petersburg, 628 pp. [In Russian]. • Tsalolikhin SY (Ed.) (1995) Keys to freshwater invertebrates of Russia and adjacent lands. 2. Zoological Institute of the Russian Academy of Sciences, St. Petersburg, 628 pp. [In Russian]. • Tsalolikhin SY (Ed.) (1997) Keys to freshwater invertebrates of Russia and adjacent lands. 3. Zoological Institute of the Russian Academy of Sciences, St. Petersburg, 439 pp. [In Russian]. • Tsalolikhin SY (Ed.) (1997) Keys to freshwater invertebrates of Russia and adjacent lands. 3. Zoological Institute of the Russian Academy of Sciences, St. Petersburg, 439 pp. [In Russian]. • Tsalolikhin SY (Ed.) (2001) Keys to freshwater invertebrates in Russia and adjacent territories. 5. Nauka, St. Petersburg, 836 pp. [In Russian]. The fauna of aquatic invertebrates in the river impacted by wastewaters ... 17 • Ostroumov SA (2002) Biodiversity protection and quality of water: role of ecosystems. Doklady Biological Sciences 382 (1-6): 18‑21. • Patova EN, Kondratenok BM, Sivkov MD, Kostrova SN (2021) Water quality of the Vychegda River under the conditions of the receipt of treated wastewater from the pulp and paper industry. Contemporary Problems of Ecologу 28 (6). • Rylov VM (1948) Fauna of the USSR. Crustaceans. Cyclopoida of the fresh waters. III. USSR Academy of Sciences Publishing House, Moscow, Leningrad, 320 pp. [In Russian]. • Schmidt-Kloiber A, Bremerich V, Wever AD, Jahnig SC, Martens K, Strackbein J, Tockner K, Hering D (2019) The Freshwater Information Platform: a global online network providing data, tools and resources for science and policy support. Hydrobiologia 838: 1‑11. https://doi.org/10.1007/s10750-019-03985-5 Hydrobiologia 838: 1‑11. https://doi.org/10.1007/s10750-019-03985-5 • Segers H (1995) Rotifera, Volume 2: The Lecanidae (Monogononta). SPB Academic Publishing BV, Hague, 226 pp. • Shashkov MP, Chadin IF, Ivanova NV (2017) Guide to best practices on a data standardization for publication via the global portal GBIF.ORG and preparing data paper. Trudy Kol'skogo Nauchnogo Centra Rossijskoj Akademii Nauk 8 (5-6): 22‑35. [In Russian]. • Shashkov MP, Chadin IF, Ivanova NV (2017) Guide to best practices on a data standardization for publication via the global portal GBIF.ORG and preparing data paper. Trudy Kol'skogo Nauchnogo Centra Rossijskoj Akademii Nauk 8 (5-6): 22‑35. [In Russian]. • Shiel RJ, Koste W (1992) Rotifera from Australian inland waters. VIII. Trichocercidae (Rotifera: Monogononta). Transactions of the Royal Society of South Australia 116: 1‑2. • Smirnov NN (1971) Chydoridae of the fauna of the world. Nauka, Leningrad, 531 pp. [In Russian] • Shiel RJ, Koste W (1992) Rotifera from Australian inland waters. VIII. Trichocercidae (Rotifera: Monogononta). Transactions of the Royal Society of South Australia 116: 1‑2. • Shiel RJ, Koste W (1992) Rotifera from Australian inland waters. VIII. Trichocercidae (Rotifera: Monogononta). Transactions of the Royal Society of South Australia 116: 1‑2. (Rotifera: Monogononta). Transactions of the Royal Society of South Australia 116: 1‑2. • Smirnov NN (1971) Chydoridae of the fauna of the world. Nauka, Leningrad, 531 pp. [In Russian]. • Smirnov NN (1971) Chydoridae of the fauna of the world. Nauka, Leningrad, 531 pp. [In Russian]. • Smirnov NN (1976) Macrotricidae and Moinidae fauna of the world. Nauka, Leningrad, 238 pp. [In Russian]. Smirnov NN (1976) Macrotricidae and Moinidae fauna of the world. Nauka, Leningrad, 238 pp. [In Russian]. The fauna of aquatic invertebrates in the river impacted by wastewaters ... [In Russian]. • Morduhaj-Boltovskoj FD (Ed.) (1975) Methodology for the study of biogeocenoses of inland waters. Nauka, Moscow, 240 pp. [In Russian]. • Morduhaj-Boltovskoj FD (Ed.) (1975) Methodology for the study of biogeocenoses of inland waters. Nauka, Moscow, 240 pp. [In Russian]. • Narchuk EP, Tumanov DV (Eds) (2000) Keys to freshwater invertebrates in Russia andadjacent territories. Diptera. Nauka, St. Petersburg, 997 pp. • Narchuk EP, Tumanov DV (Eds) (2000) Keys to freshwater invertebrates in Russia andadjacent territories. Diptera. Nauka, St. Petersburg, 997 pp. • Nogrady T, Pourriot R, Segers H (1995) Rotifera. Volume 3: The Notommatidae and the Scaridiidae. SPB Academic Publishing, Amsterdam, 102 pp. • Nogrady T, Pourriot R, Segers H (1995) Rotifera. Volume 3: The Notommatidae and the Scaridiidae. SPB Academic Publishing, Amsterdam, 102 pp. • Nogrady T, Segers H (2002) Rotifera. Volume 6: Asplanchnidae, Gastropodidae, Lindiidae, Microcodidae, Synchaetidae, Trochosphaeridae and Filinia. Backhuys Publishers, Leiden, 264 pp. • Nogrady T, Segers H (2002) Rotifera. Volume 6: Asplanchnidae, Gastropodidae, Lindiidae, Microcodidae, Synchaetidae, Trochosphaeridae and Filinia. Backhuys Publishers, Leiden, 264 pp. The fauna of aquatic invertebrates in the river impacted by wastewaters ... The fauna of aquatic invertebrates in the river impacted by wastewaters ... [ISBN 5-02-026162-9] • Tsalolikhin SY (Ed.) (2001) Keys to freshwater invertebrates in Russia and adjacent territories. 5. Nauka, St. Petersburg, 836 pp. [In Russian]. [ISBN 5-02-026162-9] 18 Baturina M et al • Tsalolikhin SY (Ed.) (2016) Key to identification of zooplankton and zoobenthos of fresh water of European Russia. 2. KMK Scientific Press, Moscow, St. Petersburg, 457 pp. [In Russian]. [ISBN 978-5-9907572-4-0] • Ueda H, Reid JW (Eds) (2003) Copepoda: Cyclopoida: Genera Mesocyclops and Thermocyclops. 20. Bachhuys Publishers, Leide, 318 pp. • Wilke T, Wilko HA, Bininda-Emonds OR (2019) A weighted taxonomic matrix key for species of the rotifer genus Synchaeta (Rotifera, Monogononta, Synchaetidae). ZooKeys 871: 1‑40. https://doi.org/10.3897/zookeys.871.36435 • Zinchenko TD, et al. (2014) Saline rivers provide arid landscapes with a considerableamount of biochemically valuable production of chironomid (Diptera) larvae. Hydrobiologia 722: 115‑128. https://doi.org/10.1007/s10750-013-1684-5 • Tsalolikhin SY (Ed.) (2016) Key to identification of zooplankton and zoobenthos of fresh water of European Russia. 2. KMK Scientific Press, Moscow, St. Petersburg, 457 pp. [In Russian]. [ISBN 978-5-9907572-4-0] • Ueda H, Reid JW (Eds) (2003) Copepoda: Cyclopoida: Genera M Thermocyclops. 20. Bachhuys Publishers, Leide, 318 pp. Thermocyclops. 20. Bachhuys Publishers, Leide, 318 pp. • Wilke T, Wilko HA, Bininda-Emonds OR (2019) A weighted taxonomic matrix key for species of the rotifer genus Synchaeta (Rotifera, Monogononta, Synchaetidae). ZooKeys 871: 1‑40. https://doi.org/10.3897/zookeys.871.36435 • Zinchenko TD, et al. (2014) Saline rivers provide arid landscapes with a considerableamount of biochemically valuable production of chironomid (Diptera) larvae. Hydrobiologia 722: 115‑128. https://doi.org/10.1007/s10750-013-1684-5 Thermocyclops. 20. Bachhuys Publishers, Leide, 318 pp. • Wilke T, Wilko HA, Bininda-Emonds OR (2019) A weighted taxonomic matrix key for species of the rotifer genus Synchaeta (Rotifera, Monogononta, Synchaetidae). ZooKeys 871: 1‑40. https://doi.org/10.3897/zookeys.871.36435 • Zinchenko TD, et al. (2014) Saline rivers provide arid landscapes with a considerableamount of biochemically valuable production of chironomid (Diptera) • Zinchenko TD, et al. (2014) Saline rivers provide arid landscapes with a considerableamount of biochemically valuable production of chironomid (Diptera) larvae. Hydrobiologia 722: 115‑128. https://doi.org/10.1007/s10750-013-1684-5
https://openalex.org/W2886064957	https://europepmc.org/articles/pmc6161027?pdf=render	English	null	In Vitro Stimulation of Multidrug Resistance-Associated Protein 2 Function Is Not Reproduced In Vivo in Rats	Pharmaceutics	2,018	cc-by	13,080	Received: 29 June 2018; Accepted: 31 July 2018; Published: 8 August 2018 Abstract: Previously we reported that coproporphyrin-I (CP-I) is an optimal probe substrate for multidrug resistance-associated protein 2 (MRP2), and stimulation of MRP2-mediated transport is probe substrate-dependent. In the present investigation, we assessed if the in vitro stimulation is physiologically relevant. Similar to human MRP2 transport, CP-I was transported by rat Mrp2 in a typical Michaelis-Menten kinetics with apparent Km and Vmax values of 15 ± 6 µM and 161 ± 20 pmol/min/mg protein, respectively. In vivo Mrp2 functions were monitored by biliary and renal secretion of CP-I and its isomer CP-III, in bile-duct cannulated rats before and after treatment with mitoxantrone, progesterone, and verapamil. These compounds stimulated Mrp2-mediated CP-I transport in vitro. No signiﬁcant increase in biliary or renal clearances, as well as in the cumulative amount of CP-I or CP-III eliminated in bile, were detected following treatment with the in vitro stimulators, indicating an in vitro to in vivo disconnect. In presence of 10 µM bilirubin, the in vitro stimulation was suppressed. We concluded that the in vitro stimulation of CP-I transport mediated by Mrp2 is not translatable in vivo, and proposed that the presence of endogenous compounds such as bilirubin in the liver may contribute to the in vitro to in vivo disconnect. Keywords: ABCC2; MRP2; in vitro stimulation; drug transporter; coproporphyrin Ravindranath Reddy Gilibili 1,†, Vishwanath Kurawattimath 1,†, Bokka Venkata Murali 1, Yurong Lai 2, T. Thanga Mariappan 1, Hong Shen 3 and Sagnik Chatterjee 1,* 1 Pharmaceutical Candidate Optimization, Biocon Bristol-Myers Squibb R&D Center (BBRC), Syngene International Ltd., Bangalore 560100, India; ravindra.reddy@syngeneintl.com (R.R.G.); Vishwanath.Kurawattimath@syngeneintl.com (V.K.); Venkata.murali@syngeneintl.com (B.V.M.); Thanga.Mariappan@syngeneintl.com (T.T.M.) 1 Pharmaceutical Candidate Optimization, Biocon Bristol-Myers Squibb R&D Center (BBRC), Syngene International Ltd., Bangalore 560100, India; ravindra.reddy@syngeneintl.com (R.R.G.); Vishwanath.Kurawattimath@syngeneintl.com (V.K.); Venkata.murali@syngeneintl.com (B.V.M.); Thanga.Mariappan@syngeneintl.com (T.T.M.) 1 Pharmaceutical Candidate Optimization, Biocon Bristol-Myers Squibb R&D Center (BBRC), Syngene International Ltd., Bangalore 560100, India; ravindra.reddy@syngeneintl.com (R.R.G.); Vishwanath.Kurawattimath@syngeneintl.com (V.K.); Venkata.murali@syngeneintl.com (B.V.M.); Thanga.Mariappan@syngeneintl.com (T.T.M.) 2 Department of drug metabolism, Gilead Sciences Inc., Foster City, CA 94404, USA; Yurong.Lai@gilead.com 3 Pharmaceutical Candidate Optimization, Bristol-Myers Squibb Company, 3551 Lawrenceville Road, Princeton, NJ 08540, USA; hong.shen1@bms.com 2 Department of drug metabolism, Gilead Sciences Inc., Foster City, CA 94404, USA; Yurong.Lai@gilead.com 3 Pharmaceutical Candidate Optimization, Bristol-Myers Squibb Company, 3551 Lawrenceville Road, Princeton, NJ 08540, USA; hong.shen1@bms.com g * Correspondence: sagnik.chatterjee@syngeneintl.com; Tel.: +91-80-6633-4084 † These authors contributed equally to this work. † These authors contributed equally to this work. In Vitro Stimulation of Multidrug Resistance-Associated Protein 2 Function Is Not Reproduced In Vivo in Rats Ravindranath Reddy Gilibili 1,†, Vishwanath Kurawattimath 1,†, Bokka Venkata Murali 1, Yurong Lai 2, T. Thanga Mariappan 1, Hong Shen 3 and Sagnik Chatterjee 1,* pharmaceutics pharmaceutics pharmaceutics pharmaceutics Pharmaceutics 2018, 10, 125; doi:10.3390/pharmaceutics10030125 www.mdpi.com/journal/pharmaceutics 1. Introduction It is becoming increasingly evident that multidrug resistance-associated protein 2 (MRP2/Mrp2, denoted by gene ABCC2/Abcc2) plays an important role in disposition and elimination to regulate pharmacokinetics and pharmacodynamic of xenobiotics [1,2]. MRP2 is located in the canalicular membrane of hepatocytes but also in the apical (luminal) membrane of enterocytes and renal proximal tubule cells [3]. In hepatocytes, MRP2 facilitates the efﬂux of glucuronide and glutathione conjugates of endobiotics, such as bilirubin conjugates and dianionic bile acids [4], as well as few anionic xenobiotics, such as methotrexate [5]. The inhibition of MRP2 activity is therefore important for drug hepatic disposition and elimination, and can cause potential drug-drug interactions (DDI) and hepatotoxicity [6]. For example, a two-fold increase in methotrexate plasma levels due to MRP2 polymorphism has been reported [7]. However, changes in MRP2 activity in vivo is usually associated Pharmaceutics 2018, 10, 125; doi:10.3390/pharmaceutics10030125 www.mdpi.com/journal/pharmaceutics 2 of 15 Pharmaceutics 2018, 10, 125 with profound changes in liver exposure, and minor/no effect on plasma exposure, and therefore it usually remains undetected [8]. Recently, it was shown that atorvastatin liver concentration increases 1.64-fold, in the presence of metformin, an Mrp2 inhibitor, without affecting the plasma exposure in Sprague Dawley (SD) rats [9]. Interactions with MRP2 are commonly characterized in vitro, in an inverted membrane vesicle expressing MRP2/Mrp2 in the presence of ATP. Decrease or increase in ATP-dependent transport of probe substrates in the presence of an interacting compound suggest either inhibition or stimulation. To investigate if a compound is an inhibitor of MRP2/Mrp2, estradiol 17-β-glucuronide (E17βG) is widely used as a probe substrate [1]. However, many compounds such as indomethacin, diclofenac, and benzbromarone are reported to “modulate” the transport of E17βG [10–13]. Modulation refers to stimulation at lower concentrations of the compound, followed by inhibition at higher concentration, providing a “bell-shaped” curve in a percent (%) transport against interacting compound concentration plots. The compounds that impart similar proﬁle to the transport of probe substrates are henceforth termed as modulators or stimulators interchangeably in this article. The presence of a modulation site, in addition to active transport site in the MRP2 protein, is proposed to explain the modulation and “bell-shaped” curve [10,14,15]. There are two manuscripts that explored the in vivo relevance of the in vitro stimulation of Mrp2, in rats. Heredi-Szabo et al. used indomethacin as a stimulator and reported that indomethacin modestly decreases the half-life (40%) and increases E17βG biliary clearance [16]. 1. Introduction In addition, Ito et al. used benzyl penicillin as a stimulator and found that biliary clearance of glutathione (GSH) and bile ﬂow were transiently enhanced (about four-fold for GSH and two-fold for bile ﬂow) [17]. However, in vivo relevance of the substrate-dependence of in vitro MRP2 stimulation remains unknown. In addition, there are no reports of any clinical drug-drug interaction or toxicity that has been attributed to this stimulation. This raises questions on the in vivo relevance of in vitro stimulation of MRP2-mediated probe substrate transport. Recently, we demonstrated that coproporphyrin-I (CP-I) is an optimal probe substrate for MRP2 [10]. CP-I is a byproduct of heme biosynthesis and its physiological role is not yet clear. The uptake of CP-I in MRP2-overexpressing membrane vesicles follows a typical Michaelis-Menten equation with a low Km value of 7.7 µM. When CP-I is used as a MRP2 probe for in vitro characterization of MRP2 inhibition, a lower percentage of compounds were found to stimulate MRP2 mediated transport, compared to E17βG [10]. In our study using 97 compounds, we found 51 (53%) to stimulate MRP2-mediated E17βG transport, while only 8 out of 47 (17%) compounds stimulated the transport of CP-I, while 30 compounds were found to inhibit CP-I transport [10]. The results suggest that the stimulation appeared to be a probe substrate-dependent phenomenon, rather than an intrinsic property of the transporter. Furthermore, the stimulators shared the basic backbone structures of physiological steroids, which raises a question of potential in vivo physiological relevance of in vitro stimulation of MRP2-mediated CP-I transport [10]. The aim of the current study was to investigate whether the compounds that stimulated Mrp2-mediated CP-I transport in vitro, also stimulate Mrp2-mediated CP-I transport in vivo. Mrp2-mediated in vivo transport is appreciated by measuring biliary clearance of CP-I and its isomer CP-III in rats, in the presence of selected stimulators. The stimulation of CP-I transport in rat Mrp2 transfected membrane vesicles was conﬁrmed and a mechanistic understanding for the perceived in vitro to in vivo disconnect put forward with potential in vitro evidence. 2.2. Animals Male Sprague Dawley rats weighing 300 to 350 g (10–12 weeks of age) were obtained from Vivo Bio Tech Ltd., Hyderabad (India). All animal experiments were conducted in the animal research facility of Syngene International Ltd., Bangalore, India, after obtaining approval of the Institutional Animal Ethics Committee (Approval details: Title: Investigation of Metabolism and Elimination pathways of New Chemical Entities (NCEs) in rats; SYNGENE/IAEC/858/07-2017; Principal Investigator/Research Scholar: Mr. Vishwanath K M). The committee was registered for the Purpose of Control and Supervision on Experiments on Animals and accredited by Association for Assessment and Accreditation of Laboratory Animal Care International. The animals were fed a standard laboratory rodent diet (Tetragon Chemie Pvt. Ltd., Bangalore, India) and housed at room temperature (22 ± 3 ◦C) and relative humidity of 50 ± 20% on a 12-h light and dark cycle. Water was provided ad libitum throughout the study. An intravenous solution formulation of mitoxantrone (6 and 15 mg/mL) and verapamil (10 mg/mL) were formulated as a solution by using 100% saline. Progesterone (1 mg/mL) was formulated using a solution of dimethylacetamide: Solutol: PEG400: water (5:5:75:15% v/v). 2.1. Materials/Chemicals Mitoxantrone hydrochloride, coproporphyrin-I dihydrochloride (CP-I), progesterone, verapamil, ritonavir and testosterone were procured from Sigma-Aldrich (Sigma-Aldrich Chemie GmbH, Munich, Germany). d4-CP-I (15N4 deuterated) sodium bisulfate salt procured from Toronto Research Chemicals (Toronto, ON, Canada). MultiScreen Solvinert ﬁlter plates (0.45 µm, low binding hydrophilic 3 of 15 Pharmaceutics 2018, 10, 125 polytetraﬂuoroethylene) were purchased from Millipore (Tullagreen, Ireland). High-performance liquid chromatography–grade methanol was purchased from Merck (Mumbai, India). Formic acid was purchased from Fluka (Fluka Chemie, GmbH, Munich, Germany) and Milli-Q water from Milli-Q system (Millipore SAS, Molsheim, France). Polyethylene-10 and -50 tubing was purchased from Smiths Medical (ASD incorporation, Dublin, OH, USA), and 22-gauge needles and syringes were purchased from Becton Dickinson India Pvt Ltd. (Bangalore, India). Rat Mrp2-expressing inside-out membrane vesicles (protein concentration 4 mg/mL) derived from Sf9 insect cells were in-house prepared. Reaction incubation plates (96-well, ultra-low attachment, polystyrene, ﬂat bottom, clear), Filter plate, 96 well (0.25 mm Glass Fiber/1.2 µm PES) and assay plates (96 well, black, ﬂat bottom, polystyrene) for ﬂuorescence measurement purchased from Corning® Costar® (Kennebunk, ME, USA). Krebs-Henseleit buffer purchased from Bioreclamation IVT (Bangalore, India). Stock solutions were prepared in dimethyl sulphoxide and stored at −70 ◦C. 2.3. Vesicular Transport Assay The vesicular transport assay in membrane vesicles overexpressing rat Mrp2 was conducted using the methods reported previously [10]. Brieﬂy, rat Mrp2 membrane vesicles were diluted to an appropriate concentration in buffer-A containing 50 mM MOPS-Tris (pH 7.0), 70 mM KCl, 7.5 mM MgCl2. Membrane vesicles (20 µL, 50 µg protein) were co-incubated with 0.5 µL of test substrates (CP-I) at 37 ◦C for 3 min. Then, the reaction was initiated by adding pre-warmed buffer-A (29.5 µL) premixed with 4 mM MgATP or 4 mM MgAMP and 2 mM glutathione. Following incubation for designated times at 37 ◦C on a rotary shaker (Innova 40, New Brunswick Scientiﬁc Co., Inc., Enﬁeld, CT, USA) at 100 rpm, the reaction was stopped by adding 150 µL of cold wash with buffer-B containing 40 mM MOPS-Tris (pH 7.0), 70 mM KCl. The reaction mixture was then transferred into a pre-wet, 96-well ﬁlter plate which was placed onto a ﬁltration device (FiltrEX™96-Well Filter Plates, Corning Technologies India Pvt Ltd., Pune, India), and ﬁltered rapidly via a connected vacuum pump (MultiScreen®HTS Vacuum, Manifold, MA, USA). All wells were washed 5 times, each time with 200 µL of ice-cold wash buffer to remove excess CP-I. After the ﬁnal wash, the ﬁlter plate was dried at room temperature for 1 h. Membrane vesicles were digested using an extraction solvent (100 µL of 0.5% SDS dissolved in milliQ water, and the plate was kept on a microplate shaker (VWR, Radnor, PA, USA) for 15 min at 230 rpm. Then, the ﬁlter plate was centrifuged (Eppendorf, Hauppauge, NY, USA) for 2 min at 2000 rpm along with a receiver plate attached to collect the ﬁltrate. The ﬁltrate was further used to quantify the CP-I levels using a ﬂuorimeter. Fluorescence measurements were conducted with a microplate reader (SpectraMax® M2e, molecular device, San Jose, CA, USA), using 401 and 595 nm as excitation and emission wavelengths, respectively. ATP-dependent net transport was calculated by 4 of 15 Pharmaceutics 2018, 10, 125 subtracting the AMP values from those of ATP values. All experiments related to CP-I were conducted in reduced light to minimize ﬂorescent bleaching. Mitoxantrone, testosterone, progesterone, and verapamil were selected to assess their modulatory effects on rat Mrp2-mediated CP-I transport in membrane vesicles. The ﬁnal assay concentration of CP-I was 5 µM. 2.3. Vesicular Transport Assay The concentrations tested for each modulator were 1, 10, 50, 100, 250, 500, and 1000 µM, except verapamil (25, 50, and 100 µM). 2.4. In Vivo Studies with Bile Duct Cannulated Rats Bile duct cannulation on Sprague-Dawley (SD) rats was conducted [18] in a cross-over study design. The freely moving bile duct cannulated rats were kept in metabolic cages. Following 48 h acclimatization, rats were dosed with saline (dosed in all the groups as a control at 1 mL/kg dose volume) and bile, then plasma and urine were collected at speciﬁed time points. Ten minutes after the last bile, plasma, and urine sample collection, the same rats were dosed with mitoxantrone, verapamil, and progesterone via the jugular vein as an intravenous infusion at 1 mL/kg dose volume by using a single syringe model ‘11’ Pico plus pump. The doses used were 6 and 15 mg/kg for mitoxantrone, 10 mg/kg for verapamil, and 1 mg/kg for progesterone. The blood samples (200 µL) were collected at 0.17, 0.25, 0.5, 0.75, 1, 1.25, 1.5, 2, 2.5, and 3 h and bile samples were collected in pre-weighed amber color tubes over 15 min intervals up to 1.5 h and over 30 min intervals up to 3 h after dosing mitoxantrone. The blood samples (200 µL) were collected at 0.17, 0.25, 0.5, 0.75, 1, 1.5, and 2 h and bile samples were collected in pre-weighed amber color tubes over 10 to 15 min intervals up to 1 h, and over 30 min intervals up to 2 h after dosing verapamil and progesterone. In addition, 0–3 h urine samples and terminal liver samples were collected from 3 rats using serial sampling. Blood was collected through the jugular vein in tubes containing 2% w/v potassium ethylenediaminetetraacetic acid solution, and bile was collected in pre-weighed amber tubes, rats were euthanized using carbon dioxide inhalation, liver samples were harvested and homogenized using 4 volumes of water. The bile, plasma, and liver homogenate samples were stored at −80 ◦C until further analysis. 2.6. LC-MS/MS Method LC-MS/MS method was developed with adequate sensitivity for the quantiﬁcation of CP-I, CP-III, mitoxantrone, verapamil, and progesterone using a UPLC system (Waters Corporation, Milford, MA, USA) coupled with a Triple Quad 5500 System (AB Sciex, Framingham, MA, USA), operated in positive electrospray ionization. A combined internal standard solution consisting of d4-CP-I (15N4 deuterated) sodium bisulfate salt and ritonavir was used during the analysis. d4-CP-I was used as internal standard in the analysis of CP-I and CP-III, whereas ritonavir was used as internal standard (IS) for the analysis of progesterone, mitoxantrone, and verapamil. The analytes and internal standards were separated on an Ace excel 2 C18-150 × 2.1 mm column (Advanced Chromatography Technologies Ltd., Aberdeen, Scotland) using a binary gradient elution with 10 mM ammonium formate in water with 0.1% formic acid (solvent-A); and acetonitrile with 0.1% formic acid (solvent-B). The gradient started with 40% solvent B and maintained for 0.5 min; in 2.5 min, solvent B was increased from 40 to 55% and then to 95% in 2.8 min and remained constant at 95% solvent B for 1.10 min. Then in 3.80 min solvent B was decreased from 95 to 40%, and it remained constant for 0.2 min, accounting to a total gradient time of 4 min. The ﬂow rate was set at 0.5 mL/min. The mass spectrometric conditions adopted in the analysis were as follows: Capillary voltage, 5500 V; drying gas temperature, 450 ◦C; and nebulizer gas pressure, 50 psi (both nebulizer and drying gas were high-purity nitrogen). The MRMs monitored for the compounds were as follows: CP-I (655.5 →596.4), CP-III (655.5 → 596.4), mitoxantrone (455.1 →358.3), verapamil (455.2 →165.1), progesterone (315.3 →297.5), d4-CP-I (659.3 →600.3), and ritonavir (721.2 →296.2). Respective analyte concentrations in study samples were obtained from the calibration curves, plotted as analyte to internal standard peak area ratios against respective analyte concentrations. Analyst version 1.6.2 was used for system control and data processing. A total of 15 calibration standards were prepared in 1% bovine serum albumin (BSA) by serial dilution. All the bile samples and urine samples were diluted in 1% BSA (~10 fold) prior to analysis. Then 100 µL of the sample was quenched with 800 µL of acidiﬁed acetonitrile (1% formic acid) containing IS. 2.6. LC-MS/MS Method The samples were vortexed on a spinix mixer (The Scientiﬁc House, Chennai, India), and centrifuged in a refrigerated centrifuge (Minispin plus, Eppendorf AG-22331, Hamburg, Germany) at 14,000× g, 4 ◦C for 5 min. A supernatant of approximately 150 µL was transferred into a 96-well plate and subjected to LC-MS/MS analysis. 2.5. Liver Free Fraction Determination for Mitoxantrone, Verapamil and Progesterone The free fraction of the test compounds (Mitoxantrone, Verapamil, and Progesterone) in liver was determined using a rapid equilibrium dialysis (RED) device. Ritonavir spiked in human plasma at 2 µM concentration was used as a positive control for this purpose. The fu (free fraction) was calculated using the formula given by Riccardi et al. [19]. Brieﬂy, blank liver homogenates were prepared at two different set of dilutions (50-times and 100-times) in 100 mM phosphate buffer using homogenizer (POLYTRON®, PT4000, Kinematica AG, Luzern, Switzerland). The test compounds were then individually spiked into these liver homogenates to achieve a target concentration of 10 µM. RED device (Thermo Fisher Scientiﬁc, Hudson, NH, USA) was set up as per the manufacturer’s instructions (Instructions, RED Device inserts, 89809, Thermo Scientiﬁc, Marietta, OH, USA). An aliquot of 200 µL spiked samples were added into the donor side in triplicate, and 350 µL of 100 mM potassium phosphate buffer into the respective receiver side and allowed to equilibrate for 6 h in a HEPA class 100, CO2 incubator (Steri-cycle, Model 370, Thermo Scientiﬁc, Marietta, OH, USA) maintained at 37 ◦C with 5% CO2. At the end of the incubation, 25 µL incubate from the donor side was aspirated and mixed with 25 µL of 100 mM phosphate buffer; similarly, 25 µL from receiver side was mixed with 25 µL of respective control liver homogenates in a collection plate. Then 150 µL of acetonitrile containing internal standard (Propranolol at 150 nM) was added to the plate and mixed with a multi-channel pipette (Thermo Scientiﬁc, FINNPIPETTE F, MH-14040, and Finland). The contents were then transferred to Solvinert ﬁlter plate (Merck Millipore Ltd., Tullagreen, Ireland) ﬁtted with 1 mL capacity 96-well collection plate (Waters Corporation, Milford, MA, USA). The plate was centrifuged at 3220× g for 5 min and the ﬁltrate was injected on LC-MS/MS. Analyte concentrations in study samples were then obtained from the calibration standards prepared similarly, and by plotting the analyte to internal standard peak area ratios against respective analyte concentrations. Analyst version 5 of 15 Pharmaceutics 2018, 10, 125 1.6.2 was used for system control and data processing. The reported % fu was calculated as average of fu obtained from both 50-times and 100-times dilutions. 1.6.2 was used for system control and data processing. The reported % fu was calculated as average of fu obtained from both 50-times and 100-times dilutions. 2.7. Pharmacokinetic Analysis CP-I and CP-III area under the plasma concentration-time proﬁles (AUC) were calculated by non-compartmental analysis and mixed log linear method using Kinetica software (Version 5.5.1; Thermo Electron Corporation, Waltham, MA, USA). Cumulative amounts of CP-I and CP-III were calculated by using Microsoft excel. The biliary (CLb) and renal (CLr) clearances of CP-I and CP-III were estimated by the following equations: CLb = Xb AUC CLr = Xr AUC where, Xb and Xr are cumulative amounts of CP-I and CP-III excreted in bile and urine, respectively. 2.8. Statistical Analysis 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results In time linearity studies, Mrp2-mediated CP-I transport was linear with time up to 30 min, and a 15 min time point was used for all incubations unless speciﬁed. The transport kinetics of CP-I were then characterized by assessing concentration-dependent transport of CP-I in membrane vesicles overexpressing rat Mrp2 protein. ATP over AMP ratio of CP-I uptake is 23, with negligible uptake in the presence of AMP. As depicted in Figure 1, the concentration-dependent CP-I uptake appeared to follow a hyperbolic relationship. The kinetic curve could be adequately modelled by the standard Michaelis-Menten equation. The apparent Km and Vmax values (mean ± SD) for CP-I were 15 ± 6 µM and 161 ± 20 pmol/min/mg protein, respectively. 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles In time linearity studies, Mrp2-mediated CP-I transport was linear with time up to 30 min, and a 15 min time point was used for all incubations unless specified. The transport kinetics of CP-I were then characterized by assessing concentration-dependent transport of CP-I in membrane vesicles overexpressing rat Mrp2 protein. ATP over AMP ratio of CP-I uptake is 23, with negligible uptake in the presence of AMP. As depicted in Figure 1, the concentration-dependent CP-I uptake appeared to follow a hyperbolic relationship. The kinetic curve could be adequately modelled by the standard Michaelis-Menten equation. The apparent Km and Vmax values (mean ± SD) for CP-I were 15 ± 6 µM and 161 ± 20 pmol/min/mg protein, respectively. Figure 1. Concentration-dependent uptake transport of coproporphyrin-I (CP-I) in rat multidrug resistance-associated protein 2 (Mrp2) membrane vesicles. Rat Mrp2 membrane vesicles were incubated with CP-I at 37 °C for 15 min. ATP-dependent CP-I uptake is measured by subtracting the uptake in the presence of AMP from that of ATP. For more experimental details refer to Section 2.3. All data values are presented as mean and SD of a single experiment performed in triplicate wells. Fi u e 2 ho the o e t atio de e de t effe t of a iou odulato ( ito a t o e Figure 1. Concentration-dependent uptake transport of coproporphyrin-I (CP-I) in rat multidrug resistance-associated protein 2 (Mrp2) membrane vesicles. Rat Mrp2 membrane vesicles were incubated with CP-I at 37 ◦C for 15 min. ATP-dependent CP-I uptake is measured by subtracting the uptake in the presence of AMP from that of ATP. For more experimental details refer to Section 2.3. 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results All data values are presented as mean and SD of a single experiment performed in triplicate wells. Figure 1. Concentration-dependent uptake transport of coproporphyrin-I (CP-I) in rat multidrug resistance-associated protein 2 (Mrp2) membrane vesicles. Rat Mrp2 membrane vesicles were incubated with CP-I at 37 °C for 15 min. ATP-dependent CP-I uptake is measured by subtracting the uptake in the presence of AMP from that of ATP. For more experimental details refer to Section 2.3. All data values are presented as mean and SD of a single experiment performed in triplicate wells. Figure 1. Concentration-dependent uptake transport of coproporphyrin-I (CP-I) in rat multidrug resistance-associated protein 2 (Mrp2) membrane vesicles. Rat Mrp2 membrane vesicles were incubated with CP-I at 37 ◦C for 15 min. ATP-dependent CP-I uptake is measured by subtracting the uptake in the presence of AMP from that of ATP. For more experimental details refer to Section 2.3. All data values are presented as mean and SD of a single experiment performed in triplicate wells. progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2-overexpressing membrane vesicles. These modulators were selected based on the previous results found with human MRP2 vesicles [10]. As expected, all four modulators showed increased uptake of CP-I (stimulation) up to a specific concentrations and then showed inhibition of uptake at higher concentrations in rat Mrp2 vesicles, leading to a bell-shaped curve. This trend of stimulation and inhibition is common for all the four modulators, although at different concentration ranges, which agree with the results observed in human MRP2 vesicles (Gilibili et al., 2017). The interactions between CP-I uptake in rat Mrp2 vesicles and the above stimulators were further conducted in the presence of 10 µM bilirubin. As shown in Figure 3, the presence of bilirubin suppressed the stimulation of CP-I transport in rat Mrp2 vesicles by 200 µM mitoxantrone, 100 µM progesterone, and 100 µM verapamil. The suppression reached a maximum (about 50%) at 100 µM mitoxantrone. Figure 2 shows the concentration-dependent effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2-overexpressing membrane vesicles. These modulators were selected based on the previous results found with human MRP2 vesicles [10]. As expected, all four modulators showed increased uptake of CP-I (stimulation) up to a speciﬁc concentrations and then showed inhibition of uptake at higher concentrations in rat Mrp2 vesicles, leading to a bell-shaped curve. 2.8. Statistical Analysis Statistical analysis was performed using the GraphPad Prism program version 5.02 (GraphPad software, San Diego, CA, USA). Mean and standard deviations are calculated in Microsoft Excel 6 of 15 Pharmaceutics 2018, 10, 125 2013. Student t test was conducted to compare the cumulative amount of CP-I and CP-III obtained at each time point when administered saline and different modulators (mitoxantrone, verapamil, and progesterone); p < 0.05 was regarded as signiﬁcant. No statistical test was conducted in in vitro experiments, as they represent triplicate data from only one experiment. 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results In the inset, same data set is represented in curve to show the bell shape of the concentration-dependent stimulation and inhibition (also referred as modulation in the text). Figure 2. Effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2 membrane vesicles. Rat Mrp2-mediated CP-I transport was measured in inside-out membrane vesicles. Mrp2 vesicle protein (50 µg/well) were preincubated with modulators at above reported concentrations at 37 °C and reaction was started by adding CP-I (5 µM) and incubated for 15 min. NP; not performed. For more on experimental details refer to Section 2.3. Dotted line represents basal level (100%) of CP-I uptake in control wells in the presence of ATP with no modulators added. All data values presented as mean and SD of single experiment performed in triplicate wells. In the inset, same data set is represented in curve to show the bell shape of the concentration-dependent stimulation and inhibition (also referred as modulation in the text). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results This trend of stimulation and inhibition is common for all the four modulators, although at different concentration ranges, which agree with the results observed in human MRP2 vesicles (Gilibili et al., 2017). The interactions between CP-I uptake in rat Mrp2 vesicles and the above stimulators were further conducted in the presence of 10 µM bilirubin. As shown in Figure 3, the presence of bilirubin suppressed the stimulation of CP-I transport in rat Mrp2 vesicles by 200 µM mitoxantrone, 100 µM progesterone, and 100 µM verapamil. The suppression reached a maximum (about 50%) at 100 µM mitoxantrone. 7 of 15 7 of 15 Pharmaceutics 2018, 10, 125 Ph ti 2018 10 FOR P maceutics 2018, 10, x FOR PEER REVIEW 7 of 1 Figure 2. Effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2 membrane vesicles. Rat Mrp2-mediated CP-I transport was measured in inside-out membrane vesicles. Mrp2 vesicle protein (50 µg/well) were preincubated with modulators at above reported concentrations at 37 °C and reaction was started by adding CP-I (5 µM) and incubated for 15 min. NP; not performed. For more on experimental details refer to Section 2.3. Dotted line represents basal level (100%) of CP-I uptake in control wells in the presence of ATP with no modulators added. All data values presented as mean and SD of single experiment performed in triplicate wells. In the inset, same data set is represented in curve to show the bell shape of the concentration-dependent stimulation and inhibition (also referred as modulation in the text). Figure 2. Effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2 membrane vesicles. Rat Mrp2-mediated CP-I transport was measured in inside-out membrane vesicles. Mrp2 vesicle protein (50 µg/well) were preincubated with modulators at above reported concentrations at 37 ◦C and reaction was started by adding CP-I (5 µM) and incubated for 15 min. NP; not performed. For more on experimental details refer to Section 2.3. Dotted line represents basal level (100%) of CP-I uptake in control wells in the presence of ATP with no modulators added. All data values presented as mean and SD of single experiment performed in triplicate wells. In the inset, same data set is represented in curve to show the bell shape of the concentration-dependent stimulation and inhibition (also referred as modulation in the text). 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results Pharmaceutics 2018, 10, x FOR PEER REVIEW 7 of 15 Figure 2. Effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2 membrane vesicles. Rat Mrp2-mediated CP-I transport was measured in inside-out membrane vesicles. Mrp2 vesicle protein (50 µg/well) were preincubated with modulators at above reported concentrations at 37 °C and reaction was started by adding CP-I (5 µM) and incubated for 15 min. NP; not performed. For more on experimental details refer to Section 2.3. Dotted line represents basal level (100%) of CP-I uptake in control wells in the presence of ATP with no modulators added. All data values presented as mean and SD of single experiment performed in triplicate wells. In the inset, same data set is represented in curve to show the bell shape of the concentration-dependent stimulation and inhibition (also referred as modulation in the text). Pharmaceutics 2018, 10, x FOR PEER REVIEW Figure 2. Effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2 membrane vesicles. Rat Mrp2-mediated CP-I transport was measured in inside-out membrane vesicles. Mrp2 vesicle protein (50 µg/well) were preincubated with modulators at above reported concentrations at 37 °C and reaction was started by adding CP-I (5 µM) and incubated for 15 min. NP; not performed. For more on experimental details refer to Section 2.3. Dotted line represents basal level (100%) of CP-I uptake in control wells in the presence of ATP with no modulators added. All data values presented as mean and SD of single experiment performed in triplicate wells. In the inset, same data set is represented in curve to show the bell shape of the concentration-dependent stimulation and inhibition (also referred as modulation in the text). Figure 2. Effect of various modulators (mitoxantrone, progesterone, testosterone, and verapamil) on the uptake of CP-I in rat Mrp2 membrane vesicles. Rat Mrp2-mediated CP-I transport was measured in inside-out membrane vesicles. Mrp2 vesicle protein (50 µg/well) were preincubated with modulators at above reported concentrations at 37 ◦C and reaction was started by adding CP-I (5 µM) and incubated for 15 min. NP; not performed. For more on experimental details refer to Section 2.3. Dotted line represents basal level (100%) of CP-I uptake in control wells in the presence of ATP with no modulators added. All data values presented as mean and SD of single experiment performed in triplicate wells. 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). and modulators for 3 min at 37 ◦C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 ◦C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 ◦C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. 3.1. Mrp2-Mediated CP-I Transport in Membrane Vesicles 3. Results All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 °C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). Figure 3. Effect of bilirubin on the transport of CP-I using rat Mrp2 vesicles in the presence and absence of modulators. Rat Mrp2 vesicles (50 µg/well) were preincubated with and without bilirubin (10 µM) and modulators for 3 min at 37 ◦C. % Net decrease in ATP-dependent CP-I uptake in presence of bilirubin was calculated by taking CP-I uptake values in presence of modulators as control. All data values presented as mean of single experiment performed in duplicate wells. The duplicate values are: 54%, and 40% (mitoxantrone 100 µM), 34% and 42% (mitoxantrone 200 µM), 17% and 12% (progesterone 100 µM), and 21, and 21% (verapamil 100 µM). p µg p (10 µM) and modulators for 3 min at 37 °C. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To understand whether the observed in vitro effect of modulators on transport of CP-I is translated in in vivo systems, studies were conducted using bile duct cannulated rats. In this model, the modulators were administered as an infusion through the jugular vein. Plasma and bile levels of I d III d d ff l d h f l d To understand whether the observed in vitro effect of modulators on transport of CP-I is translated in in vivo systems, studies were conducted using bile duct cannulated rats. In this model, the modulators were administered as an infusion through the jugular vein. Plasma and bile levels of CP-I and CP-III were measured at different time intervals during 2 h (for verapamil and To understand whether the observed in vitro effect of modulators on transport of CP-I is translated in in vivo systems, studies were conducted using bile duct cannulated rats. In this model, the modulators were administered as an infusion through the jugular vein. Plasma and bile levels of 8 of 15 Pharmaceutics 2018, 10, 125 CP-I and CP-III were measured at different time intervals during 2 h (for verapamil and progesterone) and 3 h (for mitoxantrone) pre- and post-dose treatment with modulators. The basal levels of CP-I and CP-III excreted in 2 h were found to be about 600 and 1200 pmoles in rat bile, respectively. The plasma AUCs (0–2 h) were found to be about 1.2 and 9.5 nM.h for CP-I and CP-III, respectively. Figure 4 shows the amount CP-I and -III excreted in bile, in the presence and absence of the modulators at different time durations. As can be observed from Figure 4, there was a signiﬁcant increase in biliary excretion of CP-I and III between 0.25 and 0.5 h following dosing of mitoxantrone 6 mg/kg (Figure 4A1,D1). However, the same changes were not observed with 15 mg/kg mitoxantrone (Figure 4A2,D2). At the higher dose of mitoxantrone (15 mg/kg), the biliary secretion of CP-I was decreased from 0.33 to 1.25 h. In contrast, with administration of progesterone at 1 mg/kg, a signiﬁcant decrease in biliary secretion of CP-III from 0.17 to 1.5 h was observed. A similar decrease of CP-I biliary clearance was also observed from 0 to 0.75 h, following 10 mg/kg verapamil treatment. Pharmaceutics 2018, 10, x FOR PEER REVIEW 8 of 15 progesterone) and 3 h (for mitoxantrone) pre- and post-dose treatment with modulators. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 9 of 15 9 of 15 Pharmaceutics 2018, 10, 125 Ph ti 2018 10 FO harmaceutics 2018, 10, x FOR PEER REVIEW 9 of (C) (F) Figure 4. Amount of CP-I and CP-III excreted in bile at diﬀerent time intervals in the presence (□) and absence (■) of modulators. Biliary excretion profile of endogenous CP-I and CP-III in male Sprague Dawley (SD) rats following intravenous administration of various modulators. All animals received saline (0.9% NaCl) first then modulator after collection of bile at designated time points. (A1,D1,A2,D2); represent biliary excretion profile of CP-I and CP-III in male SD rats treated with mitoxantrone at 6 and 15 mg/kg, respectively. (B,E) represents biliary excretion profile of CP-I and CP-III in male SD rats treated with progesterone at 1 mg/kg. (C,F) represents biliary excretion profile of CP-I and CP-III in male SD rats treated with verapamil at 10 mg/kg. All the data values presented in mean and SD of single experiment performed in three animals. 0 100 200 Vehicle Verapamil 10 mg/kg 0 200 400 600 Vehicle Verapamil 10 mg/kg Time interval Figure 4. Amount of CP-I and CP-III excreted in bile at different time intervals in the presence (□) and absence (■) of modulators. Biliary excretion proﬁle of endogenous CP-I and CP-III in male Sprague Dawley (SD) rats following intravenous administration of various modulators. All animals received saline (0.9% NaCl) ﬁrst then modulator after collection of bile at designated time points. (A1,D1,A2,D2); represent biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with mitoxantrone at 6 and 15 mg/kg, respectively. (B,E) represents biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with progesterone at 1 mg/kg. (C,F) represents biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with verapamil at 10 mg/kg. All the data values presented in mean and SD of single experiment performed in three animals. (C) (F) 0 100 200 Vehicle Verapamil 10 mg/kg 0 200 400 600 Vehicle Verapamil 10 mg/kg (C) (F) Time interval Figure 4. Amount of CP-I and CP-III excreted in bile at diﬀerent time intervals in the presence (□) and absence (■) of modulators. Biliary excretion profile of endogenous CP-I and CP-III in male Sprague Dawley (SD) rats following intravenous administration of various modulators. All animals received saline (0.9% NaCl) first then modulator after collection of bile at designated time points. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats All animals received saline (0.9% NaCl) ﬁrst then modulator after collection of bile at designated time points. (A1,D1,A2,D2); represent biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with mitoxantrone at 6 and 15 mg/kg, respectively. (B,E) represents biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with progesterone at 1 mg/kg. (C,F) represents biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with verapamil at 10 mg/kg. All the data values presented in mean and SD of single experiment performed in three animals. The biliary (Table 1) and renal clearances (Supplementary Table S2) over the 2 or 3 h time periods, with and without the modulators treatment, were calculated as described in materials and methods. The average biliary clearance of CP-I is four-fold higher than CP-III in the control experiments, while urinary clearance of CP-III is three-fold higher than CP-I. No significant increase in biliary or urinary clearance of CP-I or CP-III is observed at pre- and post-dosing with modulators (p > 0.05). Supplementary Table S1 lists the cumulative amount of CP-I and CP-III excreted in bile, in the 2 or 3 h time interval. The cumulative amount of CP-I excreted in bile and the plasma AUC of CP- I were found to be similar in the presence and absence of all the three modulators. A similar trend was also observed in CP-III levels. The biliary (Table 1) and renal clearances (Supplementary Table S2) over the 2 or 3 h time periods, with and without the modulators treatment, were calculated as described in materials and methods. The average biliary clearance of CP-I is four-fold higher than CP-III in the control experiments, while urinary clearance of CP-III is three-fold higher than CP-I. No signiﬁcant increase in biliary or urinary clearance of CP-I or CP-III is observed at pre- and post-dosing with modulators (p > 0.05). Supplementary Table S1 lists the cumulative amount of CP-I and CP-III excreted in bile, in the 2 or 3 h time interval. The cumulative amount of CP-I excreted in bile and the plasma AUC of CP-I were found to be similar in the presence and absence of all the three modulators. A similar trend was also observed in CP-III levels. Table 1. Biliary clearances coproporphyrin-I (CP-I) and CP-III in male Sprague Dawley (SD) rats treated with vehicle or modulator. Table 1. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats The basal levels of CP-I and CP-III excreted in 2 h were found to be about 600 and 1200 pmoles in rat bile, respectively. The plasma AUCs (0–2 h) were found to be about 1.2 and 9.5 nM.h for CP-I and CP-III, respectively. Figure 4 shows the amount CP-I and -III excreted in bile, in the presence and absence of the modulators at different time durations. As can be observed from Figure 4, there was a significant increase in biliary excretion of CP-I and III between 0.25 and 0.5 h following dosing of mitoxantrone 6 mg/kg (Figure 4A1,D1). However, the same changes were not observed with 15 mg/kg mitoxantrone (Figure 4A2,D2). At the higher dose of mitoxantrone (15 mg/kg), the biliary secretion of CP-I was decreased from 0.33 to 1.25 h. In contrast, with administration of progesterone at 1 mg/kg, a significant decrease in biliary secretion of CP-III from 0.17 to 1.5 h was observed. A similar decrease of CP-I biliary clearance was also observed from 0 to 0.75 h, following 10 mg/kg verapamil treatment. (A1) (D1) (A2) (D2) (B) (E) 0 50 100 150 200 Vehicle Mitoxantrone 6 mg/kg Coproporphyrin-I 0 200 400 600 Vehicle Mitoxantrone 6 mg/kg Coproporphyrin-III 0 50 100 150 200 Vehicle Mitoxantrone 15 mg/kg 0 200 400 600 Vehicle Mitoxantrone 15 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg Figure 4. Cont. (A1) (D1) 0 50 100 150 200 Vehicle Mitoxantrone 6 mg/kg Coproporphyrin-I 0 200 400 600 Vehicle Mitoxantrone 6 mg/kg Coproporphyrin-III Coproporphyrin-III (A1) (D1) (A2) (D2) (B) (E) 0 50 100 150 0 200 400 Mitoxantrone 6 mg/kg 0 50 100 150 200 Vehicle Mitoxantrone 15 mg/kg 0 200 400 600 Vehicle Mitoxantrone 15 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg Figure 4. Cont. (D1) (A1) (D1) (A2) (D2) 0 50 100 150 200 Vehicle Mitoxantrone 15 mg/kg 0 200 400 600 Vehicle Mitoxantrone 15 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg (A2) 0 50 100 150 200 Vehicle Mitoxantrone 15 mg/kg Mitoxantrone 15 mg/kg (A2) (D2) (B) (E) 0 200 400 600 Vehicle Progesterone 1 mg/kg 0 200 400 600 Vehicle Progesterone 1 mg/kg Figure 4. Cont. (B) 0 200 400 600 Vehicle Progesterone 1 mg/kg (B) (E) Figure 4. Cont. The total and free liver concentrations of the stimulators followed by i.v infusion w #: All data values are presented in Mean ± SD, collected from three animal. For experimental design refer to Section 2.4. Saline (0.9% NaCl) was used as vehicle. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats (A1,D1,A2,D2); represent biliary excretion profile of CP-I and CP-III in male SD rats treated with mitoxantrone at 6 and 15 mg/kg, respectively. (B,E) represents biliary excretion profile of CP-I and CP-III in male SD rats treated with progesterone at 1 mg/kg. (C,F) represents biliary excretion profile of CP-I and CP-III in male SD rats treated with verapamil at 10 mg/kg. All the data values presented in mean and SD of single experiment performed in three animals. Figure 4. Amount of CP-I and CP-III excreted in bile at different time intervals in the presence (□) and absence (■) of modulators. Biliary excretion proﬁle of endogenous CP-I and CP-III in male Sprague Dawley (SD) rats following intravenous administration of various modulators. All animals received saline (0.9% NaCl) ﬁrst then modulator after collection of bile at designated time points. (A1,D1,A2,D2); represent biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with mitoxantrone at 6 and 15 mg/kg, respectively. (B,E) represents biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with progesterone at 1 mg/kg. (C,F) represents biliary excretion proﬁle of CP-I and CP-III in male SD rats treated with verapamil at 10 mg/kg. All the data values presented in mean and SD of single experiment performed in three animals. Figure 4. Amount of CP-I and CP-III excreted in bile at diﬀerent time intervals in the presence (□) and absence (■) of modulators. Biliary excretion profile of endogenous CP-I and CP-III in male Sprague Dawley (SD) rats following intravenous administration of various modulators. All animals received saline (0.9% NaCl) first then modulator after collection of bile at designated time points. (A1,D1,A2,D2); represent biliary excretion profile of CP-I and CP-III in male SD rats treated with mitoxantrone at 6 and 15 mg/kg, respectively. (B,E) represents biliary excretion profile of CP-I and CP-III in male SD rats treated with progesterone at 1 mg/kg. (C,F) represents biliary excretion profile of CP-I and CP-III in male SD rats treated with verapamil at 10 mg/kg. All the data values presented in mean and SD of single experiment performed in three animals. Figure 4. Amount of CP-I and CP-III excreted in bile at different time intervals in the presence (□) and absence (■) of modulators. Biliary excretion proﬁle of endogenous CP-I and CP-III in male Sprague Dawley (SD) rats following intravenous administration of various modulators. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats Biliary clearances coproporphyrin-I (CP-I) and CP-III in male Sprague Dawley (SD) rats treated with vehicle or modulator. Table 1. Biliary clearances coproporphyrin-I (CP-I) and CP-III in male Sprague Dawley (SD) rats treated with vehicle or modulator. Table 1. Biliary clearances coproporphyrin-I (CP-I) and CP-III in male Sprague Dawley (SD) rats treated with vehicle or modulator. treated with vehicle or modulator. Dose Group Modulator Dose (mg/kg, i.v) Biliary Clearance # (mL/min) CP-I CP-III Vehicle - 7.6 ± 1.5 2.3 ± 1.3 Mitoxantrone 6 5.9 ± 0.3 1.9 ± 1.7 Vehicle - 11 ± 0.1 3.3 ± 0.03 Mitoxantrone 15 16 ± 9.5 3.4 ± 2.1 Vehicle - 14 ± 6.0 2.3 ± 0.2 Progesterone 1 11 ± 4.6 2.3 ± 0.3 Vehicle - 5.8 ± 2.1 1.6 ± 0.5 Verapamil 10 3.8 ± 0.8 1.2 ± 0.6 #: All data values are presented in Mean ± SD, collected from three animal. For experimental design refer to Section 2.4. Saline (0.9% NaCl) was used as vehicle. The total and free liver concentrations of the stimulators followed by i.v infusion w i d Th li i f i ( 15 /k ) d with vehicle or modulator. Dose Group Modulator Dose (mg/kg, i.v) Biliary Clearance # (mL/min) CP-I CP-III Vehicle - 7.6 ± 1.5 2.3 ± 1.3 Mitoxantrone 6 5.9 ± 0.3 1.9 ± 1.7 Vehicle - 11 ± 0.1 3.3 ± 0.03 Mitoxantrone 15 16 ± 9.5 3.4 ± 2.1 Vehicle - 14 ± 6.0 2.3 ± 0.2 Progesterone 1 11 ± 4.6 2.3 ± 0.3 Vehicle - 5.8 ± 2.1 1.6 ± 0.5 Verapamil 10 3.8 ± 0.8 1.2 ± 0.6 #: All data values are presented in Mean ± SD, collected from three animal. For experimental design refer to Section 2.4. Saline (0.9% NaCl) was used as vehicle. 10 of 15 Pharmaceutics 2018, 10, 125 The total and free liver concentrations of the stimulators followed by i.v infusion were determined. The liver concentrations of mitoxantrone (at 15 mg/kg), progesterone, and verapamil were 950, 197, and 1040 µM, respectively, resulting in free liver concentration (Cmax, unbound, liver) of mitoxantrone, progesterone, and verapamil 0.28, 1.86, and 25.8 µM, respectively (Table 2). Table 2. Liver unbound concentrations of modulators. 4. Discussion In vitro stimulation of drug metabolizing enzymes and transporters has long been recognized. However, in vivo or clinical impact of stimulation of drug metabolizing enzymes or transporters are scarce [20,21]. Therefore, the translation of in vitro stimulation to in vivo has not been fully understood. For example, a two-fold stimulation of ﬂurbiprofen metabolism in liver microsome in presence of dapsone yields only 10% increase in ﬂubiprofen clearance in humans, which appears not to be clinically relevant [22,23]. In vitro stimulation of MRP2-mediated transport of different probe substrates, in the presence of a chemically diverse set of compounds, have been widely reported [10,13,14,24]. Previously, we reported that 53% compound of a list of 97 compounds stimulated MRP2-mediated E17βG transport [10]. We found that the stimulation is substrate dependent, as lesser percentage (17%) of compounds stimulated CP-I transport [10]. So far, two previous reports investigated in vivo stimulation of Mrp2- mediated probe substrate transport by measuring the probe compounds, E17βG and GSH, respectively, in rat bile, following treatment with stimulators [16,17]. However, as previously explained, MRP2/Mrp2-mediated E17βG transport is known to be stimulated by many compounds with diverse chemical motifs. We have previously shown that the stimulators used in the study of Heredi-Szabo et al., benzbromarone and indomethacin, are actually inhibitors of MRP2, when CP-I is used as a probe substrate [10]. Therefore, the stimulation effect was re-investigated with CP-I as endogenous biomarker. CP-I, being endogenous, provided us with the advantage that external administration of a radioactive/labelled probe compound could be avoided. In agreement with the results found in human MRP2 vesicles, rat Mrp2 mediated CP-I transport also displayed conventional Michaelis-Menten kinetics, with Km of 15 µM. The Km values of CP-I in rat Mrp2 vesicles were comparable to that in human (15 vs. 7.7 µM for rat and human MRP2, respectively), suggesting the interspecies differences in afﬁnity of CP-I towards MRP2/Mrp2-meditated biliary secretion is minimum. This is in contrast to E17βG, as signiﬁcant differences in kinetics of E17βG are reported between different species, such as Michaelis-Menten reported for rat, while sigmoidal kinetics reported for human [14,16]. Four stimulators, mitoxantrone, verapamil, testosterone, and progesterone were selected as tool compounds for the rat in vitro and in vivo investigations, based on their signiﬁcant stimulation of CP-I transport observed in human MRP2 vesicles and the structural features [10]. 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats To u de ta d hethe the ob e ed i it o effe t of odulato 3.2. Plasma Level and Biliary Secretion of Endogenous CP-I and CP-III in Rats Dose Group Modulator Dose (mg/kg, i.v) In Vivo Parameters # In Vitro Parameter Liver Unbound Concentrations $ (µM) Liver to Plasma Ratio Cmax (µM) Liver Total Concentrations (µM) Fu,liver Mitoxantrone 6 199 4.3 854 0.0003 0.26 Mitoxantrone 15 202 4.7 950 0.28 Progesterone 1 1.2 165 197 0.0094 1.86 Verapamil 10 63 16.6 1044 0.0247 25.8 #: Livers were collected at the terminal time point (2 or 3 h) of the respective study, and modulator concentrations were determined in liver homogenate by LC-MS/MS method, refer to Section 2.5 for more details. Fu,liver: modulator free fraction in liver was determined using rapid equilibrium dialysis method, using blank liver homogenate as matrices. $: Liver unbound concentrations were determined by multiplying Fu,liver with total liver concentrations. Table 2. Liver unbound concentrations of modulators. #: Livers were collected at the terminal time point (2 or 3 h) of the respective study, and modulator concentrations were determined in liver homogenate by LC-MS/MS method, refer to Section 2.5 for more details. Fu,liver: modulator free fraction in liver was determined using rapid equilibrium dialysis method, using blank liver homogenate as matrices. $: Liver unbound concentrations were determined by multiplying Fu,liver with total liver concentrations. 4. Discussion All four compounds (structures in supplementary Figure S2) stimulated CP-I transport in membrane vesicles overexpressing rat Mrp2, with mitoxantrone being the strongest stimulator. Mitoxantrone showed a bell-shaped curve: stimulation of Mrp2 mediated CP-I transport at lower concentrations (up to 250 µM) followed by inhibition at higher (>250 µM) concentrations of mitoxantrone (Figure 2). Therefore, two doses of mitoxantrone (6 mg/kg and 15 mg/kg) were administered in bile duct cannulated rats, to assess the impact on CP-I biliary clearance. Mitoxantrone, verapamil (6 mg/kg), and progesterone 11 of 15 Pharmaceutics 2018, 10, 125 (1 mg/kg) were investigated by a cross-over study design, so that the drug-pretreatment CP-I and CP-III biliary clearance values can be used as controls to compare with post drug-treatment biliary clearances. Collection of bile was initiated at early time points, as previous reports suggest that the stimulation of Mrp2 mediated transported appeared to occur within the ﬁrst hour of dosing of indomethacin or benzylpenicillin [16,17]. Contrary to in vitro ﬁndings, no signiﬁcant difference in either biliary clearances or cumulative amount of CP-I or CP-III eliminated in bile were detected before and after treatment of mitoxantrone (Table 1 and Figure 4, Supplementary Table S1). Similarly, neither biliary clearance nor cumulative amount of CP-I excreted of in bile were increased, when dosed with progesterone or verapamil. Although there was a signiﬁcant increase in CP-I and CP-III levels in bile between 15–30 min, following the 6 mg/kg mitoxantrone dose (Figure 3), the same effect was not observed at the higher dose, nor at following or earlier time intervals in the same dose, suggesting that could be an experimental error. To know if the lack of in vivo translatability of in vitro stimulation is because the compounds are not attaining sufﬁcient concentrations in the liver, free liver concentrations of mitoxantrone, progesterone, and verapamil were determined. In vitro, 1 µM progesterone and 25 µM verapamil stimulated the Mrp2 mediated transport of CP-I to >1.5-fold and >1.3-fold, respectively (Figure 2). The free liver exposures obtained following dosing progesterone and verapamil are 1.86 and 25.8 µM respectively (Table 2). Hence, we can conﬁdently conclude that even after reaching relevant concentrations in liver, progesterone and verapamil did not stimulate Mrp2-mediated CP-I transport in rats. Mitoxantrone dosing was limited by its observed toxicity in rats. Therefore, we could not reach free liver concentration beyond 0.28 µM, while in vitro stimulations started from 1 µM onwards. 4. Discussion In vitro, in presence of vesicles, the “free” concentration of the compound is expected to be lower than 1 µM, because of compounds binding to vesicles. Thus, ‘vesicle binding’ will further lower in vitro concentration and we can comfortably assume that “free in vitro concentration” will be in similar range as of “free in vivo concentration”. Furthermore, apart from direct measurement of CP-I in bile, there are two indirect evidences of not observing any in vivo stimulation of Mrp2 function by mitoxantrone, progesterone, and verapamil. First, the bile ﬂow is not enhanced following compound treatment. Glutathione excretion in bile via Mrp2 mediates bile salt-independent bile ﬂow [25,26]. We assessed the bile ﬂow before and after treatment with respective stimulators and did not observe any increase in bile ﬂow following treatment with the stimulators (Supplementary Table S3). The second indirect proof comes from urinary coproporphyrin ratio (UCP-I/UCP-(I + III)), which is often used as a surrogate marker of MRP2 function and has been used to assess methotrexate PK variability [27]. The urinary coproporphyrin ratio did not decrease following dosing the stimulators (Supplementary Table S4. Collectively, we concluded that the in vitro stimulation of MRP2/Mrp2-mediated CP-I transport is not translatable in vivo. Plasma and urinary coproporphyrins (CP-I and III) have been suggested as in vivo and clinical biomarkers for organic anion transporting polypeptide (OATP)-mediated drug interactions [28–30]. The plasma and urinary levels of CPs are shown to be sensitive markers of Oatp-mediated DDI in cynomolgus monkeys and mice [30]. CP-I uptake in isolated rat hepatocytes was shown to be rifamycin sensitive (Supplementary Figure S1), suggesting a similar role of Oatp transporters in CP dispositions among mice, rats, monkeys, and humans [30]. This opens the possibility of using rats as an investigational model to understand CP-I as an in vivo marker for Oatp inhibition. In addition to hepatocyte canalicular membranes, Mrp2 is also expressed on the apical side of renal proximal tubules. Thus, one can argue that functional activity of Mrp2 can be monitored via urinary levels of CPs. However, other MRP/Mrp isoforms on the apical side of proximal tubule cells (such as Mrp4), with overlapping substrate speciﬁcity, can interfere with Mrp2 function estimation via renal secretion of CPs. Therefore, urinary elimination of CPs is not assessed to understand the in vivo impact of stimulation. Author Contributions: Conceptualization: S.C., Y.L., R.R.G., T.T.M., H.S., V.K.; Methodology: R.R.G., V.K., B.V.M., S.C.; Resources: R.R.G., V.K., B.V.M, S.C.; Data Curation: S.C., R.R.G., T.T.M., V.K., B.V.M.; Writing-original draft preparation: S.C., Y.L., R.R.G., V.K., B.V.M., T.T.M.; Writing-review and editing: S.C., Y.L., H.S.; Footnote: R.R.G. and V.K. equally contributed to the work. 4. Discussion In the present investigation, we found the basal plasma levels of CP-III are three-four times higher than CP-I levels, which is similar to what is reported in mice, and is in 12 of 15 Pharmaceutics 2018, 10, 125 contrast to monkeys and humans, where CP-I levels are higher in plasma ([30] and Supplementary Table S1). In addition, the biliary clearance of CP-I is about four-fold higher than CP-III, while the urinary clearance of CP-III is three-fold higher than CP-I. This suggests there can be preferential afﬁnity of CP-I towards biliary transporters while CP-III towards urinary transporters. Higher renal clearance of CP-III, compared to CP-I, is observed for monkeys as well [30]. To further investigate the possible mechanism for the in vitro-in vivo disconnect of stimulation of CP-I transport via Mrp2, the effect of endogenous compounds such as bile acids and bilirubin on CP-I transport were evaluated. Bile acids did not decrease stimulation by the stimulators (data not shown), while 10 µM bilirubin suppressed the stimulation of Mrp2-mediated CP-I transport to about 50% and 30% of what is observed with 100 and 200 µM mitoxantrone, respectively. To a lesser extent, bilirubin also decreased the stimulation of CP-I transport by 100 µM progesterone and verapamil. Due to the unavailability of bilirubin glucuronides, we were not be able to further assess its effect on the stimulation of Mrp2-mediated CP-I transport by these stimulators. This is the ﬁrst time that the effect of bilirubin on the in vitro stimulation of Mrp2-mediated transport is reported, to the best of our knowledge. Thus, we hypothesize that endogenous compounds may act on the postulated modulator or transport site of MRP2/Mrp2 and contribute to the in vitro-in vivo discrepancy. This can add to the previous assumptions that the differences in membrane composition between the in vitro and in vivo systems [31] may lead to perceived in vitro to in vivo disconnect in transporter function. The possibility to add bilirubin in vitro to minimize the stimulation effects on MRP2-mediated transport could be further explored to improve the in vitro tools in assessing MRP2 inhibition. Further investigation with altering the in vitro assay can provide conﬁrmatory data in this regard. Funding: This research received no external funding. References 1. Hillgren, K.; Keppler, D.; Zur, A.; Giacomini, K.; Stieger, B.; Cass, C.; Zhang, L. Emerging transporters of clinical importance: An update from the International Transporter Consortium. Clin. Pharmacol. Ther. 2013, 94, 52–63. [CrossRef] [PubMed] 1. Hillgren, K.; Keppler, D.; Zur, A.; Giacomini, K.; Stieger, B.; Cass, C.; Zhang, L. Emerging transporters of clinical importance: An update from the International Transporter Consortium. Clin. Pharmacol. Ther. 2013, 94, 52–63. [CrossRef] [PubMed] 2. Lai, Y.; Hsiao, P. Beyond the ITC White Paper: Emerging sciences in drug transporters and opportunities for drug development. Curr. Pharm. Des. 2014, 20, 1577–1594. [CrossRef] [PubMed] 2. Lai, Y.; Hsiao, P. Beyond the ITC White Paper: Emerging sciences in drug transporters and opportunities for drug development. Curr. Pharm. Des. 2014, 20, 1577–1594. [CrossRef] [PubMed] 3. Keppler, D.; Konig, J.; Buchler, M. The canalicular multidrug resistance protein, cMRP/MRP2, a novel conjugate export pump expressed in the apical membrane of hepatocytes. Adv. Enzym. Regul. 1997, 37, 321–333. [CrossRef] 4. Keppler, D.; Leier, I.; Jedlitschky, G. Transport of glutathione conjugates and glucuronides by the multidrug resistance proteins MRP1 and MRP2. Biol. Chem. 1997, 378, 787–791. [PubMed] 4. Keppler, D.; Leier, I.; Jedlitschky, G. Transport of glutathione conjugates and glucuronides by the multidrug resistance proteins MRP1 and MRP2. Biol. Chem. 1997, 378, 787–791. [PubMed] 5. Borst, P.; Evers, R.; Kool, M.; Wijnholds, J. A family of drug transporters: The multidrug resistance-associated proteins. J. Natl. Cancer Inst. 2000, 92, 1295–1302. [CrossRef] [PubMed] 5. Borst, P.; Evers, R.; Kool, M.; Wijnholds, J. A family of drug transporters: The multidrug resistance-associated proteins. J. Natl. Cancer Inst. 2000, 92, 1295–1302. [CrossRef] [PubMed] 6. Yoshida, K.; Maeda, K.; Sugiyama, Y. Hepatic and intestinal drug transporters: Prediction of pharmacokinetic effects caused by drug-drug interactions and genetic polymorphisms. Annu. Rev. Pharmacol. Toxicol. 2013, 53, 581–612. [CrossRef] [PubMed] 6. Yoshida, K.; Maeda, K.; Sugiyama, Y. Hepatic and intestinal drug transporters: Prediction of pharmacokinetic effects caused by drug-drug interactions and genetic polymorphisms. Annu. Rev. Pharmacol. Toxicol. 2013, 53, 581–612. [CrossRef] [PubMed] 7. Rau, T.; Erney, B.; Göres, R.; Eschenhagen, T.; Beck, J.; Langer, T. High-dose methotrexate in pediatric acute lymphoblastic leukemia: Impact of ABCC2 polymorphisms on plasma concentrations. Clin. Pharmacol. Ther. 2006, 80, 468–476. [CrossRef] [PubMed] 8. Toshimoto, K.; Tomoda, Y.; Chiba, K.; Sugiyama, Y. 5. Conclusions In this work, we ﬁrst showed that CP-I follows a Michaelis-menten kinetics in rat Mrp2 vesicles. Few compounds (verapamil, mitoxantrone, progesterone, and testosterone) which stimulated MRP2-mediated CP-I transport in human MRP2 vesicles also stimulated rat Mrp2-mediated CP-I transport in vitro. In order to understand the in vivo signiﬁcance of the ‘stimulation’, bile-duct cannulated rats were dosed with mitoxantrone, verapamil, and progesterone, and CP-I and CP-III levels were measured in bile and urine. Different doses of mitoxantrone, verapamil, or progesterone do not impact Mrp2-mediated CP-I and CP-III excretion in rat bile. This was contradictory to the in vitro data, where mitoxantrone, verapamil, and progesterone stimulated CP-I transport mediated by Mrp2. Our ﬁndings support non-existence of any clinical event that can be attributed to stimulation of MRP2-mediated transport. In addition, we also provide evidence in favor of the hypotheses that the presence of an endogenous compound such as bilirubin may lead to the in vitro and in vivo discrepancy. Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4923/10/3/125/s1, Figure S1: CP-I uptake in presence and absence of 1 mM rifamycin SV in suspended rat hepatocytes; Figure S2: Structures of the modulators (mitoxantrone, progesterone, testosterone and verapamil) used in this study, along with conventional probe substrate of MRP/Mrp2, estradiol b-glucuronide.; Table S1: Cumulative amount of excretion and area under plasma concentration-time (AUC) proﬁle of CP-I and CP-III following i.v injection of vehicle or modulator.; Table S2: Renal clearance of CP-I and CP-III following i.v injection of vehicle or modulator.; Table S3: Bile ﬂow rate (ml/h) before and after treatment with modulators. The cumulative volume is divided by the total time span of bile collection, to obtain bile ﬂow rate; Table S4: Urinary coproporphyrin (UCP-I/(UCP-I+UCP-III)) ratios with vehicle and compound dosing. Author Contributions: Conceptualization: S.C., Y.L., R.R.G., T.T.M., H.S., V.K.; Methodology: R.R.G., V.K., B.V.M., S.C.; Resources: R.R.G., V.K., B.V.M, S.C.; Data Curation: S.C., R.R.G., T.T.M., V.K., B.V.M.; Writing-original draft preparation: S.C., Y.L., R.R.G., V.K., B.V.M., T.T.M.; Writing-review and editing: S.C., Y.L., H.S.; Footnote: R.R.G. and V.K. equally contributed to the work. Funding: This research received no external funding. Funding: This research received no external funding. 13 of 15 13 of 15 Pharmaceutics 2018, 10, 125 Acknowledgments: We would like to thank Punit Marathe, and Mike Sinz during various aspects of this work, e.g., data analysis, experimental set up, reviewing manuscript. 5. Conclusions We would like to thank Prabhakar Rajanna for helping bioanalysis, Anoop Kumar for providing freshly isolated rat hepatocytes, Ram Vishwanath for assisting in in vivo experiments and Harbeer Kaur for helping in data analysis. Conﬂicts of Interest: Authors declare no conﬂict of interest. S.C., R.R.G., V.K., B.V.M., T.T.M. are working in Biocon Bristol Myers-Squibb Research Center for last 3 years or more. H.S. is working in BMS, NJ for more than 5 years. Y.L. was working in BMS, NJ till 2017, currently he is working in Gilead, Foster City. Abbreviations Abbreviations E17βG estradiol-17β-glucuronide MRP2 multidrug-resistance associated protein 2 OATP/Oatp organic anion transporting polypeptide CP-I coproporphyrin-I CP-III coproporphyrin-III LC-MS/MS, liquid chromatography–tandem mass spectrometry ATP adenosine tri-phosphate AMP adenosine mono-phosphate ABC ATP binding cassette GSH glutathione CA Cholic acid GCA Glycocholic acid TCA taurocholic acid DDI drug-drug interactions References Burden, N.; Kendrick, J.; Knight, L.; McGregor, V.; Murphy, H.; Punler, M.; van Wijk, H. Maximizing the success of bile duct cannulation studies in rats: Recommendations for best practice. Lab. Anim. 2017, 51, 457–464. [CrossRef] [PubMed] 19. Riccardi, K.; Cawley, S.; Yates, P.D.; Chang, C.; Funk, C.; Niosi, M.; Lin, J.; Di, L. Plasma protein binding of challenging compounds. J. Pharm. Sci. 2015, 104, 2627–2636. [CrossRef] [PubMed] 20. Hutzler, J.M.; Tracy, T.S. Atypical kinetic proﬁles in drug metabolism reactions. Drug Metab. Dispos. 2002, 30, 355–362. [CrossRef] [PubMed] 21. Van Waterschoot, R.A.; Rooswinkel, R.W.; Sparidans, R.W.; van Herwaarden, A.E.; Beijnen, J.H.; Schinkel, A.H. Inhibition and stimulation of intestinal and hepatic CYP3A activity: Studies in humanized CYP3A4 transgenic mice using triazolam. Drug Metab. Dispos. 2009, 37, 2305–2313. [CrossRef] [PubMed] 22. Hutzler, J.M.; Frye, R.F.; Korzekwa, K.R.; Branch, R.A.; Huang, S.-M.; Tracy, T.S. Minimal in vivo activation of CYP2C9-mediated ﬂurbiprofen metabolism by dapsone. Eur. J. Pharm. Sci. 2001, 14, 47–52. [CrossRef] 23. Korzekwa, K.; Krishnamachary, N.; Shou, M.; Ogai, A.; Parise, R.; Rettie, A.; Gonzalez, F.; Tracy, T. Evaluation of atypical cytochrome P450 kinetics with two-substrate models: Evidence that multiple substrates can simultaneously bind to cytochrome P450 active sites. Biochemistry 1998, 37, 4137–4147. [CrossRef] [PubMed] 24. Ellis, L.C.; Hawksworth, G.M.; Weaver, R.J. ATP-dependent transport of statins by human and rat MRP2/Mrp2 Toxicol Appl Pharmacol 2013 269 187 194 [CrossRef] [PubMed] 23. Korzekwa, K.; Krishnamachary, N.; Shou, M.; Ogai, A.; Parise, R.; Rettie, A.; Gonzalez, F.; Tracy, T. Evaluation of atypical cytochrome P450 kinetics with two-substrate models: Evidence that multiple substrates can simultaneously bind to cytochrome P450 active sites. Biochemistry 1998, 37, 4137–4147. [CrossRef] [PubMed] yp y p simultaneously bind to cytochrome P450 active sites. Biochemistry 1998, 37, 4137–4147. [CrossRef] [PubMed] 24. Ellis, L.C.; Hawksworth, G.M.; Weaver, R.J. ATP-dependent transport of statins by human and rat MRP2/Mrp2. Toxicol. Appl. Pharmacol. 2013, 269, 187–194. [CrossRef] [PubMed] 24. Ellis, L.C.; Hawksworth, G.M.; Weaver, R.J. ATP-dependent transport of statins by human and rat MRP2/Mrp2. Toxicol. Appl. Pharmacol. 2013, 269, 187–194. [CrossRef] [PubMed] 25. Ballatori, N.; Truong, A. Glutathione as a primary osmotic driving force in hepatic bile formation. Am. J. Physiol. Gastroint. Liver Physiol. 1992, 263, G617–G624. [CrossRef] [PubMed] 26. Johnson, D.R.; Habeebu, S.S.; Klaassen, C.D. Increase in bile ﬂow and biliary excretion of glutathione-derived sulfhydryls in rats by drug-metabolizing enzyme inducers is mediated by multidrug resistance protein 2. Toxicol. Sci. 2002, 66, 16–26. [CrossRef] [PubMed] 27. References Analysis of the change in the blood concentration-time proﬁle caused by complex drug-drug interactions in the liver considering the enterohepatic circulation: Examining whether the inhibition constants for uptake, metabolism, and biliary excretion can be recovered by the analyses using physiologically based pharmacokinetic modeling. J. Pharm. Sci. 2017, 106, 2727–2738. [CrossRef] [PubMed] 9. Shin, E.; Shin, N.; Oh, J.-H.; Lee, Y.-J. High-dose metformin may increase the concentration of atorvastatin in the liver by inhibition of multidrug resistance–associated protein 2. J. Pharm. Sci. 2017, 106, 961–967. [CrossRef] [PubMed] 10. Gilibili, R.R.; Chatterjee, S.; Bagul, P.; Mosure, K.W.; Murali, B.V.; Mariappan, T.T.; Mandlekar, S.; Lai, Y. Coproporphyrin-I: A ﬂuorescent, endogenous optimal probe substrate for ABCC2 (MRP2) suitable for vesicle-based MRP2 inhibition assay. Drug Metab. Dispos. 2017, 45, 604–611. [CrossRef] [PubMed] 14 of 15 14 of 15 Pharmaceutics 2018, 10, 125 11. Kidron, H.; Wissel, G.; Manevski, N.; Hakli, M.; Ketola, R.A.; Finel, M.; Yliperttula, M.; Xhaard, H.; Urtti, A. Impact of probe compound in MRP2 vesicular transport assays. Eur. J. Pharm. Sci. 2012, 46, 100–105. [CrossRef] [PubMed] 12. Zelcer, N.; Huisman, M.T.; Reid, G.; Wielinga, P.; Breedveld, P.; Kuil, A.; Knipscheer, P.; Schellens, J.H.; Schinkel, A.H.; Borst, P. Evidence for two interacting ligand binding sites in human multidrug resistance protein 2 (ATP binding cassette C2). J. Biol. Chem. 2003, 278, 23538–23544. [CrossRef] [PubMed] 13. Zimmermann, C.; van de Wetering, K.; van de Steeg, E.; Wagenaar, E.; Vens, C.; Schinkel, A.H. Species-dependent transport and modulation properties of human and mouse multidrug resistance protein 2 (MRP2/Mrp2, ABCC2/Abcc2). Drug Metab. Dispos. 2008, 36, 631–640. [CrossRef] [PubMed] 14. Borst, P.; Zelcer, N.; Van De Wetering, K. MRP2 and 3 in health and disease. Cancer Lett. 2006, 234, 51–61. [CrossRef] [PubMed] 15. Hirono, S.; Nakagome, I.; Imai, R.; Maeda, K.; Kusuhara, H.; Sugiyama, Y. Estimation of the three-dimensional pharmacophore of ligands for rat multidrug-resistance-associated protein 2 using ligand-based drug design techniques. Pharm. Res. 2005, 22, 260–269. [CrossRef] [PubMed] 16. Heredi-Szabo, K.; Glavinas, H.; Kis, E.; Mehn, D.; Bathori, G.; Veres, Z.; Kobori, L.; von Richter, O.; Jemnitz, K.; Krajcsi, P. Multidrug resistance protein 2-mediated estradiol-17β-D-glucuronide transport potentiation: In vitro-in vivo correlation and species speciﬁcity. Drug Metab. Dispos. 2009, 37, 794–801. [CrossRef] [PubMed] 17. Ito, K.; Koresawa, T.; Nakano, K.; Horie, T. Mrp2 is involved in benzylpenicillin-induced choleresis. Am. J. Physiol. Gastroint. Liver Physiol. 2004, 287, G42–G49. [CrossRef] [PubMed] 18. © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). p p p 1. Bastiaanse, E.L.; Höld, K.M.; Van der Laarse, A. The effect of membrane cholesterol content on ion trans processes in plasma membranes. Cardiovasc. Res. 1992, 33, 272–283. [CrossRef] References Benz-de Bretagne, I.; Zahr, N.; Le Gouge, A.; Hulot, J.S.; Houillier, C.; Hoang-Xuan, K.; Gyan, E.; Lissandre, S.; Choquet, S.; Le Guellec, C. Urinary coproporphyrin I/(I + III) ratio as a surrogate for MRP2 or other transporter activities involved in methotrexate clearance. Br. J. Clin. Pharmacol. 2014, 78, 329–342. [CrossRef] [PubMed] 28. Benz-de Bretagne, I.; Respaud, R.; Vourc’h, P.; Halimi, J.-M.; Caille, A.; Hulot, J.-S.; Andres, C.R.; Le Guellec, C. Urinary elimination of coproporphyrins is dependent on ABCC2 polymorphisms and represents a potential biomarker of MRP2 activity in humans. J. Biomed. Biotechnol. 2011, 2011, 1–9. [CrossRef] [PubMed] 29. Lai, Y.; Mandlekar, S.; Shen, H.; Holenarsipur, V.K.; Langish, R.; Rajanna, P.; Murugesan, S.; Gaud, N.; Selvam, S.; Date, O. Coproporphyrins in plasma and urine can be appropriate clinical biomarkers to recapitulate drug-drug interactions mediated by organic anion transporting polypeptide inhibition. J. Pharmacol. Exp. Ther. 2016, 358, 397–404. [CrossRef] [PubMed] 15 of 15 Pharmaceutics 2018, 10, 125 30. Shen, H.; Dai, J.; Liu, T.; Cheng, Y.; Chen, W.; Freeden, C.; Zhang, Y.; Humphreys, W.G.; Marathe, P.; Lai, Y. Coproporphyrins I and III as functional markers of OATP1B activity: In vitro and in vivo evaluation in preclinical species. J. Pharmacol. Exp. Ther. 2016, 357, 382–393. [CrossRef] [PubMed] 31. Bastiaanse, E.L.; Höld, K.M.; Van der Laarse, A. The effect of membrane cholesterol content on ion transport processes in plasma membranes. Cardiovasc. Res. 1992, 33, 272–283. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). 30. Shen, H.; Dai, J.; Liu, T.; Cheng, Y.; Chen, W.; Freeden, C.; Zhang, Y.; Humphreys, W.G.; Marathe, P.; Lai, Y. Coproporphyrins I and III as functional markers of OATP1B activity: In vitro and in vivo evaluation in preclinical species. J. Pharmacol. Exp. Ther. 2016, 357, 382–393. [CrossRef] [PubMed] 30. Shen, H.; Dai, J.; Liu, T.; Cheng, Y.; Chen, W.; Freeden, C.; Zhang, Y.; Humphreys, W.G.; Marathe, P.; Lai, Y. Coproporphyrins I and III as functional markers of OATP1B activity: In vitro and in vivo evaluation in preclinical species. J. Pharmacol. Exp. Ther. 2016, 357, 382–393. [CrossRef] [PubMed] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). © 2018 by the authors. References Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W2505331358	https://europepmc.org/articles/pmc4965049?pdf=render	English	null	Proteomic Analysis of the Excretory and Secretory Proteins of Haemonchus contortus (HcESP) Binding to Goat PBMCs In Vivo Revealed Stage-Specific Binding Profiles	PloS one	2,016	cc-by	13,071	Javaid Ali Gadahi, Shuai Wang, Gao Bo, Muhammad Ehsan, RuoFeng Yan, XiaoKai Song, LiXin Xu, XiangRui Li* College of Veterinary Medicine, Nanjing Agricultural University, Nanjing, 210095, PR China a1111 * lixiangrui@njau.edu.cn * lixiangrui@njau.edu.cn Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: The project was supported by the “National Key Basic Research Program (973 program) of China” (Grant No. 2015CB150300) and by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). Abstract Haemonchus contortus is a parasitic gastrointestinal nematode, and its excretory and secretory products (HcESPs) interact extensively with the host cells. In this study, we report the interaction of proteins from HcESPs at different developmental stages to goat peripheral blood mononuclear cells (PBMCs) in vivo using liquid chromatography-tandem mass spec- trometry. A total of 407 HcESPs that interacted with goat PBMCs at different time points were identified from a H. contortus protein database using SEQUEST searches. The L4 and L5 stages of H. contortus represented a higher proportion of the identified proteins com- pared with the early and late adult stages. Both stage-specific interacting proteins and pro- teins that were common to multiple stages were identified. Forty-seven interacting proteins were shared among all stages. The gene ontology (GO) distributions of the identified goat PBMC-interacting proteins were nearly identical among all developmental stages, with high representation of binding and catalytic activity. Cellular, metabolic and single-organism pro- cesses were also annotated as major biological processes, but interestingly, more proteins were annotated as localization processes at the L5 stage than at the L4 and adult stages. Based on the clustering of homologous proteins, we improved the functional annotations of un-annotated proteins identified at different developmental stages. Some unnamed H. con- tortus ATP-binding cassette proteins, including ADP-ribosylation factor and P-glycoprotein- 9, were identified by STRING protein clustering analysis. OPEN ACCESS Citation: Gadahi JA, Wang S, Bo G, Ehsan M, Yan R, Song X, et al. (2016) Proteomic Analysis of the Excretory and Secretory Proteins of Haemonchus contortus (HcESP) Binding to Goat PBMCs In Vivo Revealed Stage-Specific Binding Profiles. PLoS ONE 11(7): e0159796. doi:10.1371/journal.pone.0159796 Editor: Shoba Ranganathan, Macquarie University, AUSTRALIA Received: October 27, 2015 Accepted: July 9, 2016 Published: July 28, 2016 Received: October 27, 2015 Accepted: July 9, 2016 Published: July 28, 2016 Copyright: © 2016 Gadahi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. RESEARCH ARTICLE Introduction Haemonchus contortus (H. contortus) is the most important abomasal nematode of small rumi- nates. H. contortus infection causes high economic losses worldwide [1, 2]. This worm pene- trates the abomasal mucosa to feed on the blood of the host, resulting in anemia and low total plasma protein [3, 4]. H. contortus is one of the most extensively used parasitic nematodes in drug discovery, vaccine development and anthelmintic resistance research [5–8]. The Competing Interests: The authors have declared that no competing interests exist. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 1 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo development of first (L1), second (L2) and third (L3) stage H. contortus larvae occurs in the fae- ces. The infective larvae (L3) are ingested by the host with herbage. After exsheathment trig- gered by pepsin and HCl in the rumen, L3 migrates to the abomasum and develops into the L4 stage, which feeds on blood, followed by final development into adults approximately 3 weeks post ingestion [9]. Each developmental stage has different motility, sensory and hormonal reg- ulation requirements, which may require rapid transcriptional changes [10]. Excretory and secretory products (ESPs) are produced and released by parasites during in vitro cultivation [11] and in vivo [12]. ESPs contain various proteins and glycoproteins whose functions include depression of host immunity and modulate the host immune system from the early stages of infection for their survival [13–15]. H. contortus excretory and secretory products (HcESPs) contain many proteins [16] that perform diverse functions such as tissue penetration and host protein degradation [14]. A 55 kDa secretory glycoprotein was identified as an immunogenic protein that causes immune modulation by inhibiting host neutrophils [17]. The purified 66 kDa adult H. contortus excre- tory/secretory (E/S) antigen inhibits monocyte function in vitro, as confirmed by decreased production of hydrogen peroxide and nitric oxide in the culture medium [11]. One HcESP pro- tein induces eosinophil and neutrophil chemotactic activity [18]. We have also demonstrated that recombinant H. contortus galectin (rHco-gal-m) is recognized by the serum of goats natu- rally infected with H. contortus and can bind and modulate the activity of goat T cells and monocytes. rHco-gal-m inhibits the expression of MHC II molecules, decreases T cell activa- tion and proliferation, induces the apoptosis of T cells and affects several signaling cascades [19]. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Introduction In vitro studies have reported that parasitic ESPs have a direct effect on cultured cells or tissues, such as inhibiting acid secretion [20] and inducing the vacuolation and detachment of HeLa cells [21, 22]. These findings indicate that ESPs have multiple functions in vivo. In a previous analysis of HcESP, approximately 193 immunogenic spots were detected by 2D gel analysis [16], and 52 proteins were identified by MS. We identified 129 male-specific, 132 female-specific and 23 shared immunogenic proteins from adult H. contortus by MALDI- TOF [23]. The presence of antibodies against many E/S proteins in infected animals strongly indicates the presence of ESPs in the circulation of infected animals [12, 23]. Other intestinal nematodes of livestock that are very closely related to H. contortus, including Cooperia spp. [24], Ostertagia ostertagi [25], and Teladorsagia circumcincta [26], also secrete a GAL/VAL- dominated suite of ESPs. The large number of ESP molecules also suggests functional complexity. Binding to the host cell is often a prerequisite for ESP function [11, 17–19, 27]. Some ESP molecules react to the molecules on the surface of the host cell to form receptor-ligand com- plexes, similar to many other receptor-ligand systems, for example, galectin binds β-galactoside sugars in a metal-independent manner [28, 29]. Despite the large number of ESP molecules and their diverse functions, few ESP proteins have been identified and functionally characterized, particularly in vivo, and the ESP receptors on the host cell surface have not been fully characterized. Peripheral blood mononuclear cells (PBMCs) consist of several populations of immune cells, included lymphocytes (T cells, B cells, and NK cells) and monocytes that play important roles in the immune responses. Previously we reported that, HcESPs had immune suppressive potential on the goat PBMCs in vitro [30]. The present study is the first to analyze HcESPs from different developmental stages of H. con- tortus that interact with goat PBMCs in vivo using proteomics. This study will facilitate the elu- cidation of HcESP functions and the mechanisms of H. contortus immune evasion and pathogenesis. 2 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Production of anti HcESP polyclonal antibodies (IgGHcESP) To generate polyclonal antibodies against HcESP (IgGHcESP), 0.4 mg of HcESP protein was mixed with Freund’s complete adjuvant (1:1) and injected subcutaneously into 3 female Spra- gue Dawley (SD) rats [27, 32]. Rats received four doses at 2-week intervals. Ten days after the last injection, the rats were anesthetized with diethyl ether, sera containing specific anti-HcESP antibodies were collected, and the concentration of antibodies was determined by ELISA. The specific reactivity with HcESPs was confirmed by western blot analysis. Production of H. contortus excretory and secretory product (HcESP) in i To harvest ESP, the standard procedure for H. contortus described by Yatsuda et al. was used [16]. Briefly, H. contortus (Nanjing strain) adult worms were harvested from the abomasum of an experimentally infected donor goat, washed several times with PBS, and incubated for 4 h in RPMI 1640 medium (100/ml) containing antibiotics (100 IU of penicillin, 0.1 mg/ml strepto- mycin, and 5g/ml gentamicin) at 37°C under 5% CO. The medium was then removed, and the parasites were incubated in new medium containing 2% glucose overnight. The supernatant was collected, centrifuged, filter-sterilized (0.2 m), concentrated, and desalted (10 mMTris, NaCl pH7.4) using 3 kDa filters (Centriprep YM-3, Millipore). The protein concentration was determined by the Bradford assay [31]. Western blot analysis of the specificity of IgGHcESP Purified HcESP (20 μg) were resolved by 10% SDS-PAGE and transferred to Hybond-C extra nitrocellulose membranes (Amersham Biosciences, UK). Non-specific binding was blocked by incubating the membranes in 5% skim milk in Tris-buffered saline (TBST) for 1 h at room temperature. The membranes were then washed 5 times (5 min each) with TBS containing 0.1% Tween-20 (TBST), followed by incubation with the primary antibodies (IgGHcESP) for 1 h at 37°C (1:100 dilution in TBST). After washing 5 times with TBST, the membranes were incu- bated with HRP-conjugated rabbit anti-rat IgG (Sigma, USA) for 1 h at 37°C (diluted 1:2000 in TBST). Finally, the immunoreaction was visualized after incubation with freshly prepared dia- minobenzidine (DAB, Sigma) as a chromogenic substrate for 5 min. Materials and Methods Ethics Statement Animal experiments were conducted following the guidelines of the Animal Ethics Committee, Nanjing Agricultural University, China. All experimental protocols were approved by the Sci- ence and Technology Agency of Jiangsu Province. The approval ID is SYXK (SU) 2010–0005. Co-immunoprecipitation of HcESPPBMC-interacting proteins Co-IP was performed using the Protein A/G PLUS-Agarose Immunoprecipitation Kit (Santa Cruz Biotechnology, USA) according to the manufacturer’s instructions. Briefly, 4× 107 PBMCs were collected from experimentally infected goats (in vivo) were pelleted and lysed with 3 mL of NP-40 lysis buffer (50 mM Tris pH 7.4, 150mM NaCl, 1% NP-40) containing pro- tease inhibitor cocktail (Merck, USA). Cellular debris was pelleted by centrifugation at 10,000 x g for 10 min at 4°C, and the supernatant was transferred to a new tube. The cell lysate was pre- cleared by incubation with 1 μg of rat normal IgG and 20 μL of Protein A/G PLUS-Agarose beads at 4°C for 30 min. After pelleting the beads by centrifugation at 1,000 ×g for 5 min at 4°C, the protein concentration of the supernatant (cell lysate for IP) was determined using the PierceTM BCATM Protein Assay (Thermo Fisher Scientific, USA). A 1-mL aliquot of the above lysate was incubated with IgGHcESP overnight at 4°C. Immune complexes were isolated using 20 μL of protein A/G plus agarose. Immunoprecipitates were collected by centrifugation at 2,500 rpm for 5 min at 4°C. The supernatant was carefully aspi- rated and discarded, and the pellet was washed four times with RIPA buffer. After the final wash, the pellet was resuspended in 1X SDS buffer. Collection of PBMCs from goats experimentally infected with H. contortus To identify HcESPPBMC-interacting proteins, three male Boer goats (2 years old) were raised under nematode-free conditions for the in vivo experiment. Infective stage larvae (L3) of H. contortus were produced in vitro, and 8000 L3 were administered to the nematode-free goats. The goats were monitored during the entire experimental period. To confirm H. contortus infection, fecal samples were collected from the rectum of each infected goat twice each week and checked for the presence of H. contortus eggs. Food and water were provided to all animals ad libitum. Twenty milliliters of heparinized blood was collected from each goat after 7 (L4 3 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo stage), 15 (L5 stage), 40 (early adult stage) and 60 days (late adult stage) by vein puncture. PBMCs were separated by the standard Ficoll-Hypaque (GE Healthcare, USA) gradient centri- fugation method [33], and isolated PBMCs were used to identify HcESP/PBMC-interacting proteins by co-immunoprecipitation (Co-IP), western blot and liquid chromatography–tan- dem mass spectrometry (LC-MS/MS) analyses. Confirmation of the proteins of HcESP interacted with PBMCs in vivo by Western blot The immunoprecipitates obtained by Co-IP were used to confirm HcESP interation in vivo by western blot using IgGHcESP as the primary antibody as described in the previous section. Functional annotation improvement by STRING protein clustering analysis STRING DB (version 9.1) was used to improve the functional annotation and analyze the func- tional networks among protein families [35, 36]. The sequences of unassigned proteins were retrieved from the UNIPROT-KB and subjected to protein clustering analysis to identify func- tional protein association networks using the STRING tool (http://string.embl.de/) [37]. Orthologous protein groups matching our queries were used for functional association net- works and gene ontology annotation. A STRING conservative score threshold of 0.4 was applied to calculate a confidence score on the basis of the conserved gene neighborhood, gene fusion events, and significant co-occurrence and co expression. Gene Ontology (GO) Gene ontology (GO) annotation was performed using BLAST2GO (version 2.7.2). The sequence alignment software NCBI BLAST + (ncbi-blast-2.2.28 + -win32.ext) was used to com- pare the identified protein sequences and the protein sequence NCBI nr database. According to the principle of similarity, functional information for homologous proteins can be used for the functional annotation of target proteins. Only results in the top10 and with an E value 1 e-3 ratio in subsequent sequence analysis were retained. A resulting ratio of similarity of 42– 100% was considered. Database search Data were searched against an in-house H. contortus sequence Uniprot database (21,722 pro- tein entries) based on the recently published Haemonchus genome [34] using the search engine Mascot (v.2.2, Matrix Science, London, UK), allowing a maximum of two missed cleavages. Carbamidomethyl (C) was specified as a fixed modification and oxidation (M) as a variable modification. In-solution trypsin digestion and liquid chromatography–tandem mass spectrometry (LC-MS/MS) In-solution trypsin digestion and LC-MS/MS of immunoprecipitates were performed at Shang- hai Applied Protein Technology, Co. Ltd. MS data for protein identification were obtained using Q Exactive (ThermoFinnigan, San Jose, CA). Approximately 30μg of sample was boiled with 30μL of STD buffer in a water bath for 5 min and cooled to room temperature. A 200-μL aliquot of UA buffer (8 M Urea, 150mMTris-HCl, pH 8.5) was added, followed by 30 kDa ultrafiltration centrifugation. After centrifugation, the filtrate was discarded, and 100μL of IAA (50mM IAA in UA) was added. After oscillation for 1 min, the sample was incubated at room temperature in the dark for 30 min, centrifugation was repeated as above, and the filtrate was discarded. Then, 100 μL of UA buffer was added, and the sample was centrifuged twice. Finally, 100μL of 25mM NH4HCO3 was added and centrifuged twice as described above. The solution was then digested with 40μL of trypsin overnight at 37°C. Dried peptides were dissolved in 40μL of 0.1% formic acid (FA), and a 20μL aliquot was desalted for 10 min on a C-18 pre-column (Zorbax 300SB-C18 peptide traps, Agilent Technol- ogies, Wilmington) pre-equilibrated with 0.1% FA. Separation was performed by capillary high-performance liquid chromatography (0.15 X 150mm RP C18 analytical column, Column Technology Inc.) at 200°C using a chromatographic gradient of 0.1% FA in H2O (A) to 0.1% aqueous FA in 84% ACN (B) over 60 min (liquid linear gradient of solution A: 1–4% (1–50 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 4 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo min), 4 to 50% (50–54 min) and linear gradient of solution B from 50–100% (54–60 min); B was maintained at 100%). PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Table 1. Oligonucleotide primer sequences for PCR. Name Accession No. Sequences(5’- 3’) Haemonchus contortus 14-3-3 (Hc-ftt) CDJ94531 GGATCCATGGCTGACAATAAGGATG GAATTCCAATTTGCACCTTCTCCTT Haemonchus contortusADP-ribosylation factor (Hc-arf) CDJ89627 AAAGGATCCATGGGTAACATTTTCGG GCGCTCGAGTTATCCTCTGTTTTTCA Haemonchus contortusserine/ threonine protein kinase (Hc-stk) AF457202 AAGCTTATGGTTCCGGCCTCTTATCAGA GAATTCTCGACTGACCGGCAGGAGCTTG Haemonchus contortus SCP-like extracellular-domain-containing protein (Hc-scp) CDJ81443 GAATTCATGTGTCCAGACACCAATGGTA AAGCTTTTATGGGGCAATACAGAGAGCT / Table 1. Oligonucleotide primer sequences for PCR. Name purity of the protein preparation was determined by SDS-PAGE. Protein concentrations were determined by the Bradford method. Endotoxins were removed from the recombinant proteins using the ToxinEraserTM Endotoxin Removal kit (GeneScript, USA). Polyclonal antibodies against the recombinant proteins were produced as described above. Production and analysis of the specificity of anti-HcESP polyclonal antibodies (IgGHcESP) IgGHcESP was produced by injection of SD rats with HcESP protein mixed with Freund’s com- plete adjuvant, and the specificity of IgGHcESP was confirmed by western blot using HcESP as the antigen. Normal rat serum was used as a control. Bands from 13 to 180 kDa were detected by IgGHcESP, and no bands were recognized by the normal rat serum (Fig 1). Interaction of recombinant proteins with PBMCs Heparinized blood was collected by vein puncture from dewormed healthy goats. PBMCs were separated as described in the previous section and washed twice in Ca2+2+/Mg- free PBS pH 7.4. Cell viability assessed by means of the trypan blue exclusion test was consistently >95%. The PBMC were resuspended to a final density of 1×105 cells/ml in RPMI 1640 medium con- taining 10% heat inactivated fetal calf serum (FCS), 100 U/ml penicillin and 100 mg/ml strep- tomycin (gibco, Life Technology). PBMCs were incubated in the presence and absence of recombinant proteins (5μg/ml) for 1 h at 37°C. Confirmation of interaction was determined by an immunofluorescence assay (IFA) as described by Yuan et al. [40]. Briefly, washed cells (105 / ml) were fixed with 4% paraformaldehyde on a poly-L-lysine-coated glass slide. The cells were then treated with blocking solution (4% BSA in PBS) for 30 min to minimize background stain- ing. After sequential incubation with rat anti-recombinant protein IgG (1:100) for 2 h and a secondary antibody (1:300) coupled to the fluorescent dye Cy3 (Beyotime, Jiangsu, China), nuclear staining with 2-(4-amidinophenyl)-6-indole carbamidinedihydrochloride (DAPI, 1.5μM; Sigma, MO, USA) was performed for 6 min. Then, protein localization was determined by observing the staining patterns with a 100× oil objective lens on a laser scanning confocal microscope (L SM710, Zeiss, Jena, Germany). Digital images were captured using the Zeiss microscope software package ZEN 2012 (Zeiss, Jena, Germany). Validation of proteomic data by interaction analysis of recombinant proteins identified at different developmental stages To validate the proteomic data, we confirmed the interaction of 6 recombinant proteins identi- fied at different developmental stages to goat PBMCs. The genes encoding 14-3-3 (Hc-ftt), ADP-ribosylation factor (Hc-arf), SCP-like extracellular-domain-containing protein (Hc-scp) and serine threonine kinase (Hc-stp) were cloned using specific reverse and forward primers (Table 1). The genes encoding actin [38], and glyceraldehyde-3-phosphate dehydrogenase (HcGPDH) [39] were previously cloned in our laboratory. Briefly, the ORF of each gene was amplified by RT-PCR and cloned into pMD-19T (Takara Biotechnology). After double diges- tion with the corresponding restriction enzymes, DNA fragments were recovered and success- fully sub-cloned into the pET32a (+) expression vector. Escherichia coli BL21 cells containing the recombinant gene expression plasmid were cultured in Luria-Bertani medium with ampi- cillin (100 μg/mL), and expression of the recombinant proteins (rHc-ftt, rHc-arf, rHc-scp, rHc-stp, rHc-act, rHc-GPDH) was induced by IPTG. The histidine-tagged fusion protein was purified from the bacterial lysates using the His-Bind Resin Chromatography kit (Novagen) and dialyzed in phosphate buffered saline (PBS, pH 7.4) to remove imidazole. The purified recombinant proteins were dissolved in PBS (pH 8.0) containing 0.1mM DTT (PBS/DTT). The PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 5 / 26 Confirmation of the interaction of HcESPs with goat PBMCs in vivo PBMCs collected from the experimentally infected goats were used to confirm the interaction of HcESP with PBMCs in vivo. Protein extracted from the infected goat PBMCs was PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 6 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 1. Protein profile of H. contortus ESP (HcESP) and western blot analysis of HcESP probed with rat anti-HcESP. (c) Control using normal rat serum as the primary antibody. doi:10.1371/journal.pone.0159796.g001 Fig 1. Protein profile of H. contortus ESP (HcESP) and western blot analysis of HcESP probed with rat anti-HcESP. (c) Control using normal rat serum as the primary antibody. doi:10.1371/journal.pone.0159796.g001 doi:10.1371/journal.pone.0159796.g001 concentrated by Co-IP. Western blot analysis of the immunoprecipitate using IgGHcESP as the primary antibody confirmed the interaction of HcESPs with the goat PBMCs collected from the experimentally infected goats on day 7 (L4 stage), 15 (L5 stage), 40 (early adult stage) and 60 (late adult stage) post infection (Fig 2). PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Analysis of HcESP interacting proteins in vivo by LC-MS/MS The interaction of HcESPs with goat PBMCs in vivo at different stages of H. contortus develop- ment was analyzed by LC-MS/MS after the interacting proteins were concentrated by Co-IP (S1 Fig). A total of 407 interacting proteins in vivo were identified from the H. contortus pro- tein database via SEQUEST searches. Of these proteins, 47 (11.54%) proteins were common to all developmental stages (S1 Table) including actin, heat shock protein 70, glycoside hydrolase, glyceraldehyde-3-phosphate dehydrogenase, zinc finger, peptidase, Ras domain, serine threo- nine protein kinase (STK) and 14-3-3. A total of 94 (23.09%) interacting proteins were common to both the L4 and L5 developmen- tal stages, including elongation factor 1-alpha, tropomyosin, immunoglobulin I-set and fibro- nectin, transcription factor E2F dimerization partner (TDP), tenascin-like and cytochrome b5. L5 and the early adult stage shared 76 (18.67%) interacting proteins, including major sperm protein (MSP), ribosomal proteins (S8, S5 and L2), Mbt repeat, NADH: ubiquinone oxidore- ductase and dynein light intermediate chain. Fifty-nine (14.49%) proteins were shared between PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 7 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 2. SDS PAGE and western blot analysis of the Co-IP of in vivo interacting HcESPs with goat PBMCs at different developmental stages. Marker (M), Immunoprecipitates (IP) and normal rat serum Control (C). (A) L4, (B) L5, (C) early adult and (D) late adult. doi:10.1371/journal.pone.0159796.g002 Fig 2. SDS PAGE and western blot analysis of the Co-IP of in vivo interacting HcESPs with goat PBMCs at different developmental stages. Marker (M), Immunoprecipitates (IP) and normal rat serum Control (C). (A) L4, (B) L5, (C) early adult and (D) late adult. doi:10.1371/journal.pone.0159796.g002 doi:10.1371/journal.pone.0159796.g002 the early and late adult stage. The distribution of the interacting proteins at different develop- mental stages is summarized in S1 Table and Fig 3. The identified proteins were further categorized according to stage-specific (S2 Table); 98 (24%) interacting proteins were identified at L4 including heat shock protein 90 (HSP90), alde- hyde dehydrogenase, nematode cuticle collagen, carbohydrate kinase, glucose-methanol-cho- line oxidoreductase and eukaryotic translation initiation factor 3. At the L5 developmental stage, 107 (26.3%) HcESP stage-specific interacting proteins were identified, including enolase, acyltransferase choActase, phosphotyrosyl phosphatase activator, myosin-10, glutamine ami- dotransferase, annexin, saposin type B and telomerase activating protein Est1. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Analysis of HcESP interacting proteins in vivo by LC-MS/MS Only 13 inter- acting proteins were stage-specific in the early adult stage; these proteins included alanine racemase, amino acid transporter domain-containing protein, aminotransferase, and conden- sation and AMP-dependent synthetase ligase. In the present study, 48 (11.54%) late adult stage-specific proteins were identified, these proteins included CK1/WORM6 protein kinase, protein synthesis factor and translation elongation factor EFTu EF1A and translation PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 8 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 3. Venn diagram of the interacting proteins shared among different developmental stages in vivo. Fig 3. Venn diagram of the interacting proteins shared among different developmental stages in vivo. doi:10.1371/journal.pone.0159796.g003 doi:10.1371/journal.pone.0159796.g003 elongation factor EFG EF2, selectin-like protein, short-chain dehydrogenase reductase SDR, EVL-14, transcription factor jumonji 1 and tyrosine protein kinase. elongation factor EFG EF2, selectin-like protein, short-chain dehydrogenase reductase SDR, EVL-14, transcription factor jumonji 1 and tyrosine protein kinase. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Gene Ontology (GO) analysis The GO signatures of 234 of the 407 proteins identified in vivo were available in the database. To further understand the functions of the proteins identified in this study, we queried the InterPro databases. The identified proteins were classified by molecular function, biological process and cellular component according to the GO hierarchy using a Web Gene Ontology Annotation Plot (WEGO). Among the 47 shared proteins by all developmental stages, 41 were annotated based on molecular function, and 5 terms were identified. Most were assigned to binding and catalytic activity. For biological processes, 29 proteins were associated with 9 terms. Most of the proteins were related to metabolic, cellular and single organism processes. Among the cellular compo- nent annotation, 16 proteins were assigned to 5 cellular component terms, and16 proteins were located in the cell as well as in organelles (Fig 4). GO analysis of the interacting proteins shared PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 9 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 4. Comparative molecular function, biological process and cellular component GO terms of stage-specific HcESP interacting proteins at different developmental stages and shared among all developmental stages. doi:10.1371/journal.pone.0159796.g004 LOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 10 / Fig 4. Comparative molecular function, biological process and cellular component GO terms of stage-specific HcESP interacting proteins at different developmental stages and shared among all developmental stages. doi:10.1371/journal.pone.0159796.g004 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 10 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo between the L4 and L5 stages resulted in the annotation of 7 molecular functions; most of the proteins were assigned to binding and catalytic activity. For biological processes, 10 terms were identified, primarily metabolic, cellular and single organism processes. An identical distribution of GO terms was observed for the proteins shared between the L5 and early adult stages. Eight terms were identified and binding and catalytic activities were the major molecular functions. Among the proteins shared by the early and late adult stages, 39 were assigned to binding and 31 to catalytic activity. The distribution of GO terms was nearly identical for the early and late adult stages (Fig 5). Among the 98 L4 stage-specific proteins, 35 were assigned to 8 molecular function terms. The most abundant terms were catalytic and binding activity. For biological processes, 11 terms were identified. Gene Ontology (GO) analysis Cellular, metabolic, single-organism and localization were the most abundant terms. According to the cellular component annotation, 7 cellular locations were assigned to 21 proteins. Cell part, membrane, membrane part and macromolecular complexes were annotated as the major cellular components. The results of the GO analysis for the 108 L5 stage-specific proteins revealed that 35 proteins were annotated with 9 molecular functions. Binding and catalytic activity were the major terms. For biological processes, 39 proteins were assigned 8 terms; metabolic, cellular process, single-organism process and localization were the most abundant biological processes. In the case of cellular component annotation, 28 proteins were annotated with 8 cellular locations, and most of the annotated proteins were membrane, cell part and membrane part. Among the 13 early adult stage-specific interacting proteins, only 2 terms related to molecular function were annotated, and 3 proteins were assigned to binding and catalytic activity. Metabolic processes were a major biological process at this stage, and only one protein was annotated as a membrane part. GO analysis of 48 late adult stage-specific proteins revealed a molecular function annotation for 18 proteins. The same pattern of func- tional distribution described above was observed, and 5 GO terms related to molecular func- tion were identified. Catalytic and binding activity were the most abundant terms. Seventeen proteins were annotated as biological process, and 7 terms were attained. Metabolic, cellular and single-organism processes were the major terms. For cellular components, 7 terms were identified for 6 proteins. Protein complex and cell part were highly represented terms (Fig 4). PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Improvement of functional annotation by protein clustering The sequences of 173 unassigned interacting proteins were retrieved from UNIPROT-KB and subjected to protein clustering analysis to determine their functional association network in the STRING database. In the STRING database, 118/173 (68.20%) proteins were available, and 80 functional associations were predicted. Fig 6 illustrates the functional interaction between protein orthologues in the nearest organism (Caenorhabditis elegans) and their predicted func- tional partners. The resultant orthologous protein groups were used to predict function. The results of the protein clusters and corresponding functional information are summarized in S3 Table and Fig 7. Based on the protein clustering, we discovered new HcESP proteins previously annotated as hypothetical proteins in the database. An unnamed protein (U6NP15) identified at the L4 and L5 stages matched WHiTe (Drosophila)-related ABC transporter family member (wht-8) with 56% similarity. An unnamed protein (W6NHX8) identified at the L4 stage matched the P-gly- coprotein subclass of the ATP-binding cassette (ABC) transporter super family with 66% homology. ADP-ribosylation factor (arf-1.2) family matched hypothetical protein (U6PBJ7). The identified protein had 97% similarity with C. elegans. Protein cluster analysis enabled the functional characterization of 42 additional proteins based on GO analysis of homologous proteins. Binding activity was a highly represented 11 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 5. Comparative molecular function, biological process and cellular component GO terms of proteins shared between the L4 and L5, L5 and early adult, and early adult and late adult stages. doi:10.1371/journal.pone.0159796.g005 Stage-Specific Binding of HcESPs to Goat PBMCs In Fig 5. Comparative molecular function, biological process and cellular component GO terms of proteins shared between the L4 and L5, L5 and early adult, and early adult and late adult stages. doi:10.1371/journal.pone.0159796.g005 doi:10.1371/journal.pone.0159796.g005 12 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 6. Improvement of functional annotation based on protein clustering. (A) STRING functional protein association network of the predicted associations of unassigned proteins. Nodes of different colors indicate clustering proteins matching our queries. (B) Co-expression graph from C. elegans. (C) Predicted protein functional partners. Fig 6. Improvement of functional annotation based on protein clustering. (A) STRING functional protein association network of the predicted associations of unassigned proteins. Nodes of different colors indicate clustering proteins matching our queries. (B) Co-expression graph from C. elegans. (C) Predicted protein functional partners. Improvement of functional annotation by protein clustering doi:10.1371/journal.pone.0159796.g006 molecular function including ion binding, cytoskeleton protein binding and DNA binding. Thirty-six biological process terms were identified for 56 proteins. Anatomical structure devel- opment, embryo development, reproduction, locomotion and transport were highly repre- sented terms. Discussion Excretory and secretory products (ESPs) are produced and released by parasites during in vitro cultivation [11] and in vivo [12]. H. contortus excretory and secretory products (HcESPs) con- tain many proteins [16] and performed diverse functions such as tissue penetration and host protein degradation [14]. It was previously reported that early expression of TH2 cytokine IL-4 and upregulation of genes that recruit neutrophils (CXCL1) and macrophages (MCP1) was associated with the resistance to H. contortus [41]. Estrada-Reyes et al. [42] reported the high regulation of IL-5 at 2, 7 and 14 days post-infection (PI) and IL-6 at 14 days PI of H. contortus. In our previous study, HcESPs displayed suppressive potential on the goat PBMCs in vitro. They inhibited the productions of IL-4, IFN-γ, nitric oxide, cell proliferation, whereas increased the production of suppressive cytokine IL-10, inflammatory modulator IL-17 and cell migration [30]. However, the protein or proteins of HcESPs that govern the functions of HcESPs in vitro or in vivo are unknown. In the present study, the interactions of HcESP at dif- ferent developmental stages with goat PBMCs in vivo were evaluated by Co-IP followed by LC-MS/MS. A total 407 non-redundant H. contortus proteins that interacted with goat PBMCs in vivo were identified by searching the H. contortus Uniprot database. This study is the first to report the in vivo identification of HcESP interacting proteins from the L4 to adult stage of worms living in the goat host. We purified 6 recombinant HcESP interacting proteins to vali- date the proteomic data and the interaction of these proteins with goat PBMCs by IFA. Excretory and secretory products (ESPs) are produced and released by parasites during in vitro cultivation [11] and in vivo [12]. H. contortus excretory and secretory products (HcESPs) con- tain many proteins [16] and performed diverse functions such as tissue penetration and host protein degradation [14]. It was previously reported that early expression of TH2 cytokine IL-4 and upregulation of genes that recruit neutrophils (CXCL1) and macrophages (MCP1) was associated with the resistance to H. contortus [41]. Estrada-Reyes et al. [42] reported the high regulation of IL-5 at 2, 7 and 14 days post-infection (PI) and IL-6 at 14 days PI of H. contortus. In our previous study, HcESPs displayed suppressive potential on the goat PBMCs in vitro. In our study, 47 HcESP interacting proteins were common among all developmental stages. Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 7. Distribution of Gene Ontology terms improved by protein clustering of 173 unassigned HcESP proteins that bound goat PBMCs at different developmental stages. Fig 7. Distribution of Gene Ontology terms improved by protein clustering of 173 unassigned HcESP proteins that bound goat PBMCs at different developmental stages Fig 7. Distribution of Gene Ontology terms improved by protein clustering of 173 unassigned HcESP proteins that bound goat PBMCs at different developmental stages. doi:10.1371/journal.pone.0159796.g007 Validation of proteomic data by assessing recombinant HcESPs interacted with PBMCs To validate the proteomic data, we purified 6 recombinant proteins (rHc-ftt, rHc-arf, rHc- scp, rHc-stp, rHc-act and rHcGPDH) identified at different developmental stages as inter- acting proteins with goat PBMCs. Interaction of the recombinant proteins with goat PBMCs was confirmed by immunofluorescence. Nuclei were stained with DAPI (blue fluo- rescence), and confocal microscopy images revealed that the recombinant proteins were interacted with the cell surface (red fluorescence). In the control group, no red fluorescence was observed (Fig 8). 13 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Discussion Among them, 4 peptidase proteins including cysteine peptidase (C46), serine peptidases (S9 & S28) and metalloprotease (M13) were identified. Significant expression of all three classes of peptidases in the parasitic L4 to adult stages has been reported [43]. Schwarz et al. reported more than 120 upregulated peptidase genes in H. contortus parasitic stages, and various clans were predicted to be secreted peptidases, including metallopeptidases (M 12A, M01, M13, M12A, M10A), aspartic peptidases (A01A) and cysteine peptidases (CA01A) [10]. The identifi- cation of secreted proteins related to the peptidase family in the current research supports these reports, which indicated that these proteins play a crucial part in the catabolism of globin PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 14 / 26 contortus PKs to goat PBMCs sug- gests that this kinase might also play a significant role in the functional regulation of goat PBMCs. We identified the secreted H. contortus 14-3-3 protein as a goat PBMC-interacting protein in all parasitic stages. Gene sequences of 14-3-3 have been reported for several protozoan and metazoan parasites, including Plasmodium falciparum, Trypanosoma cruzi, Toxoplasma gondii, Neosporacaninum, Eimeria tenella, Schistosoma japonicum, Echinococcus granulosus, Meloido- gyne incognita [53–59]. 14-3-3 proteins are phosphoserine-binding proteins that control the actions of a wide range of targets via direct protein–protein interactions. In animal cells, the majority of the known targets of 14-3-3 proteins are involved in signal transduction, transcrip- tion and proliferation [60–62]. Our findings indicate that the H. contortus 14-3-3 protein may act in signal transduction. The cytoskeletal protein actin was identified as an interacting protein in all stages. Actin has been detected in different helminths [63, 64] and is involved in very important cellular func- tions, including cell division, secretion, signaling, cellular shape and volume regulation, move- ment and phagocytosis [65, 66]. The effects of the interaction of H. contortus actin with host PBMCs merits further study. We observed that H. contortus HSP70 interacted with goat PBMCs in vivo in all develop- mental stages. HSP70 proteins are molecular chaperones that play important roles in the pro- cess of invasion, response to stress and survival in nematodes. HSP70 has been identified in several parasitic nematodes [67–69], but there is minimal information available for H. contor- tus [70, 71]. The transition from the L3 to L4 stage is key to the establishment of parasitism by H. contor- tus. ESPs play a very important role in pathogenesis and induce immune modulation at the early stage of infection. L4 is the first blood-feeding stage of H. contortus, and at this stage, genes related to motor activity and metabolism occur in the parasite [72]. Here, we observed high proteomic complexity of the HcESPs that interacted with goat PBMCs at different devel- opmental stages. HcESP interacting proteins were more abundant at the L4 and L5 stages than at the early and late adult stages [73]. A previous study reported that 234 proteins were upregu- lated in the L4 stage compared to L3 [10]. In our study, 209 and 217 proteins were identified at the L4 and L5 stages, respectively, and 94 interacting proteins were shared between the 2 stages. Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Fig 8. Interaction of recombinant proteins (A): rHc-ftt, (B): rHc-arf, (C): rHc-scp, (D): rHc-stp, (E): rHc-act and (F): rHcGPDH) to PBMC. PBMCs were untreated as controls or treated with the recombinant proteins. The cell nuclei were visualized by DAPI (blue) staining. Staining of the target proteins (red) was visualized by a Cy3-conjugated secondary antibody. Fig 8. Interaction of recombinant proteins (A): rHc-ftt, (B): rHc-arf, (C): rHc-scp, (D): rHc-stp, (E): rHc-act and (F): rHcGPDH) to PBMC. PBMCs were untreated as controls or treated with the recombinant proteins. The cell nuclei were visualized by DAPI (blue) staining. Staining of the target proteins (red) was visualized by a Cy3-conjugated secondary antibody. doi:10.1371/journal.pone.0159796.g008 doi:10.1371/journal.pone.0159796.g008 doi:10.1371/journal.pone.0159796.g008 by hemoglobin cleavage [1]. Development from L3 to L4 in vitro leads to the release of a metal- loprotease enzyme that inhibits blood clotting and thus facilitates blood feeding [1, 44]. We confirmed the above finding in vivo and report for the first time the presence of these peptidase proteins in the blood circulation. These findings support previous genomic studies reporting that genes encoding peptidase proteins are transcribed at a higher level in the host compared to free-living stages [10, 16]. In the present study, a key glycolytic enzyme, glyceraldehyde-3-phosphate dehydrogenase (GPDH), was identified as interacting protein with goat PBMCs in vivo in all developmental stages. GPDH plays an important role in host invasion of the worm in addition to its glycolytic activity [45, 46]. This protein could play a key role in immune modulation by binding comple- ment (C3) and thus inhibiting complement activity [2]. GPDH is essential for inducing the T helper (Th1 and Th17) immune response during natural infection [47]. Our findings suggested that GPDH plays an important role in immune modulation and is one of the most important HcESP proteins. In the present study, the serine/threonine kinases (PKs) were also identified in all stages. This finding is in agreement with the results of previous studies [10, 48]. PKs play a vital role in PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 15 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo the cellular signaling transduction involved in cell proliferation, differentiation, cell-cycle pro- gression, transcription, DNA replication, metabolic processes, phosphorylation, apoptosis, autophagy and inflammation [49–52]. The interaction of H. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Most of the shared proteins were related to binding (n = 51), catalytic activity (n = 41) and metabolism (n = 51).Our findings suggest the active involvement of these proteins in parasit- ism and immunemodulation. SCP-like extracellular-domain-containing protein (vap-1) was identified as an interacting protein in the L4 and L5 stages. Vap-1 encodes a predicted secreted protein that is similar to the venom allergen-like proteins reported in a number of invertebrates, including parasitic nema- todes [74–76]. Schwarz et al. identified 82 genes related to SCP proteins including 54 upregu- lated genes in the parasitic stages [10]. Previously, two proteins related to the SCP-like proteins Hc24 and Hc40 were reported in the ESPs of adult H. contortus [12, 77]. Our findings support these previous results, and the interaction of these proteins with goat PBMCs at multiple stages in vivo suggests a critical role of SCP-like proteins in infection and may be immunomodulatory factors. Elongation factor-1α protein is involved in signaling activity and was identified in our study at L4 and L5. EF-1α is highly conserved and ubiquitously expressed in all eukaryotic cells [78]. EF-1α proteins has been reported in parasites including Cryptosporidium hominis, Trichomonas vaginalis, Trypanosoma brucei, Clonorchis sinensis and Brugia malayi [79–83] Trichomonas vaginalis, Trypanosoma brucei, Clonorchis sinensis and Brugia malayi [79–83] 16 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo Functionally, EF-1α transfers aminoacylated tRNAs to the ribosome A site in a GTP-depen- dent reaction [84]. In addition, EF-1α appears to have a number of other functions associated with cell growth, motility, protein turnover, and signal transduction [85, 86], DNA replication/ repair protein networks [87] and apoptosis [88]. The interaction of EF-1α with host PBMCs indicates its active role as an immune depressant and warrants further investigation. A total of 102 proteins were identified at the early adult stage, and 76 were shared between the L5 and early adult stages. Fewer HcESPs interacted at the adult stage compared to the L4 and L5 stages. Thus, the parasite releases more ESPs in the early stages of infection to modulate immune function for parasite survival. Studies of expressed sequence tag (EST) data have pro- vided transcriptional and genomic insights on the different developmental stages of H. contor- tus [10, 59, 72, 89]. We identified 3 ribosomal proteins (S8, S5 and L2) at the L5 and early adult stages. Zamanian et al. [83] identified 14% ribosomal proteins of the B. malayi exosome-like vesicles (ELVs) released from the infective L3 stage. Ribosomal proteins actively participate in cellular processes other than protein biosynthesis and can act as components of the translation apparatus, cell proliferation and apoptosis [90]. Cantacessi et al. employed an in silico subtrac- tion approach to identify H. contortus L3 and xL3 genes and predicted that H. contortus L3-spe- cific genes encoding ribosomal proteins [91] were required for phagocytosis [91, 92]. Our findings confirm these previous reports, and the participation of ribosomal proteins in cellular processes should be further investigated. We identified 137 proteins at the late adult stage. Fifty-nine proteins were shared between the early and late stages. Among them, serine/threonine-protein phosphatase (STPs) was iden- tified as an important interacting protein. STPs from various parasites have been functionally characterized [93–95]. Protein phosphatases are involved in major biological processes such as cell division, apoptosis and exocytosis [96]. STPs are often involved in signal transduction and transcriptional activation [97–99]. Our findings suggest that H. contortus STPs might be involved in various biological processes, particularly signal transduction. We identified 98 stage-specific HcESP interacting proteins in the L4 developmental stage, including heat shock protein 90, extracellular ligand-binding receptor, aldehyde dehydroge- nase, carbohydrate kinase, myosin-4, aldehyde dehydrogenase and glucose-methanol-choline oxidoreductase. A total of 107 L5 stage-specific proteins were identified, including enolase, saposin type B, myosin-10, aromatic amino acid beta-eliminating lyase threonine aldolase and annexin. Genes or ESTs transcribed during different developmental stages have been investi- gated previously [10, 72]. Hartman et al. reported that the cysteine protease, Hc42, Hc60 and vitellogenin genes were transcribed at the adult stage, whereas glutathione peroxidase, alpha- tubulin, Hc43 and Hc38 were transcribed at the L3 and adult stages [44]. The immunological involvement of these stage-specific HcESP proteins remains to be further characterized. Scaffold proteins are essential components of signaling functions such as the trafficking, anchoring and multimerization of glutamate receptors and act as adhesion molecules [100– 102]. We identified 53 HcESP scaffold proteins that interacted with goat PBMCs at different developmental stages in vivo, including 27 in L4 and 28 in L5. Interestingly, only 5 and 10 scaf- fold proteins were observed at the early and late adult stages. Low concentrations of scaffold proteins increase the output of cascades, but as the concentration of the scaffold proteins increases, the output of the cascade decreases. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 At an elevated concentration of scaffold proteins, one molecule can bind only one kinase molecule, and thus the output of the signaling cascade is also very low [103, 104]. The high concentration of scaffold proteins observed at the L4 and L5 stages in the present study could represent a mechanism of immune modulation by combi- natorial inhibition of the signaling cascade. We identified the hypothetical H. contortus protein (U6PBJ7) as a member of the ADP- ribosylation factor (arf-1.2) family by STRING protein clustering. This highly conserved family 17 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo is involved in a wide range of cell functions. ARF proteins are N-myristoylated GTPases, which are involved in membrane trafficking, actin cytoskeleton, regulation of apoptotic fate and acti- vation of phospholipase D1 (PLD1) and phosphatidylinositol 4-phosphate 5-kinase [105–107]. The present study is the first to report interaction of the HcESPARF-1.2 protein with goat PBMCs in vivo at multiple developmental stages. In our study, another hypothetical HcESP protein (W6NHX8) identified at the L5 stage was confirmed as P-glycoprotein-9 (pgp-9), part of the ATP-binding cassette (ABC) or traffic ATPase subclass, by STRING protein clustering analysis. The H. contortus Pgp gene may be involved in host-parasite interaction, particularly in eosinophil granule product detoxification [108, 109]. Nematode parasites undergo impor- tant adaptations during the transition from free-living to parasitic stages, such as evasion of the host immune reaction, metabolism and growth. Issouf et al. [108] compared the expression level of Pgps in free and parasitic stages and reported that Hco-pgp-9.2, Hco-pgp-11, Hco-pgp- 3 and Hco-pgp16 mRNAs were over expressed in the L4 and adult stages. Here we report for the first time that H. Contortus P-glycoprotein-9 (pgp-9) interacted with goat PBMCs at the L5 stage in vivo. Secreted extracellular vesicles (EVs) play an important role in parasite-host interactions. Exosomes considered highly bioactive EVs that facilitated cell to cell communication in many eukaryotes and prokaryotes [110]. Several studies on various parasites including helminths demonstrated that EVs could carry and deliver virulence factors such as proteins and sRNAs to the host [111–115]. In the present study, various HcESPs interacting proteins including annex- ins, GAPDH, actin, HSP70, HSP90, 14-3-3 proteins, tubulin, ras-related protein, histone, ATP synthase subunit alpha, HSP DnaJ, eukaryotic translation initiation factor 3, enolase, ribosomal proteins and acyltransferase ChoActase were identified at various developmental stages in vivo. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 In the previous studies these interacting proteins were recognized as members of EVs [83, 116, 117]. The interaction of these EVs related HcESPs with goat PBMCs suggested their important regulatory role in host–parasite interaction. To provide a comprehensive understanding of the roles of the H. contortus proteins that interact with host PBMCs, the identified proteins were functionally categorized based on the GO annotation of molecular functions, biological processes and cellular components. Of the 407 proteins identified in vivo, 173 (42.50%) did not have assigned GO terms. These unanno- tated proteins were further analyzed by clustering of homologous proteins via STRING data- bases to enhance the functional annotation prediction. For molecular function GO annotation, the most enriched functions of the HcESPs were related to binding activity. Proteins associated with these functions are involved in ATP bind- ing, nucleotide binding, protein binding, GTP binding, DNA binding, motor activity, transla- tion elongation factor activity, GTPase activity, protein kinase C inhibitor activity, protein hetero- and homodimerization activity, protein polymerization and signaling activity [16, 118, 119]. The biological process GO results revealed that the most represented categories were annotated as transport, metabolic, catabolic and phosphorylation processes. A nearly identical profile of biological process annotation was reported by Moreno et al. for Heligmosomoides polygyrus ESPs [118]. The functional annotation of proteins can be predicted and improved by clustering of homologous proteins. The functional annotation of parasite proteins is often constrained by the small proportion of genes with homologs in model organisms [120]. However, based on the clustering of homologous proteins [50, 121], we were able to enrich the GO annotations of 173 unannotated proteins identified at different developmental stages. Clustering the homolo- gous proteins increased the GO annotation by 24.27% for molecular function annotation, 23.36% for biological process and 22.54% for cellular components. 18 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo In conclusion, we analyzed the interaction of HcESPs with host PBMCs in vivo at different developmental stages. Many of the identified proteins were highly developmentally specific proteins. The large numbers and the complexity of the interacting proteins indicated that the HcESPs interact with the host immune cells in complex ways and result in complex regulation of the host immune cells. Our data provide a baseline for understanding the relationship between the parasite and host. S1 Table. List of identified interacting proteins of HcESP with goat PBMCs shared among different developmental stages in vivo. (DOCX) S1 Table. List of identified interacting proteins of HcESP with goat PBMCs shared among different developmental stages in vivo. (DOCX) S1 Table. List of identified interacting proteins of HcESP with goat PBMCs shared among different developmental stages in vivo. (DOCX) S2 Table. List of developmental stage specific interacting proteins of HcESP with goat PBMCs identified at different time points in vivo. (DOCX) S3 Table. List of proteins available on STRING database matched with our Query sequences of unassigned proteins analyzed for functional annotation improvement by pro- tein clustering analysis. (DOCX) However, the interacting partners and the regulatory mecha- nisms of specific proteins remain to be further investigated. The functions of the novel interact- ing proteins and the nature of the unassigned proteins also require further study. Acknowledgments The project was supported by the “National Key Basic Research Program (973 program) of China” (Grant No. 2015CB150300) and by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD). Author Contributions Conceived and designed the experiments: XL JAG. Performed the experiments: JAG SW GB ME. Analyzed the data: JAG RY XS. Contributed reagents/materials/analysis tools: LX. Wrote the paper: JAG XL. Supporting Information S1 Fig. Digestion of peptides by trypsin and MS identification using capillary high-perfor- mance liquid chromatography. Total ion flow diagram of HcESPs interacted with goat PBMCs at the (A) L4, (B) L5, (C) early adult and (D) late adult stages. (TIF) PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References 1. Nikolaou S, Gasser RB. Prospects for exploring molecular developmental processes in Haemonchus contortus. International journal for parasitology. 2006; 36(8):859–68. Epub 2006/06/09. doi: 10.1016/j. ijpara.2006.04.007 PMID: 16759659. 2. Sahoo S, Murugavel S, Devi IK, Vedamurthy GV, Gupta SC, Singh BP, et al. Glyceraldehyde-3-phos- phate dehydrogenase of the parasitic nematode Haemonchus contortus binds to complement C3 and inhibits its activity. Parasite immunology. 2013; 35(12):457–67. Epub 2013/08/10. doi: 10.1111/pim. 12058 PMID: 23927077. 3. Blackburn HD, Rocha JL, Figueiredo EP, Berne ME, Vieira LS, Cavalcante AR, et al. Interaction of parasitism and nutrition in goats: effects on haematological parameters, correlations, and other statis- tical associations. Veterinary parasitology. 1992; 44(3–4):183–97. Epub 1992/10/01. PMID: 1466129. 3. Blackburn HD, Rocha JL, Figueiredo EP, Berne ME, Vieira LS, Cavalcante AR, et al. Interaction of parasitism and nutrition in goats: effects on haematological parameters, correlations, and other statis- tical associations. Veterinary parasitology. 1992; 44(3–4):183–97. Epub 1992/10/01. PMID: 1466129. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 19 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo 4. Jasmer DP, Lahmers KK, Brown WC. Haemonchus contortus intestine: a prominent source of muco- sal antigens. Parasite immunology. 2007; 29(3):139–51. Epub 2007/02/03. doi: 10.1111/j.1365-3024. 2006.00928.x PMID: 17266741. 4. Jasmer DP, Lahmers KK, Brown WC. Haemonchus contortus intestine: a prominent source of muco- sal antigens. Parasite immunology. 2007; 29(3):139–51. Epub 2007/02/03. doi: 10.1111/j.1365-3024. 2006.00928.x PMID: 17266741. 5. Gilleard JS. Understanding anthelmintic resistance: the need for genomics and genetics. Int J Parasi- tol. 2006; 36(12):1227–39. Epub 2006/08/08. S0020-7519(06)00240-2 [pii] doi: 10.1016/j.ijpara.2006. 06.010 PMID: 16889782. 6. Redman E, Sargison N, Whitelaw F, Jackson F, Morrison A, Bartley DJ, et al. Introgression of iver- mectin resistance genes into a susceptible Haemonchus contortus strain by multiple backcrossing. PLoS Pathog. 2012; 8(2):e1002534. Epub 2012/02/24. doi: 10.1371/journal.ppat.1002534 PMID: 22359506; PubMed Central PMCID: PMC3280990. 7. Gilleard JS. Haemonchus contortus as a paradigm and model to study anthelmintic drug resistance. Parasitology. 2013; 140(12):1506–22. Epub 2013/09/04. doi: 10.1017/s0031182013001145 PMID: 23998513. 8. Kaminsky R, Ducray P, Jung M, Clover R, Rufener L, Bouvier J, et al. A new class of anthelmintics effective against drug-resistant nematodes. Nature. 2008; 452(7184):176–80. Epub 2008/03/14. doi: 10.1038/nature06722 PMID: 18337814. 9. Stoll NR. Studies with the strongyloid nematode, Haemonchus contortus. Am J Hyg. 1929; 10:384– 418. 10. Schwarz EM, Korhonen PK, Campbell BE, Young ND, Jex AR, Jabbar A, et al. The genome and developmental transcriptome of the strongylid nematode Haemonchus contortus. Genome Biology. 2013; 14(8):R89–R. References doi: 10.1186/gb-2013-14-8-r89 PMC4053716. PMID: 23985341 11. Rathore DK, Suchitra S, Saini M, Singh BP, Joshi P. Identification of a 66 kDa Haemonchus contortus excretory/secretory antigen that inhibits host monocytes. Veterinary parasitology. 2006; 138(3– 4):291–300. Epub 2006/03/07. doi: 10.1016/j.vetpar.2006.01.055 PMID: 16517075. 12. Schallig HD, van Leeuwen MA, Cornelissen AW. Protective immunity induced by vaccination with two Haemonchus contortus excretory secretory proteins in sheep. Parasite immunology. 1997; 19 (10):447–53. Epub 1998/01/07. PMID: 9372512. 13. Knox DP. Development of vaccines against gastrointestinal nematodes. Parasitology. 2000; 120 Suppl:S43–61. Epub 2000/06/30. PMID: 10874709. 14. Cox GN, Pratt D, Hageman R, Boisvenue RJ. Molecular cloning and primary sequence of a cysteine protease expressed by Haemonchus contortus adult worms. Mol Biochem Parasitol. 1990; 41(1):25– 34. Epub 1990/06/01. PMID: 2385265. 15. Sun Y, Yan R, Muleke CI, Zhao G, Xu l, Li X. Recombinant galectins of Haemonchus contortus para- site induces apoptosis in the peripheral blood lymphocytes of goat. Int J Pept Res Ther. 2006; 13 (3):387–92. doi: 10.1007/s10989-006-9045-0 16. Yatsuda AP, Krijgsveld J, Cornelissen AW, Heck AJ, de Vries E. Comprehensive analysis of the secreted proteins of the parasite Haemonchus contortus reveals extensive sequence variation and differential immune recognition. The Journal of biological chemistry. 2003; 278(19):16941–51. Epub 2003/02/11. doi: 10.1074/jbc.M212453200 PMID: 12576473. 17. Anbu KA, Joshi P. Identification of a 55 kDa Haemonchus contortus excretory/secretory glycoprotein as a neutrophil inhibitory factor. Parasite Immunol. 2008; 30(1):23–30. doi: 10.1111/j.1365-3024. 2007.00995.x PMID: 18086013 18. Reinhardt S, Scott I, Simpson HV. Neutrophil and eosinophil chemotactic factors in the excretory/ secretory products of sheep abomasal nematode parasites: NCF and ECF in abomasal nematodes. Parasitology research. 2011; 109(3):627–35. Epub 2011/03/23. doi: 10.1007/s00436-011-2305-8 PMID: 21424403. 19. Wang W, Wang S, Zhang H, Yuan C, Yan R, Song X, et al. Galectin Hco-gal-m from Haemonchus contortus modulates goat monocytes and T cell function in different patterns. Parasites & vectors. 2014; 7:342. Epub 2014/07/25. doi: 10.1186/1756-3305-7-342 PMID: 25056558; PubMed Central PMCID: PMCPmc4117971. 20. Merkelbach P, Scott I, Khalaf S, Simpson HV. Excretory/secretory products of Haemonchus contortus inhibit aminopyrine accumulation by rabbit gastric glands in vitro. Veterinary parasitology. 2002; 104 (3):217–28. Epub 2002/01/29. PMID: 11812620. 21. Przemeck S, Huber A, Brown S, Pedley KC, Simpson HV. Excretory/secretory products of sheep abomasal nematode parasites cause vacuolation and increased neutral red uptake by HeLa cells. Parasitology research. 2005; 95(3):213–7. Epub 2005/02/22. doi: 10.1007/s00436-004-1280-8 PMID: 15723269. 22. Huber A, Prosl H, Joachim A, Simpson HV, Pedley KC. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References Effects of excretory/secretory products of Haemonchus contortus on cell vacuolation. Parasitology research. 2005; 96(5):290–5. Epub 2005/05/ 27. doi: 10.1007/s00436-005-1321-y PMID: 15918071. 22. Huber A, Prosl H, Joachim A, Simpson HV, Pedley KC. Effects of excretory/secretory products of Haemonchus contortus on cell vacuolation. Parasitology research. 2005; 96(5):290–5. Epub 2005/05/ 27. doi: 10.1007/s00436-005-1321-y PMID: 15918071. 20 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo 23. Yan F, Xu L, Liu L, Yan R, Song X, Li X. Immunoproteomic analysis of whole proteins from male and female adult Haemonchus contortus. Vet J. 2010; 185(2):174–9. Epub 2009/06/30. doi: 10.1016/j.tvjl. 2009.05.021 PMID: 19560953. 24. Yatsuda AP, Eysker M, Vieira-Bressan MC, De Vries E. A family of activation associated secreted protein (ASP) homologues of Cooperia punctata. Res Vet Sci. 2002; 73(3):297–306. Epub 2002/11/ 22. PMID: 12443689. 25. Saverwyns H, Visser A, Nisbet AJ, Peelaers I, Gevaert K, Vercruysse J, et al. Identification and char- acterization of a novel specific secreted protein family for selected members of the subfamily Osterta- giinae (Nematoda). Parasitology. 2008; 135(Pt 1):63–70. Epub 2007/10/03. doi: 10.1017/ s0031182007003666 PMID: 17908364. 26. Craig H, Wastling JM, Knox DP. A preliminary proteomic survey of the in vitro excretory/secretory products of fourth-stage larval and adult Teladorsagia circumcincta. Parasitology. 2006; 132(Pt 4):535–43. Epub 2006/01/04. doi: 10.1017/s0031182005009510 PMID: 16388693. 27. Wang W, Yuan C, Wang S, Song X, Xu L, Yan R, et al. Transcriptional and proteomic analysis reveal recombinant galectins of Haemonchus contortus down-regulated functions of goat PBMC and modu- lation of several signaling cascades in vitro. J Proteomics. 2014; 98(0):123–37. http://dx.doi.org/10. 1016/j.jprot.2013.12.017. 28. Gitt MA, Wiser MF, Leffler H, Herrmann J, Xia YR, Massa SM, et al. Sequence and mapping of galec- tin-5, a beta-galactoside-binding lectin, found in rat erythrocytes. The Journal of biological chemistry. 1995; 270(10):5032–8. Epub 1995/03/10. PMID: 7890611. 29. Kasai K, Hirabayashi J. Galectins: a family of animal lectins that decipher glycocodes. Journal of bio- chemistry. 1996; 119(1):1–8. Epub 1996/01/01. PMID: 8907168. 30. Gadahi JA, Yongqian B, Ehsan M, Zhang ZC, Wang S, Yan RF, et al. Haemonchus contortus excre- tory and secretory proteins (HcESPs) suppress functions of goat PBMCs in vitro. Oncotarget. 2016. Epub 2016/05/28. doi: 10.18632/oncotarget.9589 PMID: 27229536. 31. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein uti- lizing the principle of protein-dye binding. Anal Biochem. 1976; 72:248–54. Epub 1976/05/07. PMID: 942051. 32. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References Early IL-4 gene expression in aboma- sum is associated with resistance to Haemonchus contortus in hair and wool sheep breeds. Parasite immunology. 2016; 38(6):333–9. Epub 2016/04/10. doi: 10.1111/pim.12321 PMID: 27059919. 42. Estrada-Reyes ZM, López-Reyes AG, Lagunas-Martínez A, Ramírez- Vargas G, Olazarán-Jenkins S, Hernández-Romano J, et al. Relative expression analysis of IL-5 and IL-6 genes in tropical sheep breed Pelibuey infected with Haemonchus contortus. Parasite immunology. 2015; 37(9):446–52. doi: 10.1111/pim.12211 PMID: 26094646 43. Knox D. Proteases in blood-feeding nematodes and their potential as vaccine candidates. Adv Exp Med Biol. 2011; 712:155–76. Epub 2011/06/11. doi: 10.1007/978-1-4419-8414-2_10 PMID: 21660664. 44. Hartman D, Donald DR, Nikolaou S, Savin KW, Hasse D, Presidente PJ, et al. Analysis of develop- mentally regulated genes of the parasite Haemonchus contortus. Int J Parasitol. 2001; 31(11):1236– 45. Epub 2001/08/22. PMID: 11513893. 45. Hu Y, Zhang E, Huang L, Li W, Liang P, Wang X, et al. Expression profiles of glyceraldehyde-3-phos- phate dehydrogenase from Clonorchis sinensis: a glycolytic enzyme with plasminogen binding capac- ity. Parasitol Res. 2014; 113(12):4543–53. Epub 2014/10/11. doi: 10.1007/s00436-014-4144-x PMID: 25300416. 46. Hara MR, Agrawal N, Kim SF, Cascio MB, Fujimuro M, Ozeki Y, et al. S-nitrosylated GAPDH initiates apoptotic cell death by nuclear translocation following Siah1 binding. Nat Cell Biol. 2005; 7(7):665– 74. Epub 2005/06/14. doi: 10.1038/ncb1268 PMID: 15951807. 47. El Ridi R, Tallima H. Vaccine-induced protection against murine schistosomiasis mansoni with larval excretory-secretory antigens and papain or type-2 cytokines. J Parasitol. 2013; 99(2):194–202. Epub 2012/09/19. doi: 10.1645/ge-3186.1 PMID: 22985345. 48. Nikolaou S, Hartman D, Nisbet AJ, Presidente PJ, Gasser RB. Genomic organization and expression analysis for hcstk, a serine/threonine protein kinase gene of Haemonchus contortus, and comparison with Caenorhabditis elegans par-1. Gene. 2004; 343(2):313–22. Epub 2004/12/14. doi: 10.1016/j. gene.2004.09.017 PMID: 15588586. 49. Ali AS, Ali S, El-Rayes BF, Philip PA, Sarkar FH. Exploitation of protein kinase C: a useful target for cancer therapy. Cancer Treat Rev. 2009; 35(1):1–8. Epub 2008/09/10. doi: 10.1016/j.ctrv.2008.07. 006 PMID: 18778896. 50. Breugelmans B, Jex AR, Korhonen PK, Mangiola S, Young ND, Sternberg PW, et al. Bioinformatic exploration of RIO protein kinases of parasitic and free-living nematodes. Int J Parasitol. 2014; 44 (11):827–36. Epub 2014/07/20. doi: 10.1016/j.ijpara.2014.06.005 PMID: 25038443. 51. Yuan W, Lok JB, Stoltzfus JD, Gasser RB, Fang F, Lei WQ, et al. Toward understanding the functional role of Ss-RIOK-1, a RIO protein kinase-encoding gene of Strongyloides stercoralis. PLoS Negl Trop Dis. 2014; 8(8):e3062. Epub 2014/08/08. References Han K, Xu L, Yan R, Song X, Li X. Molecular cloning, expression and characterization of enolase from adult Haemonchus contortus. Research in veterinary science. 2012; 92(2):259–65. Epub 2011/03/30. doi: 10.1016/j.rvsc.2011.03.008 PMID: 21444098. 33. Wang Y, Yang W, Cama V, Wang L, Cabrera L, Ortega Y, et al. Population genetics of Cryptosporid- ium meleagridis in humans and birds: evidence for cross-species transmission. Int J Parasitol. 2014; 44(8):515–21. Epub 2014/04/15. doi: 10.1016/j.ijpara.2014.03.003 PMID: 24727090. 34. Laing R, Kikuchi T, Martinelli A, Tsai IJ, Beech RN, Redman E, et al. The genome and transcriptome of Haemonchus contortus, a key model parasite for drug and vaccine discovery. Genome Biology. 2013; 14(8):R88–R. doi: 10.1186/gb-2013-14-8-r88 PMC4054779. PMID: 23985316 35. Ravooru N, Ganji S, Sathyanarayanan N, Nagendra HG. Insilico analysis of hypothetical proteins unveils putative metabolic pathways and essential genes in Leishmania donovani. Front Genet. 2014; 5:291. Epub 2014/09/11. doi: 10.3389/fgene.2014.00291 PMID: 25206363; PubMed Central PMCID: PMC4144268. 36. Doerks T, van Noort V, Minguez P, Bork P. Annotation of the M. tuberculosis hypothetical orfeome: adding functional information to more than half of the uncharacterized proteins. PLoS One. 2012; 7 (4):e34302. Epub 2012/04/10. doi: 10.1371/journal.pone.0034302 PMID: 22485162; PubMed Central PMCID: PMC3317503. 37. Szklarczyk D, Franceschini A, Kuhn M, Simonovic M, Roth A, Minguez P, et al. The STRING data- base in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res. 2011; 39(Database issue):D561–8. Epub 2010/11/04. doi: 10.1093/nar/gkq973 PMID: 21045058; PubMed Central PMCID: PMCPmc3013807. 38. Yan R, Wang J, Xu L, Song X, Li X. DNA vaccine encoding Haemonchus contortus actin induces par- tial protection in goats. Acta Parasitol. 2014; 59(4):698–709. Epub 2014/09/23. doi: 10.2478/s11686- 014-0298-z PMID: 25236283. 39. Han K, Xu L, Yan R, Song X, Li X. Cloning, expression and characterization of NAD+-dependent glyc- eraldehyde-3-phosphate dehydrogenase of adult Haemonchus contortus. J Helminthol. 2011; 85 (4):421–9. Epub 2011/01/06. doi: 10.1017/s0022149x10000763 PMID: 21205411. 40. Yuan C, Zhang H, Wang W, Li Y, Yan R, Xu L, et al. Transmembrane protein 63A is a partner protein of Haemonchus contortus galectin in the regulation of goat peripheral blood mononuclear cells. Para- sit Vectors. 2015; 8(1):211. Epub 2015/04/17. doi: 10.1186/s13071-015-0816-3 PMID: 25879191; PubMed Central PMCID: PMCPmc4404006. 21 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo 41. Jacobs JR, Sommers KN, Zajac AM, Notter DR, Bowdridge SA. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References Kiss JE, Gao X, Krepp JM, Hawdon JM. Interaction of hookworm 14-3-3 with the forkhead transcrip- tion factor DAF-16 requires intact Akt phosphorylation sites. Parasit Vectors. 2009; 2:21–. doi: 10. 1186/1756-3305-2-21 PMC2683825. PMID: 19393088 63. Sotillo J, Sanchez-Flores A, Cantacessi C, Harcus Y, Pickering D, Bouchery T, et al. Secreted prote- omes of different developmental stages of the gastrointestinal nematode Nippostrongylus brasiliensis. Mol Cell Proteomics. 2014; 13(10):2736–51. Epub 2014/07/06. doi: 10.1074/mcp.M114.038950 PMID: 24994561; PubMed Central PMCID: PMC4188999. 64. Morassutti AL, Levert K, Pinto PM, da Silva AJ, Wilkins P, Graeff-Teixeira C. Characterization of Angiostrongylus cantonensis excretory-secretory proteins as potential diagnostic targets. Exp Parasi- tol. 2012; 130(1):26–31. Epub 2011/10/25. doi: 10.1016/j.exppara.2011.10.003 PMID: 22019415. 65. De Loof A, Vanden J, Janssen I. Hormones and the cytoskeleton of animals and plants. Int Rev Cytol. 1996; 166:1–58. Epub 1996/01/01. PMID: 8881772. 66. Kovaleva ES, Subbotin SA, Masler EP, Chitwood DJ. Molecular characterization of the actin gene from cyst nematodes in comparison with those from other nematodes. Comp Parasitol. 2005; 72 (1):39–49. doi: 10.1654/4138 67. Fang L, Sun L, Yang J, Gu Y, Zhan B, Huang J, et al. Heat shock protein 70 from Trichinella spiralis induces protective immunity in BALB/c mice by activating dendritic cells. Vaccine. 2014; 32 (35):4412–9. Epub 2014/06/26. doi: 10.1016/j.vaccine.2014.06.055 PMID: 24962751. 68. Hartmann W, Singh N, Rathaur S, Brenz Y, Liebau E, Fleischer B, et al. Immunization with Brugia malayi Hsp70 protects mice against Litomosoides sigmodontis challenge infection. Parasite Immunol. 2014; 36(4):141–9. Epub 2013/12/24. doi: 10.1111/pim.12093 PMID: 24359133. 69. Srivastava S, Srikanth E, Liebau E, Rathaur S. Identification of Setaria cervi heat shock protein 70 by mass spectrometry and its evaluation as diagnostic marker for lymphatic filariasis. Vaccine. 2010; 28 (5):1429–36. Epub 2009/07/04. doi: 10.1016/j.vaccine.2009.06.044 PMID: 19573640. 70. Zhang H, Zhou Q, Yang Y, Chen X, Yan B, Du A. Characterization of heat shock protein 70 gene from Haemonchus contortus and its expression and promoter analysis in Caenorhabditis elegans. Parasi- tology. 2013; 140(6):683–94. Epub 2013/01/31. doi: 10.1017/S0031182012002168 PMID: 23360558. 71. van Leeuwen MA. Heat-shock and stress response of the parasitic nematode Haemonchus contortus. Parasitol Res. 1995; 81(8):706–9. Epub 1995/01/01. PMID: 8570590. 72. Laing R, Kikuchi T, Martinelli A, Tsai IJ, Beech RN, Redman E, et al. The genome and transcriptome of Haemonchus contortus, a key model parasite for drug and vaccine discovery. Genome Biol. 2013; 14(8):R88. Epub 2013/08/30. doi: 10.1186/gb-2013-14-8-r88 PMID: 23985316; PubMed Central PMCID: PMCPmc4054779. 73. References doi: 10.1371/journal.pntd.0003062 PMID: 25101874; PubMed Central PMCID: PMC4125297. 52. Geering B. Death-associated protein kinase 2: Regulator of apoptosis, autophagy and inflammation. Int J Biochem Cell Biol. 2015; 65:151–4. Epub 2015/06/10. doi: 10.1016/j.biocel.2015.06.001 PMID: 26055515. 53. Siles-Lucas M. The 14-3-3 protein: a key molecule in parasites as in other organisms. Trends Parasi- tol. 2003; 19(12):575–81. doi: 10.1016/j.pt.2003.10.003 PMID: 14642768 54. Wang W, Shakes DC. Molecular evolution of the 14-3-3 protein family. J Mol Evol. 1996; 43(4):384– 98. Epub 1996/10/01. doi: 10.1007/BF02339012 PMID: 8798343. 55. Jeanclos EM, Lin L, Treuil MW, Rao J, DeCoster MA, Anand R. The chaperone protein 14-3-3eta interacts with the nicotinic acetylcholine receptor alpha 4 subunit. Evidence for a dynamic role in sub- unit stabilization. J Biol Chem. 2001; 276(30):28281–90. Epub 2001/05/16. doi: 10.1074/jbc. M011549200 PMID: 11352901. 56. Koyama T, Ohsawa T, Shimada S, Omata Y, Xuan X, Inoue N, et al. A 14-3-3 protein homologue is expressed in feline enteroepithelial-stages of Toxoplasma gondii. Vet Parasitol. 2001; 96(1):65–74. http://dx.doi.org/10.1016/S0304-4017(00)00424-6. PMID: 11182236 57. McGonigle S, Beall MJ, Feeney EL, Pearce EJ. Conserved role for 14-3-3epsilon downstream of type I TGFbeta receptors. FEBS Lett. 2001; 490(1–2):65–9.doi: citeulike-article-id:3736418. PMID: 11172812 58. Heusden GPH, Griffiths DJF, Ford JC, Chin-A-Woeng TFC, Schrader PAT, Carr AM, et al. The 14-3-3 proteins encoded by the BMH1 and BMH2 genes are essential in the yeast Saccharomyces cerevi- siae and can be replaced by a plant homologue. Eur J Biochem. 1995; 229(1):45–53. doi: 10.1111/j. 1432-1033.1995.0045l.x PMID: 7744048 59. Hoekstra R, Visser A, Otsen M, Tibben J, Lenstra JA, Roos MH. EST sequencing of the parasitic nem- atode Haemonchus contortus suggests a shift in gene expression during transition to the parasitic 22 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo stages. Mol Biochem Parasitol. 2000; 110(1):53–68. http://dx.doi.org/10.1016/S0166-6851(00) 00255-3. PMID: 10989145 stages. Mol Biochem Parasitol. 2000; 110(1):53–68. http://dx.doi.org/10.1016/S0166-6851(00) 00255-3. PMID: 10989145 60. 60. Aghazadeh Y, Ye X, Blonder J, Papadopoulos V. Protein modifications regulate the role of 14-3- 3gamma adaptor protein in cAMP-induced steroidogenesis in MA-10 Leydig cells. J Biol Chem. 2014; 289(38):26542–53. Epub 2014/08/03. doi: 10.1074/jbc.M114.569079 PMID: 25086053; PubMed Central PMCID: PMC4176220. 61. Jeon YH, Park YH, Kwon JH, Lee JH, Kim IY. Inhibition of 14-3-3 binding to Rictor of mTORC2 for Akt phosphorylation at Ser473 is regulated by selenoprotein W. Biochim Biophys Acta. 2013; 1833 (10):2135–42. Epub 2013/05/18. doi: 10.1016/j.bbamcr.2013.05.005 PMID: 23680186. 62. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References Bennuru S, Semnani R, Meng Z, Ribeiro JM, Veenstra TD, Nutman TB. Brugia malayi excreted/ secreted proteins at the host/parasite interface: stage- and gender-specific proteomic profiling. PLoS Negl Trop Dis. 2009; 3(4):e410. Epub 2009/04/09. doi: 10.1371/journal.pntd.0000410 PMID: 19352421; PubMed Central PMCID: PMCPmc2659452. 74. Hewitson JP, Harcus Y, Murray J, van Agtmaal M, Filbey KJ, Grainger JR, et al. Proteomic analysis of secretory products from the model gastrointestinal nematode Heligmosomoides polygyrus reveals dominance of venom allergen-like (VAL) proteins. J Proteomics. 2011; 74(9):1573–94. Epub 2011/07/ 05. doi: 10.1016/j.jprot.2011.06.002 PMID: 21722761. 75. Yoshino TP, Brown M, Wu XJ, Jackson CJ, Ocadiz-Ruiz R, Chalmers IW, et al. Excreted/secreted Schistosoma mansoni venom allergen-like 9 (SmVAL9) modulates host extracellular matrix remodel- ling gene expression. Int J Parasitol. 2014; 44(8):551–63. Epub 2014/05/27. doi: 10.1016/j.ijpara. 2014.04.002 PMID: 24859313; PubMed Central PMCID: PMC4079936. 76. Murray J, Gregory WF, Gomez-Escobar N, Atmadja AK, Maizels RM. Expression and immune recog- nition of Brugia malayi VAL-1, a homologue of vespid venom allergens and Ancylostoma secreted proteins. Mol Biochem Parasitol. 2001; 118(1):89–96. Epub 2001/11/13. PMID: 11704277. 77. Rehman A, Jasmer DP. A tissue specific approach for analysis of membrane and secreted protein antigens from Haemonchus contortus gut and its application to diverse nematode species. Mol Bio- chem Parasitol. 1998; 97(1–2):55–68. Epub 1999/01/08. PMID: 9879887. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 23 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo 78. Vinkenoog R, Speranca MA, van Breemen O, Ramesar J, Williamson DH, Ross-MacDonald PB, et al. Malaria parasites contain two identical copies of an elongation factor 1 alpha gene. Mol Biochem Parasitol. 1998; 94(1):1–12. PMID: 9719506. 79. Kaur KJ, Ruben L. Protein translation elongation factor-1 alpha from Trypanosoma brucei binds cal- modulin. J Biol Chem. 1994; 269(37):23045–50. Epub 1994/09/16. PMID: 8083206. 80. Carlton JM, Hirt RP, Silva JC, Delcher AL, Schatz M, Zhao Q, et al. Draft genome sequence of the sexually transmitted pathogen Trichomonas vaginalis. Science. 2007; 315(5809):207–12. Epub 2007/01/16. doi: 10.1126/science.1132894 PMID: 17218520; PubMed Central PMCID: PMCPmc2080659. 81. Xu P, Widmer G, Wang Y, Ozaki LS, Alves JM, Serrano MG, et al. The genome of Cryptosporidium hominis. Nature. 2004; 431(7012):1107–12. Epub 2004/10/29. doi: 10.1038/nature02977 PMID: 15510150. 82. Kim T, Cho P, Na J, Hong S-J. Molecular cloning and phylogenetic analysis of Clonorchis sinensis elongation factor-1α. Parasitol Res. 2007; 101(6):1557–62. doi: 10.1007/s00436-007-0676-7 PMID: 17674047 83. Zamanian M, Fraser LM, Agbedanu PN, Harischandra H, Moorhead AR, Day TA, et al. References Release of Small RNA-containing Exosome-like Vesicles from the Human Filarial Parasite Brugia malayi. PLoS neglected tropical diseases. 2015; 9(9):e0004069. doi: 10.1371/journal.pntd.0004069 PMC4581865. PMID: 26401956 84. Riis B, Rattan SI, Clark BF, Merrick WC. Eukaryotic protein elongation factors. Trends Biochem Sci. 1990; 15(11):420–4. PMID: 2278101. 85. Ransom-Hodgkins WD. The application of expression analysis in elucidating the eukaryotic elonga- tion factor one alpha gene family in Arabidopsis thaliana. Mol Genet Genomics. 2009; 281(4):391– 405. doi: 10.1007/s00438-008-0418-2 PMID: 19132394. 86. Matsubayashi M, Teramoto-Kimata I, Uni S, Lillehoj HS, Matsuda H, Furuya M, et al. Elongation fac- tor-1alpha is a novel protein associated with host cell invasion and a potential protective antigen of Cryptosporidium parvum. J Biol Chem. 2013; 288(47):34111–20. Epub 2013/10/03. doi: 10.1074/jbc. M113.515544 PMID: 24085304; PubMed Central PMCID: PMCPmc3837153. 87. Toueille M, Saint-Jean B, Castroviejo M, Benedetto JP. The elongation factor 1A: a novel regulator in the DNA replication/repair protein network in wheat cells? Plant Physiol Biochem. 2007; 45(2):113–8. doi: 10.1016/j.plaphy.2007.01.006 PMID: 17344053. 88. Blanch A, Robinson F, Watson IR, Cheng LS, Irwin MS. Eukaryotic translation elongation factor 1- alpha 1 inhibits p53 and p73 dependent apoptosis and chemotherapy sensitivity. PLoS One. 2013; 8 (6):e66436. doi: 10.1371/journal.pone.0066436 PMID: 23799104; PubMed Central PMCID: PMC3682968. 89. Geldhof P, Whitton C, Gregory WF, Blaxter M, Knox DP. Characterisation of the two most abundant genes in the Haemonchus contortus expressed sequence tag dataset. Int J Parasitol. 2005; 35 (5):513–22. Epub 2005/04/14. doi: 10.1016/j.ijpara.2005.02.009 PMID: 15826643. 90. Rhoads RE, Dinkova TD, Korneeva NL. Mechanism and regulation of translation in C. elegans. WormBook. 2006:1–18. Epub 2007/12/01. doi: 10.1895/wormbook.1.63.1 PMID: 18050488. 91. Cantacessi C, Campbell BE, Young ND, Jex AR, Hall RS, Presidente PJ, et al. Differences in tran- scription between free-living and CO2-activated third-stage larvae of Haemonchus contortus. BMC Genomics. 2010; 11:266. Epub 2010/04/28. doi: 10.1186/1471-2164-11-266 PMID: 20420710; PubMed Central PMCID: PMCPmc2880303. 92. Reddien PW, Horvitz HR. CED-2/CrkII and CED-10/Rac control phagocytosis and cell migration in Caenorhabditis elegans. Nat Cell Biol. 2000; 2(3):131–6. Epub 2000/03/09. doi: 10.1038/35004000 PMID: 10707082. 93. Ma GX, Zhou RQ, Hu SJ, Huang HC, Zhu T, Xia QY. Molecular characterization and functional analy- sis of serine/threonine protein phosphatase of Toxocara canis. Exp Parasitol. 2014; 141:55–61. Epub 2014/03/25. doi: 10.1016/j.exppara.2014.03.019 PMID: 24657583. 94. Hu M, Abs El-Osta YG, Campbell BE, Boag PR, Nisbet AJ, Beveridge I, et al. Trichostrongylus vitrinus (Nematoda: Strongylida): Molecular characterization and transcriptional analysis of Tv-stp-1, a serine/ threonine phosphatase gene. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References Exp Parasitol. 2007; 117(1):22–34. http://dx.doi.org/10.1016/j.exppara. 2007.03.008. PMID: 17490653 95. Kutuzov MA, Andreeva AV. Protein Ser/Thr phosphatases of parasitic protozoa. Mol Biochem Parasi- tol. 2008; 161(2):81–90. http://dx.doi.org/10.1016/j.molbiopara.2008.06.008. doi: 10.1016/j. molbiopara.2008.06.008 PMID: 18619495 96. Campbell BE, Hofmann A, McCluskey A, Gasser RB. Serine/threonine phosphatases in socioeco- nomically important parasitic nematodes—prospects as novel drug targets? Biotechnol Adv. 2011; 29 (1):28–39. Epub 2010/08/25. doi: 10.1016/j.biotechadv.2010.08.008 PMID: 20732402. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 24 / 26 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo 97. Klumpp S, Krieglstein J. Serine/threonine protein phosphatases in apoptosis. Curr Opin Pharmacol. 2002; 2(4):458–62. Epub 2002/07/20. PMID: 12127881. 98. Berndt N. Roles and regulation of serine/threonine-specific protein phosphatases in the cell cycle. Prog Cell Cycle Res. 2003; 5:497–510. Epub 2003/11/05. PMID: 14593745. 99. Sarmiento M, Zhao Y, Gordon SJ, Zhang ZY. Molecular basis for substrate specificity of protein-tyro- sine phosphatase 1B. J Biol Chem. 1998; 273(41):26368–74. Epub 1998/10/03. PMID: 9756867. 100. Verpelli C, Schmeisser MJ, Sala C, Boeckers TM. Scaffold proteins at the postsynaptic density. In: Kreutz MR, Sala C, editors. Synaptic plasticity: dynamics, development and disease. New York: Springer-Verlag; 2012. p. 29–58. 101. Ferrell JE. What do scaffold proteins really do? Sci STKE. 2000; 2000:pe1. doi: 10.1126/stke. 522000pe1 102. Good MC, Zalatan JG, Lim WA. Scaffold proteins: hubs for controlling the flow of cellular information. Science. 2011; 332(6030):680–6. Epub 2011/05/10. doi: 10.1126/science.1198701 PMID: 21551057; PubMed Central PMCID: PMC3117218. 103. Bray D, Lay S. Computer-based analysis of the binding steps in protein complex formation. Proc Natl Acad Sci U S A. 1997; 94(25):13493–8. Epub 1998/02/12. PMID: 9391053; PubMed Central PMCID: PMCPmc28333. 104. Levchenko A, Bruck J, Sternberg PW. Scaffold proteins may biphasically affect the levels of mitogen- activated protein kinase signaling and reduce its threshold properties. Proc Natl Acad Sci U S A. 2000; 97(11):5818–23. Epub 2000/05/24. PMID: 10823939; PubMed Central PMCID: PMCPmc18517. 105. Nie Z, Hirsch DS, Randazzo PA. Arf and its many interactors. Curr Opin Cell Biol. 2003; 15(4):396– 404. Epub 2003/08/02. PMID: 12892779. 106. Ackema KB, Hench J, Bockler S, Wang SC, Sauder U, Mergentaler H, et al. The small GTPase Arf1 modulates mitochondrial morphology and function. EMBO J. 2014; 33(22):2659–75. Epub 2014/09/ 06. doi: 10.15252/embj.201489039 PMID: 25190516; PubMed Central PMCID: PMCPmc4282574. 107. Singhvi A, Teuliere J, Talavera K, Cordes S, Ou G, Vale RD, et al. The Arf GAP CNT-2 regulates the apoptotic fate in C. elegans asymmetric neuroblast divisions. Curr Biol. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References 2011; 21(11):948–54. Epub 2011/05/21. doi: 10.1016/j.cub.2011.04.025 PMID: 21596567; PubMed Central PMCID: PMCPmc3109113. 108. Issouf M, Guegnard F, Koch C, Le Vern Y, Blanchard-Letort A, Che H, et al. Haemonchus contortus P-glycoproteins interact with host eosinophil granules: a novel insight into the role of ABC transporters in host-parasite interaction. PLoS ONE. 2014; 9(2):e87802. Epub 2014/02/06. doi: 10.1371/journal. pone.0087802 PMID: 24498376; PubMed Central PMCID: PMCPmc3912070. 109. Williamson SM, Wolstenholme AJ. P-glycoproteins of Haemonchus contortus: development of real- time PCR assays for gene expression studies. J Helminthol. 2012; 86(2):202–8. Epub 2011/07/07. doi: 10.1017/s0022149x11000216 PMID: 21729384. 110. Valadi H, Ekstrom K, Bossios A, Sjostrand M, Lee JJ, Lotvall JO. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nature cell biol- ogy. 2007; 9(6):654–9. Epub 2007/05/09. doi: 10.1038/ncb1596 PMID: 17486113. 111. Buck AH, Coakley G, Simbari F, McSorley HJ, Quintana JF, Le Bihan T, et al. Exosomes secreted by nematode parasites transfer small RNAs to mammalian cells and modulate innate immunity. Nature communications. 2014; 5:5488. Epub 2014/11/26. doi: 10.1038/ncomms6488 PMID: 25421927; PubMed Central PMCID: PMCPmc4263141. 112. Barteneva NS, Maltsev N, Vorobjev IA. Microvesicles and intercellular communication in the context of parasitism. Frontiers in cellular and infection microbiology. 2013; 3:49. Epub 2013/09/14. doi: 10. 3389/fcimb.2013.00049 PMID: 24032108; PubMed Central PMCID: PMCPmc3764926. 113. Silverman JM, Clos J, de'Oliveira CC, Shirvani O, Fang Y, Wang C, et al. An exosome-based secre- tion pathway is responsible for protein export from Leishmania and communication with macro- phages. Journal of cell science. 2010; 123(Pt 6):842–52. Epub 2010/02/18. doi: 10.1242/jcs.056465 PMID: 20159964. 114. Nowacki FC, Swain MT, Klychnikov OI, Niazi U, Ivens A, Quintana JF, et al. Protein and small non- coding RNA-enriched extracellular vesicles are released by the pathogenic blood fluke Schistosoma mansoni. Journal of extracellular vesicles. 2015; 4:28665. Epub 2015/10/08. doi: 10.3402/jev.v4. 28665 PMID: 26443722; PubMed Central PMCID: PMCPmc4595467. 115. Sotillo J, Pearson M, Becker L, Mulvenna J, Loukas A. A quantitative proteomic analysis of the tegu- mental proteins from Schistosoma mansoni schistosomula reveals novel potential therapeutic targets. International journal for parasitology. 2015; 45(8):505–16. Epub 2015/04/26. doi: 10.1016/j.ijpara. 2015.03.004 PMID: 25910674. 25 / 26 PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 Stage-Specific Binding of HcESPs to Goat PBMCs In Vivo 116. Buck AH, Coakley G, Simbari F, McSorley HJ, Quintana JF, Le Bihan T, et al. Exosomes secreted by nematode parasites transfer small RNAs to mammalian cells and modulate innate immunity. Nature communications. PLOS ONE \| DOI:10.1371/journal.pone.0159796 July 28, 2016 References 2014; 5:5488. doi: 10.1038/ncomms6488 PMC4263141. PMID: 25421927 116. Buck AH, Coakley G, Simbari F, McSorley HJ, Quintana JF, Le Bihan T, et al. Exosomes secreted by nematode parasites transfer small RNAs to mammalian cells and modulate innate immunity. Nature communications. 2014; 5:5488. doi: 10.1038/ncomms6488 PMC4263141. PMID: 25421927 117. Marcilla A, Trelis M, Cortes A, Sotillo J, Cantalapiedra F, Minguez MT, et al. Extracellular vesicles from parasitic helminths contain specific excretory/secretory proteins and are internalized in intestinal host cells. PLoS One. 2012; 7(9):e45974. Epub 2012/10/03. doi: 10.1371/journal.pone.0045974 PMID: 23029346; PubMed Central PMCID: PMCPmc3454434. 118. Moreno Y, Gros PP, Tam M, Segura M, Valanparambil R, Geary TG, et al. Proteomic analysis of excretory-secretory products of Heligmosomoides polygyrus assessed with next-generation sequenc- ing transcriptomic information. PLoS Negl Trop Dis. 2011; 5(10):e1370. Epub 2011/11/01. doi: 10. 1371/journal.pntd.0001370 PMID: 22039562; PubMed Central PMCID: PMCPmc3201918. 119. Morphew RM, Wright HA, LaCourse EJ, Woods DJ, Brophy PM. Comparative proteomics of excre- tory-secretory proteins released by the liver fluke Fasciola hepatica in sheep host bile and during in vitro culture ex host. Mol Cell Proteomics. 2007; 6(6):963–72. Epub 2007/02/20. doi: 10.1074/mcp. M600375-MCP200 PMID: 17308300. 120. Salavati R, Najafabadi HS. Sequence-based functional annotation: what if most of the genes are unique to a genome? Trends Parasitol. 2010; 26(5):225–9. Epub 2010/03/10. doi: 10.1016/j.pt.2010. 02.001 PMID: 20211583. 121. Mangiola S, Young ND, Korhonen P, Mondal A, Scheerlinck JP, Sternberg PW, et al. Getting the most out of parasitic helminth transcriptomes using HelmDB: implications for biology and biotechnology. Biotechnol Adv. 2013; 31(8):1109–19. Epub 2012/12/26. doi: 10.1016/j.biotechadv.2012.12.004 PMID: 23266393. 26 / 26
https://openalex.org/W3010503674	https://www.atlantis-press.com/article/125935249.pdf	English	null	Dynamic Systems with Baseline Exponential Distribution Based on Sequential Order Statistics Under a Power Trend for Hazard Rates	Journal of statistical theory and applications	2,020	cc-by	7,323	1. INTRODUCTION Let X1, ⋯, Xn be independent and identically distributed (i.i.d.) random variables with a common distribution function (DF), say F, and abbreviated by X1, ⋯, Xn i.i.d. ∼F. Denote in magnitude order of X1, ⋯, Xn by X1∶n ≤⋯≤Xn∶n, called order statistics (OSs). The theory of OSs has been widely assumption in literature. For example, in system reliability analyses, lifetimes of r-out-of- n systems coincide to Xr∶n where X1, ⋯, Xn stand for component lifetimes; For more information, see Barlow and Proschan [1] and David and Nagaraja [2] and references therein. In this setting, failure of a component does not effect the surviving components. There are various practical situations in which this investi- gated and applied does not hold. For illustration, suppose that there exist n power generators with nominal capacity G1, ⋯, Gn. The system perform satisfactory if the total generated power is at least a given threshold, say A. If a generator fails, then the remaining generators have to generate much powers to fulfill the nominal level A and hence more loads cause more pressure on the working generators. Therefore, the generator lifetimes are decreased. For statistical modeling these kinds of systems, some generalizations of OSs such as fractional OSs and generalized OSs have been introduced in literature. They are useful for providing more flexible tools and also a setting to unify simi- lar results (David and Nagaraja [2], p. 21). In this paper, we deal with another unified concept, called the sequential order statistics (SOS), which has also a motivation in reliability analyses as discussed above. Specifically, when the component lifetimes are i.i.d., the OSs are suit- able for describing the r-out-of- n system lifetime. Here failing a component does not effect the DFs of lifetimes of surviving components. Motivated by Cramer and Kamps [3], the failure of a component may result in a higher load on the surviving components and hence causes the lifetime distributions change. More precisely, suppose that Fj, for j = 1, ⋯n, denotes the common DF of the component lifetimes when n −j + 1 components are jointly working. Then, the components begin to work independently at time t = 0 with the common DF F1. If at time x1, the first component failure occurs, then the remaining n −1 components are working with the (left truncated) common DF F2 at x1. Journal of Statistical Theory and Applications Vol. 19(1), March 2020, pp. 1–9 DOI: https://doi.org/10.2991/jsta.d.200224.005; ISSN 1538-7887 https://www.atlantis-press.com/journals/jsta Journal of Statistical Theory and Applications Vol. 19(1), March 2020, pp. 1–9 DOI: https://doi.org/10.2991/jsta.d.200224.005; ISSN 1538-7887 https://www.atlantis-press.com/journals/jsta Journal of Statistical Theory and Applications Vol. 19(1), March 2020, pp. 1–9 DOI: https://doi.org/10.2991/jsta.d.200224.005; ISSN 1538-7887 https://www.atlantis-press.com/journals/jsta Journal of Statistical Theory and Applications Vol. 19(1), March 2020, pp. 1–9 DOI: https://doi.org/10.2991/jsta.d.200224.005; ISSN 1538-7887 https://www.atlantis-press.com/journals/jsta Corresponding author. Email: ma.hashempour@hormozgan.ac.ir Pdf_Folio:1 Keywords Proportional hazard rates Hypotheses testing Sequential order statistics Estimation Mathematics Subject Classification: 62P30; 62N05; 62G30. © 2020 The Authors. Published by Atlantis Press SARL. © 2020 The Authors. Published by Atlantis Press SARL. This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/). © 2020 The Authors. Published by Atlantis Press SARL. This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/). y This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/) ABSTRACT ART IC L E I N F O Article History Received 06 Jan 2018 Accepted 19 Aug 2019 Keywords Proportional hazard rates Hypotheses testing Sequential order statistics Estimation Mathematics Subject Classification: 62P30; 62N05; 62G30. ART IC L E I N F O Article History Received 06 Jan 2018 Accepted 19 Aug 2019 Keywords Proportional hazard rates Hypotheses testing Sequential order statistics Estimation This paper deals with analyzing dynamic engineering systems consisting of independent components. The failure of a compo- nents causes more load on the surviving components. This property is modeled by a power trend conditionally proportional hazard rates. For modeling system lifetimes, the theory of sequential order statistics can be used. Sequential order statistics coming from heterogeneous exponential distributions are considered. The maximum likelihood and Bayesian estimates of the parameters are obtained in different cases. The generalized likelihood ratio and the Bayesian tests are also derived for testing homogeneity of the baseline exponential component lifetimes arising from s independent engineering systems. Keywords Proportional hazard rates Hypotheses testing Sequential order statistics Estimation Majid Hashempour1,, Mahdi Doostparast2 1 Department of Statistics, School of Sciences, University of Hormozgan, Bandar Abbas, Iran 2 Department of Statistics, School of Mathematical Sciences, Ferdowsi University of Mashhad, Mashhad, Iran 2. LIKELIHOOD ANALYSIS In the sequel, suppose that we have observed s (≥2) independent heterogeneous SOS samples. The available data may be represented as In the sequel, suppose that we have observed s (≥2) independent heterogeneous SOS samples. The ava ndependent heterogeneous SOS samples. The available data may be represented as x = [[xij]]i=1,⋯,s,j=1,⋯,r, (2) (2) where the i-th row of the matrix x in (2) denotes the SOS sample coming from the i-th population. The likelihood function (LF) for a single SOS sample is L(F1, ⋯, Fr; x) = Γ(n + 1) Γ(n −r + 1) { r−1 ∏ j=1 fj(xj) ( Fj(xj) Fj+1(xj) )n−j } fr(xr)Fr(xr)n−r. (3) (3) (3) ailable data given by (2) is then obtained from (3) as e available data given by (2) is then obtained from (3) as Therefore, the LF of the available data given by (2) is then obtained from (3) as Therefore, the LF of the available data given by (2) is then obtained from (3) as L(; x) = ( n! (n −r)! )s s ∏ i=1 ⎛ ⎜ ⎜ ⎜⎝ r−1 ∏ j=1 ⎡⎢⎢ ⎣ f [i] j (xij) ⎛ ⎜ ⎜⎝ F [i] j (xij) F [i] j+1(xij) ⎞ ⎟ ⎟⎠ n−j ⎤⎥⎥ ⎦ f [i] r (xir)F [i] r (xir)n−r ⎞ ⎟ ⎟ ⎟⎠ , (4) (4) where = {F [i] j , i = 1, ⋯, s, j = 1, ⋯, r} and for i = 1, ⋯, s, j = 1, ⋯, r, F [i] j (x) = 1 −F [i] j (x), and F [i] j calls for the common DF of component lifetimes of the i−th dynamic system. For more details, see Cramer and Kamps [6,12] and Hashempour and Doostparast [11]. Suppose that, the baseline DF of the i-th population (i = 1, ⋯, s) follows the one-parameter exponential distribution with the mean 𝜎i. Substituting Equation (1) into Equation (4) and under the earlier mentioned PTCPHR model in Section 1, the LF of the available data reads L(𝜎1, ⋯, 𝜎s, a; x) = Bs ( r ∏ j=1 a j )s s ∏ i=1 ( 1 𝜎i )r s ∏ i=1 r ∏ j=1 exp {− xijmj 𝜎i } = Bs ( a sr(r+1) 2 ) ( s ∏ i=1 𝜎i )−r exp {− s ∑ i=1 r ∑ j=1 (xijmj 𝜎i ) } , (5) (5) where B = (n! /(n −r)! 1. INTRODUCTION This process continues up to n −r + 1 (r = 1, ⋯, n) components with the common DF Fr work until the r-th failure occurs at time xr and hence the whole system fails. The mentioned system is called sequential r-out-of- n system or dynamic system and the system lifetime is then r-th observed component failure time, denoted by X⋆ (r). In the literature, (X⋆ (1), ⋯, X⋆ (n)) is called SOSs. Statistical properties of SOSs have been studied by Kamps [4,5], Cramer and Kamps [3,6], Balakrishnan et al. [7], Beutner and Kamps [8], Hashempour [9], Bedbur M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 2 [10], Hashempour and Doostparast [11], and references therein. We considered the problem of estimating the parameters on the basis of s (≥2) independent SOSs samples under a new proposed power trend conditional proportional hazard rates (PTCPHR) model, defined by Fj(t) = 1 −(1 −F0(t))a j for j = 1, ⋯, r and a > 0, where the underlying CDF F0(t) is assumed to be the exponential distribution, i.e., F0(x; 𝜎) = 1 −exp {− ( x 𝜎 ) } , x > 0, 𝜎> 0. (1) (1) In this case, the hazard rate function of the CDF Fj, defined by hj(t) = fj(t)/Fj(t) for t > 0 and j = 1, ⋯, n, is proportional to the hazard rate function of the baseline DF F0, i.e., hj(t) = ajh0(t) for t > 0. In this case, the hazard rate function of the CDF Fj, defined by hj(t) = fj(t)/Fj(t) for t > 0 and j = 1, ⋯, n, is proportional to the hazard rate function of the baseline DF F0, i.e., hj(t) = ajh0(t) for t > 0. In this paper, we consider the problem of comparing non-homogeneous exponential populations on the basis of independent multiply SOS samples coming from non-homogeneous exponential populations under the abovementioned PTCPHR model. Thus, this paper is organized as follows: In Section 2, via the likelihood approach, statistical procedures including estimation, either point or interval, of the parameters as well as the problem of testing homogeneity of baseline exponential populations are considered. In Section 3, Bayesian approach is used for estimating parameters of interest. Finally, concluding remarks are given in Section 4. 2. LIKELIHOOD ANALYSIS (8) (8) One may maximize the function (8) with respect to the parameter a numerically to derive the ML estimate of the parameter a. Then substituting this estimate into (29) (the appendix), the ML estimate of ̂𝜎j, for i = 1, ⋯, s, is derived. One may maximize the function (8) with respect to the parameter a numerically to derive the ML estimate of the parameter a. Then substituting this estimate into (29) (the appendix), the ML estimate of ̂𝜎j, for i = 1, ⋯, s, is derived. Consider again the hypotheses testing problem (6). It is easy to verify that the unique ML estimates of the parameters under the null hypothesis H0 are given by ̂̂𝜎0 = 1 rs s ∑ i=1 r ∑ j=1 xij ̂m0,j = ̂a0 rs s ∑ i=1 r ∑ j=1 (n −j + 1) jDij, (9) (9) and and and ̂a0 = rs ∑s i=1 ∑r j=1(n −j + 1) jDij ̂̂𝜎0, (10) ̂a0 = rs ∑s i=1 ∑r j=1(n −j + 1) jDij ̂̂𝜎0, (10) ̂a0 = rs ∑s i=1 ∑r j=1(n −j + 1) jDij ̂̂𝜎0, (10) where ̂m0,j = (n −j + 1) j ̂a0 −(n −j)( j + 1) ̂a0, with convention ̂a0(r + 1) ≡0. Therefore, the generalized likelihood ratio test (GLRT) statistic for the hypotheses testing problem (6) is where ̂m0,j = (n −j + 1) j ̂a0 −(n −j)( j + 1) ̂a0, with convention ̂a0(r + 1) ≡0. Therefore, the generalized likelihood ratio test (GLRT) statistic for the hypotheses testing problem (6) is Λ2 = r ∏ j=1 ( ̂a0 ̂a )rs s ∏ i=1 ( ̂̂𝜎i ̂̂𝜎0 )r exp { s ∑ i=1 r ∑ j=1 ( ̂mj ̂̂𝜎i − ̂m0,j ̂̂𝜎0 ) xij} , (11) (11) e ̂mj = (n −j + 1) ̂𝛼j −(n −j) ̂𝛼j+1. The logarithm of the GLRT statistic Λ2 in Equation (11) reads where ̂mj = (n −j + 1) ̂𝛼j −(n −j) ̂𝛼j+1. The logarithm of the GLRT statistic Λ2 in Equation (11) reads re ̂mj = (n −j + 1) ̂𝛼j −(n −j) ̂𝛼j+1. The logarithm of the GLRT statistic Λ2 in Equation (11) reads j) ̂𝛼j+1. 2. LIKELIHOOD ANALYSIS After some algebraic calculations, the Hessian matrix sim H = ( B11 B12 B21 B22 ) , H = ( B11 B12 B21 B22 ) , (7) (7) where where B11 = diag { r 𝜎2 i − 2 ∑r k=1 ak(n −k + 1)Dij 𝜎3 i } i=1,⋯,s , B22 = diag {−sr a2j } j=1,⋯,r , B12 = BT 21 = [[ (n −j + 1)Dij 𝜎2 i ]] i=1,⋯,s,j=1,⋯,r , B11 = diag { r 𝜎2 i − 2 ∑r k=1 ak(n −k + 1)Dij 𝜎3 i } i=1,⋯,s , B22 = diag {−sr a2j } j=1,⋯,r , B12 = BT 21 = [[ (n −j + 1)Dij 𝜎2 i ]] i=1,⋯,s,j=1,⋯,r , where diag {ci}i=1,⋯,m stands for a diagonal matrix with elements c1, ⋯, cm on its main diagonal where diag {ci}i=1,⋯,m stands for a diagonal matrix with elements c1, ⋯, cm on its main diagonal. The Hessian matrix (7) is not necessary negative definite on the parameter space. Therefore, one needs to use numerically methods for maximizing the LF (5) with respect to a and 𝜎1, ⋯, 𝜎s. Note that the ML estimates of the parameters (if exist) are obtained numerically by solving the following likelihood equations: ⎧⎪ ⎨⎪ ⎩ ̂̂𝜎i = 1 r ∑r j=1 xij ̂mj = ̂a r ∑r j=1(n −j + 1) jDij, i = 1, ⋯, s, ̂a = rs ( ∑s i=1 ∑r j=1(n −j + 1) jDij/ ̂̂𝜎i )−1 . ⎧⎪ ⎨ ̂̂𝜎i = 1 r ∑r j=1 xij ̂mj = ̂a r ∑r j=1(n −j + 1) jDij, i = 1, ⋯, s, ⎨⎪ ⎩̂a = rs ( ∑s i=1 ∑r j=1(n −j + 1) jDij/ ̂̂𝜎i )−1 . ⎨⎪ ⎩̂a = rs ( ∑s i=1 ∑r j=1(n −j + 1) jDij/ ̂̂𝜎i )− Here another approach for deriving the ML estimates of the parameters is suggested which is numerically simple. Recall that the profile LF for the parameter a is derived by substituting ̂𝜎i given by (29), for i = 1, ⋯, s, into the LF (5) instead of 𝜎i, i.e., Lp(a; x) = L( ̂𝜎1, ⋯, ̂𝜎s, a; x). The logarithm of the profile LF is lp(a; x) = log Lp(a; x) = sr(r + 1) 2 log(a) −r s ∑ i=1 log (∑r j=1(n −j + 1)a jDij r ) −rs. 2. LIKELIHOOD ANALYSIS ) and mj = (n −j + 1)a j −(n −j)a j+1 with convention (n −r)ar+1 ≡0. For sake of brevity, we assumed that the proportional parameter a are the same among the s sequential r-out-of- n systems. In this section, we consider the problem of homogeneity testing on the basis of independent SOS samples from different exponential populations, i.e., ⎧ ⎨ ⎩ H0 ∶𝜎1 = ⋯= 𝜎s H1 ∶𝜎i ≠𝜎j ∃i ≠j. ⎧ ⎨ ⎩ H0 ∶𝜎1 = ⋯= 𝜎s H1 ∶𝜎i ≠𝜎j ∃i ≠j. (6) ⎧ ⎨ ⎩ H0 ∶𝜎1 = ⋯= 𝜎s H1 ∶𝜎i ≠𝜎j ∃i ≠j. (6) ⎨ ⎩H1 ∶𝜎i ≠𝜎j ∃i ≠j. In sequel, two cases are considered: (i) a known, and (ii) a unknown. If a is known, the results of Cramer and Kamps ([6], Chapters 4 and 5) can be used. Also, see Schenk [13]. A summarize for this case is given in the appendix A. Pdf_Folio:2 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 3 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 3 3 Suppose that the parameter a in Equation (5) is unknown. In this case, calculations are complicated. Here, the Hessian matrix (H) is defined by Suppose that the parameter a in Equation (5) is unknown. In this case, calculations are complicated. Here, the Hessian matrix (H) is defined by H = [[(𝜕2 log(L)/𝜕𝜃i𝜕𝜃j)1≤i,j≤s+1]], H = [[(𝜕2 log(L)/𝜕𝜃i𝜕𝜃j)1≤i,j≤s+1]], H = [[(𝜕2 log(L)/𝜕𝜃i𝜕𝜃j)1≤i,j≤s+1]], H = [[(𝜕2 log(L)/𝜕𝜃i𝜕𝜃j)1≤i,j≤s+1]], H = [[(𝜕2 log(L)/𝜕𝜃i𝜕𝜃j)1≤i,j≤s+1]], where 𝜃i = 𝜎i(1 ≤i ≤s) and 𝜃s+1 = a. After some algebraic calculations, the Hessian matrix simplifies to 1 = a. After some algebraic calculations, the Hessian matrix simplifies to where 𝜃i = 𝜎i(1 ≤i ≤s) and 𝜃s+1 = a. After some algebraic calculations, the Hessian matrix simplifies to where 𝜃i = 𝜎i(1 ≤i ≤s) and 𝜃s+1 = a. 2. LIKELIHOOD ANALYSIS The logarithm of the GLRT statistic Λ2 in Equation (11) reads log Λ2 = rs log ( ̂a0 ̂a ) + r s ∑ i=1 log ( ̂̂𝜎i ̂̂𝜎0 ) + s ∑ i=1 r ∑ j=1 ( ̂mj ̂̂𝜎i − ̂m0,j ̂̂𝜎0 ) xij. (12) (12) M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 4 The null hypothesis H0 rejects if (13) −2 log Λ2 > c, −2 log Λ2 > c, where c is a positive constant and determined on the basis of the test level (Lehmann and Romano [14]). where c is a positive constant and determined on the basis of the test level (Lehmann and Romano [14]). where c is a positive constant and determined on the basis of the test level (Lehmann and Romano [14]). Exact distribution of the statistic −2 log Λ2 in Equation (13) under the null hypothesis H0 is complicated and we could not obtained an explicit expression. This remains as an open problem. In practice, one may use numerical methods such as Monte Carlo simulation study to derive the threshold c in the rejection region (13). 3. SOS-BASED BAYES ANALYSIS Bayesian statistical inference on the basis of homogeneous multiply SOS samples have been considered in literature. For example, Mohie El-Din et al. [15] considered the problem of SOS-based Bayesian estimation and one-sample prediction under the CPHR model for Weibull and Pareto distributions. Schenk et al. [16] derived Bayes estimates of the parameters under the CPHR model when the baseline population follows the one- and two-parameter exponential distributions. Also, the Bayesian estimation and the two-sample Bayesian prediction based on homogeneous multiply SOS samples was studied with more details by Shafay et al. [17]. See also Hashempour and Doostparast [11]. In this section, we consider the problems of estimation, either point or interval, of parameters and hypotheses testing (6) on the basis of s independent heterogeneous SOS samples arising from the exponential populations under the PTCPHR model with known a. When the parameter a is unknown, the mentioned problems are complicated and works in this direction are under investigation and we hope to report findings in future. 3.1. Bayesian Point Estimates We here consider the problem of estimating unknown parameters via a strict Bayesian approach. To do this, we assume that a is known and suggest the conjugate prior distributions for the scale parameters 𝜎i, i = 1, ⋯, s, i.e., 𝜎i ∼IG(di, ci), i = 1, ⋯, s, (14) (14) 𝜎i ∼IG(di, ci), i = 1, ⋯, s, be independent random variables where IG(d, c) calls for the inverse gamma distribution with paramet be independent random variables where IG(d, c) calls for the inverse gamma distribution with parameters d and c and density function e IG(d, c) calls for the inverse gamma distribution with parameters d and c and density function be independent random variables where IG(d, c) calls for the inverse gamma distribution with parameters d and c and density function 𝜋(𝜎) = cd Γ(d)𝜎−(d+1) exp {− ( c 𝜎 ) } , 𝜎> 0, d > 0, c > 0. (15) (15) From Equation (14) and the LF (5), the joint posterior density function of 𝜎1, … , 𝜎s is readily obtained as From Equation (14) and the LF (5), the joint posterior density function of 𝜎1, … , 𝜎s is readily obtained as 𝜋(𝜎1, … , 𝜎s ∣x) ≡ s ∏ i=1 ( Bsa r(r+1) 2 c di i Γ(di)𝜎 −(di+r)−1 i ) × s ∏ i=1 ( exp {− (∑r j=1(n −j + 1)ajDij + ci 𝜎i ) } ) , (16) (16) which implies 𝜎i ∣x ∼IG ( di + r, r ∑ j=1 (n −j + 1)ajDij + ci ) , i = 1, ⋯, s. (17) (17) As we expected given x, the parameter 𝜎i are independent. Under the squared error loss function, the Bayes estimate of the parameter is the mean of the associate posterior distribution. Here, the Bayes estimate of 𝜎i is As we expected given x, the parameter 𝜎i are independent. Under the squared error loss function, the Bayes estimate of the parameter is the mean of the associate posterior distribution. Here, the Bayes estimate of 𝜎i is ̂𝜎i,B = ∑r j=1(n −j + 1)ajDij + ci di + r −1 = r ̂𝜎i + ci di + r −1, i = 1, ⋯, s, (18) (18) where ̂𝜎i is the ML estimate of 𝜎i given by Equation (29). Since the SOS samples are independent, the Bayes estimates in Equation (18) depend only on the respective samples. 3.2. Bayesian Test Under the null hypothesis H0 ∶𝜎1 = ⋯= 𝜎s, we assume that the common value of 𝜎i (i = 1, ⋯, s), say 𝜎, follows the IG(d0, c0)-distribution where d0 and c0 are known positive hyper parameters. Therefore, the Bayes factor (BF) is derived as (Berger [18]) Under the null hypothesis H0 ∶𝜎1 = ⋯= 𝜎s, we assume that the common value of 𝜎i (i = 1, ⋯, s), say 𝜎, follows the IG(d0, c0)-distribution where d0 and c0 are known positive hyper parameters. Therefore, the Bayes factor (BF) is derived as (Berger [18]) BF = c a0 0 Γ(sr + d0) Γ(d0) ( s ∏ i=1 (Ti + ci)di+1 dic di i ) ( s ∑ i=1 Ti + c0 )−(sr+d0) . (21) (21) Proof. Under the null hypothesis H0 in (6), the LF (5) reduces to L(𝜍; x) = ( n! (n −r)! )s ( r ∏ j=1 aj )s 𝜍−sr exp {−1 𝜍 s ∑ i=1 r ∑ j=1 xijmj} , and and 𝜋(x ∣H0) = ( n! (n −r)! )s ( r ∏ j=1 aj )s ∫ ∞ 0 𝜍−sr exp {− (E + c0 𝜍 ) } c d0 0 Γ(d0)𝜍−(d0+1)d𝜍 = ( n! (n −r)! )s ( r ∏ j=1 aj )s c d0 0 Γ(d0) ∫ ∞ 0 exp {− (E + c0 𝜍 ) } 𝜍−(sr+d0+1)d𝜍 where E = ∑s i=1 ∑r j=1 xijmj. If we use change of variable x = 1/𝜍, then where E = ∑s i=1 ∑r j=1 xijmj. If we use change of variable x = 1/𝜍, then where E = ∑s i=1 ∑r j=1 xijmj. If we use change of variable x = 1/𝜍, then where E = ∑s i=1 ∑r j=1 xijmj. If we use change of variable x = 1/𝜍, then 𝜋(x ∣H0) = ( n! (n −r)! )s ( r ∏ j=1 aj )s c d0 0 Γ(d0) ∫ ∞ 0 xsr+d0−1 exp {−(E + c0)x} dx = ( n! (n −r)! )s ( r ∏ j=1 aj )s c d0 0 Γ(d0) Γ(sr + d0) (E + c0)sr+d0 . (22) Under the alternative hypothesis H1 in (6), we have Under the alternative hypothesis H1 in (6), we have Under the alternative hypothesis H1 in (6), we have Under the alternative hypothesis H1 in (6), we have L(𝜍1, ..., 𝜍s; x) = ( n! (n −r)! 3.1. Bayesian Point Estimates • For r = n and a = 1, ̂𝜎i,n = ∑n j=1 xij/n and ̂𝜎i,B = ( ∑n j=1 xij + ci ) /(di + n −1), which are, respectively, the well-known ML and the Bayes estimates of the exponential parameters on the basis of a random sample of size n. 3.1. Bayesian Point Estimates (19) (19) • The Bayes estimates (18) is a weighted mean of the mean of the prior (14) and the ML estimate (29); i.e., ̂𝜎i,B = E (𝜎i) wi + (1 −wi) ̂𝜎i, where wi = (di −1) / (di + r −1); • The Bayes estimates (18) is a weighted mean of the mean of the prior (14) and the ML estimate (29); i.e., ̂𝜎i,B = E (𝜎i) wi + (1 −wi) ̂𝜎i, where wi = (di −1) / (di + r −1); • The Bayes estimate (18) is biased, i.e., Bias( ̂𝜎i,B) = (𝜎i(di −1) −ci) / (di + r −1); • The Bayes estimate (18) is biased, i.e., Bias( ̂𝜎i,B) = (𝜎i(di −1) −ci) / (di + r −1); • The Bayes estimate (18) is biased, i.e., Bias( ̂𝜎i,B) = (𝜎i(di −1) −ci) / (di + r −1); • The risk function of the Bayes estimates (18) is R( ̂𝜎i,B, 𝜎i) = 𝜎2 i ( r + (1 −di)2) + 2ci(1 −di)𝜎i + c2 i (di + r −1)2 ; (20) (20) The minimum of R( ̂𝜎i,B, 𝜎i), as a function of 𝜎i, occurs at point bi(di −1)/[(1 −di)2 + r]; • The minimum of R( ̂𝜎i,B, 𝜎i), as a function of 𝜎i, occurs at point bi(di −1)/[(1 −di)2 + r]; • The minimum of R( ̂𝜎i,B, 𝜎i), as a function of 𝜎i, occurs at point bi(di −1)/[(1 −di)2 + r]; • The minimum of R( ̂𝜎i,B, 𝜎i), as a function of 𝜎i, occurs at point bi(di −1)/[(1 −di)2 + r]; • For r = n and a = 1, ̂𝜎i,n = ∑n j=1 xij/n and ̂𝜎i,B = ( ∑n j=1 xij + ci ) /(di + n −1), which are, respectively, the well-known ML and the Bayes estimates of the exponential parameters on the basis of a random sample of size n. For r = n and a = 1, ̂𝜎i,n = ∑n j=1 xij/n and ̂𝜎i,B = ( ∑n j=1 xij + ci ) /(di + n −1), which are, respectively, Bayes estimates of the exponential parameters on the basis of a random sample of size n. 3.1. Bayesian Point Estimates Pdf_Folio:4 where ̂𝜎i is the ML estimate of 𝜎i given by Equation (29). Since the SOS samples are independent, the Bayes estimates in Equation (18) depend only on the respective samples. Pdf_Folio:4 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 5 5 Remark 3.1. For i = 1, ⋯, s: Remark 3.1. For i = 1, ⋯, s: Remark 3.1. For i = 1, ⋯, s: • The mode of the posterior distribution (17) is • The mode of the posterior distribution (17) is • The mode of the posterior distribution (17) is ̂̂𝜎i,B = ∑r j=1(n −j + 1)ajDij + ci di + r + 1 = r ̂𝜎i + ci di + r + 1. (19) ̂̂𝜎i,B = ∑r j=1(n −j + 1)ajDij + ci di + r + 1 = r ̂𝜎i + ci di + r + 1. 3.2. Bayesian Test (23) From Equations (22) and (23), the Bayes factor is derived as From Equations (22) and (23), the Bayes factor is derived as From Equations (22) and (23), the Bayes factor is derived as BF = c a0 0 Γ(sr + d0) Γ(d0) ⎛ ⎜ ⎜⎝ s ∏ i=1 (Ti + ci)di+1 dic di i ⎞ ⎟ ⎟⎠ ( s ∑ i=1 Ti + c0 )−(sr+d □ Under the “0 −Ki” loss function (Berger [18]), the Bayes test rejects the null hypothesis H0 in (6) if Under the “0 −Ki” loss function (Berger [18]), the Bayes test rejects the null hypothesis H0 in (6) if Under the “0 −Ki” loss function (Berger [18]), the Bayes test rejects the null hypothesis H0 in (6) if BF < (K0 k1 ) (𝜋1 𝜋0 ) , BF < (K0 k1 ) (𝜋1 𝜋0 ) , BF < (K0 k1 ) (𝜋1 𝜋0 ) , where 𝜋i and Ki, for i = 1, 2, are prior probability for the hypothesis Hi and the loss of the accepting Hi when Hj(j ≠i) is correct, respectively. where 𝜋i and Ki, for i = 1, 2, are prior probability for the hypothesis Hi and the loss of the accepting Hi when Hj(j ≠i) is correct, respectively. 3.2. Bayesian Test )s ( r ∏ j=1 𝛼j )s s ∏ i=1 { 1 𝜍i exp {− r ∑ j=1 xijmj 𝜍i }} = ( n! (n −r)! )s ( r ∏ j=1 aj )s s ∏ i=1 { 1 𝜍i exp {−Ti 𝜍i }} , 6 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 6 where Ti = ∑r j=1 xijmj, and Let ui = 1/𝜍i, then Let ui = 1/𝜍i, then 𝜋(x ∣H1) = ( n! (n −r)! )s ( r ∏ j=1 aj )s ⎛ ⎜ ⎜⎝ s ∏ i=1 c di i Γ(di) ⎞ ⎟ ⎟⎠ s ∏ i=1 × ∫ ∞ 0 ⋯∫ ∞ 0 u di+1−1 i exp {−(Ti + ci)} du1 ⋯dus = ( n! (n −r)! )s ( r ∏ j=1 aj )s ⎛ ⎜ ⎜⎝ s ∏ i=1 c di i Γ(di) ⎞ ⎟ ⎟⎠ s ∏ i=1 ( Γ(di + 1) (Ti + ci)di+1 ) = ( n! (n −r)! )s ( r ∏ j=1 aj )s s ∏ i=1 ( dic ai i (Ti + ci)adi+1 ) . 𝜋(x ∣H1) = ( n! (n −r)! )s ( r ∏ j=1 aj )s ⎛ ⎜ ⎜⎝ s ∏ i=1 c di i Γ(di) ⎞ ⎟ ⎟⎠ s ∏ i=1 × ∫ ∞ 0 ⋯∫ ∞ 0 u di+1−1 i exp {−(Ti + ci)} du1 ⋯dus = ( n! (n −r)! )s ( r ∏ j=1 aj )s ⎛ ⎜ ⎜⎝ s ∏ i=1 c di i Γ(di) ⎞ ⎟ ⎟⎠ s ∏ i=1 ( Γ(di + 1) (Ti + ci)di+1 ) = ( n! (n −r)! )s ( r ∏ j=1 aj )s s ∏ i=1 ( dic ai i (Ti + ci)adi+1 ) . 3.3. HPD Credible Sets nding highest posterior density (HPD) credible set is considered. To do this, from Equation (16), we see tha nding highest posterior density (HPD) credible set is considered. To do this, from Equation (16), we see tha n this section, the problem of finding highest posterior density (HPD) credible set is considered. To do this Ri = 2 (∑r j=1(n −j + 1)ajDij + ci 𝜎i ) \|\|x ∼𝜒2 2(di+r) . (24) Ri = 2 (∑r j=1(n −j + 1)ajDij + ci 𝜎i ) \|\|x ∼𝜒2 2(di+r) . (24) (24) equi-tailed credible set at level 𝛾for 𝜎i (i = 1, ⋯, s) is obtained as Therefore, an equi-tailed credible set at level 𝛾for 𝜎i (i = 1, ⋯, s) is obtained as Therefore, an equi-tailed credible set at level 𝛾for 𝜎i (i = 1, ⋯, s) is obtained as Ci(𝛾) = (∑r j=1(n −j + 1)ajDij + ci 𝜒2 2(di+r),(1+𝛾)/2 , ∑r j=1(n −j + 1)ajDij + ci 𝜒2 2(di+r),(1−𝛾)/2 ) . (25) (25) The HPD credible set for 𝜎i (i = 1, ⋯, s) is {𝜎i ∶𝜎 −(di+r)−1 i exp {− ∑r j=1(n −j + 1)ajDij + ci 𝜎i } > k} , (26) {𝜎i ∶𝜎 −(di+r)−1 i exp {− ∑r j=1(n −j + 1)ajDij + ci 𝜎i } > k} , (26) (26) M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 7 M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 7 7 where k is a constant determined by the condition on the level of the credible set, i.e., determined by the condition on the level of the credible set, i.e., where k is a constant determined by the condition on the level of the credible set, i.e., P ( 𝜎 −(di+r)−1 i exp {− (∑r j=1(n −j + 1)ajDij + ci 𝜎i ) } > k ∣x ) = 𝛾. 3.3. HPD Credible Sets (27) Hence, the HPD credible set for 𝜎i (i = 1, ⋯, s) in Equation (26) is an interval, say (Li, Ui), where Li and Ui are derived from the following non-linear equations: Hence, the HPD credible set for 𝜎i (i = 1, ⋯, s) in Equation (26) is an interval, say (Li, Ui), where Li and Ui are derived from the following non-linear equations: ⎧ ⎪⎪ ⎨ ⎪⎪ ⎩ (Ui Li )di+r+1 = exp { (Ui −Li UiLi ) Ψi(x, a)} F𝜒2 2(di+r) (2Ψi(x, a) Ui ) −F𝜒2 2(di+r) (2Ψi(x, a) Li ) = 𝛾i ( Γ(di) Γ(di + r)c −di i a r(r+1) 2 (4Ψi(x, a))di+r ) ⎨ ⎪⎪ ⎩ F𝜒2 2(di+r) (2Ψi(x, a) Ui ) −F𝜒2 2(di+r) (2Ψi(x, a) Li ) = 𝛾i ( Γ(di) Γ(di + r)c −di i a r(r+1) 2 (4Ψi(x, a))di+r ) where Ψi(x, a) = ∑r j=1(n −j + 1)ajDij + ci, and F𝜒𝜈(t) stands for the CDF of the 𝜒𝜈-distribution. For the proof see Appendix B. In practice, one may use numerical methods for solving the above equations. 4. CONCLUSIONS In this paper, based on independent SOSs coming from non-homogeneous exponential populations, the ML and the Bayes estimates of parameters were obtained on the basis of multiple SOS samples. The GLRT and Bayesian test were also derived for testing homogeneity of the exponential baseline populations. In Bayesian inference, when the parameter a is unknown, the mentioned problems are complicated and works in this direction are under investigation and we hope to report findings in future. The results of this paper may be extended in some directions. For example, the Bayes estimates may by obtained under some asymmetric loss functions such as linear-exponential loss function. REFERENCES 1. R.E. Barlow, F. Proschan, Statistical Theory of Reliability and Life Testing: Probability Models, second ed., Springer, New York, USA, 1981. 2. H.A. David, H.N. Nagaraja, Order Statistics, John Wiley & Sons, Inc., Hoboken, New Jersey, USA, 2003. h ( ) 1. R.E. Barlow, F. Proschan, Statistical Theory of Reliability and Life Testing: Probability Models, second ed., Springer, New York, USA, 1981. 2. H.A. David, H.N. Nagaraja, Order Statistics, John Wiley & Sons, Inc., Hoboken, New Jersey, USA, 2003. H.A. David, H.N. Nagaraja, Order Statistics, John Wiley & 3. E. Cramer, U. Kamps, Ann. Inst. Stat. Math. 48 (1996), 535–549. 4. U. Kamps, A Concept of Generalized Order Statistics, Teubner, Wiesbaden, Germany, 1995. 5. U. Kamps, J. Stat. Plan. Inference. 48 (1995), 1–23. 5. U. Kamps, J. Stat. Plan. Inference. 48 (1995), 1–23. 6. E. Cramer, U. Kamps, Ann. Inst. Stat. Math. 53 (2001), 307–324. 7. N. Balakrishnan, U. Kamps, M. Kateri, Ann. Inst. Stat. Math. 64 (2012), 302–318. 7. N. Balakrishnan, U. Kamps, M. Kateri, Ann. Inst. Stat. Math. 64 (2012), 302–318. 8. E. Beutner, U. Kamps, J. Stat. Plan. Inference. 139 (2009), 2963–2969. p J ( ) 9. M. Hashempour, Classical, Bayesian and Evidential Inferences Based on Sequential Order Statistics, PhD Thesis in Mathematical Statistics, Department of Statistics, Ferdowsi University of Mashhad, Mashhad, Iran, 2017. p 9. M. Hashempour, Classical, Bayesian and Evidential Inferences Based on Sequential Order Statistics, PhD Thesis in Mathe ayesian and Evidential Inferences Based on Sequential Order Statistics, PhD Thesis in Mathematical Statistics owsi University of Mashhad, Mashhad, Iran, 2017. 9. M. Hashempour, Classical, Bayesian and Evidential Inferences Based on Sequential Order Statistics, PhD Thesis in Mathematical Statistics, 9. M. Hashempour, Classical, Bayesian and Evidential Inferences Based on Sequent Department of Statistics, Ferdowsi University of Mashhad, Mashhad, Iran, 2017. p y q Department of Statistics, Ferdowsi University of Mashhad, Mashhad, Iran, 2017. 10. S. Bedbur, J. Stat. Plan. Inference. 140 (2010), 2520–2530. 11. M. Hashempour, M. Doostparast, Commun. Stat. Theor. Methods. 46 (2016), 8086–8100. 12. E. Cramer, U. Kamps, in: N. Balakrishnan, E. Rao (Eds.), Handbook of Statistics, Advances in Reliability, Elsevier, New York, USA, vol. 20, chap. 12, 2001, pp. 301–372. 12. E. Cramer, U. Kamps, in: N. Balakrishnan, E. Rao (Eds.), Handbook of Statistics, Advances in Reliability, Elsevier, New York, USA, vol. 20, chap. 12, 2001, pp. 301–372. 12. E. Cramer, U. Kamps, in: N. Balakrishnan, E. REFERENCES Rao (Eds.), Handbook of Statistics, Advances in Reliabili 12, 2001, pp. 301–372. pp 13. N. Schenk, Point Estimation with Sequential Order Statistics Based on Exponential Distributions, PhD Thesis, University of Oldenburg, Old- enburg, Germany, 2001. 13. N. Schenk, Point Estimation with Sequential Order Statistics Based on Exponential Distributions, PhD Thesis, University of Oldenburg, Old- enburg, Germany, 2001. 15. M.M. Mohie El-Din, M.M. Amein, M.S. Moham, J. Adv. Res. Appl. Math. 5 (2013), 80–96. 15. M.M. Mohie El-Din, M.M. Amein, M.S. Moham, J. Adv. Res. Appl. Math. 5 (2013), 80–96. 16. N. Schenk, M. Burkschat, E. Cramer, U. Kamps, J. Stat. Plan. Inference. 141 (2011), 1575–1587. 17. A.R. Shafay, N. Balakrishnan, K.S. Sultan, J. Stat. Comput. Simul. 84 (2014), 526–544. 8. J. Berger, Statistical Decision Theory and Baysian Analysis, second ed., Springer, New York, USA, 1985. 19. E.L. Lehmann, G. Casella, Theory of Point Estimation, second ed., Springer, New York, USA, 1998. M. Hashempour and M. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 8 APPENDIX A. Suppose that the parameter a in Equation (5) is known. Under the null hypothesis H0 in (6) (Cramer and Kamps [12]), the unique ML estimate of the common mean of the s exponential populations, say 𝜎0, is ̂𝜎0 = 1 rs s ∑ i=1 r ∑ j=1 (n −j + 1)ajDij = 1 rs s ∑ i=1 r ∑ j=1 xijmj, (28) (28) where Dij = xij −xi,j−1, for j = 1, ⋯, r. Here xi0 ∶= 0 for i = 1, ⋯, s. When the baseline exponential populations are heterogeneous, from Equation (28), the unique ML estimate of 𝜎i (i = 1, ⋯, s) is derived as ̂𝜎i = 1 r r ∑ j=1 (n −j + 1)ajDij = 1 r r ∑ j=1 xijmj. (29) ̂𝜎i = 1 r r ∑ j=1 (n −j + 1)ajDij = 1 r r ∑ j=1 xijmj. (29) (29) Corollary 1. Under the PTCPHR with the one-parameter exponential baseline CDF, Corollary 1. Under the PTCPHR with the one-parameter exponential baseline CDF, Corollary 1. Under the PTCPHR with the one-parameter exponential baseline CDF, Corollary 1. Under the PTCPHR with the one-parameter exponential baseline CDF, Ti = r ∑ j=1 (n −j + 1)ajDij ∼Γ(r, 𝜎i), i = 1, ⋯, s, (30) (30) where Γ(a, b) calls for the gamma distribution with shape and scale parameters a and b, respectively. Notice that ∑r j=1 xijmj = ∑r j=1(n −j + 1) ajDij, for i = 1, ⋯, s. Now consider the problem of hypotheses testing (6). The generalized likelihood ratio test (GLRT) statistic for testing the problem (6) is Now consider the problem of hypotheses testing (6). The generalized likelihood ratio test (GLRT) stati Λ1 = s ∏ i=1 ( ̂𝜎i ̂𝜎0 )r exp { s ∑ i=1 r ∑ j=1 ( 1 ̂𝜎i −1 ̂𝜎0 ) mjxij} , (31) (31) where Ω = {(𝜎1, ⋯, 𝜎s) ∶𝜎i > 0, i = 1, ⋯, s} =∶ℝ+s is the whole parameter space and Ω0 = {(𝜎1, … , 𝜎s) ∶𝜎1 = ⋯= 𝜎s} denotes the parameter space under the null hypothesis H0. After some algebraic manipulations, the logarithm of the GLRT statistic Λ1 given by Equation (31) reduces to log Λ1 = r s ∑ i=1 log ( sTi ∑s j=1 Tj ) , (32) (32) where Ti is defined by Equation (30) and ”log” stands for the natural logarithm. APPENDIX A. Under the null hypothesis H0 in (6) and the usual regularity conditions (see Lehmann and Cassella [19]), −2 log Λ1 has asymptotically the chi-square distribution with 1 degree of freedom. Thus, for large n, the rejection region of the GLR test of size 𝛾is where Ti is defined by Equation (30) and ”log” stands for the natural logarithm. Under the null hypothesis H0 in (6) and the usual regularity conditions (see Lehmann and Cassella [19]), −2 log Λ1 has asymptotically the chi-square distribution with 1 degree of freedom. Thus, for large n, the rejection region of the GLR test of size 𝛾is −2 log Λ1 > 𝜒2 1,1−𝛾, (33) (33) where 𝜒2 𝜐,𝛾is the 𝛾-th percentile of the chi-square distribution with 𝜐degrees of freedom. Also, the actual level of the GLRT may be obtained by conducting a Monte Carlo simulation study for given a and 𝜎. where 𝜒2 𝜐,𝛾is the 𝛾-th percentile of the chi-square distribution with 𝜐degrees of freedom. Also, the actual level of the GLRT may be obtained by conducting a Monte Carlo simulation study for given a and 𝜎. The observed Fisher Information, denoted by i( ̂𝜎1, ⋯, ̂𝜎s), on the basis of available SOSs data is equal to minus of the Hessian matrix evaluated at the MLEs of the parameters, i.e., i( ̂𝜎1, ⋯, ̂𝜎s) = [[(−𝜕2 log(L)/𝜕𝜎i𝜕𝜎j)1≤i,j≤s]]\|𝜍1= ̂𝜍1,⋯,𝜍s= ̂𝜍s. It is well known that the unique MLEs have asymptotically the multivariate normal distribution with mean vector (𝜎1, ⋯, 𝜎s) and the variance–covariance matrix [i( ̂𝜎1, ⋯, ̂𝜎s)]−1. Here, the observed Fisher Information for 𝜎i is derived as It is well known that the unique MLEs have asymptotically the multivariate normal distribution with mean vector (𝜎1, ⋯, 𝜎s) and the variance–covariance matrix [i( ̂𝜎1, ⋯, ̂𝜎s)]−1. Here, the observed Fisher Information for 𝜎i is derived as i( ̂𝜎1, ⋯, ̂𝜎s) = diag [[ r ˆ𝜎ij ]] . Notice that, by Equation (30), one can see that 2r ( ̂𝜎i/𝜎i) ∼𝜒2 2r. Hence, an approximate equi-tailed confidence interval for 𝜎i is Notice that, by Equation (30), one can see that 2r ( ̂𝜎i/𝜎i) ∼𝜒2 2r. Hence, an approximate equi-tailed confidence interval for 𝜎i is ( ̂𝜎i −z𝛾/2√ ̂𝜎2 i r , ̂𝜎i + z𝛾/2√ ̂𝜎2 i r ) , (34) ( ̂𝜎i −z𝛾/2√ ̂𝜎2 i r , ̂𝜎i + z𝛾/2√ ̂𝜎2 i r ) , (34) (34) M. Hashempour and M. APPENDIX A. Doostparast / Journal of Statistical Theory and Applications 19(1) 1–9 9 where z𝛾stands for the 𝛾-percentile of the standard normal distribution. Therefore, an equal-tail exact confidence interval at level 100𝛾% for 𝜎i (i = 1, ⋯, s) is ( 2r ̂𝜎i 𝜒2 2r,(1+𝛾)/2 , 2r ̂𝜎i 𝜒2 2r,(1−𝛾)/2 ) , ( 2r ̂𝜎i 𝜒2 2r,(1+𝛾)/2 , 2r ̂𝜎i 𝜒2 2r,(1−𝛾)/2 ) , (35) ( 2r ̂𝜎i 𝜒2 2r,(1+𝛾)/2 , 2r ̂𝜎i 𝜒2 2r,(1−𝛾)/2 ) , (35) (35) where 𝜒2 𝜈,p calls for the p-th percentile of the 𝜒2 𝜈-distribution. th percentile of the 𝜒2 𝜈-distribution. where 𝜒2 𝜈,p calls for the p-th percentile of the 𝜒2 𝜈-distribution. APPENDIX B. By definition, we have ∫ Ui Li r ∏ j=1 aj c di i Γ(di)𝜎 −(di+r)−1 i exp {− (∑r j=1 xijmj + ci 𝜎i ) } d𝜎i = 𝛾i. Let ti = 2Ψ(x, a)/𝜎i, then 𝛾i = ∫ 2Ψ(x,a) Li 2Ψ(x,a) Ui r ∏ j=1 aj c di i Γ(di) (2Ψ(x, a) ti )−(di+r)−1 exp {−ti 2 } ( −2Ψ(x, a) t2 i ) dti = r ∏ j=1 aj c di i Γ(di) ∫ 2Ψ(x,a) Li 2Ψ(x,a) Ui (2Ψ(x, a))−(di+r)−1 (ti)di+r+1 exp {−ti 2 } ( −2Ψ(x, a) t2 i ) dti = −a r(r+1) 2 c di i Γ(di)(2Ψ(x, a))−(di+r) ∫ 2Ψ(x,a) Li 2Ψ(x,a) Ui t (di+r)−1 i exp {−ti 2 } dti = −a r(r+1) 2 Γ(di + r) Γ(di) d ci i ( 2 Ψ(x, a) )(di+r) ∫ 2 Ψ(x,a) Li 2 Ψ(x,a) Ui t (di+r)−1 i Γ(di + r)2di+r exp {−ti 2 } dti = −Γ(di + r) Γ(di) c di i ( 2 Ψ(x, a) )(di+r) a r(r+1) 2 ( F𝜒2(r+di) (2Ψ(x, a) Li ) −F𝜒2(r+di) (2Ψ(x, a) Ui ) After some algebraic calculations, 𝛾i is simplified to After some algebraic calculations, 𝛾i is simplified to After some algebraic calculations, 𝛾i is simplified to After some algebraic calculations, 𝛾i is simplified to a r(r+1) 2 (Γ(di + r) Γ(di) c di i (4Ψi(x, a))di+r ) ( F𝜒2 2(di+r) (2Ψ(x, a) Ui ) −F𝜒2 2(di+r) (2Ψ(x, a) Li )) = 𝛾i, and then F𝜒2 2(di+r) (2Ψi(x, a) Ui ) −F𝜒2 2(di+r) (2Ψi(x, a) Li ) = 𝛾i ( Γ(di) Γ(di + r)c −di i a r(r+1) 2 (4Ψi(x, a))di+r ) . Since 𝜋(Li ∣x) = 𝜋(Ui ∣x), then L −(di+r)−1 i exp {− (Ψi(x, a) L ) } = U −(di+r)−1 i Since 𝜋(Li ∣x) = 𝜋(Ui ∣x), then Since 𝜋(Li ∣x) = 𝜋(Ui ∣x), then L −(di+r)−1 i exp {− (Ψi(x, a) Li ) } = U −(di+r)−1 i exp {− (Ψi(x, a) Ui ) } and (Ui Li )di+r+1 = exp {Ψi(x, a) Li −Ψi(x, a) Ui } = exp {Ui −Li UiLi (Ψi(x, a)) .} . ⎧⎪ ⎨⎪ ⎩ (Ui Li )di+r+1 = exp { (Ui −Li UiLi ) Ψi(x, a)} F𝜒2 2(di+r) (2Ψi(x, a) Ui ) −F𝜒2 2(di+r) (2Ψi(x, a) Li ) = 𝛾i ( Γ(di) Γ(di + r)c −di i a r(r+1) 2 (4Ψi(x, a))di+r ) .
https://openalex.org/W4394726980	https://egusphere.copernicus.org/preprints/2024/egusphere-2024-208/egusphere-2024-208.pdf	English	null	Comment on egusphere-2024-208	null	2,024	cc-by	19,550	The cryostratigraphy of thermo-erosion gullies in the Canadian High Arctic demonstrates the resilience of permafrost Samuel Gagnon1,3, Daniel Fortier2,3, Étienne Godin3, Audrey Veillette2 1Département de Géographie, Université Laval, Québec, QC, G1V 0A6, Canada 2Département de Géographie, Université de Montréal, Montréal, QC, H2V 2B8, Canada 5 3Centre d’études nordiques, Université Laval, Québec, QC, G1V 0A6, Canada Correspondence to: Samuel Gagnon (samuel gagnon 1@gmail com) We concluded that although the formation of a TEG has profound effects on the short/medium term and leaves near 20 permanent geomorphological and hydrological scars in periglacial landscapes, on the long term, High Arctic permafrost can recover and return to geocryological conditions similar to those pre-dating the initial disturbance. This suggests that in stable environmental conditions undergoing natural variability, permafrost can persist longer than the geomorphological landforms in which it forms. The cryostratigraphy of thermo-erosion gullies in the Canadian High Arctic demonstrates the resilience of permafrost Samuel Gagnon1,3, Daniel Fortier2,3, Étienne Godin3, Audrey Veillette2 1Département de Géographie, Université Laval, Québec, QC, G1V 0A6, Canada 2Département de Géographie, Université de Montréal, Montréal, QC, H2V 2B8, Canada 5 3Centre d’études nordiques, Université Laval, Québec, QC, G1V 0A6, Canada Correspondence to: Samuel Gagnon (samuel gagnon 1@gmail com) Abstract. Thermo-erosion gullies (TEGs) are one of the most common forms of abrupt permafrost degradation. They generally form in ice-wedge polygonal networks where the interconnected troughs can channel runoff water. Although TEG generally form in ice-wedge polygonal networks where the interconnected troughs can channel runoff water. Although TEG can form within a single thawing season, it takes them several decades for their complete stabilization. While the inception 10 of TEGs has been examined in several studies, the processes of their stabilization remain poorly documented, especially the ground ice patterns that form following permafrost aggradation in stabilizing TEGs. For this study, we investigated the impacts of two TEGs in the Canadian High Arctic (Bylot Island, NU, Canada) on ground ice content, cryostratigraphic patterns, and geomorphology to examine permafrost recovery following thermal erosion in ice-wedge polygonal tundra. We can form within a single thawing season, it takes them several decades for their complete stabilization. While the inception 10 of TEGs has been examined in several studies, the processes of their stabilization remain poorly documented, especially the ground ice patterns that form following permafrost aggradation in stabilizing TEGs. For this study, we investigated the impacts of two TEGs in the Canadian High Arctic (Bylot Island, NU, Canada) on ground ice content, cryostratigraphic patterns, and geomorphology to examine permafrost recovery following thermal erosion in ice-wedge polygonal tundra. We sampled 17 permafrost cores from two TEGs – one still active (since 1999) and one stabilized (>100 years old) – to describe 15 the surface conditions, interpret the cryostratigraphic patterns, and characterize the state of permafrost after TEG stabilization. We observed that although the TEG caused discernable cryostratigraphic patterns, ground ice content and active layer thickness of the TEGs were comparable to measurements made in undisturbed conditions. We also noted that once stabilized, TEGs permanently (at the Anthropocene scale) alter landscape morphology and hydrological connectivity. We concluded that although the formation of a TEG has profound effects on the short/medium term and leaves near 20 permanent geomorphological and hydrological scars in periglacial landscapes, on the long term, High Arctic permafrost can recover and return to geocryological conditions similar to those pre-dating the initial disturbance. This suggests that in stable environmental conditions undergoing natural variability, permafrost can persist longer than the geomorphological landforms in which it forms. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. The cryostratigraphy of thermo-erosion gullies in the Canadian High Arctic demonstrates the resilience of permafrost Samuel Gagnon1,3, Daniel Fortier2,3, Étienne Godin3, Audrey Veillette2 1Département de Géographie, Université Laval, Québec, QC, G1V 0A6, Canada 2Département de Géographie, Université de Montréal, Montréal, QC, H2V 2B8, Canada 5 3Centre d’études nordiques, Université Laval, Québec, QC, G1V 0A6, Canada Correspondence to: Samuel Gagnon (samuel gagnon 1@gmail com) Thermal erosion and the formation of thermo-erosion gullies (TEGs) are one of the most common forms of abrupt permafrost degradation (Abbott et al., 2015; Kokelj and Jorgenson, 2013; Olefeldt et al., 2016; Turetsky et al., 2020). Thermal erosion, which is defined as the erosion of ice-rich permafrost by the combined thermal and mechanical action of moving water (Goudie, 2004), often occurs in ice-wedge polygonal networks due to the interconnected troughs that can channel runoff water (i.e., snowmelt, summer precipitations, melting of ground ice) along ice wedges (Fortier et al., 2007; 30 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. French, 2007; Godin and Fortier, 2012b). TEGs in polygonal networks are initiated by the infiltration of runoff water (e.g., snowmelt, rainfall) into open cracks and cavities in the active layer and can develop over a single thawing season (e.g., Fortier et al. 2007). Over time, TEG can merge and form extensive gully systems which can eventually turn into thermo- erosion valleys (Fortier et al., 2007; French, 2017; Gagnon and Allard, 2019; Morgenstern et al., 2021; Arė et al., 1979; French, 2007; Godin and Fortier, 2012b). TEGs in polygonal networks are initiated by the infiltration of runoff water (e.g., snowmelt, rainfall) into open cracks and cavities in the active layer and can develop over a single thawing season (e.g., Fortier et al. 2007). Over time, TEG can merge and form extensive gully systems which can eventually turn into thermo- erosion valleys (Fortier et al., 2007; French, 2017; Gagnon and Allard, 2019; Morgenstern et al., 2021; Arė et al., 1979; Godin and Fortier, 2012b). Although the formation of TEGs can significantly affect permafrost environments by altering 35 landscape morphology and ecosystem functions (Bowden et al., 2008; Liljedahl et al., 2007; Toniolo et al., 2009; Lacelle et al., 2010; Lantz and Kokelj, 2008; Levy et al., 2008; Morgenstern et al., 2021; Grosse et al., 2006, 2005), TEG are an integral part of landscape evolution in areas of continuous permafrost (Godin et al., 2014). However, the processes of stabilization of TEGs remain poorly documented, especially the ground ice patterns that Godin and Fortier, 2012b). Although the formation of TEGs can significantly affect permafrost environments by altering 35 landscape morphology and ecosystem functions (Bowden et al., 2008; Liljedahl et al., 2007; Toniolo et al., 2009; Lacelle et al., 2010; Lantz and Kokelj, 2008; Levy et al., 2008; Morgenstern et al., 2021; Grosse et al., 2006, 2005), TEG are an integral part of landscape evolution in areas of continuous permafrost (Godin et al., 2014). However, the processes of stabilization of TEGs remain poorly documented, especially the ground ice patterns that However, the processes of stabilization of TEGs remain poorly documented, especially the ground ice patterns that form following permafrost aggradation in stabilizing TEG. 1. Introduction 25 Thermal erosion and the formation of thermo-erosion gullies (TEGs) are one of the most common forms of abrupt permafrost degradation (Abbott et al., 2015; Kokelj and Jorgenson, 2013; Olefeldt et al., 2016; Turetsky et al., 2020). Thermal erosion, which is defined as the erosion of ice-rich permafrost by the combined thermal and mechanical action of moving water (Goudie, 2004), often occurs in ice-wedge polygonal networks due to the interconnected troughs that can Thermal erosion and the formation of thermo-erosion gullies (TEGs) are one of the most common forms of abrupt permafrost degradation (Abbott et al., 2015; Kokelj and Jorgenson, 2013; Olefeldt et al., 2016; Turetsky et al., 2020). Thermal erosion, which is defined as the erosion of ice-rich permafrost by the combined thermal and mechanical action of moving water (Goudie, 2004), often occurs in ice-wedge polygonal networks due to the interconnected troughs that can channel runoff water (i e snowmelt summer precipitations melting of ground ice) along ice wedges (Fortier et al 2007; 30 channel runoff water (i.e., snowmelt, summer precipitations, melting of ground ice) along ice wedges (Fortier et al., 2007; 30 1 Ground ice and its patterns of distribution, i.e., the 40 cryostratigraphy and cryostructures, can provide valuable information on the conditions of permafrost formation and its potential response to warming and thawing (Gilbert et al., 2016). Because they can provide insights about the ground freezing processes (Murton, 2013), cryostructures have been investigated in different environments such as coastlands (Murton and French, 1994; Burn, 1997), yedoma and drained thermokarst lake basin (Kanevskiy et al., 2011; Jongejans et form following permafrost aggradation in stabilizing TEG. Ground ice and its patterns of distribution, i.e., the 40 cryostratigraphy and cryostructures, can provide valuable information on the conditions of permafrost formation and its potential response to warming and thawing (Gilbert et al., 2016). Because they can provide insights about the ground freezing processes (Murton, 2013), cryostructures have been investigated in different environments such as coastlands (Murton and French, 1994; Burn, 1997), yedoma and drained thermokarst lake basin (Kanevskiy et al., 2011; Jongejans et al., 2018), lacustrine lowlands (Kanevskiy et al., 2014), and to study buried glacier ice (Coulombe et al., 2019). Such 45 characterization has yet to be done in TEGs. This would help understand and predict the response of permafrost to thermal erosion, i.e., its resilience, which we define permafrost resilience as the capacity of permafrost to recover to a pre- disturbance state (ground temperature below 0°C, similar ground ice contents and morphologies) following a disturbance and the associated perturbations (Chapin et al., 2009; Jorgenson et al., 2010; Piégay et al., 2020; Thoms et al., 2018). For this study, we investigated the geocryological and geomorphological processes of permafrost recovery 50 following thermal erosion in ice-wedge polygonal tundra. More specifically, this study aimed: 1) to characterize the cryostructures and cryostratigraphic patterns of two TEGs in polygonal tundra, 2) to determine the impacts of thermal erosion on the ground ice content and surface characteristics of the two TEGs and compare them with undisturbed surrounding polygons, 3) to describe the stages of TEG stabilization based on changes in surface geomorphology and 50 permafrost cryostratigraphy, and 4) to evaluate the stability and resilience of permafrost after thermal erosion based on 55 ground ice content and cryostratigraphic patterns. To our knowledge, this study is the first to report ground ice contents and cryostratigraphic data in TEGs, which were used to discuss the evolution of periglacial landscapes following thermal erosion in areas with cold ice-rich permafrost. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Figure 1: Location of the thermo-erosion gullies (TEG) in the Qarlikturvik valley, Bylot Island (NU, Canada). The cores (blue (TEG1) and green (TEG2) dots) and thaw front depth (TFD) transects (yellow dots, center of the transect) for each TEG are indicated in the bottom images. Cores were retrieved in undisturbed polygons, drained polygons, in the slopes of the gullies, and at the bottom of the gully channel (BGC) of TEG2. TEG2-TFD2 and TEG2-TFD3 are shown in the top right image, upstream of the boreholes. The Center for Northern Studies (CEN) research and meteorological stations are indicated with red dots. Sources: Esri ArcGIS Pro World Imagery basemap, Earthstar Geographics (15 m resolution; top images), and Maxar (30 cm resolution; bottom images), and the GIS User Community. Figure 1: Location of the thermo-erosion gullies (TEG) in the Qarlikturvik valley, Bylot Island (NU, Canada). The cores (blue 60 (TEG1) and green (TEG2) dots) and thaw front depth (TFD) transects (yellow dots, center of the transect) for each TEG are indicated in the bottom images. Cores were retrieved in undisturbed polygons, drained polygons, in the slopes of the gullies, and at the bottom of the gully channel (BGC) of TEG2. TEG2-TFD2 and TEG2-TFD3 are shown in the top right image, upstream of the boreholes. The Center for Northern Studies (CEN) research and meteorological stations are indicated with red dots. Sources: Esri ArcGIS Pro World Imagery basemap, Earthstar Geographics (15 m resolution; top images), and Maxar (30 cm resolution; bottom 65 images), and the GIS User Community. 60 65 The two TEGs studied for this research are located in the Qarlikturvik Valley on the southwestern plain of Bylot Island, Nunavut (Canada) (Figure 1). We selected Bylot Island for its cold ice-rich permafrost, the abundance of TEG developed in ice-wedge polygons in the Qarlikturvik Valley, and for the availability of meteorological and thermal data in the area. The 70 permafrost cryostratigraphy, and 4) to evaluate the stability and resilience of permafrost after thermal erosion based on 55 ground ice content and cryostratigraphic patterns. To our knowledge, this study is the first to report ground ice contents and cryostratigraphic data in TEGs, which were used to discuss the evolution of periglacial landscapes following thermal erosion in areas with cold ice-rich permafrost. 2 2 Based on meteorological data recorded near (73°9’21.6”N,79°57’24.8”W, 12 m a.s.l.) the Center for Northern Studies (CEN) research station on Bylot Island (Figure 1), (73 9 21.6 N,79 57 24.8 W, 12 m a.s.l.) the Center for Northern Studies (CEN) research station on Bylot Island (Figure 1), mean annual air temperature was -14.6°C for the 1995-2018 period (missing 1997, 1998, 2009, 2010, and 2014) (Climate 80 station data from Bylot Island in Nunavut, Canada, v. 1.11 (1992-2019). Nordicana D2, doi: 10.5885/45039SL- EE76C1BDAADC4890). Mean annual precipitations recorded at the Mattimatalik (Pond Inlet) Airport (72°41’N, 77°58’W, 61.6 m a.s.l.), which is about 85 km southwest of the study area, was 189 mm for the 1981-2010 period, 48% of which fell as rain between June and November (Historical Climate Data - Climatic data of Pond Inlet, Nunavut, 2022). Surface vegetation mean annual air temperature was -14.6°C for the 1995-2018 period (missing 1997, 1998, 2009, 2010, and 2014) (Climate 80 station data from Bylot Island in Nunavut, Canada, v. 1.11 (1992-2019). Nordicana D2, doi: 10.5885/45039SL- EE76C1BDAADC4890). Mean annual precipitations recorded at the Mattimatalik (Pond Inlet) Airport (72°41’N, 77°58’W, 61.6 m a.s.l.), which is about 85 km southwest of the study area, was 189 mm for the 1981-2010 period, 48% of which fell as rain between June and November (Historical Climate Data - Climatic data of Pond Inlet, Nunavut, 2022). Surface vegetation in the valley floor is mainly composed of sedges, grasses, and mosses typical of wetlands and environments with a high 85 water table (Perreault et al., 2016). The first TEG, TEG1, has been studied since its initiation in 1999 (R08 in Fortier et al., 2007; Godin and Fortier, 2012b; Rioux, 2020). TEG1 is located on the southern terrace of the Qarlikturvik Valley, about 800 m east of the CEN research station and 5.5 km from the valley mouth (Figure 1). It cuts into the north-facing terrace, which is gently sloping (2%) toward the outwash river and about 20 m a.s.l. (Godin and Fortier, 2012b). The gully showed high rates of erosion 90 during the first years following its inception that decreased over the years (from 390 m a-1 in 1999 to 5 m a-1 for the 2013- 2017 period) (Fortier et al., 2007; Godin and Fortier, 2012b, a; Rioux, 2020). 2. Study site The two TEGs studied for this research are located in the Qarlikturvik Valley on the southwestern plain of Bylot Island, Nunavut (Canada) (Figure 1). We selected Bylot Island for its cold ice-rich permafrost, the abundance of TEG developed in ice-wedge polygons in the Qarlikturvik Valley, and for the availability of meteorological and thermal data in the area. The 70 70 3 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. ~17 km-long glacial valley is oriented east-northeast west-southwest with a floor about 4-5 km wide. Two glaciers (C-79 and C-93) at the valley head feed a proglacial river that runs through braided channels in an outwash plain (Inland Waters Branch, 1969). The plain is bordered on both sides by terraces composed of over 4 to 5 m of organic-rich silty aeolian deposits and, in some areas, alluvial gravelly silty sands, overlying glaciofluvial deposits (Allard, 1996; Fortier et al., 2006). ~17 km-long glacial valley is oriented east-northeast west-southwest with a floor about 4-5 km wide. Two glaciers (C-79 and C-93) at the valley head feed a proglacial river that runs through braided channels in an outwash plain (Inland Waters Branch, 1969). The plain is bordered on both sides by terraces composed of over 4 to 5 m of organic-rich silty aeolian deposits and, in some areas, alluvial gravelly silty sands, overlying glaciofluvial deposits (Allard, 1996; Fortier et al., 2006). Well-developed networks of low-centered ice-wedge polygons that have formed syngenetically during the Late Holocene as 75 well as thaw ponds, tundra lakes and pingos cover the terraces (Fortier and Allard, 2004). 75 Well-developed networks of low-centered ice-wedge polygons that have formed syngenetically during the Late Holocene as 75 well as thaw ponds, tundra lakes and pingos cover the terraces (Fortier and Allard, 2004). The region has a high-Arctic climate and is underlain by continuous permafrost estimated to be more than 400 m in thickness in some areas (Maxwell 1980; Smith and Burgess 2002) Based on meteorological data recorded near Well-developed networks of low-centered ice-wedge polygons that have formed syngenetically during the Late Holocene as 75 well as thaw ponds, tundra lakes and pingos cover the terraces (Fortier and Allard, 2004). The region has a high-Arctic climate and is underlain by continuous permafrost estimated to be more than 400 m in thickness in some areas (Maxwell, 1980; Smith and Burgess, 2002). Based on meteorological data recorded near (73°9’21.6”N,79°57’24.8”W, 12 m a.s.l.) the Center for Northern Studies (CEN) research station on Bylot Island (Figure 1), The region has a high-Arctic climate and is underlain by continuous permafrost estimated to be more than 400 m in thickness in some areas (Maxwell, 1980; Smith and Burgess, 2002). https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. TEG1 had a main axis length of 863±4 m and formed a hydrological network of 33 458±379 m2 in 2017 (Rioux, 2020), and was still active in 2021 during fieldwork in the area. The second TEG, TEG2 (R05 in Godin and Fortier, 2012a; Veillette, 2019), is also located on the southern terrace of 90 the valley, about 2.8 km east of the CEN research station and 7.4 km from the valley mouth (Figure 1). TEG2 had a main 95 axis length of 1899 m and a surface area of 29 485 m2 in 2010 (Godin and Fortier, 2012a). The head TEG2 is located at the junction between two alluvial fans that formed from sediments eroded in the deep ravines incising the plateau of sedimentary rocks (Veillette, 2019). TEG2 cuts through the polygonal terrace (1.2% slope) before reaching the proglacial river. Aerial photos (1958, 1961, 1972, 1982) and satellite imagery (2007, 2010) showed that TEG2 was already stabilized in 1958 ant therefore much older than 65 years old (1958-2023) (Godin and Fortier, 2012a; Veillette, 2019). Based on stabilization rates 100 by Godin and Fortier (2012a), the gully is likely >100 years old. therefore much older than 65 years old (1958-2023) (Godin and Fortier, 2012a; Veillette, 2019). Based on stabilization rates 100 by Godin and Fortier (2012a), the gully is likely >100 years old. 4 3.1 Fieldwork and laboratory analyses Fieldwork was carried out between June and August in 2013, 2016, 2017 and 2018. To assess Fieldwork was carried out between June and August in 2013, 2016, 2017 and 2018. To assess the impacts of the TEG on the permafrost thermal regime, thaw front depth (TFD) was measured at different locations at TEG1 and TEG2 (Figure 1). TFD 105 was obtained by probing to refusal with a graduated steel rod in 2017 from 13 June to 29 July, and in 2018 from 12 June to 16 August. Measurements were made every 2-3 days along transects of ~30 points (10 to 30 m in length) perpendicular to the TEG and averaged into three sections (shoulders, slopes, and channel; Figure 2). TFD was also calculated from temperatures (Supplementary material 1) recorded by thermistor cables installed in two boreholes in undisturbed low- g p permafrost thermal regime, thaw front depth (TFD) was measured at different locations at TEG1 and TEG2 (Figure 1). TFD 105 was obtained by probing to refusal with a graduated steel rod in 2017 from 13 June to 29 July, and in 2018 from 12 June to 16 August. Measurements were made every 2-3 days along transects of ~30 points (10 to 30 m in length) perpendicular to the TEG and averaged into three sections (shoulders, slopes, and channel; Figure 2). TFD was also calculated from temperatures (Supplementary material 1) recorded by thermistor cables installed in two boreholes in undisturbed low- centered polygons, one adjacent to TEG1 (IP-A, see Godin et al. 2016) and the other one about two kilometres west of TEG1 110 (BYLOTPD; Allard et al., 2020). The tundra polygons near the gullies were described based on their shape (low-centered, flat, high-centered) and the drainage conditions (standing water, mesic, drained). To measure ground ice content and analyse soil composition, the TEG and the surrounding polygons were drilled using a portable earth-drill, with a 10 cm-diameter core barrel. At TEG1, we drilled in 2013 along and in the TEG at three subsites of different ages (Figure 2). At each subsite, one centered polygons, one adjacent to TEG1 (IP-A, see Godin et al. 2016) and the other one about two kilometres west of TEG1 110 (BYLOTPD; Allard et al., 2020). The tundra polygons near the gullies were described based on their shape (low-centered, flat, high-centered) and the drainage conditions (standing water, mesic, drained). 3.1 Fieldwork and laboratory analyses To measure ground ice content and analyse soil composition, the TEG and the surrounding polygons were drilled using a portable earth-drill, with a 10 cm-diameter core barrel. At TEG1, we drilled in 2013 along and in the TEG at three subsites of different ages (Figure 2). At each subsite, one centered polygons, one adjacent to TEG1 (IP-A, see Godin et al. 2016) and the other one about two kilometres west of TEG1 110 (BYLOTPD; Allard et al., 2020). The tundra polygons near the gullies were described based on their shape (low-centered, flat, high-centered) and the drainage conditions (standing water, mesic, drained). To measure ground ice content and analyse soil composition, the TEG and the surrounding polygons were drilled using a portable earth-drill, with a 10 cm-diameter core barrel. At TEG1, we drilled in 2013 along and in the TEG at three subsites of different ages (Figure 2). At each subsite, one hole was drilled in an undisturbed polygon (TEG1-1Undist, TEG1-2Undist, TEG1-3Undist), one in a drained polygon adjacent to 115 the TEG (TEG1-1Drained, TEG1-2Drained, TEG1-3Drained), and one in the gully slope adjacent to the drained polygon (TEG1- 1Slope, TEG1-2Slope, TEG1-3Slope). At TEG2, we drilled at eight locations in 2016 to extract permafrost cores down to 1.15 to 2.12 m. The holes were located on a ~170 m transect going across TEG2 and sampled undisturbed polygons (TEG2-1Undist, TEG2-2Undist), drained polygons adjacent to the TEG (TEG2-1Drained, TEG2-2Drained), gully slopes (TEG2-1Slope, TEG2-2Slope, TEG2-3Slope), and the bottom of the gully channel (TEG2BGC) (Figure 1; Figure 2). All permafrost cores were kept frozen and 120 brought back to Université de Montréal for laboratory analyses. 5 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Figure 2. Conceptual cross-section of the two thermo-erosion gullies (TEG1 and TEG2), sections of the TEGs where thaw front depth was probed (only one side is annotated, but for slopes and shoulders, both sides were probed), and locations where we drilled into permafrost. The locations of the cores are relative to the cross-section of the gullies, for their exact location see Figure 1. Figure not to scale. Figure 2. Conceptual cross-section of the two thermo-erosion gullies (TEG1 and TEG2), sections of the TEGs where thaw front depth was probed (only one side is annotated, but for slopes and shoulders, both sides were probed), and locations where we drilled into permafrost. 3.2 Data analyses The weighted average VIC (VICwa) and OMC (OMCwa) were calculated for each cryostructure, cores from a same hole 140 (hereafter referred as cores), and TEG (Supplementary material 2). To facilitate comparison between cores, VICwa and OMCwa down to one meter were used (VICwa and OMCwa hereafter describe the weighted average down to 1 m). For TEG1- 1Undist and TEG1-1Drained, which were 95 cm and 96 cm deep, respectively, the bottom cryostructure/stratigraphy and associated VIC/OMC were assumed to be the same down to 100 cm deep. The weighted average VIC (VICwa) and OMC (OMCwa) were calculated for each cryostructure, cores from a same hole 140 (hereafter referred as cores), and TEG (Supplementary material 2). To facilitate comparison between cores, VICwa and OMCwa down to one meter were used (VICwa and OMCwa hereafter describe the weighted average down to 1 m). For TEG1- 1Undist and TEG1-1Drained, which were 95 cm and 96 cm deep, respectively, the bottom cryostructure/stratigraphy and associated VIC/OMC were assumed to be the same down to 100 cm deep. To produce the generalized stratigraphy of each core, we used the grain size statistics combined with the 145 photographs and CT-scans of the cores. Grain size analyses were performed with the GRADISTAT program (version 9.1) (Blott and Pye, 2001). Based on sediment size classes derived from sieving and the particle size analyzer, the program generated statistical parameters that were used to determine the textural groups and the types of sediment (Folk and Ward (1957) measures were used). We also performed statistical analyses to determine if VIC differed between the two TEG, between all sub-sites of the TEG, and between the position of the cores about the TEGs (i.e., undisturbed polygon, drained 150 polygon, gully slope, gully channel) using RStudio software (RStudio Inc., version 1.4.1103). This was done using VICwa down to 1 m of each core in a one-way analysis of variance (ANOVA) with a 95% confidence interval followed by a Tukey’s HSD analysis. between all sub-sites of the TEG, and between the position of the cores about the TEGs (i.e., undisturbed polygon, drained 150 polygon, gully slope, gully channel) using RStudio software (RStudio Inc., version 1.4.1103). This was done using VICwa down to 1 m of each core in a one-way analysis of variance (ANOVA) with a 95% confidence interval followed by a Tukey’s HSD analysis. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. and silt (≤63 µm) in each sample (Supplementary material 3). Clay (≤2 µm) was not described as it accounted for <1% in all samples. Organic matter content (OMC) was determined using the loss-on-ignition method (5-10g triplicates at 550°C for 5h, Gagnon and Allard, 2020). and silt (≤63 µm) in each sample (Supplementary material 3). Clay (≤2 µm) was not described as it accounted for <1% in all samples. Organic matter content (OMC) was determined using the loss-on-ignition method (5-10g triplicates at 550°C for 5h, Gagnon and Allard, 2020). 3.1 Fieldwork and laboratory analyses The locations of the cores are relative to the cross-section of the gullies, for their exact location see Figure Figure 2. Conceptual cross-section of the two thermo-erosion gullies (TEG1 and TEG2), sections of the TEGs where thaw front depth was probed (only one side is annotated, but for slopes and shoulders, both sides were probed), and locations where we drilled into permafrost. The locations of the cores are relative to the cross-section of the gullies, for their exact location see Figure 1. Figure not to scale. AS+128, Siemens®, resolution of 0.4 mm). The scans were used to characterize the cryostructures present in permafrost (Calmels and Allard, 2004). Cryostructures were described using a classification based on existing literature (French and Shur, 2010; Kanevskiy et al., 2013; Murton, 2013). Then, the cores were cut into samples corresponding to the main 130 cryostructures and sediment textures, and vacuum sealed in plastic bags to determine their total volume (Vt) using the water displacement method (Strauss et al., 2013). After, the samples were weighted, dried in an oven at 105°C until dry, and reweighted to determine the water content. The volume and mass of the samples were used to calculate bulk (ρbulk) and dry (ρdry) densities, and the mass of water for volumetric ice content (VIC; cm3ice cm-3soil Supplementary material 2). Soil composition was determined by calculating the proportion (% of total mass) of gravel (>2 mm), sand (>63 µm to ≤2 mm), 135 6 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. a) TEG1, near subsite TEG1-2, b) TEG1, near subsite TEG1-3. EG2, near TEG2_TFD2, about 400 m upstream of the boreholes. backpack on the pictures is about 80 cm long. Figure 3. Thermo-erosion gullies 1 (TEG1, a-b) and 2 (TEG2, c-d). a) TEG1, near subsite TEG1-2, b) TEG1, near subsite TEG1-3. c) TEG2, near the transect where the cores were drilled, and d) TEG2, near TEG2_TFD2, about 400 m upstream of the boreholes. All pictures were taken on 24 June 2018 and point South-East. The backpack on the pictures is about 80 cm long. 165 65 Figure 3. Thermo-erosion gullies 1 (TEG1, a-b) and 2 (TEG2, c-d). a) TEG1, near subsite TEG1-2, b) TEG1, near subsite TEG1-3. c) TEG2, near the transect where the cores were drilled, and d) TEG2, near TEG2_TFD2, about 400 m upstream of the boreholes. All pictures were taken on 24 June 2018 and point South-East. The backpack on the pictures is about 80 cm long. For both TEG1 and TEG2, the polygons adjacent to the TEG were drained, i.e., mesic to dry (TEG1-1-2-3Drained, , p yg j , , y ( , TEG2-1-2Drained), compared to other undisturbed polygons away from the gully incisions where standing water was common 170 (TEG1-1-2-3Undist, TEG2-1-2Undist). At TEG1, the drained polygons adjacent to the TEG were low-centered. At TEG2, the drained polygons were flat or high-centered while the undisturbed polygons were low-centered. In addition, the area encompassing the drained polygons along the TEG was also higher in elevation than the surrounding low-centered polygons (based on a normalized topographic position index, see Veillette, 2019). Over 70 thermal contraction cracks were found at TEG2-1-2Drained), compared to other undisturbed polygons away from the gully incisions where standing water was common 170 (TEG1-1-2-3Undist, TEG2-1-2Undist). At TEG1, the drained polygons adjacent to the TEG were low-centered. At TEG2, the drained polygons were flat or high-centered while the undisturbed polygons were low-centered. In addition, the area encompassing the drained polygons along the TEG was also higher in elevation than the surrounding low-centered polygons (based on a normalized topographic position index, see Veillette, 2019). Over 70 thermal contraction cracks were found at TEG2-1-2Drained), compared to other undisturbed polygons away from the gully incisions where standing water was common 170 (TEG1-1-2-3Undist, TEG2-1-2Undist). At TEG1, the drained polygons adjacent to the TEG were low-centered. 4.1 State of the TEG, polygon shape, and active layer thickness 155 The two TEG differed significantly in their surface characteristics and general shape. TEG1 was characterized by a narrow (~10 m) box-shaped main channel, i.e., high-angle slopes connecting to a flat gully channel bottom, and had many ramifications through the surrounding polygonal network. Erosion was apparent in all the channels, i.e., the sides of channels were steep and uneven with blocks of ground falling off, causing the soil to be exposed (Figure 3). It was also possible to see The two TEG differed significantly in their surface characteristics and general shape. TEG1 was characterized by a narrow (~10 m) box-shaped main channel, i.e., high-angle slopes connecting to a flat gully channel bottom, and had many ramifications through the surrounding polygonal network. Erosion was apparent in all the channels, i.e., the sides of channels were steep and uneven with blocks of ground falling off, causing the soil to be exposed (Figure 3). It was also possible to see thermokarst mounds, i.e., baydzherakhi, in the main channel of TEG1 where erosion isolated polygons. TEG2 was 160 characterized by a well-developed main channel with few ramifications. The main channel was wide (20-30 m from side to side) with a flat bottom and the slopes were smooth and at low angle (Figure 3). Both the sides and bottom of the gully channel and the sides were stabilized with a continuous vegetation cover except at the bottom of the gully channel where water was flowing. thermokarst mounds, i.e., baydzherakhi, in the main channel of TEG1 where erosion isolated polygons. TEG2 was 160 characterized by a well-developed main channel with few ramifications. The main channel was wide (20-30 m from side to side) with a flat bottom and the slopes were smooth and at low angle (Figure 3). Both the sides and bottom of the gully channel and the sides were stabilized with a continuous vegetation cover except at the bottom of the gully channel where water was flowing. 7 At TEG2, the drained polygons were flat or high-centered while the undisturbed polygons were low-centered. In addition, the area encompassing the drained polygons along the TEG was also higher in elevation than the surrounding low-centered polygons (based on a normalized topographic position index, see Veillette, 2019). Over 70 thermal contraction cracks were found at the bottom of the gully channel and on the slopes of TEG2 (Veillette, 2019). The cracks seemed to follow the same pre- 175 disturbance polygonal network as many cracks were extending from the troughs of the surrounding network. the bottom of the gully channel and on the slopes of TEG2 (Veillette, 2019). The cracks seemed to follow the same pre- 175 disturbance polygonal network as many cracks were extending from the troughs of the surrounding network. 8 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. TFD varied between the two TEG and depending on the location of the measurements in the TEG. At TEG1, TEG1_TFD1-2-3-4 averaged 26 cm, 46 cm, and 39 cm on 29 July 2017 for the shoulders, slopes, and bottom of the gully channel, respectively (Supplementary table 1). On 30 July 2018, TEG1_TFD1 and TEG1_TFD3 averaged 38 cm (shoulders), 46 cm (slopes), and 34 cm (bottom of gully channel). Although there was some variability (e.g., TEG1_TFD2 180 2018), the general trend in TFD at TEG1 was deepest on the slopes and shallowest on the shoulders, and measurements on the slopes and the gully channel were comparable to undisturbed TFD (BYLOTPD and IP-A) (Figure 4a-b-c-d-e-f). At TEG2, TEG2_TFD1-2-3 averaged 41 cm, 30 cm, and 56 cm on 30 July 2018 for the shoulders, slopes, and bottom of the gully channel, respectively (Supplementary table 1). All three locations at TEG2 showed the same trend: TFD was deepest at the bottom of the gully channel and shallowest on the slopes, which were similar to undisturbed TFD (Figure 4g-h-i). 185 Shoulder values showed more variability than at TEG1. the bottom of the gully channel and shallowest on the slopes, which were similar to undisturbed TFD (Figure 4g-h-i). 185 Shoulder values showed more variability than at TEG1. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/17 8/1/17 Active layer thickness (m) a) TEG1_TFD1 (2017) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/12/18 8/1/18 Active layer thickness (m) b) TEG1_TFD1 (2018) Shoulders Slopes Channel BYLOTPD IP-A 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/17 8/1/17 Active layer thickness (m) c) TEG1_TFD2 (2017) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/12/18 7/31/18 Active layer thickness (m) d) TEG1_TFD2 (2018) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/17 8/1/17 Active layer thickness (m) e) TEG1_TFD3 (2017) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/17 8/1/17 Active layer thickness (m) f) TEG1_TFD4 (2017) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/17 8/1/17 Active layer thickness (m) c) TEG1_TFD2 (2017) / / / / 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/12/18 7/31/18 Active layer thickness (m) d) TEG1_TFD2 (2018) e) 9 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Figure 4. a) and b) TEG1_TFD1, near subsite TEG1-3, in 2017 and 2018, respectively, c) and d) TEG1_TFD2, near subsite TEG1-2, in 2017 and 2018 respectively, e) TEG1_TFD3, about 40 m upstream of TEG1_TFD2, f) TEG1_TFD4, about 90 m upstream of TEG1_TFD2, g) TEG2_TFD1, along the transect of the cores, h) TEG2_TFD2, about 400 m upstream of TEG2 TFD1 d i) TEG2 TFD3 b t 640 t f TEG2 TFD1 S fi 1 f th iti f th TFD t t 195 Figure 4. Thaw front depth (TFD) measured in 2017 (13 June to 29 July) and 2018 (12 June to 16 August) in cross-sectional 190 transects of the thermo-erosion gullies (TEG; curves) and TFD calculated from ground temperature measurements (red and yellow dots). a) and b) TEG1_TFD1, near subsite TEG1-3, in 2017 and 2018, respectively, c) and d) TEG1_TFD2, near subsite TEG1-2, in 2017 and 2018 respectively, e) TEG1_TFD3, about 40 m upstream of TEG1_TFD2, f) TEG1_TFD4, about 90 m upstream of TEG1_TFD2, g) TEG2_TFD1, along the transect of the cores, h) TEG2_TFD2, about 400 m upstream of TEG2_TFD1, and i) TEG2_TFD3, about 640 m upstream of TEG2_TFD1. See figure 1 for the position of the TFD transects. 195 4.2 Soil profiles The generalized stratigraphy of the cores at TEG1 showed that permafrost was characterized by silty sand or sandy silt sediment, with a larger proportion of peaty material than at TEG2 (Figure 5). There were also some gravelly sediment in cores TEG1-1Undist and TEG1-1Slope. OMC values varied between 3.1% and 47.8%, and the OMC weighted averages of each core varied between 5.6% and 20.0%. At TEG2, the cores were largely dominated by sandy silt sediment, sometime with 200 peaty material (Figure 6). OMC values of the different strata varied between 4.0% and 25.0%, but the weighted averages of each core remained between 6.2% and 15.3% (Table 1). In general, there were more peaty sediment in the undisturbed and d i d l th i th l d th b tt f th TEG The generalized stratigraphy of the cores at TEG1 showed that permafrost was characterized by silty sand or sandy silt sediment, with a larger proportion of peaty material than at TEG2 (Figure 5). There were also some gravelly sediment in cores TEG1-1Undist and TEG1-1Slope. OMC values varied between 3.1% and 47.8%, and the OMC weighted averages of each core varied between 5.6% and 20.0%. At TEG2, the cores were largely dominated by sandy silt sediment, sometime with 200 The generalized stratigraphy of the cores at TEG1 showed that permafrost was characterized by silty sand or sandy silt sediment, with a larger proportion of peaty material than at TEG2 (Figure 5). There were also some gravelly sediment in cores TEG1-1Undist and TEG1-1Slope. OMC values varied between 3.1% and 47.8%, and the OMC weighted averages of each The generalized stratigraphy of the cores at TEG1 showed that permafrost was characterized by silty sand or sandy silt sediment, with a larger proportion of peaty material than at TEG2 (Figure 5). There were also some gravelly sediment in cores TEG1-1Undist and TEG1-1Slope. OMC values varied between 3.1% and 47.8%, and the OMC weighted averages of each core varied between 5.6% and 20.0%. At TEG2, the cores were largely dominated by sandy silt sediment, sometime with 200 peaty material (Figure 6). OMC values of the different strata varied between 4.0% and 25.0%, but the weighted averages of each core remained between 6.2% and 15.3% (Table 1). In general, there were more peaty sediment in the undisturbed and drained polygons than in the slopes and the bottom of the TEG. 4.2 Soil profiles core varied between 5.6% and 20.0%. At TEG2, the cores were largely dominated by sandy silt sediment, sometime with 200 peaty material (Figure 6). OMC values of the different strata varied between 4.0% and 25.0%, but the weighted averages of each core remained between 6.2% and 15.3% (Table 1). In general, there were more peaty sediment in the undisturbed and drained polygons than in the slopes and the bottom of the TEG. A total of 10 main cryostructures were identified in the cores of both TEG: interstitial (pore ice not visible to the unaided eye), interstitial visible (pore ice visible), crustal, ice veins, layered, lenticular (includes micro-lenticular 205 cryostructures and non-parallel ice lenses), isolated ice lens, organic matrix (ice-saturated peat), reticulate, and suspended (Figures 5 and 6, Table 2). Reticulate, suspended, organic matrix, and layered cryostructures accounted for 75.3% of the total length of all cores, while the seven other cryostructures covered the remaining length. Most of the cryostructures predominantly developed in sandy silt sediment (Table 2 and Supplementary table 2), except for crustal (silty sandy gravel) and interstitial/interstitial visible (silty sand) cryostructures. The poorly defined sections corresponded to transitional 210 cryostructures, often composed of organic matter unevenly mixed with mineral sediment. VIC within cryostructures varied between 28.9% and 100.0%, and the VIC range (max-min) of the cryostructures averaged 27% (Table 2). The layered and reticulate cryostructures showed the widest VIC ranges (respectively 52.2% and 46.2%). VICwa of all cores ranged between 47.0% and 81.5% and averaged 69.3% (Table 1). There was no statistical difference (p-value>0.05) in VICwa between the TEG1 and TEG2 between any of the sub sites or between the positions of the cores Cores in undisturbed polygons (TEG1 215 and interstitial/interstitial visible (silty sand) cryostructures. The poorly defined sections corresponded to transitional 210 cryostructures, often composed of organic matter unevenly mixed with mineral sediment. VIC within cryostructures varied between 28.9% and 100.0%, and the VIC range (max-min) of the cryostructures averaged 27% (Table 2). The layered and reticulate cryostructures showed the widest VIC ranges (respectively 52.2% and 46.2%). VICwa of all cores ranged between 47.0% and 81.5% and averaged 69.3% (Table 1). There was no statistical difference (p-value>0.05) in VICwa between the and interstitial/interstitial visible (silty sand) cryostructures. The poorly defined sections corresponded to transitional 210 cryostructures, often composed of organic matter unevenly mixed with mineral sediment. Thaw front depth (TFD) measured in 2017 (13 June to 29 July) and 2018 (12 June to 16 August) in cross-sectional 190 transects of the thermo-erosion gullies (TEG; curves) and TFD calculated from ground temperature measurements (red and yellow dots). a) and b) TEG1_TFD1, near subsite TEG1-3, in 2017 and 2018, respectively, c) and d) TEG1_TFD2, near subsite TEG1-2, in 2017 and 2018 respectively, e) TEG1_TFD3, about 40 m upstream of TEG1_TFD2, f) TEG1_TFD4, about 90 m upstream of TEG1_TFD2, g) TEG2_TFD1, along the transect of the cores, h) TEG2_TFD2, about 400 m upstream of TEG2_TFD1, and i) TEG2_TFD3, about 640 m upstream of TEG2_TFD1. See figure 1 for the position of the TFD transects. 195 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) g) TEG2_TFD1 (2018) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) h) TEG_TFD2 (2018) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) i) TEG2_TFD3 (2018) Fi 4 Th f t d th (TFD) d i 2017 (13 J t 29 J l ) d 2018 (12 J t 16 A t) i ti l 0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) g) TEG2_TFD1 (2018) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) h) TEG_TFD2 (2018) 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) i) TEG2_TFD3 (2018) 8 (12 J t 16 A t) i ti l 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 6/13/18 8/1/18 Active layer thickness (m) i) TEG2_TFD3 (2018) 8/1/18 FD) measured in 2017 (13 June to 29 July) and 2018 (12 June to 16 August) in cross-sectional lli (TEG ) d TFD l l t d f d t t t ( d d Figure 4. Thaw front depth (TFD) measured in 2017 (13 June to 29 July) and 2018 (12 June to 16 August) in cross-sectional 190 transects of the thermo-erosion gullies (TEG; curves) and TFD calculated from ground temperature measurements (red and yellow dots). https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. 1-2-3Undist, TEG2-1-2Undist) were mostly composed of layered, organic matrix, and reticulate cryostructures (Supplementary table 3). Cores in drained polygons (TEG1-1-2-3Drained, TEG2-1-2Drained) showed similar cryostructures, as well as some ice enrichment closer to the surface. In the slopes of the gullies, the cores (TEG1-1-2-3Slope, TEG2-1-2-3Slope) often displayed inclined layers of sediment/ice/air bubbles and were dominated by the reticulate cryostructure. 1-2-3Undist, TEG2-1-2Undist) were mostly composed of layered, organic matrix, and reticulate cryostructures (Supplementary table 3). Cores in drained polygons (TEG1-1-2-3Drained, TEG2-1-2Drained) showed similar cryostructures, as well as some ice enrichment closer to the surface. In the slopes of the gullies, the cores (TEG1-1-2-3Slope, TEG2-1-2-3Slope) often displayed inclined layers of sediment/ice/air bubbles and were dominated by the reticulate cryostructure. Table 1. Weighted average and range (minimum/maximum) of volumetric ice content (VIC) and organic matter content (OMC) 220 down to a depth of 1 m (excluding the surface thawed layer) for the two thermo-erosion gullies (TEG1 and TEG2) and for each borehole. Table 1. Weighted average and range (minimum/maximum) of volumetric ice content (VIC) and organic matter content (OMC) 20 down to a depth of 1 m (excluding the surface thawed layer) for the two thermo-erosion gullies (TEG1 and TEG2) and for each borehole. 4.2.1 TEG1 4.2 Soil profiles VIC within cryostructures varied between 28.9% and 100.0%, and the VIC range (max-min) of the cryostructures averaged 27% (Table 2). The layered and reticulate cryostructures showed the widest VIC ranges (respectively 52.2% and 46.2%). VICwa of all cores ranged between 47.0% and 81.5% and averaged 69.3% (Table 1). There was no statistical difference (p-value>0.05) in VICwa between the TEG1 and TEG2 between any of the sub-sites or between the positions of the cores Cores in undisturbed polygons (TEG1- 215 TEG1 and TEG2, between any of the sub-sites, or between the positions of the cores. Cores in undisturbed polygons (TEG1- 215 10 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Table 2. Weighted average volumetric ice content (VICwa), total proportion, and proportion of the dominant soil compositions of each cryostructure identified. The total proportions were calculated by dividing the total length of a cryostructure by the total length of the cryostructures, and the proportions of the dominant soil compositions were calculated by dividing the total length of a type of sediment within a cryostructure by the total length of the cryostructure. Only VIC values falling within one cryostructure (i.e., not sections overlapping two or more cryostructures) were used in the total proportion to obtain representative values. 235 240 In drained polygons, the stratigraphic layers were mostly composed of silty sand and sandy silt, sometimes peaty (Figure 5b). TEG1-1Drained was composed of sandy silt and silty sand layers with peaty inclusions at the surface and at 66-79 cm. Most of the core was characterized by interstitial and organic matrix cryostructures, except for an ice-rich layer (suspended cryostructure) at 55-66 cm. TEG1-2Drained showed an ice-rich layer (suspended, reticulate cryostructures) at 44-88 cm with peaty (organic matrix cryostructure) sediment directly below it and at the surface of the core. The sediment in the 245 suspended cryostructure was almost entirely composed of silt. TEG1-3Drained was mostly composed of silty sand. Organic layers (organic matrix cryostructure) were observed both at the top and the bottom of the core. In the center, the only section composed of sandy silt was characterized by a layered cryostructure. cm with peaty (organic matrix cryostructure) sediment directly below it and at the surface of the core. The sediment in the 245 suspended cryostructure was almost entirely composed of silt. TEG1-3Drained was mostly composed of silty sand. Organic layers (organic matrix cryostructure) were observed both at the top and the bottom of the core. In the center, the only section composed of sandy silt was characterized by a layered cryostructure. Cores from the gully slopes showed contrasting stratigraphy and cryostratigraphy Cores from the gully slopes showed contrasting stratigraphy and cryostratigraphy (Figure 5c). TEG1-1Slope was composed of three main sections: a top section (4-28 cm) with peaty-silty-sandy sediments, a middle section (28-68 cm) with 250 silty sand sediment showing some cross-stratification, and a bottom section with coarser gravelly-silty-sandy sediments. 4.2.1 TEG1 Cores at had VICwa values ranging between 47.9% and 75.4% and averaged 65.5% (Figure 5 and Table 1). The cores 225 extracted from undisturbed polygons varied in soil composition, stratigraphy, and cryostratigraphy (Figure 5a). TEG1-1Undist contrasted with the other undisturbed cores due with its top layer of sandy gravel and by having the lowest VICwa (47.9%). The gravelly sediments were characterized by crustal cryostructures, and ice casts of some of the stones were visible, directly below them (e.g., ice lens at 47-48 cm deep). TEG1-2Undist was composed of layers of sandy silt and silty sand with a peaty inclusion (organic matrix cryostructure) at 63-78 cm. There were ice-rich layers above and below the peaty inclusion at 31- 230 63 cm and 88-115 cm. TEG1-3Undist was characterized by a top half organic layer (organic matrix cryostructure) in silty sand sediment overlying a bottom half of mineral sediment with a layered cryostructure. The ice layers of this cryostructure increased in thickness as the sediment changed from silty sand (62 to 90 cm) to sandy silt (>90 cm). 11 The top section was characterized by an organic matrix cryostructure, the middle section by mostly interstitial visible cryostructures, and the bottom section by mostly reticulate cryostructures. TEG1-2Slope displayed a thick ice lens at 110-120 cm separating the bottom silty sand sediment from the sandy silt sediment above and including a few oblique streaks of composed of three main sections: a top section (4-28 cm) with peaty-silty-sandy sediments, a middle section (28-68 cm) with 250 silty sand sediment showing some cross-stratification, and a bottom section with coarser gravelly-silty-sandy sediments. The top section was characterized by an organic matrix cryostructure, the middle section by mostly interstitial visible cryostructures, and the bottom section by mostly reticulate cryostructures. TEG1-2Slope displayed a thick ice lens at 110-120 cm separating the bottom silty sand sediment from the sandy silt sediment above and including a few oblique streaks of bubbles (<0.5 cm). The bottom sandy sediment was characterized by horizontal and cross-stratification at 120-135 cm 255 (interstitial visible), and inclined stratification (>135 cm) with layered cryostructure aligned with the stratification plane. Above the ice lens and up to 78 cm deep, the sediments and the cryostructures were inclined. TEG1-3Slope contrasted with the two other cores by being more uniform in soil composition and cryostructures. Except at the top of the core (<36 cm), bubbles (<0.5 cm). The bottom sandy sediment was characterized by horizontal and cross-stratification at 120-135 cm 255 (interstitial visible), and inclined stratification (>135 cm) with layered cryostructure aligned with the stratification plane. Above the ice lens and up to 78 cm deep, the sediments and the cryostructures were inclined. TEG1-3Slope contrasted with the two other cores by being more uniform in soil composition and cryostructures. Except at the top of the core (<36 cm), 12 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. TEG1-3Slope was entirely composed of reticulate cryostructures in sandy silt sediment. Starting at 76 cm, the sediment/ice layers were inclined. TEG1-3Slope was entirely composed of reticulate cryostructures in sandy silt sediment. Starting at 76 cm, the sediment/ice layers were inclined. 260 layers were inclined. 260 260 13 13 Figure 5. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %silt), soil stratigraphy, CT-scan imagery, soil cryostratigraphy, volumetric ice content (VIC; %), and geocryological layers for the cores at TEG1. a) Cores in undisturbed 5 polygons, b) cores in drained polygons, and c) cores in slopes of TEG1. While delineations were used to mark the transitions between the frozen part of the active layer, the transient layer, and the intermediate layer, color gradients were also used to denote the uncertainty in the limits between these layers. Refer to figure 6 for the legend. Figure 5. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %silt), soil stratigraphy, CT-scan imagery, soil cryostratigraphy, volumetric ice content (VIC; %), and geocryological layers for the cores at TEG1. a) Cores in undisturbed 65 polygons, b) cores in drained polygons, and c) cores in slopes of TEG1. While delineations were used to mark the transitions between the frozen part of the active layer, the transient layer, and the intermediate layer, color gradients were also used to denote the uncertainty in the limits between these layers. Refer to figure 6 for the legend. Figure 5. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %silt), soil stratigraphy, CT-scan imagery, soil cryostratigraphy, volumetric ice content (VIC; %), and geocryological layers for the cores at TEG1. a) Cores in undisturbed 265 polygons, b) cores in drained polygons, and c) cores in slopes of TEG1. While delineations were used to mark the transitions between the frozen part of the active layer, the transient layer, and the intermediate layer, color gradients were also used to denote the uncertainty in the limits between these layers. Refer to figure 6 for the legend. Figure 5. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %si 265 14 4.2.1 TEG2 Cores at TEG2 had VICwa values ranging between 64.2% and 81.4% and averaged 74.2% (Figure 6; Table 1). TEG2-1Undist 270 and TEG2-2Undist were mostly composed of sandy silt with some layers of peaty sandy silt (Figure 6a). TEG2-1Undist was characterized by a succession of organic layers (organic matrix) underlain by ice-rich layers (suspended, reticulate, layered). Similarly, TEG2-2Undist was dominated by an organic matrix cryostructure in the top section (13-52 cm) underlain by an ice- rich (suspended cryostructure) layer (61-81 cm) in sandy silt above a layered cryostructure with thin ice lenses in silty sand. Cores at TEG2 had VICwa values ranging between 64.2% and 81.4% and averaged 74.2% (Figure 6; Table 1). TEG2-1Undist 270 and TEG2-2Undist were mostly composed of sandy silt with some layers of peaty sandy silt (Figure 6a). TEG2-1Undist was characterized by a succession of organic layers (organic matrix) underlain by ice-rich layers (suspended, reticulate, layered). Similarly, TEG2-2Undist was dominated by an organic matrix cryostructure in the top section (13-52 cm) underlain by an ice- rich (suspended cryostructure) layer (61-81 cm) in sandy silt above a layered cryostructure with thin ice lenses in silty sand. In drained polygons (TEG2-1Drained, TEG2-2Drained), the stratigraphic layers were mostly composed of sandy silt, 275 sometimes peaty (Figure 6b). Both cores were characterized by an ice-rich layer (ice lens, suspended, reticulate, layered cryostructures) near the top of permafrost, a middle section with organic-rich sediment (organic matrix cryostructure), and a bottom section characterized by more developed cryostructures (layered, reticulate, ice lens) and ice lenses. In drained polygons (TEG2-1Drained, TEG2-2Drained), the stratigraphic layers were mostly composed of sandy silt, 275 sometimes peaty (Figure 6b). Both cores were characterized by an ice-rich layer (ice lens, suspended, reticulate, layered cryostructures) near the top of permafrost, a middle section with organic-rich sediment (organic matrix cryostructure), and a bottom section characterized by more developed cryostructures (layered, reticulate, ice lens) and ice lenses. 15 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. In the gully slopes, while TEG2-2Slope and TEG2-3Slope were entirely composed of sandy silt, TEG2-1Slope also contained peaty sections (Figure 6c). In TEG2-1Slope, it was possible to see alternation of peaty sediment (organic matrix 280 cryostructure) and ice-rich (suspended, reticulate, layered cryostructures) sandy silt sediment. TEG2-2Slope was dominated by ice-rich cryostructures (layered, reticulate, suspended) transitioning from one to the other. The reticulate and layered cryostructures were well-developed, i.e., had high VIC values and were characterized by thick ice inclusions (and vice versa for poorly-developed). TEG2-3Slope was dominated by reticulate cryostructures, though they were often poorly-developed. The layered cryostructures at the bottom of the core were very clearly defined. Oblique streaks of bubbles (<1.5 cm) were 285 found in the suspended and reticulate cryostructures of the slope cores, i.e., parallel to the slope surface. Finally, the stratigraphy of core at the bottom of the gully channel (TEG2BGC) contrasted with the other cores of TEG2 with more sandy sediments and some gravel at the bottom of the core (Figure 6d). The core had the lowest VICwa of all cores at TEG2 (Table 1), though this only included the top 40 cm of the core, which was not representative of deeper layers where ice-rich cryostructures (reticulate, suspended, layered) were found. Starting at 192 cm deep, inclined layers of 290 coarser sediments formed inclined cryostructures. The layered cryostructures at the bottom of the core were very clearly defined. Oblique streaks of bubbles (<1.5 cm) were 285 found in the suspended and reticulate cryostructures of the slope cores, i.e., parallel to the slope surface. Finally, the stratigraphy of core at the bottom of the gully channel (TEG2BGC) contrasted with the other cores of TEG2 with more sandy sediments and some gravel at the bottom of the core (Figure 6d). The core had the lowest VICwa of all cores at TEG2 (Table 1), though this only included the top 40 cm of the core, which was not representative of deeper layers where ice-rich cryostructures (reticulate, suspended, layered) were found. Starting at 192 cm deep, inclined layers of 290 coarser sediments formed inclined cryostructures. 16 17 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. 295 Figure 6. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %silt), soil stratigraphy, CT-scan imagery, soil cryostratigraphy, volumetric ice content (VIC; %), and geocryological layers for the boreholes at TEG2. a) Cores in undisturbed polygons, b) cores in drained polygons, c) cores in slopes of TEG2, and d) core at the bottom of the gully channel of TEG2. While delineations were used to mark the transitions between the frozen part of the active layer, the transient layer, and the intermediate layer, color gradients were also used to denote the uncertainty in the limits between these layers. 300 95 Figure 6. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %silt), soil stratigraphy, CT-scan imagery, soil cryostratigraphy, volumetric ice content (VIC; %), and geocryological layers for the boreholes at TEG2. a) Cores in 295 Figure 6. Organic matter content (OMC; %), soil composition (SC; %gravel, %sand, %silt), soil stratigraphy, CT-scan imagery, soil cryostratigraphy, volumetric ice content (VIC; %), and geocryological layers for the boreholes at TEG2. a) Cores in undisturbed polygons, b) cores in drained polygons, c) cores in slopes of TEG2, and d) core at the bottom of the gully channel of TEG2. While delineations were used to mark the transitions between the frozen part of the active layer, the transient layer, and the intermediate layer, color gradients were also used to denote the uncertainty in the limits between these layers. 00 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. 18 Suspended cryostructures were often found with grading cryostructures (e.g., reticulate, layered) before and/or after them leading to lower VIC. This reflect the surface conditions (e.g., water availability) during their formation (Paquette et al., 2022). 320 Except for the isolated ice lenses, the suspended and organic matrix cryostructures had the highest VICwa. The suspended cryostructure predominantly (>80%) formed in silty sediment, which is consistent with ice segregation during periods of rapid sedimentation of fine sediments and sustained water availability (Murton, 2013). Suspended cryostructures were often found with grading cryostructures (e.g., reticulate, layered) before and/or after them leading to lower VIC. This suggests an acceleration and/or slowing down of ice segregation and thus of sedimentation during their formation, or that 325 two freezing fronts contributed to the formation of the cryostructures. Organic matrix cryostructures were found in similar proportions in peaty sandy silt (60%) and peaty silty sand (40%), thus making the presence of organic material rather than a specific sediment size the main factor of formation. suggests an acceleration and/or slowing down of ice segregation and thus of sedimentation during their formation, or that 325 two freezing fronts contributed to the formation of the cryostructures. Organic matrix cryostructures were found in similar proportions in peaty sandy silt (60%) and peaty silty sand (40%), thus making the presence of organic material rather than a specific sediment size the main factor of formation. 5.1.1 Cryostructures Our results show that the development of cryostructures was largely influenced by soil composition and the stratigraphic sequences. The cold climate and permafrost at our study site also significantly affected freezing processes and thus the 305 formation of the cryostructures. Here were present a more detailed description of four out of 11 cryostructures as those four accounted for 76% of all cryostructures occurrences while most of the other ones accounted for <5%. 305 19 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. The reticulate cryostructure was the most abundant one and formed almost exclusively in sandy silt, resulting in VICwa similar to what others have reported (Kanevskiy et al., 2013; Shur et al., 2021; Thévenin, 2023). Since permafrost in the study area is syngenetic, reticulate cryostructures likely formed by the hydraulic fracturing of the sediment ahead of an 310 advancing freezing front followed by infilling of the resulting fractures with expelled and pressurized pore water (McRoberts and Nixon, 1975). Layered cryostructures formed in both predominantly silty (62%) and sandy (38%) sediments, though more well-developed (thicker) ice layers were generally found in silty sediment (e.g., TEG1-3Undist), reflecting how ice segregation was facilitated in finer sediments. They likely formed due to prolonged periods of ice segregation when the thaw front was stationary and water was available, mirroring sediment deposition patterns, and consistent with syngenetic 315 permafrost (French and Shur, 2010). Both the reticulate and layered cryostructures had a wide VIC range. While all cryostructures showed some variability in VIC, the reticulate and layered cryostructures were the most affected because of the variability in the thickness of the ice lattices (reticulate) or lenses (layered). Depending on the saturation of the host sediments, those cryostructures can be observed as much in ice-rich sediments as in sediment-poor ice in permafrost and fl t th f diti ( t il bilit ) d i th i f ti (P tt t l 2022) 320 front was stationary and water was available, mirroring sediment deposition patterns, and consistent with syngenetic 315 permafrost (French and Shur, 2010). Both the reticulate and layered cryostructures had a wide VIC range. While all cryostructures showed some variability in VIC, the reticulate and layered cryostructures were the most affected because of the variability in the thickness of the ice lattices (reticulate) or lenses (layered). Depending on the saturation of the host sediments, those cryostructures can be observed as much in ice-rich sediments as in sediment-poor ice in permafrost and reflect the surface conditions (e.g., water availability) during their formation (Paquette et al., 2022). 320 Except for the isolated ice lenses, the suspended and organic matrix cryostructures had the highest VICwa. The suspended cryostructure predominantly (>80%) formed in silty sediment, which is consistent with ice segregation during periods of rapid sedimentation of fine sediments and sustained water availability (Murton, 2013). 5.1.2 Cryostratigraphic patterns Most of the cryostructures in this study developed with syngenetic permafrost aggradation, irrespective of their position with 330 regards to the TEG. In general, we observed that the inclusion of organic matter in the soil profiles largely affected the cryostratigraphy due to the peaty material promoting ice aggradation below it. This was observed in cores from both TEG where it was common to see suspended, reticulate, layered, and isolated ice lens cryostructures either directly or 10-20 cm below the organic matter inclusion. This suggests that when organic matter accumulated, the top of permafrost not only moved up due to the new material accumulating (syngenetic permafrost growth), but also because of the low thermal 335 conductivity of organic matter compared to mineral sediment (Andersland and Ladanyi, 2004), which promoted ice enrichment in the layers below. Organic layers at depth also promoted ice enrichment in the mineral sediment above them by slowing down the progression of the thaw front and thus facilitating ice segregation. The cryostratigraphy of the cores in undisturbed polygons (except TEG1-1Undist) reflected the conditions during permafrost formation, i.e., reticulate, layered, and suspended cryostructures forming in fine-grained aeolian sediment 340 20 It is also possible that permafrost recovery leads to ice enrichment in more concentrated layers instead of throughout the entire permafrost column near the surface. However, new deeper cores would be necessary to confirm this. Second, drainage of the polygons bordering the TEG could also cause ice enrichment, specifically near the surface. This patterns deserve attention. First, looking at the cryostratigraphy beyond 1 m deep shows that re-aggradation of cold 360 permafrost following a disturbance may lead to ice enrichment, i.e., the recovery from the disturbance may lead to more ice than initially. It is also possible that permafrost recovery leads to ice enrichment in more concentrated layers instead of throughout the entire permafrost column near the surface. However, new deeper cores would be necessary to confirm this. Second, drainage of the polygons bordering the TEG could also cause ice enrichment, specifically near the surface. This could be a stabilization mechanism preventing further TEG enlargement. Finally, the slope cores had less peaty/organic 365 material than at other positions because it was likely eroded away during the formation of the TEG. Although permafrost re- aggradation promoted ice enrichment in the mineral sediment, the slopes are now prone to greater thaw subsidence as organic material requires an external load to cause thaw settlement comparable to mineral soil with similar VIC (Kanevskiy et al., 2012). could be a stabilization mechanism preventing further TEG enlargement. Finally, the slope cores had less peaty/organic 365 material than at other positions because it was likely eroded away during the formation of the TEG. Although permafrost re- aggradation promoted ice enrichment in the mineral sediment, the slopes are now prone to greater thaw subsidence as organic material requires an external load to cause thaw settlement comparable to mineral soil with similar VIC (Kanevskiy et al., 2012). could be a stabilization mechanism preventing further TEG enlargement. Finally, the slope cores had less peaty/organic 365 material than at other positions because it was likely eroded away during the formation of the TEG. Although permafrost re- aggradation promoted ice enrichment in the mineral sediment, the slopes are now prone to greater thaw subsidence as organic material requires an external load to cause thaw settlement comparable to mineral soil with similar VIC (Kanevskiy et al., 2012). https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. coupled with layers of organic layers promoting ice enrichment below them. Some of the cores in drained polygons (TEG1- 3Drained, TEG2-1Drained, TEG2-2Drained) also showed an ice-rich layer closer to the surface than in undisturbed polygons, which we attribute to permafrost aggradation after polygon drainage. This is supported by our TFD measurements at TEG1 showing shallower TFD on the shoulders of the TEG, by Godin et al. (2016) who noted that drained polygons had a thinner active layer causing upward (syngenetic) permafrost aggradation, and by Veillette’s (2019) results showing general uplifting 345 of the drained polygons along TEG2. In the slopes, the high values of TFD at TEG1 showed that after formation of the TEG, the slopes experience greater thawing than in natural conditions. This can be explained by the lack of vegetation protecting the ground surface from air convection and solar radiation which, after vegetation colonization, is substantially reduced. This is supported by the TFD at TEG2 where the slopes have values similar to undisturbed conditions. The oblique streaks of 345 bubbles and inclined ice layers indicated that the permafrost table moved up following an inclined plane likely parallel to the 350 slope surface during permafrost aggradation. For the core at the bottom of the gully channel, the stratigraphy and cryostratigraphy suggest that after the initial accumulation of coarser alluvial sediment, finer alluvial sediment accumulated relatively quickly, which led to rapid ice segregation. Based on the type of cryostructure (e.g., suspended, well-developed reticulate) and their high VIC, we estimate that permafrost was able to re-aggrade relatively quickly and with sufficient water following stabilization of the TEG. We 355 attribute this to the low permafrost temperatures of the study site (Allard et al., 2020), which promoted permafrost aggradation upward following TEG stabilization. There was ice enrichment in all sections affected by the TEG (drained, slope, bottom of the gully channel), which led to ice contents similar to undisturbed conditions. Though there was no statistical difference in VICwa between the position of cores in the TEG, between the TEG or between the sub-sites, some patterns deserve attention. First, looking at the cryostratigraphy beyond 1 m deep shows that re-aggradation of cold 360 permafrost following a disturbance may lead to ice enrichment, i.e., the recovery from the disturbance may lead to more ice than initially. They form a protective layer against annual thawing that would require a strong disturbance at the 380 surface to be degraded (Kanevskiy et al., 2016, 2017) and thus contribute to the long-term stabilization of the TEG. Finally, buried intermediate layers were interpreted in most of the TEG2 cores, but only in TEG1-2Undist, TEG1- 2Slope, and TEG1-2Drained at TEG1. Those intermediate layers were interpreted as buried because they represented different sequences of permafrost aggradation. In most cases, the buried intermediate layers were located under a layer of organic sequences of permafrost aggradation. In most cases, the buried intermediate layers were located under a layer of organic sediment, which promoted ice-enrichment in the mineral sediment below. Then, new mineral sediment accumulated over the 385 organic layer and were part of the active layer, until organic matter started to accumulate again. This resulted in the gradual decrease in active layer thickness, i.e., permafrost aggradation and formation of aggradational ice (Mackay, 1972; Burn, 1988; French and Shur, 2010), which led to the formation of a new intermediate layer in the mineral sediment below the newly accumulated organic matter. sediment, which promoted ice-enrichment in the mineral sediment below. Then, new mineral sediment accumulated over the 385 organic layer and were part of the active layer, until organic matter started to accumulate again. This resulted in the gradual decrease in active layer thickness, i.e., permafrost aggradation and formation of aggradational ice (Mackay, 1972; Burn, 1988; French and Shur, 2010), which led to the formation of a new intermediate layer in the mineral sediment below the newly accumulated organic matter. 5.1.3 Geocryological layers 370 All cores were characterized by a transient layer, which we expected because it is the result of the interannual variations in the ALT and all sites experience it (Shur, 1988; Shur et al., 2005; French and Shur, 2010). Most of the cores were interpreted to have an intermediate layer, except for TEG1-1Undist and TEG1-1Slope. Those two cores were both characterized by coarser 21 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. sediments than in other cores, resulting in higher hydraulic conductivities and therefore less water to form ice-rich layers. The average depth of the top of the intermediate layers from cores in drained polygons and slopes was shallower than in 375 undisturbed conditions, especially in the slope cores where the intermediate layers were also thicker. In drained polygons, this is consistent with the permafrost table moving up after the formation of a TEG causing drainage of the bordering polygons and their uplifting. For the slopes and TEG2BGC, the increased thickness of the combined transient and intermediate layers suggests that permafrost recovery following TEG stabilization can increase permafrost resilience to thawing compared The average depth of the top of the intermediate layers from cores in drained polygons and slopes was shallower than in 375 undisturbed conditions, especially in the slope cores where the intermediate layers were also thicker. In drained polygons, this is consistent with the permafrost table moving up after the formation of a TEG causing drainage of the bordering polygons and their uplifting. For the slopes and TEG2BGC, the increased thickness of the combined transient and intermediate layers suggests that permafrost recovery following TEG stabilization can increase permafrost resilience to thawing compared to natural conditions. They form a protective layer against annual thawing that would require a strong disturbance at the 380 surface to be degraded (Kanevskiy et al., 2016, 2017) and thus contribute to the long-term stabilization of the TEG. Finally, buried intermediate layers were interpreted in most of the TEG2 cores, but only in TEG1-2Undist, TEG1- 2Slope, and TEG1-2Drained at TEG1. Those intermediate layers were interpreted as buried because they represented different sequences of permafrost aggradation. In most cases, the buried intermediate layers were located under a layer of organic to natural conditions. They form a protective layer against annual thawing that would require a strong disturbance at the 380 surface to be degraded (Kanevskiy et al., 2016, 2017) and thus contribute to the long-term stabilization of the TEG. Finally, buried intermediate layers were interpreted in most of the TEG2 cores, but only in TEG1-2Undist, TEG1- 2Slope, and TEG1-2Drained at TEG1. Those intermediate layers were interpreted as buried because they represented different sequences of permafrost aggradation. In most cases, the buried intermediate layers were located under a layer of organic to natural conditions. 5.2 Landscape evolution following thermo-erosion gullying 390 However, hydrological 400 disturbances and changes in environmental conditions can reactivate geomorphological processes and cause further stage I degradation as the lack of a developed intermediate layer reduces the capacity of permafrost to withstand disturbances. If stabilization persists, the TEG enters a new state of the periglacial landscape. At stage III, the TEG is stabilized, and even though the permafrost thermal regime and ground ice content have recovered to conditions similar to stage 0, the TEG has caused long- lasting geomorphological and hydrological changes to the landscape. 405 Figure 7. Conceptual model of thermo-erosion gully (TEG) evolution in cold permafrost. At stage 0, the polygonal network has well-developed ice wedges that can withstand some thermal erosion, until a threshold is surpassed and causes a TEG to form. At stage I, the thermal regime and geomorphological processes are in disequilibrium, which causes high rates of degradation. As the TEG reaches new thermal, hydrological, and geomorphological equilibria, geomorphological processes slow down, and stabilization begins. At stage II, stabilization takes place and the channel angularity gradually decreases. However, hydrological 400 disturbances and changes in environmental conditions can reactivate geomorphological processes and cause further stage I degradation as the lack of a developed intermediate layer reduces the capacity of permafrost to withstand disturbances. If stabilization persists, the TEG enters a new state of the periglacial landscape. At stage III, the TEG is stabilized, and even though the permafrost thermal regime and ground ice content have recovered to conditions similar to stage 0, the TEG has caused long- lasting geomorphological and hydrological changes to the landscape. 405 400 lasting geomorphological and hydrological changes to the landscape. 405 5.2 Landscape evolution following thermo-erosion gullying 390 By investigating a recent and a stable TEG in the High Arctic, we were able to identify and characterize the main processes of formation and stabilization of a TEG and the long-term impacts of thermal erosion in ice-rich polygonal tundra. We summarize these processes in a conceptual model describing three main stages of TEG formation and stabilization in our study area (Figure 7). The following sections describe each stage in detail. 22 22 5 Figure 7. Conceptual model of thermo-erosion gully (TEG) evolution in cold permafrost. At stage 0, the polygonal network has well-developed ice wedges that can withstand some thermal erosion, until a threshold is surpassed and causes a TEG to form. At stage I, the thermal regime and geomorphological processes are in disequilibrium, which causes high rates of degradation. As the https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. of thermo-erosion gully (TEG) evolution in cold permafrost. At stage 0, the polygonal network has t can withstand some thermal erosion, until a threshold is surpassed and causes a TEG to form. At nd geomorphological processes are in disequilibrium, which causes high rates of degradation. As the hydrological, and geomorphological equilibria, geomorphological processes slow down, and II, stabilization takes place and the channel angularity gradually decreases. However, hydrological environmental conditions can reactivate geomorphological processes and cause further stage I developed intermediate layer reduces the capacity of permafrost to withstand disturbances. If G enters a new state of the periglacial landscape. At stage III, the TEG is stabilized, and even though e and ground ice content have recovered to conditions similar to stage 0, the TEG has caused long- hydrological changes to the landscape. phere 2024 208 14 March 2024 .0 License. 395 Figure 7. Conceptual model of thermo-erosion gully (TEG) evolution in cold permafrost. At stage 0, the polygonal network has well-developed ice wedges that can withstand some thermal erosion, until a threshold is surpassed and causes a TEG to form. At stage I, the thermal regime and geomorphological processes are in disequilibrium, which causes high rates of degradation. As the TEG reaches new thermal, hydrological, and geomorphological equilibria, geomorphological processes slow down, and stabilization begins. At stage II, stabilization takes place and the channel angularity gradually decreases. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. and intermediate layers act as buffers against disturbances due to the high latent heat of fusion of ice (Jorgenson et al., 2010; Andersland and Ladanyi, 2004), and cold permafrost promotes rapid recovery. However, a change in surface conditions (e.g., polygon drainage), a disturbance (e.g., heavy rainfalls, above average snow accumulation leading to increased snowmelt, catastrophic lake drainage), or a shift in climate conditions (e.g., more precipitations, warmer air temperatures increasing permafrost temperature) can trigger additional runoff and increase the power of thermal erosion, eventually surpassing a threshold (e.g., tunnel collapse) and forming a TEG (Figure 7, Stage I). 415 5.2.2 Stage I: Degradation and disequilibrium This first stage of thermo-erosion gullying is characterized by major geomorphic, hydrolog This transitional stage is unstable and undergoes rapid changes through positive feedbacks caused by thermal erosion of ice 420 wedges and of the channel, thaw slumping, collapse of active layer and tunnel overhangs, slope failure, and mudflows to reach a new equilibrium (Godin and Fortier, 2012b; Godin et al., 2016; Fortier et al., 2007). This translates into a box-shaped channel with unstable slopes and multiple ramifications, sometime forming baydzherakhi. Thermal erosion also progresses vertically into ice wedges and permafrost until it reaches its base level or enough sediment is deposited at the bottom of the newly formed stream (French, 2007). Then, lateral thermal erosion occurs and contributes to enlarge the TEG channels 425 (Morgenstern et al., 2021). Thermokarst mounds become increasingly affected by thermal erosion and thermokarst degradation due to the increased surface area when isolated and are eventually completely eroded from the main channel if erosion persists. newly formed stream (French, 2007). Then, lateral thermal erosion occurs and contributes to enlarge the TEG channels 425 (Morgenstern et al., 2021). Thermokarst mounds become increasingly affected by thermal erosion and thermokarst degradation due to the increased surface area when isolated and are eventually completely eroded from the main channel if erosion persists. During this stage, sediment export is high because of the active erosion and collap During this stage, sediment export is high because of the active erosion and collapsed material being remobilized downstream in low gradient areas of the TEG channel (Godin and Fortier, 2012b; Rioux, 2020). The uncovered ground 430 material in the TEG becomes exposed to solar radiation and atmospheric heat, which both contribute to erosion, even without running water (i.e., thermal erosion) (Fortier et al., 2007). In addition, the thaw front can penetrate deeper into permafrost due to the lack of protective cover, causing degradation of the transient and intermediate layers of the slopes and bottom of the gully channel. 5.2.1 Stage 0: Undisturbed polygonal network In areas with cold continuous syngenetic permafrost where thick transient and intermediate layers have developed, permafrost is well-protected against interannual climatic variations and meteorological events (Kanevskiy et al., 2017; Shur et al., 2005). Indeed, the infiltration of surface runoff water from snowmelt, summer precipitation, or thawing permafrost In areas with cold continuous syngenetic permafrost where thick transient and intermediate layers have developed, permafrost is well-protected against interannual climatic variations and meteorological events (Kanevskiy et al., 2017; Shur et al., 2005). Indeed, the infiltration of surface runoff water from snowmelt, summer precipitation, or thawing permafrost (French, 2007) into open thermal contraction cracks and other types of cavities followed by thermal erosion of permafrost 410 and ice wedges may be occurring on an annual basis, but not necessarily lead to gullying (Figure 7, Stage 0). The transient et al., 2005). Indeed, the infiltration of surface runoff water from snowmelt, summer precipitation, or thawing permafrost (French, 2007) into open thermal contraction cracks and other types of cavities followed by thermal erosion of permafrost 410 and ice wedges may be occurring on an annual basis, but not necessarily lead to gullying (Figure 7, Stage 0). The transient (French, 2007) into open thermal contraction cracks and other types of cavities followed by thermal erosion of permafrost 410 and ice wedges may be occurring on an annual basis, but not necessarily lead to gullying (Figure 7, Stage 0). The transient 23 It is important to note that not all sections of a TEG are at the same stage simultaneously and that sections of a TEG can go 450 back and forth between stages I and II. Once the slopes of the TEG are vegetated and the gully no longer experiences erosion, the permafrost environment supporting the TEG reaches a new state. 5.2.3 Stage II: Stabilization and transition to a new state 435 While TEG inception can occur within a single thawing season, it often takes several decades for erosion, mass movement, thermokarst, and sediment export to decrease, at which point stabilization can begin (Figure 7, Stage II). Over time, slope angularity of the TEG decreases (Godin and Fortier, 2010), which contributes to reduce mass movements. In addition, alluvial sediment deposited on the sides of main channel of the TEG can form alluvial levees that reduce gully wall While TEG inception can occur within a single thawing season, it often takes several decades for erosion, mass movement, thermokarst, and sediment export to decrease, at which point stabilization can begin (Figure 7, Stage II). Over time, slope angularity of the TEG decreases (Godin and Fortier, 2010), which contributes to reduce mass movements. In addition, alluvial sediment deposited on the sides of main channel of the TEG can form alluvial levees that reduce gully wall mechanical erosion by water (Godin and Fortier, 2012b). As the slopes become stable and erosion winds down, permafrost 440 re-aggradation takes place and vegetation starts to colonize exposed ground in the TEG. This creates a negative feedback on t d t f th f t d ti b hi ldi th d f f i i l di ti alluvial sediment deposited on the sides of main channel of the TEG can form alluvial levees that reduce gully wall mechanical erosion by water (Godin and Fortier, 2012b). As the slopes become stable and erosion winds down, permafrost 440 re-aggradation takes place and vegetation starts to colonize exposed ground in the TEG. This creates a negative feedback on mass movement and promotes further permafrost aggradation by shielding the ground surface from incoming solar radiation and insulating it from atmospheric heat. The breaching and collapsing of the rims of the polygons encompassing the TEG mechanical erosion by water (Godin and Fortier, 2012b). As the slopes become stable and erosion winds down, permafrost 440 re-aggradation takes place and vegetation starts to colonize exposed ground in the TEG. This creates a negative feedback on mass movement and promotes further permafrost aggradation by shielding the ground surface from incoming solar radiation and insulating it from atmospheric heat. The breaching and collapsing of the rims of the polygons encompassing the TEG 24 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. also contribute to TEG stabilization (Godin and Fortier, 2012b; Godin et al., 2016). Active-layer and pond drainage of breached low-centered polygons creates a dry porous top layer with reduced thermal conductivity (Fortier et al., 2007) 445 resulting in a thinning of the active layer on the shoulders of the gully. This causes a shift in vegetation (Perreault et al., 2016), from wet to mesic, which also contributes to decrease thermal conductivity and active layer thickness, resulting in thin transient and intermediate layers form closer to the surface than in undisturbed conditions. This ice enrichment contributes to increase the thermal resistance of the encompassing polygons and thus to reduce further widening of the TEG. also contribute to TEG stabilization (Godin and Fortier, 2012b; Godin et al., 2016). Active-layer and pond drainage of breached low-centered polygons creates a dry porous top layer with reduced thermal conductivity (Fortier et al., 2007) 445 resulting in a thinning of the active layer on the shoulders of the gully. This causes a shift in vegetation (Perreault et al., 2016), from wet to mesic, which also contributes to decrease thermal conductivity and active layer thickness, resulting in thin transient and intermediate layers form closer to the surface than in undisturbed conditions. This ice enrichment contributes to increase the thermal resistance of the encompassing polygons and thus to reduce further widening of the TEG. breached low-centered polygons creates a dry porous top layer with reduced thermal conductivity (Fortier et al., 2007) 445 resulting in a thinning of the active layer on the shoulders of the gully. This causes a shift in vegetation (Perreault et al., 2016), from wet to mesic, which also contributes to decrease thermal conductivity and active layer thickness, resulting in thin transient and intermediate layers form closer to the surface than in undisturbed conditions. This ice enrichment contributes to increase the thermal resistance of the encompassing polygons and thus to reduce further widening of the TEG. It is important to note that not all sections of a TEG are at the same stage simultaneously and that sections of a TEG can go 450 back and forth between stages I and II. Once the slopes of the TEG are vegetated and the gully no longer experiences erosion, the permafrost environment supporting the TEG reaches a new state. 5.2.4 Stage III: New state of the periglacial landscape and permafrost resilience The formation of TEG has profound impacts on ice-wedge polygonal networks, permanen The formation of TEG has profound impacts on ice-wedge polygonal networks, permanently (at the Anthropocene time scale) altering landscape morphology and ecosystem functions (Figure 7, Stage III). The stabilized TEG has now evolved 455 into a thermo-erosional valley (Morgenstern et al., 2021) characterized by a broad, densely vegetated main channel with stable low-angle slopes and limited ramifications. Despite alluvial sedimentation at the bottom of the gully channel, the base level of the gully is now lower than the surrounding polygons, thus affecting local hydrological connectivity. The low- centered polygons along the margins of the TEG have changed into raised flat/high-centered ones due to their drainage. The changes in their thermal properties caused a thinning of the active layer and ground ice aggradation, ultimately raising the 460 polygons and further affecting local hydrological patterns. In this new state, even though the TEG has caused long-lasting geomorphological and hydrological changes to the landscape, the permafrost affected by the TEG can recover to conditions similar to undisturbed permafrost pre-dating the disturbance, i.e., similar ground ice content and temperature, and thus is resilient. However, permafrost re-aggradation now , , g p , , p gg leads to the formation of thicker intermediate layers in areas that have been the most disturbed, i.e., the slopes and bottom of 465 the gully channel. Thermal contraction cracking can resume following the same pre-disturbance network, causing ice-wedge formation or rejuvenation (Lewkowicz, 1994; Gagnon and Allard, 2019). While permafrost shows resilience, the local landscape does not because it has surpassed thresholds of recovery to pre-disturbance conditions. From a landscape- evolution perspective, this suggests that in stable environmental conditions undergoing natural variability, permafrost is able leads to the formation of thicker intermediate layers in areas that have been the most disturbed, i.e., the slopes and bottom of 465 the gully channel. Thermal contraction cracking can resume following the same pre-disturbance network, causing ice-wedge formation or rejuvenation (Lewkowicz, 1994; Gagnon and Allard, 2019). While permafrost shows resilience, the local landscape does not because it has surpassed thresholds of recovery to pre-disturbance conditions. From a landscape- evolution perspective, this suggests that in stable environmental conditions undergoing natural variability, permafrost is able to persist longer than the geomorphological landforms in which it forms. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. 6. Conclusion 490 rich geocryological layers and cold ground temperatures played a significant role in the inception and stabilization of the 485 TEG by both buffering against smaller disturbances and promoting permafrost aggradation following thermal erosion. Once stabilized, the TEG was in a new state characterized by long-lasting geomorphological and hydrological changes to the landscape, but is able to recover to pre-disturbance geocryological and thermal conditions, showing that permafrost is more resilient to thermal erosion disturbances than surface geomorphology. This suggests that in stable environmental conditions undergoing natural variability, permafrost can persist longer than the geomorphological landforms in which it forms. 490 The contact author has declared that none of the authors has any competing interests. The contact author has declared that none of the authors has any competing interests. 8. Acknowledgements We thank Karine Rioux and Stéphanie Coulombe for their help during fieldwork. We also thank the Centre for Northern Studies for providing access and logistical support to its research station. We are also grateful to the Natural Sciences and 495 Engineering Research Council of Canada, Parks Canada, Fonds de recherche du Québec - Nature et technologie, Northern Scientific Training Program and the Polar Continental Shelf Program (PCSP). Abbott, B. W., Jones, J. B., Godsey, S. E., Larouche, J. R., and Bowden, W. B.: Patterns and persistence of hydrologic carbon and nutrient export from collapsing upland permafrost, Biogeosciences, 12, 3725–3740, https://doi.org/10.5194/bg- 500 12-3725-2015, 2015. 5.2.4 Stage III: New state of the periglacial landscape and permafrost resilience This adds to the stabilizing effect of thermal 470 erosion on tundra landscapes, which levels gradients and limits thermokarst processes by enhancing hydrological connectivity (Morgenstern et al., 2021). to persist longer than the geomorphological landforms in which it forms. This adds to the stabilizing effect of thermal 470 erosion on tundra landscapes, which levels gradients and limits thermokarst processes by enhancing hydrological connectivity (Morgenstern et al., 2021). to persist longer than the geomorphological landforms in which it forms. This adds to the stabilizing effect of thermal 470 erosion on tundra landscapes, which levels gradients and limits thermokarst processes by enhancing hydrological connectivity (Morgenstern et al., 2021). 25 6. Conclusion This study investigated the impacts of two TEGs in the Canadian High Arctic on ground ice content, cryostratigraphic patterns, and geomorphology to examine permafrost recovery following thermal erosion in ice-wedge polygonal tundra. Our 475 results concurred with other research showing that thermal erosion affects hydrological connectivity, snow accumulation patterns and snowmelt dynamics, changes vegetation, causes the remobilization of soil material and nutrients, and showed that once stabilized, TEG permanently (at the Anthropocene scale) alter landscape morphology and ecosystem functions. In addition to the gully channel, the polygons bordering the TEG evolved from wet low-centered to dry/high-centered, causing changes in the ground thermal properties and ultimately leading to ice aggradation and raising the polygons. This 480 cryostratigraphic pattern, in addition to vegetation colonization and decrease in slope angularity, were interpreted as stabilization mechanisms leading to permafrost aggradation and TEG stabilization. The study also reported the first VIC values in the slopes and at the bottom of the gully channel of TEG. Although the TEG caused discernable cryostratigraphic patterns, the ice contents in the cores affected by the TEG were comparable to undisturbed polygons. The presence of ice- changes in the ground thermal properties and ultimately leading to ice aggradation and raising the polygons. This 480 cryostratigraphic pattern, in addition to vegetation colonization and decrease in slope angularity, were interpreted as stabilization mechanisms leading to permafrost aggradation and TEG stabilization. The study also reported the first VIC values in the slopes and at the bottom of the gully channel of TEG. Although the TEG caused discernable cryostratigraphic patterns, the ice contents in the cores affected by the TEG were comparable to undisturbed polygons. The presence of ice- rich geocryological layers and cold ground temperatures played a significant role in the inception and stabilization of the 485 TEG by both buffering against smaller disturbances and promoting permafrost aggradation following thermal erosion. Once stabilized, the TEG was in a new state characterized by long-lasting geomorphological and hydrological changes to the landscape, but is able to recover to pre-disturbance geocryological and thermal conditions, showing that permafrost is more resilient to thermal erosion disturbances than surface geomorphology. This suggests that in stable environmental conditions undergoing natural variability, permafrost can persist longer than the geomorphological landforms in which it forms. 9. References Abbott, B. W., Jones, J. B., Godsey, S. E., Larouche, J. R., and Bowden, W. B.: Patterns and persistence of hydrologic carbon and nutrient export from collapsing upland permafrost, Biogeosciences, 12, 3725–3740, https://doi.org/10.5194/bg- 500 12-3725-2015, 2015. 26 and Allard, M.: Ice segregation and gas distribution in permafrost using tomodensitometric analysis, Permafrost Periglac. Process., 15, 367–378, https://doi.org/10.1002/ppp.508, 2004. Climate station data from Bylot Island in Nunavut, Canada, v. 1.11 (1992-2019). Nordicana D2, doi: 10.5885/45039SL- 520 EE76C1BDAADC4890: http://www.cen.ulaval.ca/nordicanad/dpage.aspx?doi=45039SL-EE76C1BDAADC4890. Climate station data from Bylot Island in Nunavut, Canada, v. 1.11 (1992-2019). Nordicana D2, doi: 10.5885/45039SL- 520 EE76C1BDAADC4890: http://www.cen.ulaval.ca/nordicanad/dpage.aspx?doi=45039SL-EE76C1BDAADC4890. Chapin, F. S., Kofinas, G. P., and Folke, C. (Eds.): Principles of Ecosystem Stewardship: Resilience-Based Natural Resource Management in a Changing World, Springer New York, New York, NY, https://doi.org/10.1007/978-0-387-73033-2, 2009. Coulombe, S., Fortier, D., Lacelle, D., Kanevskiy, M., and Shur, Y.: Origin, burial and preservation of late Pleistocene-age glacier ice in Arctic permafrost (Bylot Island, NU, Canada), The Cryosphere, 13, 97–111, https://doi.org/10.5194/tc-13-97- 525 2019, 2019. Historical Climate Data - Climatic data of Pond Inlet, Nunavut: https://climate.weather.gc.ca/index_e.html, last access: 15 June 2022. Folk, R. L. and Ward, W. C.: Brazos River bar [Texas]; a study in the significance of grain size parameters, Journal of Sedimentary Research 27 3 26 https://doi org/10 1306/74D70646 2B21 11D7 8648000102C1865D 1957 530 Climate station data from Bylot Island in Nunavut, Canada, v. 1.11 (1992-2019). Nordicana D2, doi: 10.5885/45039SL- 520 EE76C1BDAADC4890: http://www.cen.ulaval.ca/nordicanad/dpage.aspx?doi=45039SL-EE76C1BDAADC4890. Chapin, F. S., Kofinas, G. P., and Folke, C. (Eds.): Principles of Ecosystem Stewardship: Resilience-Based Natural Resource Management in a Changing World, Springer New York, New York, NY, https://doi.org/10.1007/978-0-387-73033-2, 2009. Chapin, F. S., Kofinas, G. P., and Folke, C. (Eds.): Principles of Ecosystem Stewardship: Resilience-Based Natural Resource Management in a Changing World, Springer New York, New York, NY, https://doi.org/10.1007/978-0-387-73033-2, 2009. Coulombe, S., Fortier, D., Lacelle, D., Kanevskiy, M., and Shur, Y.: Origin, burial and preservation of late Pleistocene-age glacier ice in Arctic permafrost (Bylot Island, NU, Canada), The Cryosphere, 13, 97–111, https://doi.org/10.5194/tc-13-97- 525 2019, 2019. Historical Climate Data Climatic data of Pond Inlet Nunavut: https://climate weather gc ca/index e html last access: 15 Coulombe, S., Fortier, D., Lacelle, D., Kanevskiy, M., and Shur, Y.: Origin, burial and preservation of late Pleistocene-age glacier ice in Arctic permafrost (Bylot Island, NU, Canada), The Cryosphere, 13, 97–111, https://doi.org/10.5194/tc-13-97- 525 2019, 2019. Historical Climate Data - Climatic data of Pond Inlet, Nunavut: https://climate.weather.gc.ca/index_e.html, last access: 15 June 2022. Folk, R. L. and Ward, W. C.: Brazos River bar [Texas]; a study in the significance of grain size parameters, Journal of Sedimentary Research, 27, 3–26, https://doi.org/10.1306/74D70646-2B21-11D7-8648000102C1865D, 1957. 530 Sedimentary Research, 27, 3–26, https://doi.org/10.1306/74D70646-2B21-11D7-8648000102C1865D, 1957. 530 Fortier, D. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Allard, M.: Geomorphological Changes and Permafrost Dynamics: Key Factors in Changing Arctic Ecosystems. An Example from Bylot Island, Nunavut, Canada, Geoscience Canada, 23, 205–212, 1996. Allard, M., Sarrazin, D., and L’Hérault, E.: Borehole and near-surface ground temperatures in northeastern Canada, v. 1.5 (1988-2019). Nordicana D8, doi: 10.5885/45291SL-34F28A9491014AFD., 2020. Andersland, O. B. and Ladanyi, B.: Frozen Ground Engineering, 2nd edition., Wiley, 384 pp., 200 Arė, F. Ėrnestovich., Balobaev, V. Tikhonovich., Bosikov, N. P., and Cold Regions Research and Engineering Laboratory (U.S.): Characteristics of the reshaping of shorelines of thermokarst lakes of central Yakutia., 1979. Blott, S. J. and Pye, K.: GRADISTAT: a grain size distribution and statistics package for the analysis of unconsolidated sediments, Earth Surf. Process. Landforms, 26, 1237–1248, https://doi.org/10.1002/esp.261, 2001. 0 Blott, S. J. and Pye, K.: GRADISTAT: a grain size distribution and statistics package for the analysis of unconsolidated sediments, Earth Surf. Process. Landforms, 26, 1237–1248, https://doi.org/10.1002/esp.261, 2001. 510 Bowden, W. B., Gooseff, M. N., Balser, A., Green, A., Peterson, B. J., and Bradford, J.: Sediment and nutrient delivery from thermokarst features in the foothills of the North Slope, Alaska: Potential impacts on headwater stream ecosystems: Thermokarst Impacts on Stream Ecosystems, J. Geophys. Res., 113, n/a-n/a, https://doi.org/10.1029/2007JG000470, 2008. Bowden, W. B., Gooseff, M. N., Balser, A., Green, A., Peterson, B. J., and Bradford, J.: Sediment and nutrient delivery from thermokarst features in the foothills of the North Slope, Alaska: Potential impacts on headwater stream ecosystems: Thermokarst Impacts on Stream Ecosystems, J. Geophys. Res., 113, n/a-n/a, https://doi.org/10.1029/2007JG000470, 2008. Burn, C. R.: The development of near-surface ground ice during the Holocene at sites near Mayo, Yukon Territory, Canada, Journal of Quaternary Science, 3, 31–38, https://doi.org/10.1002/jqs.3390030106, 1988. 515 , p g g y , y, , Journal of Quaternary Science, 3, 31–38, https://doi.org/10.1002/jqs.3390030106, 1988. 515 Burn, C. R.: Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada, Can. J. Earth Sci., 34, 912–925, https://doi.org/10.1139/e17-076, 1997. Calmels, F. and Allard, M.: Ice segregation and gas distribution in permafrost using tomodensitometric analysis, Permafrost Periglac. Process., 15, 367–378, https://doi.org/10.1002/ppp.508, 2004. Burn, C. R.: Cryostratigraphy, paleogeography, and climate change during the early Holocene warm interval, western Arctic coast, Canada, Can. J. Earth Sci., 34, 912–925, https://doi.org/10.1139/e17-076, 1997. Calmels, F. Sedimentary Research, 27, 3–26, https://doi.org/10.1306/74D70646-2B21-11D7-8648000102C1865D, 1957. 530 Fortier, D. and Allard, M.: Late Holocene syngenetic ice-wedge polygons development, Bylot Island, Canadian Arctic Archipelago, 41, 16, 2004. Fortier, D., Allard, M., and Pivot, F.: A late-Holocene record of loess deposition in ice-wedge polygons reflecting wind activity and ground moisture conditions, Bylot Island, eastern Canadian Arctic, The Holocene, 16, 635–646, https://doi.org/10.1191/0959683606hl960rp, 2006. 535 Fortier, D., Allard, M., and Pivot, F.: A late-Holocene record of loess deposition in ice-wedge polygons reflecting wind activity and ground moisture conditions, Bylot Island, eastern Canadian Arctic, The Holocene, 16, 635–646, https://doi.org/10.1191/0959683606hl960rp, 2006. 535 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Fortier, D., Allard, M., and Shur, Y.: Observation of rapid drainage system development by thermal erosion of ice wedges on Bylot Island, Canadian Arctic Archipelago, Permafrost Periglac. Process., 18, 229–243, https://doi.org/10.1002/ppp.595, 2007. French, H. and Shur, Y.: The principles of cryostratigraphy, Earth-Science Reviews, 101, 190–206, https://doi.org/10.1016/j.earscirev.2010.04.002, 2010. 540 https://doi.org/10.1016/j.earscirev.2010.04.002, 2010. 540 French, H. M.: The Periglacial Environment, 3rd edition., John Wiley & Sons Ltd, Sussex, England, 458 pp., 2007. French, H. M.: The periglacial environment, Fourth edition., Wiley, Blackwell, Hoboken, NJ, 1 pp., 2017. Gagnon, S. and Allard, M.: Changes in ice‐wedge activity over 25 years of climate change near Salluit, Nunavik (northern Québec, Canada), Permafrost and Periglac Process, 31, 69–84, https://doi.org/10.1002/ppp.2030, 2019. French, H. M.: The Periglacial Environment, 3rd edition., John Wiley & Sons Ltd, Sussex, England, 458 pp., 2007. French, H. M.: The Periglacial Environment, 3rd edition., John Wiley & Sons Ltd, Sussex, England, 458 pp., 2007. French, H. M.: The periglacial environment, Fourth edition., Wiley, Blackwell, Hoboken, NJ, 1 pp., 2017. French, H. M.: The Periglacial Environment, 3rd edition., John Wiley & Sons Ltd, Sussex, Engla g y g pp French, H. M.: The periglacial environment, Fourth edition., Wiley, Blackwell, Hoboken, NJ, 1 pp., 2017. Gagnon, S. and Allard, M.: Changes in ice‐wedge activity over 25 years of climate change near Salluit, Nunavik (northern Québec, Canada), Permafrost and Periglac Process, 31, 69–84, https://doi.org/10.1002/ppp.2030, 2019. French, H. M.: The periglacial environment, Fourth edition., Wiley, Blackwell, Hoboken, NJ, 1 pp., 2017. Gagnon, S. and Allard, M.: Changes in ice‐wedge activity over 25 years of climate change near Salluit, Nunavik (northern Québec Canada) Permafrost and Periglac Process 31 69 84 https://doi org/10 1002/ppp 2030 2019 French, H. M.: The periglacial environment, Fourth edition., Wiley, Blackwell, Hoboken, NJ, 1 pp., 2017. Gagnon, S. and Allard, M.: Changes in ice‐wedge activity over 25 years of climate change near Salluit, Nunavik (northern Québec, Canada), Permafrost and Periglac Process, 31, 69–84, https://doi.org/10.1002/ppp.2030, 2019. Gagnon, S. and Allard, M.: Geomorphological controls over carbon distribution in permafrost soils: the case of the 545 Narsajuaq river valley, Nunavik (Canada), Arctic Science, 6, 509–528, https://doi.org/10.1139/as-2019-0026, 2020. Gilbert, G. L., Kanevskiy, M., and Murton, J. and Allard, M.: Late Holocene syngenetic ice-wedge polygons development, Bylot Island, Canadian Arctic Archipelago, 41, 16, 2004. Fortier, D., Allard, M., and Pivot, F.: A late-Holocene record of loess deposition in ice-wedge polygons reflecting wind activity and ground moisture conditions, Bylot Island, eastern Canadian Arctic, The Holocene, 16, 635–646, https://doi.org/10.1191/0959683606hl960rp, 2006. 535 Fortier, D. and Allard, M.: Late Holocene syngenetic ice-wedge polygons development, Bylot Island, Canadian Arctic Archipelago, 41, 16, 2004. Fortier, D., Allard, M., and Pivot, F.: A late-Holocene record of loess deposition in ice-wedge polygons reflecting wind activity and ground moisture conditions, Bylot Island, eastern Canadian Arctic, The Holocene, 16, 635–646, https://doi.org/10.1191/0959683606hl960rp, 2006. 535 27 B.: Recent Advances (2008-2015) in the Study of Ground Ice and Cryostratigraphy: Recent Advances in the Study of Ground Ice and Cryostratigraphy, Permafrost and Periglac. Process., 27, 377–389, https://doi.org/10.1002/ppp.1912, 2016. Gagnon, S. and Allard, M.: Geomorphological controls over carbon distribution in permafrost soils: the case of the 45 Narsajuaq river valley, Nunavik (Canada), Arctic Science, 6, 509–528, https://doi.org/10.1139/as-2019-0026, 2020. Gilbert, G. L., Kanevskiy, M., and Murton, J. B.: Recent Advances (2008-2015) in the Study of Ground Ice and Cryostratigraphy: Recent Advances in the Study of Ground Ice and Cryostratigraphy, Permafrost and Periglac. Process., 27, 377–389, https://doi.org/10.1002/ppp.1912, 2016. Godin, E. and Fortier, D.: Geomorphology of thermo-erosion gullies – case study from Bylot Island, Nunavut, Canada, 550 https://doi.org/10.13140/2.1.4498.9120, 2010. Godin, E. and Fortier, D.: Fine Scale Spatio-Temporal Monitoring of Multiple Thermo-Erosion Gullies Development on Bylot Island, Eastern Canadian Archipelago, 7, 2012a. Godin, E. and Fortier, D.: Geomorphology of thermo-erosion gullies – case study from Bylot Island, Nunavut, Canada, 550 https://doi.org/10.13140/2.1.4498.9120, 2010. https://doi.org/10.13140/2.1.4498.9120, 2010. Godin, E. and Fortier, D.: Fine Scale Spatio-Temporal Monitoring of Multiple Thermo-Erosion Gullies Development on Bylot Island, Eastern Canadian Archipelago, 7, 2012a. Godin, E. and Fortier, D.: Fine Scale Spatio-Temporal Monitoring of Multiple Thermo-Erosion Gullies Development on Bylot Island, Eastern Canadian Archipelago, 7, 2012a. Godin, E. and Fortier, D.: Geomorphology of a thermo-erosion gully, Bylot Island, Nunavut, Canada, Can. J. Earth Sci., 49, 979–986, https://doi.org/10.1139/e2012-015, 2012b. 555 Godin, E., Fortier, D., and Coulombe, S.: Effects of thermo-erosion gullying on hydrologic flow networks, discharge and soil loss, Environ. Res. Lett., 9, 105010, https://doi.org/10.1088/1748-9326/9/10/105010, 2014. Godin, E., Fortier, D., and Lévesque, E.: Nonlinear thermal and moisture response of ice-wedge polygons to permafrost disturbance increases heterogeneity of high Arctic wetland, Biogeosciences, 13, 1439–1452, https://doi.org/10.5194/bg-13- 1439-2016, 2016. 560 Goudie, A. (Ed.): Encyclopedia of geomorphology, Routledge : International Association of Geomorphologists, London ; New York, 2 pp., 2004. Grosse, G., Schirrmeister, L., Kunitsky, V. V., and Hubberten, H.-W.: The use of CORONA images in remote sensing of periglacial geomorphology: an illustration from the NE Siberian coast, Permafrost and Periglacial Processes, 16, 163–172, https://doi.org/10.1002/ppp.509, 2005. 565 https://doi.org/10.1002/ppp.509, 2005. 565 Grosse, G., Schirrmeister, L., and Malthus, T. J.: Application of Landsat-7 satellite data and a DEM for the quantification of thermokarst-affected terrain types in the periglacial Lena–Anabar coastal lowland, Polar Research, 25, 51–67, https://doi.org/10.3402/polar.v25i1.6238, 2006. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Jongejans, L. L., Strauss, J., Lenz, J., Peterse, F., Mangelsdorf, K., Fuchs, M., and Grosse, G.: Organic matter characteristics in yedoma and thermokarst deposits on Baldwin Peninsula, west Alaska, Biogeosciences, 15, 6033–6048, https://doi.org/10.5194/bg-15-6033-2018, 2018. Jorgenson, M. T., Romanovsky, V., Harden, J., Shur, Y., O’Donnell, J., Schuur, E. A. G., Kanevskiy, M., and Marchenko, S.: Resilience and vulnerability of permafrost to climate change, Can. J. For. Res., 40, 1219–1236, https://doi.org/10.1139/X10-060, 2010. 575 S.: Resilience and vulnerability of permafrost to climate change, Can. J. For. Res., 40, 1219–1236, 575 https://doi.org/10.1139/X10-060, 2010. Kanevskiy, M., Shur, Y., Fortier, D., Jorgenson, M. T., and Stephani, E.: Cryostratigraphy of late Pleistocene syngenetic permafrost (yedoma) in northern Alaska, Itkillik River exposure, Quat. res., 75, 584–596, https://doi.org/10.1016/j.yqres.2010.12.003, 2011. Kanevskiy, M., Shur, Y., Fortier, D., Jorgenson, M. T., and Stephani, E.: Cryostratigraphy of late Pleistocene syngenetic permafrost (yedoma) in northern Alaska, Itkillik River exposure, Quat. res., 75, 584–596, https://doi.org/10.1016/j.yqres.2010.12.003, 2011. Kanevskiy, M., Shur, Y., Connor, B., Dillon, M., Stephani, E., and O’Donnell, J.: Study of the Ice-Rich Syngenetic 580 Permafrost for Road Design (Interior Alaska), in: Proceedings of the Tenth International Conference on Permafrost, Tenth International Conference on Permafrost, Boulders, Colorado, US, journalAbbreviation: Proceedings of the Tenth International Conference on Permafrost, 2012. Kanevskiy, M., Shur, Y., Jorgenson, M. T., Ping, C.-L., Michaelson, G. J., Fortier, D., Stephani, E., Dillon, M., and Tumskoy, V.: Ground ice in the upper permafrost of the Beaufort Sea coast of Alaska, Cold Regions Science and 585 Technology, 85, 56–70, https://doi.org/10.1016/j.coldregions.2012.08.002, 2013. Kanevskiy, M., Shur, Y., Jorgenson, M. T., Ping, C.-L., Michaelson, G. J., Fortier, D., Stephani, E., Dillon, M., and Tumskoy, V.: Ground ice in the upper permafrost of the Beaufort Sea coast of Alaska, Cold Regions Science and 585 Technology, 85, 56–70, https://doi.org/10.1016/j.coldregions.2012.08.002, 2013. Kanevskiy, M., Jorgenson, T., Shur, Y., O’Donnell, J. A., Harden, J. W., Zhuang, Q., and Fortier, D.: Cryostratigraphy and Permafrost Evolution in the Lacustrine Lowlands of West-Central Alaska: Cryostratigraphy and Permafrost Evolution in the Kanevskiy, M., Shur, Y., Jorgenson, M. T., Ping, C.-L., Michaelson, G. J., Fortier, D., Stephani, E., Dillon, M., and Tumskoy, V.: Ground ice in the upper permafrost of the Beaufort Sea coast of Alaska, Cold Regions Science and 585 Technology, 85, 56–70, https://doi.org/10.1016/j.coldregions.2012.08.002, 2013. Inland Waters Branch: Glacier atlas of Canada, Bylot Island area, 46201, Inland Waters Branch., Ottawa,Ontario,Canada, Grosse, G., Schirrmeister, L., and Malthus, T. J.: Application of Landsat-7 satellite data and a DEM for the quantification of thermokarst-affected terrain types in the periglacial Lena–Anabar coastal lowland, Polar Research, 25, 51–67, https://doi.org/10.3402/polar.v25i1.6238, 2006. Inland Waters Branch: Glacier atlas of Canada, Bylot Island area, 46201, Inland Waters Branch., Ottawa,Ontario,Canada, 1969. 570 28 Kanevskiy, M., Jorgenson, T., Shur, Y., O’Donnell, J. A., Harden, J. W., Zhuang, Q., and Fortier, D.: Cryostratigraphy and Permafrost Evolution in the Lacustrine Lowlands of West-Central Alaska: Cryostratigraphy and Permafrost Evolution in the Lacustrine Lowlands, Alaska, Permafrost and Periglac. Process., 25, 14–34, https://doi.org/10.1002/ppp.1800, 2014. Kanevskiy, M., Jorgenson, T., Shur, Y., O’Donnell, J. A., Harden, J. W., Zhuang, Q., and Fortier, D.: Cryostratigraphy and Permafrost Evolution in the Lacustrine Lowlands of West-Central Alaska: Cryostratigraphy and Permafrost Evolution in the Lacustrine Lowlands, Alaska, Permafrost and Periglac. Process., 25, 14–34, https://doi.org/10.1002/ppp.1800, 2014. Kanevskiy, M., Shur, Y., Walker, D., Buchhorn, M., Jorgenson, T., Matyshak, G., Raynolds, M., Peirce, J., and Wirth, L.: 590 Evaluation of Risk of Ice-Wedge Degradation, Prudhoe Bay Oilfield, AK., 2016. Kanevskiy, M., Shur, Y., Jorgenson, T., Brown, D. R. N., Moskalenko, N., Brown, J., Walker, D. A., Raynolds, M. K., and Buchhorn, M.: Degradation and stabilization of ice wedges: Implications for assessing risk of thermokarst in northern Alaska, Geomorphology, 297, 20–42, https://doi.org/10.1016/j.geomorph.2017.09.001, 2017. Kanevskiy, M., Shur, Y., Walker, D., Buchhorn, M., Jorgenson, T., Matyshak, G., Raynolds, M., Peirce, J., and Wirth, L.: 590 Evaluation of Risk of Ice-Wedge Degradation, Prudhoe Bay Oilfield, AK., 2016. Kanevskiy, M., Shur, Y., Jorgenson, T., Brown, D. R. N., Moskalenko, N., Brown, J., Walker, D. A., Raynolds, M. K., and Buchhorn, M.: Degradation and stabilization of ice wedges: Implications for assessing risk of thermokarst in northern Alaska, Geomorphology, 297, 20–42, https://doi.org/10.1016/j.geomorph.2017.09.001, 2017. Kokelj, S. V. and Jorgenson, M. T.: Advances in Thermokarst Research: Recent Advances in Research Investigating 595 Thermokarst Processes, Permafrost and Periglac. Process., 24, 108–119, https://doi.org/10.1002/ppp.1779, 2013. Lacelle, D., Bjornson, J., and Lauriol, B.: Climatic and geomorphic factors affecting contemporary (1950-2004) activity of retrogressive thaw slumps on the Aklavik Plateau, Richardson Mountains, NWT, Canada: Climatic and Geomorphic Factors affecting Thaw Slump Activity, Permafrost Periglac. Process., 21, 1–15, https://doi.org/10.1002/ppp.666, 2010. Kokelj, S. V. and Jorgenson, M. T.: Advances in Thermokarst Research: Recent Advances in Research Investigating 595 Thermokarst Processes, Permafrost and Periglac. Process., 24, 108–119, https://doi.org/10.1002/ppp.1779, 2013. Lacelle, D., Bjornson, J., and Lauriol, B.: Climatic and geomorphic factors affecting contemporary (1950-2004) activity of retrogressive thaw slumps on the Aklavik Plateau, Richardson Mountains, NWT, Canada: Climatic and Geomorphic Factors affecting Thaw Slump Activity, Permafrost Periglac. Process., 21, 1–15, https://doi.org/10.1002/ppp.666, 2010. Lantz, T. C. and Kokelj, S. V.: Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., 600 Canada, Geophys. Res. Lett., 35, L06502, https://doi.org/10.1029/2007GL032433, 2008. Levy, J. S., Head, J. W., and Marchant, D. R.: The role of thermal contraction crack polygons in cold-desert fluvial systems, Antarctic Science, 20, 565–579, https://doi.org/10.1017/S0954102008001375, 2008. Lantz, T. C. and Kokelj, S. V.: Increasing rates of retrogressive thaw slump activity in the Mackenzie Delta region, N.W.T., 600 Canada, Geophys. Res. Lett., 35, L06502, https://doi.org/10.1029/2007GL032433, 2008. Levy, J. S., Head, J. W., and Marchant, D. R.: The role of thermal contraction crack polygons in cold-desert fluvial systems, Antarctic Science, 20, 565–579, https://doi.org/10.1017/S0954102008001375, 2008. Lewkowicz, A. G.: Ice-wedge rejuvenation, fosheim peninsula, ellesmere Island, Canada, Permafrost Periglac. Process., 5, 251–268, https://doi.org/10.1002/ppp.3430050405, 1994. 605 251–268, https://doi.org/10.1002/ppp.3430050405, 1994. 605 Liljedahl, A., Hinzman, L., Busey, R., and Yoshikawa, K.: Physical short-term changes after a tussock tundra fire, Seward Peninsula, Alaska, J. Geophys. Res., 112, F02S07, https://doi.org/10.1029/2006JF000554, 2007. Liljedahl, A., Hinzman, L., Busey, R., and Yoshikawa, K.: Physical short-term changes after a tussock tundra fire, Seward Peninsula, Alaska, J. Geophys. Res., 112, F02S07, https://doi.org/10.1029/2006JF000554, 2007. 29 Earth Sci., 59, 759–771, https://doi.org/10.1139/cjes- 2020-0134, 2022. 625 g p p y p g j 2020-0134, 2022. 625 Perreault, N., Lévesque, E., Fortier, D., and Lamarque, L. J.: Thermo-erosion gullies boost the transition from wet to mesic tundra vegetation, Biogeosciences, 13, 1237–1253, https://doi.org/10.5194/bg-13-1237-2016, 2016. Piégay, H., Chabot, A., and Le Lay, Y.-F.: Some comments about resilience: From cyclicity to trajectory, a shift in living and nonliving system theory, Geomorphology, 367, 106527, https://doi.org/10.1016/j.geomorph.2018.09.018, 2020. Perreault, N., Lévesque, E., Fortier, D., and Lamarque, L. J.: Thermo-erosion gullies boost the transition from wet to mesic tundra vegetation, Biogeosciences, 13, 1237–1253, https://doi.org/10.5194/bg-13-1237-2016, 2016. Piégay, H., Chabot, A., and Le Lay, Y.-F.: Some comments about resilience: From cyclicity to trajectory, a shift in living and nonliving system theory, Geomorphology, 367, 106527, https://doi.org/10.1016/j.geomorph.2018.09.018, 2020. Rioux, K.: Impacts de la dégradation du pergélisol par thermo-érosion sur les processus hydrologiques et les flux de 630 matières, M.Sc. géographie, Université de Montréal, Montréal, QC, Canada, 83 pp., 2020. Shur, Y., Hinkel, K. M., and Nelson, F. E.: The transient layer: implications for geocryology and climate-change science, Permafrost Periglac. Process., 16, 5–17, https://doi.org/10.1002/ppp.518, 2005. Rioux, K.: Impacts de la dégradation du pergélisol par thermo-érosion sur les processus hydrologiques et les flux de 630 matières, M.Sc. géographie, Université de Montréal, Montréal, QC, Canada, 83 pp., 2020. Shur, Y., Hinkel, K. M., and Nelson, F. E.: The transient layer: implications for geocryology and climate-change science, Permafrost Periglac. Process., 16, 5–17, https://doi.org/10.1002/ppp.518, 2005. Shur, Y., Jones, B. M., Kanevskiy, M., Jorgenson, T., Jones, M. K. W., Fortier, D., Stephani, E., and Vasiliev, A.: Fluvio‐ thermal erosion and thermal denudation in the yedoma region of northern Alaska: Revisiting the Itkillik River exposure, 635 P f t d P i l P 32 277 298 htt //d i /10 1002/ 2105 2021 Rioux, K.: Impacts de la dégradation du pergélisol par thermo-érosion sur les processus hydrologiques et les flux de 630 matières, M.Sc. géographie, Université de Montréal, Montréal, QC, Canada, 83 pp., 2020. Rioux, K.: Impacts de la dégradation du pergélisol par thermo-érosion sur les processus hydrologiques et les flux de 630 matières, M.Sc. géographie, Université de Montréal, Montréal, QC, Canada, 83 pp., 2020. Shur, Y., Hinkel, K. M., and Nelson, F. E.: The transient layer: implications for geocryology and climate-change science, Shur, Y., Hinkel, K. M., and Nelson, F. E.: The transient layer: implications for geocryology and climate-change science, Permafrost Periglac. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Mackay, J. R.: The World of Underground Ice, Annals of the Association of American Geographers, 62, 1–22, 1972. Maxwell, J. B.: The climate of the Canadian Arctic islands and adjacent waters / Le climat des îles arctiques et des eaux adjacentes du canada, Environment Canada, Atmospheric Environment Service., Hull, QC, Canada, 1980. 610 McRoberts, E. C. and Nixon, J. F.: Reticulate Ice Veins in Permafrost, Northern Canada: Discussion, Can. Geotech. J., 12, 159–162, https://doi.org/10.1139/t75-017, 1975. Morgenstern, A., Overduin, P. P., Günther, F., Stettner, S., Ramage, J., Schirrmeister, L., Grigoriev, M. N., and Grosse, G.: Thermo‐erosional valleys in Siberian ice‐rich permafrost, Permafrost and Periglac Process, 32, 59–75, https://doi.org/10.1002/ppp.2087, 2021. 15 615 Murton, J. B.: 8.14 Ground Ice and Cryostratigraphy, in: Treatise on Geomorphology, Elsevier, 173–201, https://doi.org/10.1016/B978-0-12-374739-6.00206-2, 2013. Murton, J. B. and French, H. M.: Cryostructures in permafrost, Tuktoyaktuk coastlands, western arctic Canada, Can. J. Earth Sci., 31, 737–747, https://doi.org/10.1139/e94-067, 1994. Murton, J. B.: 8.14 Ground Ice and Cryostratigraphy, in: Treatise on Geomorphology, Elsevier, 173–201, https://doi.org/10.1016/B978-0-12-374739-6.00206-2, 2013. Murton J B and French H M : Cryostructures in permafrost Tuktoyaktuk coastlands western arctic Canada Can J Earth Murton, J. B.: 8.14 Ground Ice and Cryostratigraphy, in: Treatise on Geomorphology, Elsevier, 173–201, https://doi.org/10.1016/B978-0-12-374739-6.00206-2, 2013. Murton, J. B. and French, H. M.: Cryostructures in permafrost, Tuktoyaktuk coastlands, western arctic Canada, Can. J. Earth Sci., 31, 737–747, https://doi.org/10.1139/e94-067, 1994. Olefeldt, D., Goswami, S., Grosse, G., Hayes, D., Hugelius, G., Kuhry, P., McGuire, A. D., Romanovsky, V. E., Sannel, A. 620 B. K., Schuur, E. A. G., and Turetsky, M. R.: Circumpolar distribution and carbon storage of thermokarst landscapes, Nat Commun, 7, 13043, https://doi.org/10.1038/ncomms13043, 2016. Olefeldt, D., Goswami, S., Grosse, G., Hayes, D., Hugelius, G., Kuhry, P., McGuire, A. D., Romanovsky, V. E., Sannel, A. 620 B. K., Schuur, E. A. G., and Turetsky, M. R.: Circumpolar distribution and carbon storage of thermokarst landscapes, Nat Commun, 7, 13043, https://doi.org/10.1038/ncomms13043, 2016. Paquette, M., Fortier, D., and Lamoureux, S. F.: Cryostratigraphical studies of ground ice formation and distribution in a High Arctic polar desert landscape, Resolute Bay, Nunavut1, Can. J. Earth Sci., 59, 759–771, https://doi.org/10.1139/cjes- Paquette, M., Fortier, D., and Lamoureux, S. F.: Cryostratigraphical studies of ground ice formation and distribution in a High Arctic polar desert landscape, Resolute Bay, Nunavut1, Can. J. Process., 16, 5–17, https://doi.org/10.1002/ppp.518, 2005. Shur, Y., Jones, B. M., Kanevskiy, M., Jorgenson, T., Jones, M. K. W., Fortier, D., Stephani, E., and Vasiliev, A.: Fluvio‐ thermal erosion and thermal denudation in the yedoma region of northern Alaska: Revisiting the Itkillik River exposure, 635 Permafrost and Periglac Process, 32, 277–298, https://doi.org/10.1002/ppp.2105, 2021. Shur, Y. L.: The upper horizon of permafrost soils, in: Proceedings of the fifth international conference on permafrost, 867– 871, 1988. Smith, S. L. and Burgess, M. M.: A digital database of permafrost thickness in Canada, https://doi.org/10.4095/213043, 2002. 640 640 30 https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. https://doi.org/10.5194/egusphere-2024-208 Preprint. Discussion started: 14 March 2024 c⃝Author(s) 2024. CC BY 4.0 License. Strauss, J., Schirrmeister, L., Grosse, G., Wetterich, S., Ulrich, M., Herzschuh, U., and Hubberten, H.: The deep permafrost carbon pool of the Yedoma region in Siberia and Alaska, Geophys. Res. Lett., 40, 6165–6170, https://doi.org/10.1002/2013GL058088, 2013. Thévenin, E.: Les lithalses : étude d’un pergélisol marginal en dégradation dans la vallée Ä’ä ̨y Chù, au Sud-Ouest du Yukon, Mémoire de maîtrise, Université de Montréal, Montréal, QC, Canada, 2023. 645 Thoms, M. C., Piégay, H., and Parsons, M.: What do you mean, ‘resilient geomorphic systems’?, Geomorphology, 305, 8– 19, https://doi.org/10.1016/j.geomorph.2017.09.003, 2018. Toniolo, H., Kodial, P., Hinzman, L. D., and Yoshikawa, K.: Spatio-temporal evolution of a thermokarst in Interior Alaska, Cold Regions Science and Technology, 56, 39–49, https://doi.org/10.1016/j.coldregions.2008.09.007, 2009. Thoms, M. C., Piégay, H., and Parsons, M.: What do you mean, ‘resilient geomorphic systems’?, Geomorphology, 305, 8– 19, https://doi.org/10.1016/j.geomorph.2017.09.003, 2018. Toniolo, H., Kodial, P., Hinzman, L. D., and Yoshikawa, K.: Spatio-temporal evolution of a thermokarst in Interior Alaska, Cold Regions Science and Technology, 56, 39–49, https://doi.org/10.1016/j.coldregions.2008.09.007, 2009. Turetsky, M. R., Abbott, B. W., Jones, M. C., Anthony, K. W., Olefeldt, D., Schuur, E. A. G., Grosse, G., Kuhry, P., 0 Hugelius, G., Koven, C., Lawrence, D. M., Gibson, C., Sannel, A. B. K., and McGuire, A. D.: Carbon release through abrupt permafrost thaw, Nat. Geosci., 13, 138–143, https://doi.org/10.1038/s41561-019-0526-0, 2020. Turetsky, M. R., Abbott, B. W., Jones, M. C., Anthony, K. W., Olefeldt, D., Schuur, E. A. G., Grosse, G., Kuhry, P., 650 Hugelius, G., Koven, C., Lawrence, D. M., Gibson, C., Sannel, A. B. K., and McGuire, A. D.: Carbon release through abrupt permafrost thaw, Nat. Geosci., 13, 138–143, https://doi.org/10.1038/s41561-019-0526-0, 2020. Veillette, A.: Stabilisation du paysage périglaciaire suite à un épisode de ravinement par thermo-érosion : implication pour la structure et la stabilité thermique du pergélisol de surface, M.Sc. géographie, Université de Montréal, Montréal, QC, Canada, 91 pp., 2019. 655 31
W4380360451.txt	https://ejournal.iainbatanghari.ac.id/index.php/attasyrih/article/download/128/73	en		Principal Leadership as a Quality Culture Motivator	At-Tasyrih: Jurnal Pendidikan dan Hukum Islam/At-Tasyrih	2,023	cc-by-sa	5,555	At-Tasyrih Jurnal Pendidikan dan Hukum Islam, 7(1) - June 2021 38-50 Principal Leadership as a Quality Culture Motivator Mustajab 1, Chuen Lee 2, Jansee 3 1 Institut Agama Islam Nahdlatul Ulama, Indonesia 2 The Education University of Hong Kong, Hong Kong 3 University of Santo Thomas, Country Corresponding Author: Mustajab, E-mail; musta.dewey@gmail.com Received: June 01, 2021 Revised: June 05, 2021 Accepted: June 11, 2021 Online: June 11, 2023 ABSTRACT Quality improvement is becoming increasingly important for an educational institution as an effort to gain better control through its actions. The principal is a leader with the authority to regulate everything in a school institution. For the duties and responsibilities of school leaders to become real, presumably, school principals need to understand, explore and apply several concepts of management science first these have been developed by thinkers in the business world, namely TQM (Total Quality Management) or Integrated Quality Management - later applied in education. Implementation of education needs to pay attention to the characteristics, aspirations, and needs of the community. Education should be able to provide a contextual response according to the global context. In other words, various efforts to bring together multiple parties with interests or stakeholders so that access to the formulation of policies and decision-making concerning equity and expansion of services, quality, relevance, and efficiency of education management. Keywords: Education, Principal, Quality Journal Homepage https://ejournal.iainbatanghari.ac.id/index.php/attasyrih/index This is an open-access article under the CC BY SA license. https://creativecommons.org/licenses/by-sa/4.0/ How to cite: Mustajab, Mustajab., Lee, C., Jansee, Jansee. (2021). Principal Leadership as a Quality Culture Motivator. At-Tasyrih: Jurnal Pendidikan dan Hukum Islam, 7(1), 38-50. https://doi.org/10.55849/attasyrih.v7i1.128 Published by: Pusat Penelitian dan Pengabdian Masyarakat, Institut Agama Islam Nusantara Batang Hari INTRODUCTION Today's development of school management thinking leads to a management system called Total Quality Management or TQM. Quality improvement is becoming increasingly important for an educational institution as an effort to gain better control through its efforts (Viner dkk., 2020). Especially now that there is a demand for educational institutions to be able to provide quality educational services for students. In addition to the problems above (Qureshi dkk., 2022), education should be able to provide a contextual response following the orientation of regional development (Xia dkk., 2019). In other words, various efforts to bring multiple parties with interests or stakeholders closer together so that access to policy formulation and decision-making related to equity and expansion of services, quality, relevance, and efficiency of education management is very justified. At-Tasyrih: Jurnal Pendidikan dan Hukum Islam Principal Leadership as a Quality Culture Motivator In the era of school independence and School-Based Management (SBM), the school principal is the first and foremost task and responsibility (Kartel dkk., 2022). Because the principal, as a leader, has the authority to regulate everything in a school institution (Keshav dkk., 2022). For the duties and responsibilities of school leaders to become real, presumably, school principals need to understand, explore and apply several concepts of management science first - these have been developed by thinkers in the business world, namely TQM (Total Quality Management) or Integrated Quality Management - later applied in education. For a school institution, the principal is not everything, but everything requires the role of a school principal. But what is more profound in this paper is to see the figure of a quality school principal. For this reason, where to find/build a quality school principal will be known. Not then the 'fake' principal who is present in educational institutions. RESEARCH METHODOLOGY The research method used in this study is a qualitative descriptive analysis method and qualitative data analysis procedures, namely collecting data from several research journals that focus on research topics (Dewi S dkk., 2022). The success of scientific research can be determined by the validity of the data and the methods used in the study (Hikmah dkk., 2022). We use literature research and a descriptive qualitative approach in the form of problem finding and detailed development of certain events through research results or theories based on philosophical analysis studies (Anoum dkk., 2022). The data collection technique is a secondary data collection technique in which data can be quoted or obtained from several sources through books, research results, journals, articles, and other documents related to this research. RESULT AND DISCUSSION The concept of a Quality Principal In the formal sense that the principal is defined as a functional teacher who is given the task of leading a school where the teaching-learning process is held or a place where there is interaction between the teacher who gives the lesson and the students who receive the address (Ilham dkk., 2022). At the same time, Quality is the nature of objects and services (McGinnis dkk., 2019). According to Arcaro, that quality is a structured process to improve the output produced. Quality education creates an environment for educators, parents, government officials, community representatives, and business people to work together to provide opportunities and hopes for students' futures. Leaders need to have personal characteristics that include drive, motivation to lead, honesty and integrity (Hartini dkk., 2022), self-confidence, initiative, creativity/originality, adaptability/flexibility, cognitive ability, knowledge, and charisma (Nopiana dkk., 2022). Leaders' managerial qualities must inspire all levels of Management to demonstrate the leadership qualities needed to develop a TQM culture (Hwangbo dkk., 2019). Therefore, the direct involvement of leaders of educational institutions is significant. 39 Principal Leadership as a Quality Culture Motivator Long-term thinking refers to the vision and mission of the institution (Rahmah dkk., 2022). The vision and mission of the institution must be formulated and achieved jointly by all elements within the institution, where the institution's direction will be for the long term (Nadya dkk., 2022). An activity of staff or anyone in the institution must be able to trace what and how much it contributes to the achievement of the vision and mission of the institution (Cui dkk., 2019). This is where, to implement TQM (Ji dkk., 2019), it is necessary to have a leader who has a long-term vision, can work hard, is diligent and steadfast in carrying out the mission, is disciplined, and has a good attitude of service, for example: caring for staff, polite and virtuous, patient, wise, friendly and willing help others in the organization. Awareness of quality in educational institutions depends on intangible factors (Wolf dkk., 2020), especially the attitude of top Management (heads of secondary elementary education institutions, school principals, and leaders of tertiary institutions/rectors) towards the quality of education services (Nemati dkk., 2020). Attaining a quality level is not the result of short-term implementation to increase competitiveness (Shaji dkk., 2021) but rather through implementing TQM, which requires continuous leadership. Leadership and Management are inseparable units. Leaders must have a far-reaching strategic vision and great strength of will to create conducive organizational dynamics. In contrast, managers must have intrinsic knowledge and in-depth professional judgment to organize, direct and control corporate members. The position of the Principal in TQM According to Hadari Nawari, Integrated Quality Management (TQM) is functional Management with an approach that is continuously focused on improving quality (Hirsch dkk., 2020) so that the product conforms to the quality standards of the people served in the implementation of public service and community development tasks (Baldwin dkk., 2019). The concept departs from Management as a process or series of activities integrating owned resources, which must also be combined with the phasing of the implementation of management functions so that work is realized as a quality production activity (Pizzi dkk., 2020). Every piece in integrated quality management must be carried out through the stages of planning, preparation (including materials and tools), and technical implementation using effective and efficient working methods/methods to produce products in the form of goods or services that are beneficial to society. In carrying out management or management functions must be someone who is given authority. This authority is assigned to the principal in an educational institution or school. It is this role that the principal carries out in striving for himself and his institution to build quality education. And also the change of leadership into an upside-down organization, as follows: The student Teaching team and support staff Leader 40 Principal Leadership as a Quality Culture Motivator CommunicationThe application of TQM is not just a management program aimed at complementing or sweetening activities but is indeed needed to improve the quality of work productivity. TQM must be goal-oriented so that organizational performance is more effective. In this case, according to Thomas Packard that eight parts must be implemented in the organization's activities which are elements that function to support each other to form a TQM building, as below: award Training teamwork leadership Ethics, Integrity, and Trust Leadership and teamwork complement each other in implementing an activity (Dirani dkk., 2020). Loss of communication factors between sections and executors can be a burden or obstacle to the performance of activities (Bartsch dkk., 2020). The provision of appropriate awards for program implementers must be realized immediately after they have finished implementing and achieving their work performance. Through implementing TQM (Raut dkk., 2019), good quality work will be obtained, create a comfortable working atmosphere, share mutually beneficial knowledge, and receive awards according to performance. Integrated Quality Management in a non-profit organization, including education, can only be realized if it is supported by the availability of resources to discover the quality of the process and results to be achieved. In a healthy organizational environment (Hao dkk., 2021), various quality sources can support the full implementation of TQM. According to Hadari Nawawi (Hao dkk., 2021), there are several sources of quality, one of which is the commitment of the top leadership (principal) to quality (Liu dkk., 2019). This commitment is significant because it directly affects every decision and policy making (Marpaung dkk., 2019), selection and implementation of programs and projects, empowerment of human resources, and implementation of controls. This commitment is necessary to create and develop the performance of management functions that are oriented towards the quality of products and public services. The foundation group comprises ethics, integrity, and trust sections. Through ethics, integrity, and honesty, TQM will form a foundation for good work. These three parts will open and show the potential of TQM in an activity. The explanation of the three sections is as follows: Ethics Ethics is a discipline regarding good and evil under any circumstances. Two subjects need to be considered in ethics; the first is organizational ethics, namely organizational ethics, which will determine the organization's condition through the rules that the 41 Principal Leadership as a Quality Culture Motivator workforce must obey to produce a good performance. The second is individual ethics, which relates to the right and wrong behavior of each individual who is part of the actor or executor of an activity within the organization. Integrity Integrity embodies honesty, morals, values, fairness, obedience, and sincerity to the organization's reality (Kayama dkk., 2020). These conditions will show a particular characteristic, namely what (internal and external) is expected and entitled to be obtained by the customer (Dziallas & Blind, 2019). The opposite of integrity is duplicity (not loyal or two-faced). TQM will only work if there is consistency in the work environment. Trust Trust is a product of integrity and ethics. Trust is necessary for performance in TQM to be built or realized (van Pinxteren dkk., 2019). Trust will foster the full participation of all members of the organization. In addition to the commitment to implementing activities, trust is needed to determine decisions at every level in the organization and bring every individual to take risks to improve their performance to help improve the organization. Trust is a factor that supports and guarantees user (customer) satisfaction, so it can be interpreted that trust can build environmental cooperation to support TQM. Communication Communication is an interactive activity between the sender and receiver of information (Lin dkk., 2020). The success of TQM requires communication among all members of the organization, providers or suppliers, and users of products or services. Supervisors or assistants must open the atmosphere so activity implementers can receive and send information about the TQM process (Donahue dkk., 2019). Two-way communication and sharing correct statements are expected conditions (Li dkk., 2019). For this communication to be trusted, the message must be clear, and the recipient must interpret it according to the flow of the sender's message. Recognition Awards are the final part of a TQM system and must be given to teams and individuals who have implemented and achieved their activity objectives. Any individual involved in an activity has shown his performance as himself and in a group. Knowing the success of a job is the task of a companion or supervisor in the field. Appreciation can improve or increase self-confidence, productivity, and quality in handling a job. Rewards must be given immediately because it is the right after someone does a job. Applying TQM in program implementation requires high commitment from members or individuals involved. Through integration in the TQM section, it is hoped that guarantees for the success of its activities will be obtained; the role of the school principal is enormous because it must be able to develop, train and apply these eight sections to the implementation of educational programs in schools. Using TQM with a foundation of ethics, integrity, and trust will work. Training for both TQM empowerment and specific technical skills is the key to implementing a program because it will form a work environment that follows program objectives. 42 Principal Leadership as a Quality Culture Motivator The Head of School as TQM Mover In principle, this TQM management system supervises all organization members (school residents) on school activities. So that the application here means that all school members are responsible for the quality of education. The application of TQM requires the involvement of all parties in the educational process, starting from school principals, heads of administration, teachers, students, and employees to school committees, who must understand the nature and purpose of an education which is its vision and mission. In other words, every individual involved must understand the meaning of education. With a thorough understanding of the individuals involved, it is possible to implement TQM. According to Hersey and Blanchard, as quoted by Tobroni, that leadership is the process of influencing the activities of a person or group of people to achieve goals in certain situations. Gibson added that leadership affects the motivation or competence of individuals in a group. While relation to TQM -according to Goetsch and Davis, that leadership is the ability to arouse the enthusiasm of other people to be willing and have total responsibility for efforts to achieve or exceed organizational goals. Can be described as follows: Intangible: Personal Leaders, Critical attitude toward organizational decisions Organization Leader Beliefs, Values, Vision, and self organizational quality Tangible: Organizational structure, facility management system, formal statements. Picture. Series of Leaders in Forming Institutional Quality Based on the picture above, it is explained that the quality of an institution is influenced by the extent to which leadership can direct organizational activities, both tangible (tangible) and intangible (actual), to build strong quality. Here a leader is required to manage the quality elements of the institution as a manifestation in the form of beliefs, values , and all member behavior. Leadership or leadership is the most essential part of TQM. Leadership in TQM requires a leader who has a vision, can make strategic directions for activities understood by all implementers, and can provide guidance on organizational performance. Supervisors in the activity process must understand TQM, have a commitment, and always excel in performance as role model for other implementers. In addition, a supervisor or activity assistant must ensure that the activity's strategy, philosophy, values , and objectives are conveyed throughout the organization to maintain focus and clear direction. According to Dheeraj Mehrotra, a leader must know what students, service users, or consumers want or will get, whether as a supervisor, teacher, school principal, or 43 Principal Leadership as a Quality Culture Motivator activity companion. Commitment from all individuals involved is needed, from the leadership down to the lowest level, to form and disseminate high-quality work values to achieve the sustainability of activity objectives. Building a Quality Culture in Schools A school has components of teachers, students, and administrative staff, each with a specific task in carrying out the program. As a formal educational institutions, schools must produce graduates with specific academic abilities, skills, attitudes, mentality, and other personalities, so they can continue to a higher level of education or work in jobs requiring expertise and skills. Quality-Based School Culture According to Mulyadi, they cite the opinion of Schein, Martin, Kilmann, and Robin that school culture is a system of meaning to foster mentality so that employees' thoughts and actions are based on moral considerations and can be accounted for. Or in other words, school culture is a set of assumptions built and shared by a school as morals in adapting to the external environment and internal integration processes. In building a quality school (TQM), school administrators' work culture, performance, and discipline (teachers, staff, and principals) are the first to be improved. Everyone must be able to view students as 'customers' who must be served as well as possible for their satisfaction. School administrators are always eager to move forward, keen to continue increasing their abilities and skills, ultimately improving their performance in front of students. Suppose all school implementers already have a high work culture, performance, and discipline. In that case, the implementation of TQM can work. It will make educational organizations (schools) more advanced, exist, have a higher brand image, and, in the end, be able to create cadres - national cadres who are qualified and can be aligned with other nations. In Total Quality Management (TQM) or Integrated Quality Management in education, the ultimate goal is to improve quality and competitiveness for output (graduates) with indicators of competency, both intellectual and skill, as well as high social competence of students/graduates. In achieving these results, implementing TQM in educational organizations (schools) must be done truthfully, not half-heartedly. By utilizing all the quality entities within the organization, our education will run differently today. The quality of our education is ranked 101st and is still below Vietnam, which incidentally is a country that has just become independent compared to the independence of our nation, Indonesia. Role of School Culture Culture has links and roles to various aspects of school life. The primary function of school culture is to adapt to the external environment and internal integration processes. This function is the force that drives and controls the behavior of its members, be it teachers, employees, students, and people involved in communicating with their environment. A strong culture plays a role in two things, namely, first, directing behavior. All elements will understand how to act and what is expected of them. Second, a strong culture gives a sense of purpose and makes them think positively about the institution. 44 Principal Leadership as a Quality Culture Motivator Socialization of School Culture Socialization of school culture is a process of adapting employees to the culture created by the school. Cultural socialization consists of two main stages, namely, learning and adaptation. The learning stage is when employees learn about the life patterns of school institutions. Employees know various rules relating to the performance of duties and institutional behavior. At the same time, the adaptation stage is when employees have adjusted to the school system. The adaptation process goes through various ways, namely the example of the leader (principal), characterizations, routines, symbols, and slogans. Build a Quality Culture Culture will shape the characteristics and build trust in an institution. Hickman and Silva suggest three things in an institution that encourage a thriving quality culture: commitment, competence and consistency. Commitment is an employee's agreement to the existence of the institution. Competence is the ability to carry out tasks to achieve organizational goals, and consistency is the stability to continuously adhere to commitments and abilities as employees responsible for the organization's sustainability. And they are also supported by cohesion. Namely, there is integration between commitment, knowledge, and consistency in implementing culture in an institution. A unit is a strong culture built by the four dimensions of the 4 K. The commitment to do the best for the institution needs to be supported by Kemam. capability Consistency commitment consistency Building a strong culture requires strong leaders with a strong vision and personality. Leaders influence, instilling the values that have been created. A leader must exemplify how subordinates carry out their duties correctly and responsibly. In the realm of implementing TQM in educational organizations, it takes work. There are obstacles in the work culture, and the performance of teachers and employees is very influential. There is no denying that civil servants' work culture, performance, and discipline in our country are shallow. This dramatically affects the effectiveness of TQM implementation. Implementing School-Based Management (SBM), which has adopted TQM principles, does not necessarily boost the performance of school administrators, the implications of which can improve the competence of our students. Integrated Quality Management (Total Quality Management) is very popular in the profit organization environment, especially in the climate of various business entities/companies and industries, which have proven their success in maintaining and developing their respective existences in competitive business conditions. Conditions like this have encouraged various parties to practice it in non-profit organizations, educational 45 Principal Leadership as a Quality Culture Motivator institutions, or schools. The principle of TQM is that all this time, schools are considered as a production unit, where students are the raw material and school graduates are the product. In TQM, schools are understood as service units, namely learning services. For quality to be maintained and the process of quality improvement to be controlled, there must be standards that regulate and are agreed upon nationally to be used as indicators for evaluating the success of quality improvement (benchmarking). This thinking has encouraged the emergence of a new approach, namely the Management of improving the quality of education, which can empower all the resources owned by schools to achieve school goals. Several mentalities need to be built into the school culture to achieve quality standards so that they can provide satisfying services for education consumers such as Reliability. Services follow what was promised in meetings or brochures by prioritizing honesty, safety, timeliness, and availability. Guarantee (assurance). This means that schools can guarantee the quality of the services provided, for example, regarding teacher/staff competence and objectivity. Appearance (tangible). That is, the school situation looks good regarding tidiness, cleanliness, order, and beauty. Attention (empathy). This means that the school gives full attention to customers. Responsiveness. This means that schools must respond quickly to customer needs. Developing a quality culture in schools is not instantaneous but through a relatively long process of struggle with various challenges and even resistance. Creating a quality culture must start with the willingness and ability of the school principal, along with staff and stakeholders, to conduct a careful and objective school review. Starting from the school review, schools must establish benchmarking and follow up with quality control. After setting the ideal performance, the principal and staff must try to control quality by comparing actual performance with perfect performance (benchmarking). If there is a discrepancy between the real version and the ideal arrangement, it is necessary to diagnose the cause of the problem and then take concrete actions on how to solve it. On the other hand, if the actual performance has reached the ideal arrangement, higher quality improvements (benchmarking) must be carried out continuously. If this is constantly done, the principal will be ready and able to guarantee the quality of his school (quality assurance). Quality assurance may still be partial because it prioritizes specific components, such as the learning process, curriculum, and teacher components or others. If a school has been able to guarantee the quality of all members, both academic and nonacademic, as well as all components of the school system, it means that the school has been able to implement TQM (Total Quality Management). If this happens, a quality culture will be achieved at the school. So a school that has implemented TQM, including a quality culture, has developed well - it doesn't matter who the principal is - then the existing components will provide the best quality service to students. CONCLUSION Understanding school institutions is very important because of the primary basis of education in school. The importance of understanding school effectiveness is related to improving the quality of teaching. It aligns with the national policy, namely the 46 Principal Leadership as a Quality Culture Motivator decentralization of education in implementing regional autonomy. With this concept, the government hopes to improve the quality of teaching and achieve equity, relevance, and efficiency in education delivery. So the implementation of education needs to pay attention to the characteristics, aspirations, and needs of the community. Education should provide a contextual response following the orientation of regional development. In other words, various efforts to bring multiple parties with interests or stakeholders closer together so that access to policy formulation and decision-making related to equity and expansion of services, quality, relevance, and efficiency of education management is very justified. All institutional activities must be oriented towards achieving quality. All members or parts of the institution must be aware and confident about the need for the quality of their respective performances. Therefore there must be a strong determination and commitment to maintain and improve the quality of their work, which supports the institution's rate. With a commitment to quality, it will be able to drive continuous efforts to enhance the quality so that it will not give up on the obstacles and difficulties that face implementing TQM in the context of continuous quality improvement. REFERENCES Anoum, P., Arifa, F., & May, C. (2022). Strategies to Increase the Motivation of Tahfidz Al-Quran. Journal International Inspire Education Technology, 1(2), 74–85. https://doi.org/10.55849/jiiet.v1i2.88 Baldwin, J. R., Reuben, A., Newbury, J. B., & Danese, A. (2019). Agreement Between Prospective and Retrospective Measures of Childhood Maltreatment: A Systematic Review and Meta-analysis. JAMA Psychiatry, 76(6), 584. https://doi.org/10.1001/jamapsychiatry.2019.0097 Bartsch, S., Weber, E., Büttgen, M., & Huber, A. (2020). Leadership matters in crisisinduced digital transformation: How to lead service employees effectively during the COVID-19 pandemic. Journal of Service Management, 32(1), 71–85. https://doi.org/10.1108/JOSM-05-2020-0160 Cui, Y., Yao, H., Hong, L., Zhang, T., Xu, Y., Xian, K., Gao, B., Qin, J., Zhang, J., Wei, Z., & Hou, J. (2019). Achieving Over 15% Efficiency in Organic Photovoltaic Cells via Copolymer Design. Advanced Materials, 31(14), 1808356. https://doi.org/10.1002/adma.201808356 Dewi S, Y., Rasyid Umar, A., Ali Khan, A., & Aziz, A. (2022). Fun Arabic Teaching With Media Song For Early Child Education School. Journal International of Lingua and Technology, 1(2), 140–156. https://doi.org/10.55849/jiltech.v1i2.80 Dirani, K. M., Abadi, M., Alizadeh, A., Barhate, B., Garza, R. C., Gunasekara, N., Ibrahim, G., & Majzun, Z. (2020). Leadership competencies and the essential role of human resource development in times of crisis: A response to Covid-19 pandemic. Human Resource Development International, 23(4), 380–394. https://doi.org/10.1080/13678868.2020.1780078 Donahue, N. D., Acar, H., & Wilhelm, S. (2019). Concepts of nanoparticle cellular uptake, intracellular trafficking, and kinetics in nanomedicine. Advanced Drug Delivery Reviews, 143, 68–96. https://doi.org/10.1016/j.addr.2019.04.008 47 Principal Leadership as a Quality Culture Motivator Dziallas, M., & Blind, K. (2019). Innovation indicators throughout the innovation process: An extensive literature analysis. Technovation, 80–81, 3–29. https://doi.org/10.1016/j.technovation.2018.05.005 Hao, Y., Hao, S., Andersen-Nissen, E., Mauck, W. M., Zheng, S., Butler, A., Lee, M. J., Wilk, A. J., Darby, C., Zager, M., Hoffman, P., Stoeckius, M., Papalexi, E., Mimitou, E. P., Jain, J., Srivastava, A., Stuart, T., Fleming, L. M., Yeung, B., … Satija, R. (2021). Integrated analysis of multimodal single-cell data. Cell, 184(13), 3573-3587.e29. https://doi.org/10.1016/j.cell.2021.04.048 Harilal, S., Jose, J., Parambi, D. G. T., Kumar, R., Mathew, G. E., Uddin, M. S., Kim, H., & Mathew, B. (2019). Advancements in nanotherapeutics for Alzheimer’s disease: Current perspectives. Journal of Pharmacy and Pharmacology, 71(9), 1370–1383. https://doi.org/10.1111/jphp.13132 Hartini, S., Alie, E., & March, J. (2022). The Relationship Between Authoritarian Parenting and Aggressive Behavior of Adolescents in Nagari Bungo Tanjung. World Psychology, 1(2), 18–26. https://doi.org/10.55849/wp.v1i2.98 Hikmah, D., Petoukhoff, G., & Papaioannou, J. (2022). The Utilization Of The Animiz Application As A Media For Arabic Language Learning On Students. Journal International of Lingua and Technology, 1(2), 157–171. https://doi.org/10.55849/jiltech.v1i2.84 Hirsch, J. S., Ng, J. H., Ross, D. W., Sharma, P., Shah, H. H., Barnett, R. L., Hazzan, A. D., Fishbane, S., Jhaveri, K. D., Abate, M., Andrade, H. P., Barnett, R. L., Bellucci, A., Bhaskaran, M. C., Corona, A. G., Chang, B. F., Finger, M., Fishbane, S., Gitman, M., … Ng, J. Hwei. (2020). Acute kidney injury in patients hospitalized with COVID-19. Kidney International, 98(1), 209–218. https://doi.org/10.1016/j.kint.2020.05.006 Hwangbo, J., Lee, J., Dosovitskiy, A., Bellicoso, D., Tsounis, V., Koltun, V., & Hutter, M. (2019). Learning agile and dynamic motor skills for legged robots. Science Robotics, 4(26), eaau5872. https://doi.org/10.1126/scirobotics.aau5872 Ilham, S., Adelir, D., & Delr, Q. (2022). The Benefits of Whatsapp as a Medium in Depositing Memorization of the Qur’an. Journal International Inspire Education Technology, 1(2), 86–98. https://doi.org/10.55849/jiiet.v1i2.90 Ji, X., Vedaldi, A., & Henriques, J. (2019). Invariant Information Clustering for Unsupervised Image Classification and Segmentation. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), 9864–9873. https://doi.org/10.1109/ICCV.2019.00996 Kartel, A., Charles, M., Xiao, H., & Sundi, D. (2022). Strategies for Parent Involvement During Distance Learning in Arabic Lessons in Elementary Schools. Journal International of Lingua and Technology, 1(2), 99–113. https://doi.org/10.55849/jiltech.v1i2.82 Kayama, H., Okumura, R., & Takeda, K. (2020). Interaction Between the Microbiota, Epithelia, and Immune Cells in the Intestine. Annual Review of Immunology, 38(1), 23–48. https://doi.org/10.1146/annurev-immunol-070119-115104 Keshav, M., Julien, L., & Miezel, J. (2022). The Role Of Technology In Era 5.0 In The Development Of Arabic Language In The World Of Education. Journal International of Lingua and Technology, 1(2), 79–98. https://doi.org/10.55849/jiltech.v1i2.85 Li, C., Liu, Q., Kong, F., Guo, D., Zhai, J., Su, M., & Sun, D. (2019). Circulation and genetic diversity of Feline coronavirus type I and II from clinically healthy and FIP 48 Principal Leadership as a Quality Culture Motivator ‐suspected cats in China. Transboundary and Emerging Diseases, 66(2), 763–775. https://doi.org/10.1111/tbed.13081 Lin, C., Tu, P., & Beitsch, L. M. (2020). Confidence and Receptivity for COVID-19 Vaccines: A Rapid Systematic Review. Vaccines, 9(1), 16. https://doi.org/10.3390/vaccines9010016 Liu, W., Wang, Z., Wang, G., Liu, G., Chen, J., Pu, X., Xi, Y., Wang, X., Guo, H., Hu, C., & Wang, Z. L. (2019). Integrated charge excitation triboelectric nanogenerator. Nature Communications, 10(1), 1426. https://doi.org/10.1038/s41467-019-09464-8 Marpaung, D., Yao, J., & Capmany, J. (2019). Integrated microwave photonics. Nature Photonics, 13(2), 80–90. https://doi.org/10.1038/s41566-018-0310-5 McGinnis, C. S., Murrow, L. M., & Gartner, Z. J. (2019). DoubletFinder: Doublet Detection in Single-Cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Systems, 8(4), 329-337.e4. https://doi.org/10.1016/j.cels.2019.03.003 Nadya, N., Ameer, A., & Zaamil, Z. (2022). Emotional Intelligence and Conflict Management Capabilities in Prisoners in Prisons. World Psychology, 1(2), 54–70. https://doi.org/10.55849/wp.v1i2.94 Nemati, M., Ebrahimi, B., & Nemati, F. (2020). Assessment of Iranian Nurses’ Knowledge and Anxiety Toward COVID-19 During the Current Outbreak in Iran. Archives of Clinical Infectious Diseases, 15(COVID-19). https://doi.org/10.5812/archcid.102848 Nopiana, N., Egie, J., & Mers, O. (2022). The Impact of Internet Addiction on Introvert Personality. World Psychology, 1(2), 1–17. https://doi.org/10.55849/wp.v1i2.97 Pizzi, S., Caputo, A., Corvino, A., & Venturelli, A. (2020). Management research and the UN sustainable development goals (SDGs): A bibliometric investigation and systematic review. Journal of Cleaner Production, 276, 124033. https://doi.org/10.1016/j.jclepro.2020.124033 Qureshi, M., Mahdiyyah, D., Mohamed, Y., & Ardchir, M. (2022). Scale For Measuring Arabic Speaking Skills In Early Children’s Education. Journal International of Lingua and Technology, 1(2), 114–130. https://doi.org/10.55849/jiltech.v1i2.81 Rahmah, A., Rouns, E., & Luck, A. (2022). The Effect of Self-Development Program for Improving Independence in Defective Students in SLB N 1 Lima Kaum Batusangkar. World Psychology, 1(2), 46–53. https://doi.org/10.55849/wp.v1i2.96 Raut, R. D., Mangla, S. K., Narwane, V. S., Gardas, B. B., Priyadarshinee, P., & Narkhede, B. E. (2019). Linking big data analytics and operational sustainability practices for sustainable business management. Journal of Cleaner Production, 224, 10–24. https://doi.org/10.1016/j.jclepro.2019.03.181 Shaji, E., Santosh, M., Sarath, K. V., Prakash, P., Deepchand, V., & Divya, B. V. (2021). Arsenic contamination of groundwater: A global synopsis with focus on the Indian Peninsula. Geoscience Frontiers, 12(3), 101079. https://doi.org/10.1016/j.gsf.2020.08.015 van Pinxteren, M. M. E., Wetzels, R. W. H., Rüger, J., Pluymaekers, M., & Wetzels, M. (2019). Trust in humanoid robots: Implications for services marketing. Journal of Services Marketing, 33(4), 507–518. https://doi.org/10.1108/JSM-01-2018-0045 Viner, R. M., Russell, S. J., Croker, H., Packer, J., Ward, J., Stansfield, C., Mytton, O., Bonell, C., & Booy, R. (2020). School closure and management practices during coronavirus outbreaks including COVID-19: A rapid systematic review. The 49 Principal Leadership as a Quality Culture Motivator Lancet Child & Adolescent Health, 4(5), 397–404. https://doi.org/10.1016/S23524642(20)30095-X Wolf, M. S., Serper, M., Opsasnick, L., O’Conor, R. M., Curtis, L., Benavente, J. Y., Wismer, G., Batio, S., Eifler, M., Zheng, P., Russell, A., Arvanitis, M., Ladner, D., Kwasny, M., Persell, S. D., Rowe, T., Linder, J. A., & Bailey, S. C. (2020). Awareness, Attitudes, and Actions Related to COVID-19 Among Adults With Chronic Conditions at the Onset of the U.S. Outbreak: A Cross-sectional Survey. Annals of Internal Medicine, 173(2), 100–109. https://doi.org/10.7326/M20-1239 Xia, X., Wang, X., Cheng, Z., Qin, W., Lei, L., Jiang, J., & Hu, J. (2019). The role of pyroptosis in cancer: Pro-cancer or pro-“host”? Cell Death & Disease, 10(9), 650. https://doi.org/10.1038/s41419-019-1883-8 Copyright Holder : © Mustajab et al (2021). First Publication Right : © At-Tasyrih: Jurnal Pendidikan dan Hukum Islam This article is under: 50
https://openalex.org/W3034315626	https://escholarship.org/content/qt54r0m1cq/qt54r0m1cq.pdf?t=qhs5i7	English	null	Refining genetic boundaries for Agassiz’s desert tortoise (Gopherus agassizii) in the western Sonoran Desert: the influence of the Coachella Valley on gene flow among populations in southern California	Frontiers of biogeography	2,020	cc-by	10,683	Title Refining genetic boundaries for Agassiz’s desert tortoise (Gopherus agassizii) in the western Sonoran Desert: the influence of the Coachella Valley on gene flow among populations in southern California UC Merced Frontiers of Biogeography Title Refining genetic boundaries for Agassiz’s desert tortoise (Gopherus agassizii) in the western Sonoran Desert: the influence of the Coachella Valley on gene flow among populations in southern California UC Merced Frontiers of Biogeography Title Refining genetic boundaries for Agassiz’s desert tortoise (Gopherus agassizii) in the western Sonoran Desert: the influence of the Coachella Valley on gene flow among populations in southern California Permalink https://escholarship.org/uc/item/54r0m1cq Journal Frontiers of Biogeography, 12(3) Authors Lovich, Jeffrey E. Edwards, Taylor Berry, Kristin H. et al. Publication Date 2020 DOI 10.21425/F5FBG46888 Copyright Information Copyright 2020 by the author(s).This work is made available under the terms of a Creative Commons Attribution License, available at https://creativecommons.org/licenses/by/4.0/ Peer reviewed UC Merced Frontiers of Biogeography Title Refining genetic boundaries for Agassiz’s desert tortoise (Gopherus agassizii) in the western Sonoran Desert: the influence of the Coachella Valley on gene flow among populations in southern California Permalink https://escholarship.org/uc/item/54r0m1cq Journal Frontiers of Biogeography, 12(3) Authors Lovich, Jeffrey E. Edwards, Taylor Berry, Kristin H. et al. Publication Date 2020 DOI 10.21425/F5FBG46888 Copyright Information Copyright 2020 by the author(s).This work is made available under the terms of a Creative Commons Attribution License, available at https://creativecommons.org/licenses/by/4.0/ Peer reviewed Authors Copyright Information Copyright 2020 by the author(s).This work is made available under the terms of a Creative Commons Attribution License, available at https://creativecommons.org/licenses/by/4.0/ Permalink https://escholarship.org/uc/item/54r0m1cq Abstract Substantial changes in the landscape of the American Southwest occurred during the last six million years (including the formation of the Gulf of California and the lower Colorado River), which shaped the distribution and genetic structuring of tortoise populations. Th h f h G lf f C lif i i i d b h • Gene flow and connectivity among populations of Agassiz’s desert tortoises, a conservation-reliant species, are influenced by landscape features including mountains, low-elevation depressions, and even roads. The area northwest of the Gulf of California is occupied by the Salton Trough, including the Coachella Valley at its northern end. Much of this area is below sea level and unsuitable as tortoise habitat, thus forming a potential barrier for gene flow. We assessed genetic relationships among three tortoise populations separated by the Coachella Valley. Two adjacent populations were on the east side of the valley in the foothills of the Cottonwood and Orocopia mountains separated by Interstate 10. The third population, Mesa, was located about 87 km away in the foothills of the San Bernardino Mountains at the far northwestern tip of the valley. The Cottonwood and Orocopia localities showed genetic affiliation with the adjacent Colorado Desert GU immediately to the east, and the Mesa population exhibited affiliation with both the Southern Mojave and Colorado Desert GUs, despite having a greater geographic distance (0.5x–1.5x greater) to the Colorado Desert GU. The genetic affiliation with the Colorado Desert GU suggests that the boundary for that GU needs to be substantially extended to the west to include the desert tortoise populations around the Coachella Valley. Their inclusion in the Colorado Desert GU may benefit these often overlooked populations when recovery actions are considered. • Sub-sea level areas northwest of the Gulf of California occupied by the Salton Trough and the Coachella Valley are unsuitable as tortoise habitat, thus forming a potential barrier for gene flow. • Comparison of tortoise populations separated by the northern Coachella Valley revealed genetic linkages with the Colorado Desert (part of the Sonoran Desert) genetic unit, despite the presence of the low elevation barrier formed by the valley. • Genetic connectivity is postulated to be through the mountains at the north end of the valley, and the Colorado Desert genetic unit should be extended to the west to reflect this connectivity. Abstract Understanding the influence of geographic features on the evolutionary history and population structure of a species can assist wildlife managers in delimiting genetic units (GUs) for conservation and management. Landscape features including mountains, low elevation depressions, and even roads can influence connectivity and gene flow among Agassiz’s desert tortoise (Gopherus agassizii) populations. Substantial changes in the landscape of the American Southwest occurred during the last six million years (including the formation of the Gulf of California and the lower Colorado River), which shaped the distribution and genetic structuring of tortoise populations. The area northwest of the Gulf of California is occupied by the Salton Trough, including the Coachella Valley at its northern end. Much of this area is below sea level and unsuitable as tortoise habitat, thus forming a potential barrier for gene flow. We assessed genetic relationships among three tortoise populations separated by the Coachella Valley. Two adjacent populations were on the east side of the valley in the foothills of the Cottonwood and Orocopia mountains separated by Interstate 10. The third population, Mesa, was located about 87 km away in the foothills of the San Bernardino Mountains at the far northwestern tip of the valley. The Cottonwood and Orocopia localities showed genetic affiliation with the adjacent Colorado Desert GU immediately to the east, and the Mesa population exhibited affiliation with both the Southern Mojave and Colorado Desert GUs, despite having a greater geographic distance (0.5x–1.5x greater) to the Colorado Desert GU. The genetic affiliation with the Colorado Desert GU suggests that the boundary for that GU needs to be substantially extended to the west to include the desert tortoise populations around the Coachella Valley. Their inclusion in the Colorado Desert GU may benefit these often overlooked populations when recovery actions are considered. • The complex geologic and climatic history of the Desert Southwest region of the United States during the last 6 million years shaped the past and present distribution, diversity, and evolution of plants and animals. Understanding the influence of geographic features on the evolutionary history and population structure of a species can assist wildlife managers in delimiting genetic units (GUs) for conservation and management. Landscape features including mountains, low elevation depressions, and even roads can influence connectivity and gene flow among Agassiz’s desert tortoise (Gopherus agassizii) populations. Powered by the California Digital Library University of California Powered by the California Digital Library University of California eScholarship.org eScholarship.org Frontiers of Biogeography 2020, 12.3, e46888 Research Article Frontiers of Biogeography the scientific journal of the International Biogeography Society Frontiers of Biogeography the scientific journal of the International Biogeography Society Refining genetic boundaries for Agassiz’s desert tortoise (Gopherus agassizii) in the western Sonoran Desert: the influence of the Coachella Valley on gene flow among populations in southern California Jeffrey E. Lovich1* , Taylor Edwards2 , Kristin H. Berry3 , Shellie R. Puffer1 , Kristy L. Cummings1 , Joshua R. Ennen4 , Mickey Agha5 , Rachel Woodard6, Kathleen D. Brundige7, and Robert W. Murphy8  Jeffrey E. Lovich1* , Taylor Edwards2 , Kristin H. Berry3 , Shellie R. Puffer1 , Kristy L. Cummings1 , Joshua R. Ennen4 , Mickey Agha5 , Rachel Woodard6, Kathleen D. Brundige7, and Robert W. Murphy8  1 U.S. Geological Survey, Southwest Biological Science Center, 2255 N. Gemini Drive, MS-9394, Flagstaff, AZ 86001-1600, USA; 2 University of Arizona Genetics Core, University of Arizona, 1657 E. Helen Street, Room 111H, Tucson, AZ 85721, USA; 3 U.S. Geological Survey, Western Ecological Research Center, 21803 Cactus Avenue, Suite F, Riverside, CA 92518, USA; 4 Tennessee Aquarium Conservation Institute, Tennessee Aquarium, 175 Baylor School Road, Chattanooga, TN 37405, USA; 5 Washington Department of Fish and Wildlife, 1111 Washington St. SE, Olympia, WA 98501, USA; 6 1500 McLean St. #143, Ridgecrest, CA 93555, USA; 7 Coachella Valley Conservation Commission, 73–710 Fred Waring Drive, Suite 200, Palm Desert, CA 92260–2516, USA; 8 Centre for Biodiversity and Conservation Biology, Royal Ontario Museum, 100 Queen’s Park, Toronto, ON, Canada M5S 2C6. *Corresponding author: Jeffrey Lovich, jeffrey_lovich@usgs.gov; https://www.usgs.gov/staff-profiles/jeffrey-e-lovich Keywords: Agassiz’s desert tortoise, Barrier, Conservation, Genetic Boundaries, Genetic Unit, Gopherus biogeography, Gulf of California, Salton Trough Agassiz’s desert tortoise, Barrier, Conservation, Genetic Boundaries, Genetic Unit, Gopherus biogeography, Gulf of California, Salton Trough https://escholarship.org/uc/fb doi:10.21425/F5FBG46888 doi:10.21425/F5FBG46888 https://escholarship.org/uc/fb Introduction The Desert Southwest region of the United States has a complex geologic and climatic history that shaped the past and present distribution, diversity, and evolution of plants and animals. Biological patterns in the region are a reflection of the effects of processes driven by tectonic, volcanic, glacial-interglacial cycling, and shorter-term climatic events across widely varying time scales (Dolby et al. 2015). Situated at the tectonic and zoologically important interface of the North American and Pacific Plates (Atwater 1970, Gottscho 2016), the Desert Southwest has been profoundly shaped and influenced by two major geological processes in the last 6 million years: the formation of the Gulf of California (Dolby et al. 2015) and the evolution of the lower Colorado River (Howard et al. 2019), both of which presented aquatic barriers to some plants and animals. Disentangling the effects of those and other geological processes on the evolution of the biota of the region is an ongoing topic of interest to researchers (Dolby et al. 2015, 2019). Vicariant events in the region caused by these and other geographic barriers to dispersal (e.g., high mountains and deep valleys) play important roles in the genetic structuring of species. Genetic analyses that assess gene flow, or the lack thereof, can serve to identify barriers to dispersal. Syntheses of geographic and genetic analyses can objectively identify critical management units for conservation, defined as genetic units (GUs; Wayne and Morin 2004). As such, genetic analyses are essential for the effective management of species, especially those that are threatened or endangered. Although GUs have been defined for G. agassizii (Murphy et al. 2007, Sánchez-Ramírez et al. 2018), these studies did not include sampling from near the southwestern limits of the species range (e.g., populations surrounding the Coachella Valley; Berry and Murphy 2019). The low density or absence of tortoises in much of the area surrounding the Coachella Valley is due to the lack of suitable habitat in such low elevation areas (much of which is below sea level) and environmental conditions that are physiologically challenging for sustaining G. agassizii populations (reviewed by Ernst and Lovich 2009, Berry and Murphy 2019). The valley is one of the driest and hottest parts of the Sonoran Desert (Barrows et al. 2010). Its short flowering season for winter annual plants challenges survival of the herbivorous tortoise (Morafka and Berry 2002). e-ISSN: 1948-6596 © the authors, CC-BY 4.0 license © the authors, CC-BY 4.0 license 1 Testing a barrier to gene flow in desert tortoises Lovich et al. effects of natural geographic barriers on movements (Britten et al. 1997, Murphy et al. 2007, Gaillard et al. 2017). For G. agassizii, significant impediments to dispersal and habitation include high mountain ranges and extremely low elevation areas with high summer temperatures (Hagerty et al. 2011), both of which are inhospitable to tortoises. In the Mojave Desert, the latter includes Eureka, Saline, and Death valleys. Similarly, tortoises in the Sonoran Desert of California (including the Colorado Desert subdivision) are all but absent from the low elevation Salton Trough, including the Coachella Valley (Dimmitt 1977, Luckenbach 1982, Berry and Murphy 2019), Cadiz valley, and the lower Colorado River region (Morafka and Berry 2002). Agassiz’s desert tortoises tend to prefer valley bottoms and bajadas at moderate elevations (e.g., 300–920 m; Luckenbach 1982) as habitat. Thus, mountain ranges over about 2,000 m, like the New York (2,296 m) and Providence (2,183 m) mountains, are barriers to gene flow (Hagerty and Tracy 2010). At the local scale, variables such as slope and roads may influence tortoise movements and, thus, gene flow (Latch et al. 2011). Not surprisingly, geographic distance also influences gene flow (Murphy et al. 2007, Hagerty and Tracy 2010, Averill-Murray and Hagerty 2014).i Introduction Further, large areas of the Coachella Valley are affected by urban development and agriculture (Beatley 1992), rendering it even less suitable for desert tortoises and other wildlife.i g Agassiz’s desert tortoise (Gopherus agassizii Cooper) is a threatened species with a wide range involving large portions of both the Mojave and Sonoran deserts of Arizona, California, Nevada, and Utah (Luckenbach 1982, Germano et al. 1994). Desert tortoises are obligate herbivores that avoid some of the harshest conditions of their arid environment by constructing deep burrows and spending the majority of their annual cycle sheltering within. With delayed maturity, a long lifespan, and relatively low egg production, they are very sensitive to additive mortality from human activities and subsidized predators (see reviews in Ernst and Lovich 2009, Berry and Murphy 2019). Despite their wide distribution, declines in the densities of tortoises led to the species being listed as threatened under the U.S. Endangered Species Act in 1990. Declines have continued since then (Allison and Mcluckie 2018), despite significant efforts to recover populations (Averill-Murray et al. 2012). Because conditions in the Coachella Valley are not suitable for tortoises, we hypothesize that it acts a barrier to gene flow for disjunct populations of tortoises that occur on the upland slopes of various mountains and bajadas surrounding the valley, including the Chocolate, Orocopia, Cottonwood, Little San Bernardino, San Bernardino, San Jacinto, and Santa Rosa mountains (e.g., Lovich et al. 2015, 2018a; Berry and Murphy 2019). Comparing the genetic composition of three tortoise populations separated by the Coachella ValIey allows a preliminary test of our hypothesis. It is important to note that inferring the natural distribution of tortoises today is complicated by intentional introductions by humans (Murphy et al. 2007, Edwards et al. 2010, Edwards and Berry 2013). For example, tortoises are known to have been released in Anza Borrego Desert State Park in the uplands on i An important part of the recovery plan for the desert tortoise, developed by the U.S. Fish and Wildlife Service (USFWS 2011), is maintaining and managing for genetic variability and linkages among tortoise populations. Linkages are necessary to offset the negative effects of human-caused barriers in the desert, including roads (Latch et al. 2011) and utility-scale wind and solar energy developments (Lovich and Ennen 2011, 2013). Genetic variability of populations of G. Frontiers of Biogeography 2020, 12.3, e46888 Introduction agassizii is largely a reflection of the © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 2 Testing a barrier to gene flow in desert tortoises Lovich et al. the west side of the Salton Trough between 1971 and 1972 (Luckenbach 1982; J. Manning, Washington State University, personal communication), far from our the study sites. In addition, possible releases by Native American tribes that used tortoises for food or other cultural purposes (Schneider and Everson 1989) likely occurred throughout the range of the species. The success of those and other introductions depends on the ability of translocated tortoises to survive under local conditions, something that is not assured (Germano and Bishop 2008) and that varies among individuals (Germano et al. 2017).ii than the genetic distances these populations have to adjacent tortoise populations outside of the valley to the east or northwest. Analyses integrate data from 24 short tandem repeats (STRs; microsatellites) from these samples with homologous information from Murphy et al. (2007). the west side of the Salton Trough between 1971 and 1972 (Luckenbach 1982; J. Manning, Washington State University, personal communication), far from our the study sites. In addition, possible releases by Native American tribes that used tortoises for food or other cultural purposes (Schneider and Everson 1989) likely occurred throughout the range of the species. The success of those and other introductions depends on the ability of translocated tortoises to survive under local conditions, something that is not assured (Germano and Bishop 2008) and that varies among individuals (Germano et al. 2017).ii Study area The Salton Trough is the northern extension of the basin occupied by the Gulf of California. It is a pull-apart basin caused by subsidence from oblique extension across strike-slip faults (Brothers et al. 2009). Extending for about 225 km, the Salton Trough in California includes (from north to south) the Coachella Valley, Salton Sea, Imperial Valley, and in Mexico the Colorado River Delta (Fig. 1). The terrestrial portion of the trough extends northwest from the Gulf of California to the San Gorgonio Pass, near Palm Springs, Herein, we provide the first genetic assessment of tortoise populations surrounding the Coachella Valley. The first two sampled populations were east of the southern portion of the valley and a third was located about 85 km away at the northwestern tip of the valley. If the Coachella Valley is a barrier to tortoise dispersal and gene flow, genetic distances between sampled populations should be greater Figure 1. Map showing Agassiz’s desert tortoise sampling sites around the Coachella Valley (shaded in gray) in California, USA, the northern extension of the Salton Trough. Study sites are numbered for consistency with other figures and tables: 16=Mesa, 17=Cottonwood, 18=Orocopia. The northwestern part of the Coachella Valley between the label for study site 16 and the San Jacinto Mountains is the San Gorgonio Pass. The light-colored area around and south of the Salton Sea is a continuation of the terrestrial portion of the Salton Trough before it ends at the Gulf of California in Mexico. County lines are shown for reference. The horizontal line is the southern boundary of Riverside County and the vertical line separates San Diego (left) from Imperial counties (right). Dashed line demarcates the approximate high stand of Lake Cahuilla at approximately 12 m ASL during the Pleistocene obtained from https://www.arcgis.com/home/item. html?id=ac2b6de1149047b9af934acd4d01fdca Figure 1. Map showing Agassiz’s desert tortoise sampling sites around the Coachella Valley (shaded in gray) in California, USA, the northern extension of the Salton Trough. Study sites are numbered for consistency with other figures and tables: 16=Mesa, 17=Cottonwood, 18=Orocopia. The northwestern part of the Coachella Valley between the label for study site 16 and the San Jacinto Mountains is the San Gorgonio Pass. The light-colored area around and south of the Salton Sea is a continuation of the terrestrial portion of the Salton Trough before it ends at the Gulf of California in Mexico. County lines are shown for reference. Study area The horizontal line is the southern boundary of Riverside County and the vertical line separates San Diego (left) from Imperial counties (right). Dashed line demarcates the approximate high stand of Lake Cahuilla at approximately 12 m ASL during the Pleistocene obtained from https://www.arcgis.com/home/item. html?id=ac2b6de1149047b9af934acd4d01fdca © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 3 Testing a barrier to gene flow in desert tortoises Lovich et al. California. At its lowest point in the Salton Sink, the depression is about 85 m below sea level, although some of that area is now inundated by the Salton Sea with a surface elevation of about 71 meters below sea level. Approximately 5,400 km2 of the trough (including the Eastern Coachella Valley) is below sea level, and it is surrounded by mountains on all sides except in the south near the Gulf of California (Waters 1983). The Coachella Valley covers about 777 km2 in the northernmost portion of the Salton Trough (Beatley 1992). As mentioned earlier, the low elevation habitats of the Salton Trough, including the Coachella Valley, are unsuitable as tortoise habitat (Nussear et al. 2009) and expected to be barriers to dispersal and gene flow. The Mesa study site, near Palm Springs, California, is located in the foothills (600–900 m) of the San Bernardino Mountains just above the northwestern tip of the Coachella Valley, on the extreme western edge of the Sonoran Desert ecosystem. The site has been extensively developed for wind energy production (Lovich et al. 2011). Vegetation at the site is a combination of coastal, montane, and desert plant species in a fire-prone landscape (Lovich et al. 2018b). Additional details of the study site and a history of research conducted there are presented by Lovich and Daniels (2000), Agha et al. (2015), and Lovich and Ennen (2017). The other two study sites (both near Chiriaco Summit, California) are located in the uplands about 28 km due east of the edge of the Coachella Valley floor and about 85 km southeast of Mesa. The Cottonwood study site is located in the southern part of Joshua Tree National Park along the base of the Cottonwood Mountains. The site is characterized by sloping bajadas and desert washes originating in the steep foothills (520–780 m) of the Cottonwood Mountains to the north. Further details of the vegetation in the area are provided by Lovich et al. (2018a). Study area The Orocopia study site is relatively flat, with elevations (480–620m) increasing to the south and east, and a gentle uphill grade to the Orocopia Mountains to the south. This site is characterized by creosote scrub vegetation (Larrea tridentata Coville) with widely-scattered ocotillos (Fouquieria splendens Engelm.) and blue palo verde trees (Parkinsonia florida Watson). The surface is characterized by areas of desert pavement (Wood et al. 2005) interspersed with sandy to gravelly soil and numerous tank and jeep track scars from World War II training activities (Prose 1985) associated with nearby Camp Young in the early 1940’s. California. At its lowest point in the Salton Sink, the depression is about 85 m below sea level, although some of that area is now inundated by the Salton Sea with a surface elevation of about 71 meters below sea level. Approximately 5,400 km2 of the trough (including the Eastern Coachella Valley) is below sea level, and it is surrounded by mountains on all sides except in the south near the Gulf of California (Waters 1983). The Coachella Valley covers about 777 km2 in the northernmost portion of the Salton Trough (Beatley 1992). As mentioned earlier, the low elevation habitats of the Salton Trough, including the Coachella Valley, are unsuitable as tortoise habitat (Nussear et al. 2009) and expected to be barriers to dispersal and gene flow. p p gl The Mesa study site, near Palm Springs, California, is located in the foothills (600–900 m) of the San Bernardino Mountains just above the northwestern tip of the Coachella Valley, on the extreme western edge of the Sonoran Desert ecosystem. The site has been extensively developed for wind energy production (Lovich et al. 2011). Vegetation at the site is a combination of coastal, montane, and desert plant species in a fire-prone landscape (Lovich et al. 2018b). Additional details of the study site and a history of research conducted there are presented by Lovich and Daniels (2000), Agha et al. (2015), and Lovich and Ennen (2017). Population analyses p We mixed whole blood with lysis buffer and incubated overnight with proteinase K at 550C, followed by robotic extraction using a QIAGEN BioSprint 96 robotic magnetic-particle purification system (Qiagen; Valencia, California, USA) and Aline Biosciences Buccal Swab gDNA Kit (Aline Biosciences; Woburn, Massachusetts, USA). Scute samples were extracted using a Qiagen DNeasy tissue kit. We quantified recovered DNA using a BioTEK Synergy HT (BioTEK; Vermont, USA). We analyzed an approximately 1,100 base pair portion of mitochondrial ND3/ND4 for samples from Orocopia and Mesa to establish a baseline for the populations and to help identify native vs. translocated individuals (Edwards and Berry 2013). We genotyped all samples for 24 previously described short tandem repeats (STRs) following the methods of Edwards and Berry (2013). Fragment analysis and DNA sequencing were performed following standard procedures by UAGC. We compared data from our three study sites (n=53 samples) around the Coachella Valley (described above) to a reference database of 709 samples of G. agassizii collected from throughout the species’ range (Murphy et al. 2007) as updated by Edwards and Berry (2013) for additional loci that match our 24 STR loci. The samples in these previous analyses clustered into 15 groups (our sites 17 and 18 combined into one new group for this analysis) according to Murphy et al. (2007), which we assigned to 5 genetic units (GUs) based on the results of Sánchez-Ramírez et al. (2018) (Table 1). We generated an FST matrix among all groups using ARLEQUIN. Relationships between genetic and geographic distances were considered qualitatively only because geographic features affected linear travel by tortoises (Dutcher et al. 2020). Thus, relatedness among populations relied on assignment tests using Table 1. Desert tortoise study sites and groups modified from Murphy et al. 2007, including our three study sites around the Coachella Valley (sites 16, 17, 18). Sites 17 and 18 were combined into one group due to their proximity. All are grouped into 5 genetic units for G. agassizii as described in Sánchez-Ramírez et al. (2018). Modified sample set for Southern Clade California samples excludes Northern Clade groups 11, 12 and 15. Abbreviations are as follows: NTC = National Training Center, MCAGCC = Marine Corp Air Ground Combat Center, CMAGR = Chocolate Mountain Aerial Gunnery Range. Laboratory Processing A total of 55 samples from our three study sites around the Coachella Valley were processed at the University of Arizona Genetics Core (UAGC), Tucson, AZ, USA. Of these, 31 samples based on scute tissue were received from the University of Southern Mississippi as extracted DNA, and 24 blood samples were extracted and processed at UAGC. Usable DNA was obtained from 53 of the 55 samples.f Descriptive statistics We calculated diversity indices for each population based on polymorphic STR loci. We used Arlequin (v. 3.5.1.2; Excoffier and Lischer 2010) to detect significant departures from Hardy-Weinberg expectations and FSTAT v.2.9.3.2 (Goudet 1995) to generate estimates of gene diversity and allelic richness per locus. We used GENEPOP v.4.2 (Raymond 1995, Rousset 2008) to estimate inbreeding coefficients (FIS; Weir and Cockerham 1984). We used default parameters in FSTAT, ARLEQUIN, and GENEPOP for all Markov-chain tests and permutations. Field techniques We used the subcarapacial venipuncture technique (Hernandez-Divers et al. 2002, Drake et al. 2012) to collect blood and lymph from tortoises, obtaining up to 0.5 ml samples from each adult tortoise using a 23-gauge needle irrigated with sodium heparin. g g g p For samples from hatchling tortoises at the Mesa site collected in 2000, we clipped small amounts of keratinous scute tissue from the marginal scutes as part of our individual marking system (Cagle 1939). © the authors, CC-BY 4.0 license 5 Field techniques From 2000 to 2018, we collected DNA samples from Agassiz’s desert tortoises at sites that are outside but hydrologically part of the Coachella Valley (Fig. 2). Figure 2. Map showing centroids of locations of groups listed in Table 1. Group 16 is the Mesa study site and Groups 17 and 18 are Cottonwood and Orocopia study sites, respectively. The other sites are listed in Table 1. Shaded areas with fine dashed line borders show the adjacent genetic units modified from Sánchez-Ramírez et al. (2018). From left to right they are: Western Mojave GU, Central Mojave GU (upper sliver), Southern Mojave GU (lower), and Colorado Desert GU. The southern boundary of the Colorado Desert GU is further modified from Sánchez-Ramírez based on our data supporting an extension (left of the bold dashed line) of the GU going northwest from Groups 17 and 18 to Group 16. Figure 2. Map showing centroids of locations of groups listed in Table 1. Group 16 is the Mesa study site and Groups 17 and 18 are Cottonwood and Orocopia study sites, respectively. The other sites are listed in Table 1. Shaded areas with fine dashed line borders show the adjacent genetic units modified from Sánchez-Ramírez et al. (2018). From left to right they are: Western Mojave GU, Central Mojave GU (upper sliver), Southern Mojave GU (lower), and Colorado Desert GU. The southern boundary of the Colorado Desert GU is further modified from Sánchez-Ramírez based on our data supporting an extension (left of the bold dashed line) of the GU going northwest from Groups 17 and 18 to Group 16. © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 Testing a barrier to gene flow in desert tortoises Lovich et al. We scored the STR fragment data using Genemarker v.1.85 (SoftGenetics; State College, Pennsylvania, USA). We aligned mtDNA sequences using CLC DNA Workbench v.5.7.1 (CLC Bio; Denmark) to established reference haplotypes in Genbank (Murphy et al. 2007). We used the subcarapacial venipuncture technique (Hernandez-Divers et al. 2002, Drake et al. 2012) to collect blood and lymph from tortoises, obtaining up to 0.5 ml samples from each adult tortoise using a 23-gauge needle irrigated with sodium heparin. For samples from hatchling tortoises at the Mesa site collected in 2000, we clipped small amounts of keratinous scute tissue from the marginal scutes as part of our individual marking system (Cagle 1939). Population analyses Frontiers of Biogeography 2020, 12.3, e46888 © the authors, CC-BY 4.0 license 5 p y g Genetic Unit Study Site # of Samples Group # of Samples in Group Western Mojave Desert Tortoise Natural Area 56 1 60 Fremont-Valley 4 1 Hinkley 12 2 83 Kramer 3 2 Edwards Air Force Base 57 2 Fremont-Kramer 11 2 Central Mojave Superior-Cronese 10 3 52 Fort Irwin (Goldstone) 9 3 Fort Irwin (NTC) 33 3 Fort Irwin (Tiefort) 31 4 31 Fort Irwin (Control Site) 33 5 46 Fort Irwin (Eastgate 2) 13 5 Testing a barrier to gene flow in desert tortoises Lovich et al. Genetic Unit Study Site # of Samples Group # of Samples in Group Southern Mojave Lucerne Valley 11 6 25 Ord-Rodman 14 6 MCAGCC (Emerson) 9 7 70 MCAGCC (Sand Hill) 61 7 Daggett 72 8 72 MCAGCC (Lavic Lake) 8 9 27 MCAGCC (Maumee Mine) 7 9 MCAGCC (Sunshine Park) 12 9 MCAGCC (Bullion) 16 10 19 MCAGCC (Lava) 3 10 Colorado Desert Fenner 4 12 31 Goffs 27 12 Chemhuevi 7 13 17 Upper Ward Valley 10 13 Chuckwalla 17 14 36 CMAGR 19 14 Northern Mojave Ivanpah 33 11 59 Ivanpah (site 14) 23 11 Shadow Valley 3 11 Upper Virgin River. UT 28 15 28 unassigned Mesa 30 16 30 (Coachella Valley) Cottonwood 10 17 23 Orocopia 13 18 Table 1. Continued... California by Murphy et al. (2007), but excluded haplogroup MOJ_B (groups 11, 12, 15 in Table 1) from the northeastern Mojave Desert. WHICHRUN (Ver. 4.1; Banks and Eichert 2000), which calculates the likelihood of a given individual originating from ≥ 2 candidate populations on the basis of its multilocus STR genotype. We assessed stringency of population allocation based on the log of the odds ratio (LOD) for the two most likely source populations. Assignments with a LOD ratio of ≥ 2 had a ≤ 0.01 chance of type I error. Results Mesa became distinct only when evaluated for K = 7 (Fig. 3). population was equidistant in relatedness between locations in the Southern Mojave and Colorado Desert GUs, despite having a greater geographic distance (0.5x–1.5x greater) to the former locations. Assignment tests associated most (21/23) Cottonwood and Orocopia individuals to the Colorado Desert GU, whereas individuals from Mesa were split between the Colorado Desert GU and the Southern Mojave GU, and two individuals were assigned further west to the Western Mojave GU (Table 4). population was equidistant in relatedness between locations in the Southern Mojave and Colorado Desert GUs, despite having a greater geographic distance (0.5x–1.5x greater) to the former locations. Assignment tests associated most (21/23) Cottonwood and Orocopia individuals to the Colorado Desert GU, whereas individuals from Mesa were split between the Colorado Desert GU and the Southern Mojave GU, and two individuals were assigned further west to the Western Mojave GU (Table 4). j ( ) When the Coachella Valley samples were analyzed against the Southern Clade using STRUCTURE Table 2. Mean diversity indices based on 24 microsatellite (STR) loci: n = number of individuals genotyped; richness and diversity estimated using FSTAT. Randomization tests for Hardy–Weinberg equilibrium calculated with ARLIQUIN, where Obs Het = observed heterozygosity and Exp Het = expected heterozygosity. FIS, inbreeding coefficient (Weir and Cockerham 1984) estimated using GENEPOP Italicized values indicate standard deviation of the mean 1984) estimated using GENEPOP. Italicized values indicate standard deviation of the mean. Pop n # alleles Allelic range Allelic richness Gene diversity Obs Het Exp Het FIS Mesa 30 6.60 13.85 2.52 0.65 0.655 0.647 0.000 (3.59) (9.82) (0.61) (0.21) (0.259) (0.221) Cottonwood 10 6.28 13.17 2.81 0.74 0.698 0.739 0.059 (2.82) (10.59) (0.54) (0.15) (0.216) (0.158) Orocopia 13 5.80 12.05 2.56 0.65 0.561 0.649 0.147 (3.24) (10.14) (0.73) (0.24) (0.274) (0.244) Table 3. Population pairwise genetic distance (FST; below diagonal) and geographic distance (Km; above diagonal) between unassigned populations around the Coachella Valley and adjacent genetic units to the east and north-northwest. Group 16 = Mesa, Group 17 = Cottonwood, and Group 18 = Orocopia. Orocopia and Cottonwood locations combined for analysis Table 3. Population pairwise genetic distance (FST; below diagonal) and geographic distance (Km; above diagonal) between unassigned populations around the Coachella Valley and adjacent genetic units to the east and north-northwest. Group 16 = Mesa, Group 17 = Cottonwood, and Group 18 = Orocopia. Results For the three study sites around the Coachella Valley, 34 individuals were sequenced for mtDNA, including 9 from Orocopia and 25 from Mesa; all individuals had haplotype MOJ_A01 (identified by Murphy et al. 2007 as a widely distributed haplotype of haplogroup MOJ_A in California). STRs genotypes were obtained for 53 individuals, including 30 from Mesa, 10 from Cottonwood, and 13 from Orocopia. All populations exhibited heterozygosity consistent with each other (Table 2) and other reported populations of G. agassizii (Edwards and Harrison 2014). Allelic richness and gene diversity were relatively low, but this was not unexpected for small populations that are relatively isolated from a geographic perspective (Table 2).i We used STRUCTURE v.2.3.4 (Pritchard et al. 2000) to assess associations of the Coachella Valley samples with the database without prior population assignments. Analyses were run for K = 1–12 with 10 trials per K, and each run for 500,000 iterations following a burn-in period of 50,000 MCMC iterations under the admixture model, assuming allele frequencies were correlated between populations. We used STRUCTURE HARVESTER Online (Earl and von Holdt 2012) to evaluate the results. Independent STRUCTURE runs were compiled and visualized using the Greedy K algorithm in CLUMPAK (Kopelman et al. 2015). We performed analyses two ways: 1) using all samples shown in Table 1, and 2) just the “Southern Clade”. The latter included all locations with mitochondrial “haplogroup” MOJ_A, found throughout most of g g p p pi ( ) Data for populations at the nearby study sites Cottonwood and Orocopia were combined, and they were most similar (Table 3) to their closest Colorado Desert GU (Group 14; Chuckwalla and CMAGR) in our comparison with the reference database. The Mesa © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 6 Testing a barrier to gene flow in desert tortoises Lovich et al. (Figs. 3–4), the best fit was K = 5, when evaluated using the DeltaK method of Evanno et al. (2005). In all iterations, Cottonwood, Orocopia, and Mesa samples primarily clustered with the “Colorado Desert” (group 14), with a few exceptions (Table 5). Analysis using all samples in the reference database did not change the clustering. As K was increased, Daggett (group 8, Table 1) fell out as a cluster before any of the unassigned Coachella Valley samples broke away from the Colorado Desert GU (group 14). © the authors, CC-BY 4.0 license 7 Results Orocopia and Cottonwood locations combined for analysis d t i it Frontiers of Biogeography 2020, 12.3, e46888 © the authors, CC-BY 4.0 license 7 due to proximity. Genetic Unit Southern Mojave Colorado Desert Unassigned (Coachella Valley) Group 6 7 8 9 10 14 16 17&18 Southern Mojave 6 0 49 32 24 77 173 81 144 7 0.014 0 79 38 35 120 51 81 8 0.020 0.041 0 52 104 202 100 166 9 0.014 0.012 0.023 0 55 156 79 124 10 0.026 0.017 0.047 0.020 0 107 88 74 Colorado Desert 14 0.067 0.047 0.085 0.059 0.045 0 124 36 Unassigned 16 0.056 0.060 0.069 0.060 0.065 0.068 0 87 17&18 0.068 0.058 0.079 0.060 0.050 0.011 0.054 0 Table 4. Population assignment of three tortoise populations around the Coachella Valley to a reference database of 709 G. agassizii samples designated into five genetic units based on the genetic structure reported by Sánchez-Ramírez et al. (2018). Parenthetical values are assignments with LOD > 2. n Genetic unit assignment Western Mojave Southern Mojave Colorado Desert Mesa 30 2 (2) 12 (11) 16 (15) Cottonwood 10 0 1 (1) 9 (9) Orocopia 13 0 1 (1) 12 (11) Frontiers of Biogeography 2020, 12.3, e46888 © the authors, CC-BY 4.0 license 7 Table 4. Population assignment of three tortoise populations around the Coachella Valley to a reference database of 709 G. agassizii samples designated into five genetic units based on the genetic structure reported by Sánchez-Ramírez et al. (2018). Parenthetical values are assignments with LOD > 2. n Genetic unit assignment Western Mojave Southern Mojave Colorado Desert Mesa 30 2 (2) 12 (11) 16 (15) Cottonwood 10 0 1 (1) 9 (9) Orocopia 13 0 1 (1) 12 (11) on assignment of three tortoise populations around the Coachella Valley to a reference database of 709 ples designated into five genetic units based on the genetic structure reported by Sánchez-Ramírez et al. etical values are assignments with LOD > 2. Table 4. Population assignment of three tortoise populations around the Coachella Valley to a reference database of 709 G. agassizii samples designated into five genetic units based on the genetic structure reported by Sánchez-Ramírez et al. (2018). Parenthetical values are assignments with LOD > 2. n Genetic unit assignment Western Mojave Southern Mojave Colorado Desert Table 4. Frontiers of Biogeography 2020, 12.3, e46888 Results Population assignment of three tortoise populations around the Coachella Valley to a reference database of 709 G. agassizii samples designated into five genetic units based on the genetic structure reported by Sánchez-Ramírez et al. (2018). Parenthetical values are assignments with LOD > 2. n Genetic unit assignment Western Mojave Southern Mojave Colorado Desert Mesa 30 2 (2) 12 (11) 16 (15) Cottonwood 10 0 1 (1) 9 (9) Orocopia 13 0 1 (1) 12 (11) Testing a barrier to gene flow in desert tortoises Lovich et al. Table 5. STRUCTURE analysis Q-values for K = 5 clusters averaged across 10 iterations. Standard deviations in italics. Population Colorado Desert GU Southern Mojave GU Southern Mojave GU (Daggett cluster) Western Mojave GU Central Mojave GU Mesa Ave 0.752 0.100 0.064 0.063 0.021 SD 0.230 0.137 0.096 0.103 0.026 Orocopia Ave 0.893 0.042 0.025 0.019 0.019 SD 0.086 0.057 0.019 0.014 0.019 Cottonwood Ave 0.899 0.039 0.023 0.015 0.025 SD 0.049 0.036 0.011 0.007 0.025 Figure 3. STRUCTURE analyses for Agassiz’s desert tortoise without prior population assignments consolidated and visualized using CLUMPAK for genetic clusters (K) 4–7 for three Coachella Valley populations (16–18) and Southern Clade genetic units; Western Mojave GU (groups 1–2), Central Mojave GU (groups 3–5), Southern Mojave GU (groups 6–10) and Colorado Desert GU (groups 13–14). STRUCTURE analyses histograms consolidated and visualized using CLUMPAK for genetic clusters K = 4−7. Figure 3. STRUCTURE analyses for Agassiz’s desert tortoise without prior population assignments consolidated and visualized using CLUMPAK for genetic clusters (K) 4–7 for three Coachella Valley populations (16–18) and Southern Clade genetic units; Western Mojave GU (groups 1–2), Central Mojave GU (groups 3–5), Southern Mojave GU (groups 6–10) and Colorado Desert GU (groups 13–14). STRUCTURE analyses histograms consolidated and visualized using CLUMPAK for genetic clusters K = 4−7. Discussion change in the region. For example, prehistoric aquatic barriers in the lower Colorado River basin drove divergence between Mojave and Sonoran desert biota (Bell et al. 2010, Dolby et al. 2015). The first barrier, a marine incursion of the Gulf of California, extended northwestward into what is now the Salton Trough as early as 6.3 Ma (Dolby et al. 2019). Fossil‑rich marine sediments show that the Gulf extended as far north as San Gorgonio Pass near Palm Springs, California, and possibly east to Parker, Arizona along what is now the Colorado River (Dolby et al. 2015, 2019). The Colorado River extensional corridor of the Miocene left a series of basins that were flooded in the Pliocene by the precursor to the lower river. Water from the Our analyses did not detect deep, lineage divergence among tortoise populations separated by the Coachella Valley, as would be expected if it was a significant barrier to gene flow now or in the past. Instead, the three tortoise populations surrounding the valley show evidence of continuity with the Colorado Desert GU and gene flow consistent with geographic distance, including some evidence of gene flow between Mesa at the northwestern tip of the valley and the adjacent Southern Mojave GU.fii The observed lack of differentiation is somewhat unexpected given the long history of geological © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 8 Testing a barrier to gene flow in desert tortoises Lovich et al. Figure 4. a) The mean log-likelihood of the number of populations of Agassiz’s desert tortoise for each K (ranging from 1–10), with error bars representing the standard deviation around the mean. b) DeltaK, derived through the Evanno et al. (2005) method, is shown for each K 1–10. and Walde 2006), especially G. agassizii (Woodbury and Hardy 1948) since their feet lack webbing present in many aquatic turtles (Pace et al. 2001, Ernst and Lovich 2009).i Aquatic barriers in the Salton Trough and lower Colorado River Valley affected gene flow and evolution not only of tortoises (Lamb et al. 1989), but their impact has also been demonstrated to varying degrees in phylogeographic studies of some other reptile species (e.g., Devitt 2006, Mulcahy et al. 2006), but not others (Lamb et al. 1992, Gottscho et al. Discussion 2017).i t Past and present conditions in the Salton Trough, including the Coachella Valley, present tortoises with inhospitable habitat that inhibits them from moving from one side to the other in a straight line, but this dispersal barrier does not isolate them. Tortoises may have circumvented the Coachella Valley using its mountainous northern perimeter as a corridor for gene flow. The most likely route of connectivity involves the northern edge of the valley through the foothills, canyons, and bajadas of the Little San Bernardino Mountains. Our analyses cluster together Mesa, Orocopia, and Cottonwood from across the Coachella Valley. Nowadays, Interstate 10 separates Orocopia and Cottonwood, which are only about 6 km apart, and their genetic distance to one another is substantially less than the distance to Mesa. Tortoises at Mesa exhibit evidence of historic gene flow with populations to the north-northwest, yet it still clusters with the Colorado Desert GU. Figure 4. a) The mean log-likelihood of the number of populations of Agassiz’s desert tortoise for each K (ranging from 1–10), with error bars representing the standard deviation around the mean. b) DeltaK, derived through the Evanno et al. (2005) method, is shown for each K 1–10. Our analyses suggest that the range of the Colorado Desert GU should be extended to include populations surrounding the Coachella Valley. Populations around the Coachella Valley appear to show stronger genetic affinities to the Colorado Desert GU to the east (e.g., populations in Chuckwalla and Chocolate Mountain Aerial Gunnery Range, Table 1) than they do to the Southern/Central Mojave Desert GUs to the north. The Little San Bernardino and Cottonwood mountains at the southern boundary of Joshua Tree National Park form a natural barrier for tortoises between the Coachella Valley and the Southern Mojave GU (e.g., Lucerne Valley, Ord-Rodman, MCAGCC, Table 1). The same barrier forms the boundary between the climatically different Mojave and Sonoran desert ecosystems (see Lovich and Bainbridge 1999, Barrows 2011, and references therein), so our results correspond with the zoogeography of the region. Figure 4. a) The mean log-likelihood of the number of populations of Agassiz’s desert tortoise for each K (ranging from 1–10), with error bars representing the standard deviation around the mean. b) DeltaK, derived through the Evanno et al. (2005) method, is shown for each K 1–10. Colorado River arrived in the lower basin after 5.24 Ma, forming a second barrier (Howard et al. Frontiers of Biogeography 2020, 12.3, e46888 Discussion 2019) or filter (Dolby et al. 2019) to dispersal and gene flow in what is now the Salton Trough. Mesa (Table 1, group 16) appears to have some introgression with the Southern Mojave GU, with decreasing influence moving eastward across the Colorado Desert GU (groups 17, 18 & 14: Fig. 2, Table 5). Thus, historic gene flow at Mesa likely occurred through the Morongo Valley corridor between the San Bernardino and Little San Bernardino mountains, and this is consistent with the model of isolation‑by‑distance (Murphy et al. 2007, Hagerty et al. 2011) exhibited by G. agassizii. No population has been isolated completely over time; gene flow occurs/occurred among neighboring populations. Over the last several thousand years, the Salton Trough has also been inundated periodically with freshwater from overflow of the Colorado River, forming a large waterbody known as Lake Cahuilla (Waters 1983), and it is now partially filled by the hypersaline Salton Sea. Lake Cahuilla (Fig. 1) and its earlier marine precursor would have been dispersal barriers to tortoise movements in the Salton Trough, for while they are capable of floating (Patterson 1973), sometimes for great distances (Gerlach et al. 2006), tortoises are poor swimmers (see review in Strong The shared mtDNA ancestry among the Coachella Valley tortoises suggests that the current distribution © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 9 Testing a barrier to gene flow in desert tortoises Lovich et al. may be attributed either to a recent origin or to unabated dispersal. Edwards (2015) suggested that demographic events or selection reduced diversity within California, leaving tortoises with mtDNA Haplogroup MOJ_A only. This could have involved a population bottleneck followed by a population expansion (Edwards 2003), possibly coinciding with climate change associated with glacial-interglacial periods during the Pleistocene. The estimated time to the most recent common ancestor of this haplogroup—98,268 years (SD ±48,000 years; Edwards 2003)—is consistent with this hypothesis. Desert tortoises likely experienced multiple population contractions and expansions correlating with Pleistocene glacial and interglacial events. The lower Colorado River Valley appears to have maintained more desert-like conditions during the Wisconsin glacial period (Betancourt et al. 1990), and this area may have acted as a refugium for desert‑evolved species. It is unclear how G. agassizii may have benefitted from such a refugium since they are not considered to be a desert-evolved species (Morafka and Berry 2002). Acknowledgements Research was supported by the Coachella Valley Conservation Commission. The latter included funding (in part) from the California Department of Fish and Wildlife via a Natural Community Conservation Planning Local Assistance Grant. Numerous people provided assistance with the project, especially Kristina Drake and Rachel Henderson. We thank Liz Cox and the University of Arizona Genetics Core for generating the genetic data. Erika Sanchez-Chopitea provided GIS mapping for geographic distance calculations and Sarah Sweat prepared our map figures. Research was conducted with permits from Joshua Tree National Park (JOTR-2012-SCI-0019), the U.S. Fish and Wildlife Service (TE-198910-1), the California Department of Fish and Game (SC-1639), and the Bureau of Land Management. We thank Al Muth and Chris Tracy for providing accommodations and stimulating discussions at the Philip L. Boyd Deep Canyon Desert Research Center (doi:10.21973/N3V66D) of the University of California, Riverside during our field work. We are grateful to the Institutional Animal Care and Use Committee of Northern Arizona University for reviewing and approving our research procedures. We thank Brian Kreiser of the University of Southern Mississippi for extracting DNA from scute tissues. Manuscript development benefitted greatly from discussions with Kyle House, Kristin McDougall-Reid and Brian Kreiser. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government. k The Salton Trough does not appear to be a major driver of population structure in the current distribution of desert tortoises. Instead the Little San Bernardino and Cottonwood mountains appear to limit gene flow to neighboring populations to the north, separating the Southern Mojave GU and the Colorado Desert GU. Although all G. agassizii exhibit a fairly recent shared ancestry (within maternal clade MOJ_A), enough time has passed for local adaptation to occur across the species’ large distribution (Sánchez-Ramírez et al. 2018). Tortoises around the Coachella Valley likely retain adaptations unique to the lower Colorado River Valley that allow them to persist at the edge of the species’ range. Frontiers of Biogeography 2020, 12.3, e46888 References (1939) A system of marking turtles for future identification. Copeia, 1939, 170–173. Devitt, T.J. (2006) Phylogeography of the Western Lyresnake (Trimorphodon biscutatus): testing aridland biogeographical hypotheses across the Nearctic–Neotropical transition. Molecular Ecology, 15, 4387–4407. Barrows, C.W. (2011) Sensitivity to climate change for two reptiles at the Mojave–Sonoran Desert interface. Journal of Arid Environments, 75, 629–635. Dimmitt, M.A. (1977) Status of the desert tortoise in the Colorado Desert. In: Proceedings of the 1977 Desert Tortoise Council Symposium (ed. by M. Trotter), pp. 46–53. Desert Tortoise Council, San Diego, USA. Barrows, C.W., Rotenberry, J.T. & Allen, M.F. (2010) Assessing sensitivity to climate change and drought variability of a sand dune endemic lizard. Biological Conservation, 143, 731–736. Barrows, C.W., Henen, B.T. & Karl, A.E. (2016) Identifying climate refugia: a framework to inform conservation strategies for Agassiz’s desert tortoise in a warmer future. Chelonian Conservation and Biology, 15, 2–11. Dolby, G.A., Bennett, S.E.K., Lira-Noriega, A., Wilder, B.T. & Munguia-Vega, A. (2015) Assessing the geological and climatic forcing of biodiversity and evolution surrounding the Gulf of California. Journal of the Southwest, 57, 391–455. Beatley, T. (1992) Balancing urban development and endangered species: the Coachella Valley habitat conservation plan. Environmental Management, 16, 7–19. Dolby, G.A., Dorsey, R.J. & Graham, M.R. (2019) A legacy of geo-climatic complexity and genetic divergence along the lower Colorado River: Insights from the geological record and 33 desert-adapted animals. Journal of Biogeography, 46, 2479–2505. Bell, K.C., Hafner, D.J., Leitner, P. & Matocq, M.D. (2010) Phylogeography of the ground squirrel subgenus Xerospermophilus and assembly of the Mojave Desert biota. Journal of Biogeography, 37, 363–378. Drake, K.K., Nussear, K.E., Esque, T.C., Barber, A.M., Vittum, K.M., Medica, P.A., Tracy, C.R. & Hunter, K.W. (2012) Does translocation influence physiological stress in the desert tortoise? Animal Conservation, 15, 560–570. Berry, K.C., & Murphy, R.W. (2019) Gopherus agassizii (Cooper 1861) – Mojave Desert Tortoise, Agassiz’s Desert Tortoise. In: Conservation Biology of Freshwater Turtles and Tortoises: A Compilation Project of the IUCN/SSC Tortoise and Freshwater Turtle Specialist Group (ed. by A.G.J. Rhodin, J.B. Iverson, P.P. van Dijk, C.B. Stanford, E.V. Goode, K.A. Buhlmann, P.C.H. Pritchard, and R.A. Mittermeier), pp. 109.101–109.145, Chelonian Research Monographs. Dutcher, K.E., Vandergast, A.G., Esque, T.C., Mitelberg, A., Matocq, M.D., Heaton, J.S., & Nussear, K.E. (2020). Genes in space: what Mojave desert tortoise genetics can tell us about landscape connectivity. Conservation Genetics, 21, 289–303. Earl, D.A. & von Holdt, B.M. References Agha, M., Lovich, J.E., Ennen, J.R., et al. (2015) Turbines and terrestrial vertebrates: variation in tortoise survivorship between a wind energy facility and an adjacent undisturbed wildland area in the Desert Southwest (USA). Environmental Management, 56, 332–341. p g Low-lying desert areas, currently occupied by desert tortoises, are expected to become unsuitable in a warming and drying climate scenario (Barrows 2011, Barrows et al. 2016), and there is some preliminary evidence that the transition is already underway (Lovich et al. 2014). In the context of species conservation, we cannot predict which adaptive traits will be most critical in the face of environmental change and which individuals will contribute most to the evolutionary potential of the species. Individuals living on the edges of their distribution or in marginal habitat have the potential for being better adapted to changing environmental conditions (like climate change) that are different or more challenging from conditions in the core of their distribution (Eckert et al. 2008, Palstra and Ruzzante 2008, Hardie and Hutchings 2010). Desert tortoise populations around the Coachella Valley, especially west of the Salton Trough, represent an underappreciated extension of the species’ range. Their genetic affiliation with the Colorado Desert GU and the lower Colorado River Valley, which potentially acted as a refugium during previous climate change events, suggest that it is essential to maintain the connectivity of these populations for the long-term sustainability of the species in that GU. Allison, L.J. & Mcluckie, A.M. (2018) Population trends in Mojave desert tortoises (Gopherus agassizii). Herpetological Conservation and Biology, 13, 433–452. Atwater, T. (1970) Implications of plate tectonics for the Cenozoic evolution of western North America. Geological Society of America Bulletin, 81, 3515–3536. Averill-Murray, R.C. & Hagerty, B.E. (2014) Translocation relative to spatial genetic structure of the Mojave Desert Tortoise, Gopherus agassizii. Chelonian Conservation and Biology, 13, 35–41. Averill-Murray, R.C., Darst, C.R., Field, K.J. & Allison, L.J. (2012) A new approach to conservation of the Mojave Desert tortoise. BioScience, 62, 893–899. © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 10 Testing a barrier to gene flow in desert tortoises Lovich et al. Banks, M.A. & Eichert, W. (2000) WHICHRUN (version 3.2): a computer program for population assignment of individuals based on multilocus genotype data. Journal of Heredity, 91, 87–89. Cagle, F.R. (1939) A system of marking turtles for future identification. Copeia, 1939, 170–173. Cagle, F.R. References (2012) STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conservation Genetics Resources, 4, 359–361. Betancourt, J.L., van Devender, T.R. & Martin, P.S. (1990) Packrat Middens: The Last 40,000 Years of Biotic Change. University of Arizona Press, Tucson, AZ. Eckert, C.G., Samis, K.E. & Lougheed, S.C. (2008) Genetic variation across species’ geographical ranges: the central–marginal hypothesis and beyond. Molecular Ecology, 17, 1170–1188. Britten, H.B., Riddle, B.R., Brussard, P.F., Marlow, R. & Lee, Jr., T.E. (1997) Genetic delineation of management units for the desert tortoise, Gopherus agassizii, in northeastern Mojave Desert. Copeia, 1997, 523–530. Edwards, T. (2003) Desert tortoise conservation genetics. University of Arizona, Tucson, USA. 87 pp. UMI Number: 1414228 Brothers, D.S., Driscoll, N.W., Kent, G.M., Harding, A.J., Babcock, J.M. & Baskin, R.L. (2009) Tectonic evolution of the Salton Sea inferred from seismic reflection data. Nature Geoscience, 2, 581–584. Edwards, T. (2015) A biogeographic perspective on speciation among desert tortoises in the genus Gopherus. University of Arizona, Tucson, USA. 168 pp. UMI Number 3702126 Frontiers of Biogeography 2020, 12.3, e46888 © the authors, CC-BY 4.0 license 11 Testing a barrier to gene flow in desert tortoises Lovich et al. Zones correspond with spatially concordant phylogeographic boundaries in western North America. Biological Reviews, 91, 235–254. Edwards, T. & Berry, K. (2013) Are captive tortoises a reservoir for conservation? An assessment of genealogical affiliation of captive Gopherus agassizii to local, wild populations. Conservation Genetics, 14, 649–659. Gottscho, A.D., Wood, D.A., Vandergast, A.G., Lemos- Espinal, J., Gatesy, J. & Reeder, T.W. (2017) Lineage diversification of fringe-toed lizards (Phrynosomatidae: Uma notata complex) in the Colorado Desert: Delimiting species in the presence of gene flow. Molecular Phylogenetics and Evolution, 106, 103–117. Edwards, T. & Harrison, S. (2014) Population and conservation genetics. In: Biology and conservation of North American tortoises (ed. by D.C. Rostal, E.D. McCoy, and H.R. Mushinsky), pp. 127–133. Johns Hopkins University Press, Baltimore, USA. Goudet, J. (1995) FSTAT (Version 1.2): A computer program to calculate F-statistics. Journal of Heredity, 86, 485–486. Edwards, T., Jarchow, C.J., Jones, C.A. & Bonine, K.E. (2010) Tracing genetic lineages of captive desert tortoises in Arizona. The Journal of Wildlife Management, 74, 801–807. Hagerty, B.E. & Tracy, C.R. (2010) Defining population structure for the Mojave desert tortoise. Conservation Genetics, 11, 1795–1807. Ernst, C.H. & Lovich, J.E. (2009) Turtles of the United States and Canada, Second Edition. Johns Hopkins University Press, Baltimore, USA. References 827 pp. Hagerty, B., Nussear, K., Esque, T. & Tracy, C. (2011) Making molehills out of mountains: landscape genetics of the Mojave desert tortoise. Landscape Ecology, 26, 267–280. Evanno, G., Regnaut, S. & Goudet, J. (2005) Detecting the number of clusters of individuals using the software structure: a simulation study. Molecular Ecology, 14, 2611–2620. Hardie, D.C. & Hutchings, J.A. (2010) Evolutionary ecology at the extremes of species’ ranges. Environmental Reviews, 18, 1–20. Excoffier, L. & Lischer, H.E.L. (2010) Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Molecular Ecology Resources, 10, 564–567. Hernandez-Divers, S., Hernandez-Divers, S. & Wyneken, J. (2002) Angiographic, anatomic and clinical technique descriptions of a subcarapacial venipuncture site for chelonians. Journal of Herpetological Medicine and Surgery, 12, 32–37. Gaillard, D., Ennen, J.R., Kreiser, B.R., et al. (2017) Range-wide and regional patterns of population structure and genetic diversity in the gopher tortoise. Journal of Fish and Wildlife Management, 8, 497–512. Howard, K.A., House, P.K., John, B.E., Crow, R.S. & Pearthree, P.A. (2019) A river is born: highlights of the geologic evolution of the Colorado River extensional corridor and its river: a field guide honoring the life and legacy of Warren Hamilton. In: Geologic Excursions in Southwestern North America (ed. by P.A. Pearthree), pp. 61–113. The Geological Society of America, Boulder, USA. Gerlach, J., Muir, C. & Richmond, M.D. (2006) The first substantiated case of trans-oceanic tortoise dispersal. Journal of Natural History, 40, 41–43. Germano, D.J., Bury, R.B., Esque, T.C., Fritts, T.H. & Medica, P.H. (1994) Range and habitats of the desert tortoise. In: Biology of North American Tortoises (ed. by R.B. Bury and D.J. Germano), pp. 73–84. U.S. Department of the Interior, National Biological Survey, Washington, D. C. Kopelman, N.M., Mayzel, J., Jakobsson, M., Rosenberg, N.A. & Mayrose, I. (2015) Clumpak: a program for identifying clustering modes and packaging population structure inferences across K. Molecular Ecology Resources, 15, 1179–1191. Germano, J.M. & Bishop, P.J. (2008) Suitability of amphibians and reptiles for translocation. Conservation Biology, 23, 7–15. Lamb, T., Avise, J.C. & Gibbons, J.W. (1989) Phylogeographic patterns in mitochondrial DNA of the desert tortoise (Xerobates agassizi), and evoluntionary relationships among the North American gopher tortoise. Evolution, 43, 76–87. Germano, J.M., Nafus, M.G., Perry, J.A., Hall, D.B. & Swaisgood, R.R. (2017) Predicting translocation outcomes with personality for desert tortoises. Behavioral Ecology, 28, 1075–1084. Gottscho, A.D. References (2016) Zoogeography of the San Andreas Fault system: Great Pacific Fracture Lamb, T., Jones, T.R. & Avise, J.C. (1992) Phylogeographic histories of representative herpetofauna of © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 12 Testing a barrier to gene flow in desert tortoises Lovich et al. the southwestern U.S.: mitochondrial DNA variation in the desert iguana (Dipsosaurus dorsalis) and the chuckwalla (Sauromalus obesus). Journal of Evolutionary Biology, 5, 465–480. Lovich, J.E., Puffer, S.R., Agha, M., et al. (2018a) Reproductive output and clutch phenology of female Agassiz’s desert tortoises (Gopherus agassizii) in the Sonoran Desert region of Joshua Tree National Park. Current Herpetology, 37, 40–57. Latch, E.K., Boarman, W.I., Walde, A. & Fleischer, R.C. (2011) Fine-scale analysis reveals cryptic landscape genetic structure in desert tortoises. PLoS ONE, 6, e27794. Lovich, J.E., Agha, M., Ennen, J.R., Arundel, T.R. & Austin, M. (2018b) Agassiz’s desert tortoise (Gopherus agassizii) activity areas are little changed after wind turbine-induced fires in California. International Journal of Wildland Fire, 27, 851–856. Lovich, J.E. & Bainbridge, D. (1999) Anthropogenic degradation of the southern California desert ecosystem and prospects for natural recovery and restoration. Environmental Management, 24, 309–326. Luckenbach, R.A. (1982) Ecology and management of the desert tortoise (Gopherus agassizii) in California. In: North American Tortoises: Conservation and Ecology (ed. by R.B. Bury), pp. 1–39. U.S. Fish and Wildlife Service, Washington, D.C. Lovich, J.E. & Daniels, R. (2000) Environmental characteristics of desert tortoise (Gopherus agassizii) burrow locations in an altered industrial landscape. Chelonian Conservation and Biology, 3, 714–721. Morafka, D.J. & Berry, K.H. (2002) Is Gopherus agassizii a desert-adapted tortoise, or an exaptive opportunist? Implications for tortoise conservation. Chelonian Conservation and Biology, 4, 263–287. Lovich, J.E. & Ennen, J.R. (2017) Reptiles and amphibian. In: Wildlife and Windfarms, Conflicts and Solutions, Volume 1, Onshore: Potential Effects (ed. by M. Perrow), pp. 97–118. Pelagic Publishing, Exeter. Mulcahy, D.G., Spaulding, A.W., Mendelson, J.R. & Brodie, E.D. (2006) Phylogeography of the flat-tailed horned lizard (Phrynosoma mcallii) and systematics of the P. mcallii–platyrhinos mtDNA complex. Molecular Ecology, 15, 1807–1826. Lovich, J.E. & Ennen, J.R. (2011) Wildlife conservation and solar energy development in the Desert Southwest, United States. BioScience, 61, 982–992. Lovich, J.E. & Ennen, J.R. (2013) Assessing the state of knowledge of utility-scale wind energy development and operation on non-volant terrestrial and marine wildlife. Applied Energy, 103, 52–60. Murphy, R.W., Berry, K.H., Edwards, T. & McLuckie, A.M. References (2007) A genetic assessment of the recovery units for the Mojave population of the desert tortoise, Gopherus agassizii. Chelonian Conservation and Biology, 6, 229–251. Lovich, J.E., Ennen, J.R., Madrak, S., et al. (2011) Effects of wind energy production on growth, demography, and survivorship of a Desert Tortoise (Gopherus agassizii) population in Southern California with comparisons to natural populations. Herpetological Conservation and Biology, 6, 161–174. Nussear, K.E., Esque, T.C., Inman, R.D., Gass, L., Thomas, K.A., Wallace, C.S. A., Blainey, J.B., Miller, D.M., Webb, R.H. (2009) Modeling habitat of the desert tortoise (Gopherus agassizii) in the Mojave and parts of the Sonoran Deserts of California, Nevada, Utah, and Arizona.U.S. Geological Survey, Open File Report, 2009- 1102, 1-18. Lovich, J.E., Yackulic, C.B., Freilich, J., et al. (2014) Climatic variation and tortoise survival: has a desert species met its match? Biological Conservation, 169, 214–224. Pace, C.M., Blob, R.W. & Westneat, M.W. (2001) Comparative kinematics of the forelimb during swimming in red-eared slider (Trachemys scripta) and spiny softshell (Apalone spinifera) turtles. Journal of Experimental Biology, 204, 3261–3271. Lovich, J.E., Ennen, J.R., Yackulic, C.B., Meyer-Wilkins, K., Agha, M., Loughran, C., Bjurlin, C., Austin, M. & Madrak, S.V. (2015) Not putting all their eggs in one basket: bet-hedging despite extraordinary annual reproductive output of desert tortoises. Biological Journal of the Linnean Society, 115, 399–410. Palstra, F.P. & Ruzzante, D.E. (2008) Genetic estimates of contemporary effective population size: what can they tell us about the importance © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 13 Testing a barrier to gene flow in desert tortoises Lovich et al. of genetic stochasticity for wild population persistence? Molecular Ecology, 17, 3428–3447. of genetic stochasticity for wild population persistence? Molecular Ecology, 17, 3428–3447. Strong, J.N. & Walde, A.D. (2006) Geochelone carbonaria (Red-footed tortoise). Swimming. Herpetological Review, 37, 457–458. Patterson, R. (1973) Why tortoises float. Journal of Herpetology, 7, 373–375. USFWS (US Fish and Wildlife Service). (2011) Revised recovery plan for the Mojave population of the desert tortoise (Gopherus agassizii). U.S. Fish and Wildlife Service, Pacific Southwest Region, Sacramento, CA. 222 pp. Pritchard, J.K., Stephens, M. & Donnelly, P. (2000) Inference of population structure using multilocus genotype data. Genetics, 155, 945–959. Waters, M.R. (1983) Late Holocene lacustrine chronology and archaeology of ancient Lake Cahuilla, California. Quaternary Research, 19, 373–387. Prose, D.V. (1985) Persisting effects of armored military maneuvers on some soils of the Mojave Desert. © the authors, CC-BY 4.0 license Frontiers of Biogeography 2020, 12.3, e46888 References Environmental Geology and Water Sciences, 7, 163–170. Wayne, R.K. & Morin, P.A. (2004) Conservation genetics in the new molecular age. Frontiers in Ecology and the Environment, 2, 89–97. Raymond, M. (1995) GENEPOP (version 1.2): population genetics software for exact tests and ecumenicism. Journal of Heredity, 86, 248–249. Weir, B.S. & Cockerham, C.C. (1984) Estimating F-statistics for the analysis of population structure. Evolution, 38, 1358–1370. Rousset, F. (2008) genepop’007: a complete re- implementation of the genepop software for Windows and Linux. Molecular Ecology Resources, 8, 103–106. Wood, Y.A., Graham, R.C. & Wells, S.G. (2005) Surface control of desert pavement pedologic process and landscape function, Cima Volcanic field, Mojave Desert, California. CATENA, 59, 205–230. Sánchez-Ramírez, S., Rico, Y., Berry, K.H., Edwards, T., Karl, A.E., Henen, B.T. & Murphy, R.W. (2018) Landscape limits gene flow and drives population structure in Agassiz’s desert tortoise (Gopherus agassizii). Scientific Reports, 8, 11231. Woodbury, A.M. & Hardy, R. (1948) Studies of the desert tortoise, Gopherus agassizii. Ecological Monographs, 18, 145–200. Schneider, J.S. & Everson, G.D. (1989) The desert tortoise (Xerobates agassizii) in the prehistory of the southwestern Great Basin and Adjacent areas. Journal of California and Great Basin Anthropology, 11, 175–202. Submitted: 11 February 2020 First Decision: 26 March 2020 Accepted: 11 May 2020 First Decision: 26 March 2020 Accepted: 11 May 2020 Edited by Robert J. Whittaker and Janet Franklin Frontiers of Biogeography 2020, 12.3, e46888
https://openalex.org/W2899568967	https://pressto.amu.edu.pl/index.php/rpeis/article/download/16155/15967	Polish	null	KATARZYNA PEREZ, EFEKTYWNOŚĆ FUNDUSZY INWESTYCYJNYCH. PODEJŚCIE TECHNICZNE I FUNDAMENTALNE, DIFIN, WARSZAWA 2012, SS. 440.	Ruch Prawniczy, Ekonomiczny i Socjologiczny	2,018	cc-by	1,781	299 299 Przegla˛d pis´miennictwa celowe przeznaczenie, mielibys´my do czynienia z odste˛pstwem od zasady jednos´ci materialnej budz˙etu w obre˛bie działu ,,os´wiata i wychowanie’’. Tymczasem ta cze˛s´c´ subwencji (podobnie jak pozostałe cze˛s´ci – razem z dochodami własnymi i udziałami w podatkach pan´ stwowych) stanowi jedna˛ pule˛ s´rodko´w publicznych przeznaczanych na finansowane wszystkich zadan´ własnych biez˙a˛cych i inwestycyjnych. S´wiadczy o tym zapis ustawodawcy, zgodnie z kto´rym o przeznaczeniu s´rodko´w otrzymanych z tytułu tej cze˛s´ci subwencji ogo´lnej decyduje organ stanowia˛cy jednostki samorza˛du terytorialnego2. celowe przeznaczenie, mielibys´my do czynienia z odste˛pstwem od zasady jednos´ci materialnej budz˙etu w obre˛bie działu ,,os´wiata i wychowanie’’. Tymczasem ta cze˛s´c´ subwencji (podobnie jak pozostałe cze˛s´ci – razem z dochodami własnymi i udziałami w podatkach pan´ stwowych) stanowi jedna˛ pule˛ s´rodko´w publicznych przeznaczanych na finansowane wszystkich zadan´ własnych biez˙a˛cych i inwestycyjnych. S´wiadczy o tym zapis ustawodawcy, zgodnie z kto´rym o przeznaczeniu s´rodko´w otrzymanych z tytułu tej cze˛s´ci subwencji ogo´lnej decyduje organ stanowia˛cy jednostki samorza˛du terytorialnego2. Recenzowana publikacja ma – moim zdaniem – pewne mankamenty. Najpowaz˙niejszy z nich to jej mała przejrzystos´c´. W teks´cie jest zbyt mało tabel, w kto´rych przedstawiono by inne tres´ci niz˙ dane liczbowe. Autorka nie wykorzystała rysunko´w, na kto´rych moz˙na by przedstawic´ na przykład kryteria klasyfikacji i wyro´z˙nione na ich podstawie grupy poszczego´lnych wielkos´ci budz˙etowych, z´ro´dła s´rodko´w zagranicznych niepodlegaja˛cych zwrotowi itp. Tekst ksia˛z˙ki jest dos´c´ s´cisły, brakuje w nim ,,s´wiatła’’, kto´re pozwoliłoby na szybkie znalezienie interesuja˛cych czytelnika tres´ci. Braki te nie sa˛na tyle istotne, aby zawaz˙yły na ogo´lnej ocenie recenzowanej pozycji, tym bardziej z˙e moga˛ byc´ wyeliminowane podczas prac nad drugim wydaniem. Recenzowana publikacja powinna spotkac´ sie˛ z zainteresowaniem pracowniko´w naukowo-dydak- tycznych wydziało´w ekonomicznych, prawa i administracji oraz ich studento´w. Wyraz˙am przeko- nanie, z˙e be˛dzie ona ciekawa˛ lektura˛ takz˙e dla pracowniko´w samorza˛dowych, radnych oraz przedstawicieli organo´w wykonawczych jednostek samorza˛dowych i innych oso´b zainteresowanych problematyka˛ samorza˛dowa˛. Przemawiaja˛ za tym walory tej ksia˛z˙ki, na kto´re zwro´ciłam juz˙ uwage˛. Sławomira Kan´ duła Sławomira Kan´ duła Katarzyna Perez, Efektywnos´c´ funduszy in- westycyjnych. Podejs´cie techniczne i funda- mentalne, Difin, Warszawa 2012, ss. 440. Recenzowana monografia jest kolejna˛praca˛w dorobku dr Katarzyny Perez pos´wie˛cona˛funduszom inwestycyjnym. Jest to opracowanie, w kto´rym w sposo´b kompleksowy Autorka opisała techniczne i fundamentalne podejs´cie do oceny efektywnos´ci funduszy inwestycyjnych. Ksia˛z˙ka dotyczy zagadnien´ bardzo waz˙nych i aktualnych. 2 Ustawa z 13 listopada 2003 r. o dochodach jednostek samorza˛du terytorialnego, t.jedn.: Dz. U. 2010, Nr 80, poz. 526, art. 7 ust. 3. 299 Podejs´cie techniczne w ocenie funduszy na podstawie osia˛ganych przez nie sto´p zwrotu waz˙onych ryzykiem nie wydaje sie˛ wystarczaja˛ce, aby podja˛c´ decyzje˛ o wyborze funduszu, w kto´ry warto inwestowac´. Konieczne według Autorki jest uzupełnienie analizy o czynniki fundamentalne, kto´rym pos´wie˛cony jest rozdział 4. Fundusze inwestycyjne działaja˛bowiem w s´rodowisku, w kto´rym mamy do czynienia z nieefektywnos´cia˛ informacyjna˛ rynku kapitałowego, a uczestnicy rynku kapitałowego nie zawsze zachowuja˛ sie˛ racjonalnie, jak zakłada sie˛ w teorii klasycznej i neoklasycznej, be˛da˛cych podstawa˛ konstrukcji wie˛kszos´ci mierniko´w efektywnos´ci funduszy inwestycyjnych. Pełna interpretacja osia˛ganych przez fundusze wyniko´w jest moz˙liwa tylko po uwzgle˛dnieniu takz˙e elemento´w fundamentalnych. Na wyniki funduszy inwestycyjnych maja˛ bowiem wpływ cechy charakteryzuja˛ce dany fundusz – okres jego funkcjonowania, stosowane style i strategie inwesto- wania (bezpieczne, umiarkowane czy agresywne), a przede wszystkim – cechy zarza˛dzaja˛cych funduszami (wykształcenie, dos´wiadczenie, pochodzenie, sposo´b wynagradzania itp.). Bardzo waz˙na z punktu widzenia analizy fundamentalnej jest tez˙ ocena, czy moz˙liwe jest utrzymanie (persystencja) sto´p zwrotu, kto´re fundusze wypracowały w poprzednich okresach. Historyczne stopy zwrotu sa˛ bowiem jednym z waz˙niejszych kryterio´w oceny funduszu – nie wiadomo jednak, czy zarza˛dzaja˛cy sa˛ w stanie utrzymac´ osia˛gnie˛ty ich poziom w przyszłos´ci. Autorka analizuje te wszystkie czynniki fundamentalne na tle faz cyklu z˙ycia funduszu inwestycyjnego. Ostatnia˛ cze˛s´cia˛ pracy jest rozdział 5, pos´wie˛cony kompleksowej ocenie funduszy inwestycyjnych działaja˛cych na rynku polskim. Przedmiotem badan´ empirycznych były wyniki funduszy akcji i funduszy hybrydowych zarejestrowanych w Polsce, kto´re działały w latach 2001-2010. Autorka skonfrontowała z praktyka˛ rozwaz˙ania zawarte we wczes´niejszych rozdziałach na temat teoretycz- nych podstaw analizy efektywnos´ci funduszy inwestycyjnych, wykorzystuja˛c podejs´cie techniczne i fundamentalne. Pozwoliło to zamkna˛c´ w logiczna˛, spo´jna˛całos´c´ zawartos´c´ merytoryczna˛ monografii. Doktor Perez zbadała efektywnos´c´ polskich funduszy inwestycyjnych według technicznych miar – klasycznych i nowoczesnych – oraz oceniła persystencje˛ osia˛gnie˛tych przez te fundusze historycznych sto´p zwrotu. Dopełnieniem tej analizy była ocena wpływu cech funduszu na jego wyniki, na zacho- wanie zarza˛dzaja˛cych funduszami oraz na zachowanie jego uczestniko´w. Bardzo interesuja˛cym wnioskiem, jaki wycia˛gne˛ła Autorka z badan´ przeprowadzonych na rynku funduszy inwestycyjnych w Polsce, jest stwierdzenie, z˙e ,,gło´wnym z´ro´dłem dodatnich sto´p zwrotu funduszy osia˛ganych przez ich menedz˙ero´w było w badanym okresie szcze˛s´cie, a nie faktyczne umieje˛tnos´ci doboru instrumento´w finansowych do funduszu!’’ (s. 355). Polski rynek funduszy inwestycyjnych znajduje sie˛ wcia˛z˙ w fazie wzrostu, a menedz˙erowie i uczestnicy funduszy dopiero zdobywaja˛ coraz wie˛ksze dos´wiadczenie. Udało sie˛ stworzyc´ w Polsce stabilne ramy prawne i instytucjonalne umoz˙liwiaja˛ce sprawne funkcjonowanie rynku funduszy inwestycyjnych zgodnie ze standardami s´wiatowymi, co niewa˛t- pliwie be˛dzie sprzyjac´ dalszemu rozwojowi tego rynku. 299 Chociaz˙ pierwsze fundusze inwestycyjne pojawiły sie˛ na s´wiecie juz˙ około dwies´cie lat temu, to sposo´b i metody analizy ich funkcjonowania stale ewoluuja˛, w s´lad za pojawiaja˛cymi sie˛ nowymi koncepcjami teoretycznymi, jak i na skutek ro´z˙nicowania sie˛ samych funduszy i ich strategii inwestycyjnych. W Polsce rynek funduszy inwestycyjnych powstał dwadzies´cia lat temu i cały czas sie˛ rozwija. Praca dr Katarzyny Perez ma konsekwentnie zarysowana˛, w pełni logiczna˛strukture˛. W pierwszej cze˛s´ci (rozdział 1 i 2) Autorka przedstawia ewolucje˛ teorii portfela inwestycyjnego oraz ewolucje˛ teorii rynku kapitałowego i ich znaczenie dla oceny efektywnos´ci funduszy inwestycyjnych. Ta cze˛s´c´ pracy tworzy podstawy do analizy funkcjonowania funduszy zaro´wno od strony technicznej, jak i funda- mentalnej. Przegla˛d teorii znajduja˛cy sie˛ w pracy został przygotowany z ogromnym znawstwem s´wiatowych dokonan´ w tym zakresie. Autorka wykazała sie˛ dogłe˛bna˛ znajomos´cia˛ rozwaz˙an´ teoretycznych prezentowanych w literaturze przedmiotu (bibliografia zawarta w pracy liczy blisko osiemset pozycji!). Dwa autorskie wykresy podsumowuja˛ce rozwo´j opisywanych teorii, zamieszczone na kon´ cu pierwszego i drugiego rozdziału (odpowiednio s. 88 i 134) daja˛ czytelnikowi przejrzysty obraz poszczego´lnych etapo´w tej ewolucji. Ta cze˛s´c´ pracy stanowi niezwykle cenne kompendium wiedzy na temat teorii portfela inwestycyjnego, jego dywersyfikacji, racjonalnych zachowan´ inwestoro´w i zalez˙nos´ci pomie˛dzy oczekiwana˛stopa˛zwrotu a ryzykiem ponoszonym przez inwestora. Zawiera ona takz˙e interesuja˛ce rozwaz˙ania na temat hipotezy rynku efektywnego i przyczyn obserwowanych 300 Przegla˛d pis´miennictwa anomalii rynkowych oraz sposobo´w ich wyjas´nienia. Warto podkres´lic´, z˙e w rozdziale 2 Autorka konfrontuje klasyczna˛ teorie˛ inwestycji finansowych z teoria˛ behawioralna˛. anomalii rynkowych oraz sposobo´w ich wyjas´nienia. Warto podkres´lic´, z˙e w rozdziale 2 Autorka konfrontuje klasyczna˛ teorie˛ inwestycji finansowych z teoria˛ behawioralna˛. Dwa kolejne rozdziały, kto´re stanowia˛ wyraz´nie druga˛ cze˛s´c´ ksia˛z˙ki, sa˛ pos´wie˛cone omo´wieniu metod oceny efektywnos´ci funduszy inwestycyjnych przy zastosowaniu zaro´wno podejs´cia tech- nicznego, jak i podejs´cia fundamentalnego. Poła˛czenie tych dwo´ch podejs´c´ w analizie wyniko´w funduszy inwestycyjnych, dotychczas stosowanych oddzielnie w literaturze przedmiotu i praktyce, stanowi istotny, nowatorski aspekt pracy. W rozdziale 3 Autorka prezentuje klasyczne i nowoczesne miary efektywnos´ci funduszy, kto´rych konstrukcja umoz˙liwia jednoczesne wykorzystanie informacji o osia˛ganych stopach zwrotu i ryzyku ponoszonym przez inwestora. Prezentacja poszczego´lnych mierniko´w efektywnos´ci jest uzupełniona krytyczna˛ ich analiza˛. Autorka przedstawia wady i zalety poszczego´lnych miar i wskaz´niko´w, a takz˙e opisuje gło´wne błe˛dy metodologiczne wyste˛puja˛ce przy interpretacji sto´p zwrotu wypracowywanych przez fundusze inwestycyjne: tradycyjne i fundusze hedge. To bardzo wartos´ciowy fragment pracy. 299 Recenzowana monografia Katarzyny Perez reprezentuje bardzo wysoki poziom merytoryczny i warsztatowy. Lektura pracy wskazuje, z˙e prezentowane przez Autorke˛ zagadnienia sa˛wynikiem jej głe˛bokich przemys´len´ , a nie wyła˛cznie omo´wieniem pogla˛do´w innych autoro´w – zaro´wno teoretyko´w, jak i praktyko´w. Ksia˛z˙ka jest napisana z ogromna˛ pasja˛ badawcza˛. Wszystkie omawiane problemy zwia˛zane z rozwia˛zaniami teoretycznymi i stosowanymi miarami efektywnos´ci funduszy Autorka opatruje własnym, cze˛sto krytycznym komentarzem, stosuje takz˙e oryginalne autorskie metody prezentacji. 301 Przegla˛d pis´miennictwa Ksia˛z˙ka Katarzyny Perez jest pierwsza˛ na polskim rynku wydawniczym publikacja˛ prezentuja˛ca˛ w sposo´b kompleksowy problematyke˛ efektywnos´ci funduszy inwestycyjnych. Praca zawiera ponadto wyniki obszernych własnych badan´ empirycznych Autorki na temat efektywnos´ci polskich funduszy. Recenzowana monografia jest z pewnos´cia˛ w pełni oryginalnym i nowatorskim opracowaniem. Doktor Perez jest niewa˛tpliwie bardzo utalentowanym pracownikiem naukowo-badawczym, o duz˙ej wiedzy i pracowitos´ci, czego dowodzi ogrom przeprowadzonych badan´ literaturowych i empi- rycznych, kto´re stały sie˛ podstawa˛ do napisania recenzowanej ksia˛z˙ki. Jej monografia ma charakter interdyscyplinarny: choc´ jest pos´wie˛cona gło´wnie problematyce finanso´w, to jednak wiele zagadnien´ nawia˛zuje bezpos´rednio do ekonomii, a takz˙e nauk o zarza˛dzaniu. Przy przygotowaniu tej pracy Autorka wykazała sie˛ tez˙ ogromna˛ erudycja˛, o czym s´wiadczy choc´by fragment pos´wie˛cony dywer- syfikacji portfela inwestycyjnego. Korzenie dywersyfikacji zmniejszaja˛cej ryzyko odnalazła ona nie tylko w osiemnastowiecznych tekstach ekonomicznych Daniela Bernoullego, ale takz˙e w dziełach Williama Shakespeare’a, a nawet w ksie˛gach biblijnych. Ksia˛z˙ka została przygotowana z duz˙a˛ dbałos´cia˛ o poprawnos´c´ je˛zykowa˛. Mimo z˙e praca jest pos´wie˛cona dos´c´ skomplikowanej, zaawansowanej merytorycznie i specjalistycznej problematyce, została napisana przyste˛pnym, jasnym je˛zykiem, czyta sie˛ ja˛ z duz˙a˛ przyjemnos´cia˛. Jest to niewa˛tpliwy atut pracy, gdyz˙ z powodzeniem moga˛ z niej korzystac´ nie tylko naukowcy – teoretycy i badacze akademiccy, lecz takz˙e praktycy – osoby zarza˛dzaja˛ce funduszami i jego uczestnicy. Ksia˛z˙ka moz˙e byc´ takz˙e bardzo pomocna studentom kierunko´w finansowych, ekonomicznych i biznesowych. Moz˙e byc´ poz˙yteczna˛ lektura˛ dla wszystkich, kto´rzy chcieliby sie˛ zapoznac´ z ta˛ stosunkowo nowa˛ w polskich realiach gospodarczych problematyka˛ funkcjonowania funduszy inwestycyjnych, a w szczego´lnos´ci pogłe˛bic´ wiedze˛ w zakresie pomiaru ich efektywnos´ci. Z pełnym przekonaniem moz˙na zatem zache˛cac´ do przeczytania tej wartos´ciowej monografii Aleksandra Duliniec almad@sgh.waw.pl Tomasz Sobczak, Ekonomis´ci czytani, ale nie słuchani. Działalnos´c´ organo´w doradczych przy Radzie Ministro´w w Polsce w latach 1957-2006, Wydawnictwo Key Text, Warszawa 2012, ss. 161. Transformacja roli pan´ stwa polega wspo´łczes´nie na ograniczeniu jego dotychczasowych prero- gatyw, ale nie oznacza to, z˙e jego rola zanika lub staje sie˛ mniej istotna. Paradoksalnie funkcje pan´ stwa (rza˛du i samorza˛do´w) ulegaja˛ raczej wzmocnieniu. 299 Ten pozorny paradoks znika, gdy funkcje te rozdzieli sie˛ na operacyjne i systemowe; operacyjne rzeczywis´cie ulegaja˛ ograniczeniu, nato- miast zapotrzebowanie na systemowe ros´nie. Fakt ten znajduje sie˛ u podstaw ponadczasowego (tzn. niewynikaja˛cego z koniunktur czy przejs´ciowych mo´d) zapotrzebowania na mys´l ekonomiczna˛ reprezentowana˛ przez ekonomisto´w – doradco´w kolejnych rza˛do´w. Na tym tle moz˙na stwierdzic´, z˙e problematyka pracy Tomasza Sobczaka jest bez wa˛tpienia aktualna, waz˙na, a przy tym interesuja˛ca zaro´wno z poznawczego, jak i gło´wnie praktycznego punktu widzenia. Autor pro´buje przeanalizowac´ szczego´łowo, z ro´z˙nych punkto´w widzenia, problematyke˛ funkcjonowania w latach 1957-2006 rza˛dowych organo´w doradczych. Pozytywnie i z uznaniem nalez˙y ocenic´ zdefiniowane we wprowadzeniu motywy wyboru obszaru badawczego, gło´wne załoz˙enia pracy, stadia procedury badawczej oraz cel pracy (s. 13). Jest nim przedstawienie działalnos´ci czterech rad, ze szczego´lnym uwzgle˛dnieniem aspekto´w prawno-orga- nizacyjnych, kalendarium ich posiedzen´ , składu osobowego oraz dorobku publikacyjnego. Autor przyja˛ł za mys´l przewodnia˛ stwierdzenie, z˙e rady ekonomiczne funkcjonuja˛ce w Polsce w anali- zowanych latach odgrywały role fasadowe w zakresie wpływania na kształt polityki społeczno- -gospodarczej ekip rza˛dowych. Sformułował tez˙ trzy interesuja˛ce hipotezy pomocnicze: 1) Skutecznos´c´ wpływania rad na ostateczny kształt polityki społeczno-gospodarczej poszcze- go´lnych rza˛do´w jest tym mniejsza, im wyz˙szy jest poziom zro´z˙nicowania pogla˛do´w członko´w tych organo´w. 1) Skutecznos´c´ wpływania rad na ostateczny kształt polityki społeczno-gospodarczej poszcze- go´lnych rza˛do´w jest tym mniejsza, im wyz˙szy jest poziom zro´z˙nicowania pogla˛do´w członko´w tych organo´w.
https://openalex.org/W3105534919	http://t2r2.star.titech.ac.jp/rrws/file/CTT100685672/ATD100000413/	English	null	Sufficient conditions for wave instability in three-component reaction-diffusion systems	Progress of theoretical and experimental physics	2,014	cc-by	9,899	論文 / 著書情報 Article / Book Information 論文 / 著書情報 Article / Book Information Title Sufficient conditions for wave instability in three-component reaction- diffusion systems Authors Shigefumi Hata, Hiroya Nakao, Alexander S Mikhailov Citation Progress of Theoretical and Experimental Physics, Vol. 2014, No. 1, pp. 013A01 Pub. date 2014, 1 URL http://arxiv.org/abs/1302.0683 Note This article has been accepted for publication in Progress of Theoretical and Experimental Physics Published by Oxford University Press. Creative Commons See next page. Powered by T2R2 (Tokyo Institute Research Repository) Sufﬁcient conditions for wave instability in three-component reaction-diffusion systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sufﬁcient conditions for the wave instability in general three-component reaction-diffusion sys- tems are derived. These conditions are expressed in terms of the Jacobian matrix of the uniform steady state of the system, and enable us to determine whether the wave instability can be observed as the mobility of one of the species is gradually increased. It is found that the insta- bility can also occur if one of the three species does not diffuse. Our results provide a useful criterion for searching wave instabilities in reaction-diffusion systems of various origins. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subject Index A53, A54 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. © The Author(s) 2014. Published by Oxford University Press on behalf of the Physical Society of Japan. License https://creativecommons.org/licenses/by/4.0/ Prog. Theor. Exp. Phys. 2014, 013A01 (17 pages) DOI: 10.1093/ptep/ptt102 © The Author(s) 2014. Published by Oxford University Press on behalf of the Physical Society of Japan. © The Author(s) 2014. Published by Oxford University Press on behalf of the Physical Society of Japan. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0/), © The Author(s) 2014. Published by Oxford University Press on behalf of the Physical Society of Japan. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.or 1. Introduction The wave instability provides an important mechanism for oscillatory pattern formation in nonequi- librium chemical systems. When it takes place, a critical mode corresponding to a traveling wave with a certain wavenumber and oscillation frequency begins to grow, destabilizing the uniform steady state. Although being less known, the wave instability has already been considered in 1952 by A. Turing in his pioneering publication [1], where the classical (i.e. static) Turing instability, lead- ing to the establishment of a periodic stationary pattern, has also been introduced. Therefore, it may also be appropriate to describe it as the oscillatory Turing bifurcation. Moreover, it was noticed by A. Turing [1] that at least three interacting species are needed for this instability to occur. Because of the spatial reﬂection symmetry, waves traveling in the left and right directions have the same growth rates, and both of them begin to spontaneously develop above the instability threshold. Nonlinear interactions between such modes determine whether one of the modes gets suppressed, so that a wave traveling in a certain direction is established, or standing waves, representing superposi- tions of left- and right-traveling waves, are formed instead [2,3]. The wave patterns resulting from such instability can also exhibit secondary instabilities, and wave turbulence may set on. In contrast to the classical Turing bifurcation, which has been extensively discussed for both bio- logical and chemical systems [1,4–10], the wave bifurcation has so far attracted less attention. It has been considered for special chemical models [11,12], and its existence was suggested in the exper- iments with Belousov-Zhabotinsky microemulsions [13]. There are also publications in which this instability was discussed for special ecological models [14]. PTEP 2014, 013A01 S. Hata et al. Because at least three species are needed for the wave instability to occur, the linear stability analy- sis is more complex in this case, as compared with the classical Turing bifurcation in two-component activator-inhibitor systems. The complexity of the stability analysis, which was performed sepa- rately for individual chemical systems, has probably also been responsible for the fact that the wave instability was not broadly investigated for reaction-diffusion media. In a recent study, sufﬁcient conditions for the wave instability in general three-component reaction-diffusion systems have been derived [15]. These conditions have been formulated in terms of the elements of the Jacobian matrix and the diffusion constants of the reacting species. 1. Introduction By using an essentially different method, in this article we derive another set of sufﬁcient conditions for the wave bifurcation in general three-component reaction–diffusion models. As we show, the derived sufﬁcient conditions are complementary to those obtained in Ref. [15]. Indeed, there are systems where our conditions can be used to predict the wave instability, whereas the other set of sufﬁcient conditions does not apply. On the other hand, our conditions do not work for some systems while the conditions from Ref. [15] hold. The conditions derived in the present paper tell whether the wave bifurcation is possible when the mobility of any chosen species is gradually increased, while diffusion coefﬁcients of other species are kept constant. The instability may take place even if one of the three species is immobile. 2. Three-component reaction–diffusion systems We consider reaction–diffusion systems with three chemical reactants U, V , and W. Local densities of the reactants are denoted as u = [U], v = [V ], and w = [W]. All reactants diffuse over the space and undergo local chemical reactions. Generally, such systems are described by the equations ⎧ ⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎩ du dt = f (u, v, w) + Du∇2u, dv dt = g(u, v, w) + Dv∇2v, dw dt = h(u, v, w) + Dw∇2w, (1) ⎧ ⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎩ du dt = f (u, v, w) + Du∇2u, dv dt = g(u, v, w) + Dv∇2v, dw dt = h(u, v, w) + Dw∇2w, (1) (1) where functions f , g, and h represent the local reactions. Diffusion coefﬁcients of the reactants are Du, Dv, and Dw. We assume that a uniform steady state (u, v, w) = (¯u, ¯v, ¯w) determined by f (¯u, ¯v, ¯w) = g(¯u, ¯v, ¯w) = h(¯u, ¯v, ¯w) = 0 exists and that this state is stable in the absence of diffusion. where functions f , g, and h represent the local reactions. Diffusion coefﬁcients of the reactants are Du, Dv, and Dw. We assume that a uniform steady state (u, v, w) = (¯u, ¯v, ¯w) determined by f (¯u, ¯v, ¯w) = g(¯u, ¯v, ¯w) = h(¯u, ¯v, ¯w) = 0 exists and that this state is stable in the absence of diffusion. 3. Linear stability analysis with rescaled variables We introduce small perturbations to the steady state as (u, v, w) = (¯u, ¯v, ¯w) + (δu, δv, δw). Sub- stituting this into Eqs. (1), the following linearized differential equations for the perturbations are obtained: ⎧ ⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎩ d dt δu = fuδu + fvδv + fwδw + Du∇2δu, d dt δv = guδu + gvδv + gwδw + Dv∇2δv, d dt δw = huδu + hvδv + hwδw + Dw∇2δw, (2) d dt δv = guδu + gvδv + gwδw + Dv∇2δv, (2) (2) 2/17 PTEP 2014, 013A01 S. Hata et al. S. Hata et al. where fu = ∂f/∂u\|(¯u,¯v, ¯w), fv = ∂f/∂v\|(¯u,¯v, ¯w), fw = ∂f/∂w\|(¯u,¯v, ¯w), ... are partial derivatives at the steady state. The following rescaled variables are introduced for convenience: where fu = ∂f/∂u\|(¯u,¯v, ¯w), fv = ∂f/∂v\|(¯u,¯v, ¯w), fw = ∂f/∂w\|(¯u,¯v, ¯w), ... are partial derivatives at the steady state. The following rescaled variables are introduced for convenience: δ ˜u = δu, δ ˜v = fv gu δv, δ ˜w = fv hv δw, ˜t = \| fvgu\|t. (3) (3) We substitute these variables into Eqs. (2) to obtain the set of equations We substitute these variables into Eqs. (2) to obtain the set of equations ⎧ ⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎩ d d˜t δ ˜u = m0δ ˜u + αδ ˜v + nδ ˜w + Duμ∇2δ ˜u, d d˜t δ ˜v = βδ ˜u + p0δ ˜v + qδ ˜w + Dvμ∇2δ ˜v, d d˜t δ ˜w = rδ ˜u + γ δ ˜v + s0δ ˜w + Dwμ∇2δ ˜w, (4) (4) where m0 = fu √\| fvgu\|, n = fw \|hv\| fv3gu , p0 = gv √\| fvgu\|, q = gw hv fvgu , r = hu \|hv\| fv gu , s0 = hw √\| fvgu\|, μ = 1 √\| fvgu\|. (5) (5) The coefﬁcients α, β, γ are determined by the signs of fv, gu, hv, α = sign( fv), β = sign(gu), γ = sign(hv). The perturbations (δ ˜u, δ ˜v, δ ˜w) are expanded over plane waves as The coefﬁcients α, β, γ are determined by the signs of fv, gu, hv, α = sign( fv), β = sign(gu), γ = sign(hv). 3. Linear stability analysis with rescaled variables The perturbations (δ ˜u, δ ˜v, δ ˜w) are expanded over plane waves as δ ˜u = d⃗k ˜U ⃗k exp λ ⃗k ˜t −i⃗k · ⃗x , δ ˜v = d⃗k ˜V ⃗k exp λ ⃗k ˜t −i⃗k · ⃗x , δ ˜w = d⃗k ˜W ⃗k exp λ ⃗k ˜t −i⃗k · ⃗x , (6) (6) where ⃗k = (k1, k2, . . .) is the wave vector, and λ ⃗k is the growth rate of the plane wave with wave vector ⃗k. Thus, we obtain the following equations for each wave vector ⃗k: where ⃗k = (k1, k2, . . .) is the wave vector, and λ ⃗k is the growth rate of the plane wave with wave vector ⃗k. Thus, we obtain the following equations for each wave vector ⃗k: λ(k) ⎛ ⎜⎝ ˜U (k) ˜V (k) ˜W (k) ⎞ ⎟⎠= ⎛ ⎜⎝ m0 −Duμk2 α n β p0 −Dvμk2 q r γ s0 −Dwμk2 ⎞ ⎟⎠ ⎛ ⎜⎝ ˜U (k) ˜V (k) ˜W (k) ⎞ ⎟⎠, (7) (7) where k = ⃗k is the wave number, the magnitude of the wave vector ⃗k. Because only the magnitude of the wave vector is important, we drop the vector symbols from here on. where k = ⃗k is the wave number, the magnitude of the wave vector ⃗k. Because only the magnitude of the wave vector is important, we drop the vector symbols from here on. 3/17 PTEP 2014, 013A01 S. Hata et al. The condition The condition det ⎛ ⎜⎝ m0 −Duμk2 −λ(k) α n β p0 −Dvμk2 −λ(k) q r γ s0 −Dwμk2 −λ(k) ⎞ ⎟⎠= 0 (8) (8) should be satisﬁed for Eq. (7) to have non-trivial solutions. Thus, the linear growth rate λ(k) is determined by the characteristic equation λ3 −(m + p + s)λ2 + (mp + ps + sm −nr −γ q −αβ)λ −(mps −npr + αqr −γ mq + βγ n −αβs) = 0, (9) (9) where m = m(k) = m0 −k2μDu, p = p(k) = p0 −k2μDv, s = s(k) = s0 −k2μDw. (10) (10) The growth of each plane-wave mode is determined by the real part of λ(k). The uniform steady state is stable if Re(λ(k)) is negative for all k. The instability occurs if Re(λ(k)) becomes positive for at least one wave number k = kc. 3. Linear stability analysis with rescaled variables Then the uniform steady state is destabilized, leading to sponta- neous development of wave patterns with critical wave number kc. If the imaginary part Im(λ(kc)) of the unstable mode is zero, the ﬁrst critical mode represents a stationary plane wave and the Turing instability occurs. On the other hand, if Im(λ(kc)) ̸= 0, the critical mode is oscillatory in time and periodic in space, so that the critical mode represents a traveling wave and the wave instability takes place. 4. Critical condition for the instabilities Boundary surface for the wave instability Iwav(m, p, s) = 0. The line touches the surface at a point P. Panel (b) is a close-up of (a). The parameters are ﬁxed at n = q = r = α = β = γ = 1, m0 = 0.4, p0 = 0.4, and s0 = −1.8. The diffusion constants are Du = 1, Dv = 1, and Dw = 14.521. Thus, the wave instability ﬁrst takes place when, for one wave number k = kc, the equation Thus, the wave instability ﬁrst takes place when, for one wave number k = kc, the equation v = (mps −npr + αqr −γ mq + βγ n −αβs) + (m + p + s)(mp + ps + sm −nr −γ q −αβ) = 0 (15) (15) becomes satisﬁed, where m, p, and s are given by Eqs. (10) with k = kc. The Turing instability ﬁrst takes place when, for one wave number k = kc, the equation becomes satisﬁed, where m, p, and s are given by Eqs. (10) with k = kc. The Turing instability ﬁrst takes place when, for one wave number k = kc, the equation comes satisﬁed, where m, p, and s are given by Eqs. (10) with k = kc. The Turing instability ﬁrst takes place when, for one wave number k = kc, the equation Ist = −(mps −npr + αqr −γ mq + βγ n −αβs) = 0 (16) (16) becomes satisﬁed, where the coefﬁcients m, p, and s are again given by Eqs. (10) with wave number kc. It is convenient to introduce the three-dimensional m-p-s space in order to represent these con- ditions graphically. As illustrated in Figs. 1 and 2, each of the conditions (15) and (16) deﬁnes a boundary surface and Eqs. (10) determine a straight line which is parameterized by the wave number k. If the line touches the boundary surface Iwav(m, p, s) = 0 or Ist(m, p, s) = 0, then condition (15) or (16) is satisﬁed and the corresponding instability takes place. 4. Critical condition for the instabilities The complex conjugate root theorem holds that a characteristic equation with real coefﬁcients λ3 + aλ2 + bλ + c = 0 (11) (11) has either three real roots or one real root and a pair of complex conjugate roots. In the latter case, the three roots can be written as has either three real roots or one real root and a pair of complex conjugate roots. In the latter case, the three roots can be written as λ1,2 = ψ ± iω, λ3 = φ, (12) (12) so that the coefﬁcients are represented as so that the coefﬁcients are represented as a = −(2ψ + φ), b = ψ2 + ω2 + 2ψφ, c = −(ψ2 + ω2)φ. (13) (13) Combining these three equations, we obtain Combining these three equations, we obtain c −ab = −(ψ2 + ω2)φ + (2ψ + φ)(ψ2 + ω2 + 2ψφ) = 2ψ (ψ + φ)2 + ω2 . (14) (14) At the threshold of the wave instability, we have ψ = 0, so that c −ab = 0. At the threshold of the Turing instability, we would have φ = 0, and therefore c = 0. Note that the Turing instability is also possible when the characteristic equation has three real roots and one of them becomes positive. It can be easily checked that, also in this case, the instability threshold corresponds to c = 0. At the threshold of the wave instability, we have ψ = 0, so that c −ab = 0. At the threshold of the Turing instability, we would have φ = 0, and therefore c = 0. Note that the Turing instability is also possible when the characteristic equation has three real roots and one of them becomes positive. It can be easily checked that, also in this case, the instability threshold corresponds to c = 0. 4/17 PTEP 2014, 013A01 S. Hata et al. (a) (b) Fig. 1. Boundary surface for the wave instability Iwav(m, p, s) = 0. The line touches the surface at a point P. Panel (b) is a close-up of (a). The parameters are ﬁxed at n = q = r = α = β = γ = 1, m0 = 0.4, p0 = 0.4, and s0 = −1.8. The diffusion constants are Du = 1, Dv = 1, and Dw = 14.521. (a) (a) (b) (b) (a) (b) Fig. 1. 5. Sufﬁcient conditions for wave instability The slope of the line in the x-s space is Du Dw/ Du2 + Dv2 and is nonnegative. The initial point P0 of the line is at (x0, s0) on the plane, where x0 = (m0 + p0)/(A + 1). The point (x0, s0) is in the stable region of the x-s space. Thus, two surfaces Iwav = 0 and Ist = 0 intersect the plane along the boundary curves ˆIwav = 0 and ˆIst = 0, that is is introduced on the plane in such a way that we have x = 0 when m + p = 0 holds. From the conditions p = Am + B and (18), variables m and p are represented as m = x −B/(A + 1) and p = Ax + B/(A + 1) on the plane. The slope of the line in the x-s space is Du Dw/ Du2 + Dv2 and is nonnegative. The initial point P0 of the line is at (x0, s0) on the plane, where x0 = (m0 + p0)/(A + 1). The point (x0, s0) is in the stable region of the x-s space. Thus, two surfaces Iwav = 0 and Ist = 0 intersect the plane along the boundary curves ˆIwav = 0 and ˆIst = 0, that is ˆIwav(x, s; A, B) = Iwav x − B A + 1, Ax + B A + 1, s = 0, (19) ˆIst(x, s; A, B) = Ist x − B A + 1, Ax + B A + 1, s = 0. (20) (19) (20) Below, we examine the dependences of the boundary curves ˆIwav = 0 and ˆIst = 0 on the parameters m0, p0, s0, n, q, r, α, β, and γ . This allows us to obtain the parameter conditions under which the wave instability can occur. onditions to satisfy Iwav = 0 ˆ 5.1. Conditions to satisfy Iwav = 0 5. Sufﬁcient conditions for wave instability The uniform steady state (¯u, ¯v, ¯w), corresponding to the plane-wave mode of wave number k = 0, should be stable if diffusion is absent. In terms of the coefﬁcients of the characteristic equation (11), this implies that (17) a > 0, b > 0, c > 0 and c −ab < 0, (17) which is known as the Routh-Hurwitz criteria [5]. Note that the third and fourth criteria in (17) are equivalent to the conditions Ist < 0 and Iwav < 0, respectively. Thus, the initial point P0 with coordinates (m0, p0, s0) on the line (10) should lie inside the stable region in the m-p-s space. The wave instability takes place if the ﬁrst critical mode is oscillatory. At the threshold of the wave instability, the line deﬁned by Eq. (10) should touch the boundary surface Iwav = 0 without having intersections with the surface Ist = 0. 5/17 S. Hata et al. PTEP 2014, 013A01 Fig. 2. Boundary surface for the Turing instability Ist(m, p, s) = 0. The parameters are ﬁxed at n = q = r = α = β = γ = 1. Fig. 2. Boundary surface for the Turing instability Ist(m, p, s) = 0. The parameters are ﬁxed at n = q = r = α = β = γ = 1. Fig. 2. Boundary surface for the Turing instability Ist(m, p, s) = 0. The parameters are ﬁxed at n = q = r = α = β = γ = 1. Suppose that we want to check whether the wave instability can occur when the diffusion constant Dw is varied. Let us denote A = Dv/Du and B = p0 −m0Dv/Du, and consider a plane p = Am + B parallel to the s-axis. The line always lies on this plane irrespective of Dw, because the conditions p0 = Am0 + B and Dv/Du = A are satisﬁed. A coordinate x = m + p A + 1 (18) (18) is introduced on the plane in such a way that we have x = 0 when m + p = 0 holds. From the conditions p = Am + B and (18), variables m and p are represented as m = x −B/(A + 1) and p = Ax + B/(A + 1) on the plane. 5.1. Conditions to satisfy Iwav = 0 Let us examine the shape of the boundary curve ˆIwav = 0 at large values of \|s\|. If \|s\| ≫1, we can neglect O(s0) terms and obtain ˆIwav(x, s; A, B) ≈s (1 + A)2x2 + s(1 + A)x −nr −γ q . (21) (21) 6/17 PTEP 2014, 013A01 S. Hata et al. Then the boundary curve ˆIwav = 0 is given by the equation Then the boundary curve ˆIwav = 0 is given by the equation e ˆIwav = 0 is given by the equation (1 + A)2x2 + s(1 + A)x −nr −γ q = 0, (22) (1 + A)2x2 + s(1 + A)x −nr −γ q = 0, (22) (22) and we have x = 1 2 ⎡ ⎣− s A + 1 ± s A + 1 2 + 4 nr + γ q (A + 1)2 ⎤ ⎦ ≃ s 2(A + 1) −1 ± 1 + 2nr + γ q s2 . (23) (23) From Eq. (21), the boundary curve approaches x = 0 in the limit of large \|s\| as (A + 1)xs2 = 0. (24) (24) Then the plus sign should be chosen in Eq. (23). Thus, the asymptotic boundary curve in the limit s ≫1 is the hyperbola Then the plus sign should be chosen in Eq. (23). Thus, the asymptotic boundary curve in the limit s ≫1 is the hyperbola xs = nr + γ q A + 1 . (25) (25) If the coefﬁcient of this hyperbola is negative, i.e. nr + γ q < 0, the boundary curve lies in the region of x > 0 and s < 0. In such cases, given that x0 > 0, the line always touches the boundary curve when increasing the diffusion constant Dw, as shown in Figs. 3 and 4. Therefore, we require nr + γ q < 0 and x0 > 0 as a part of the sufﬁcient conditions. Using the deﬁnitions of the model parameters (4), these requirements can be written in terms of the Jacobian matrix as fwhu + gwhv < 0, (26) fu + gv > 0. (27) (26) (27) 5.2. Conditions not to satisfy Ist = 0 5.2. Conditions not to satisfy Ist = 0 ˆ 5.2. Conditions not to satisfy Ist = 0 The boundary curve for the Turing instability ˆIst(x, s; A, B) = 0 is given by 5.2. Conditions not to satisfy Ist = 0 The boundary curve for the Turing instability ˆIst(x, s; A, B) = 0 is given by dary curve for the Turing instability ˆIst(x, s; A, B) = 0 is given by The boundary curve for the Turing instability ˆIst(x, s; A, B) = 0 is given by s(x) = −(A + 1)2(Anr + γ q)x −(A + 1)2(βγ n + αqr) + (A + 1)B(nr −γ q) A(A + 1)2x2 −(A2 −1)Bx −αβ(A + 1)2 −B2 , (28) (28) which is a single-valued function of x. which is a single-valued function of x. Let us assume that all three reactants diffuse over the space, Du ̸= 0, Dv ̸= 0, and Dw ̸= 0, which leads to A ̸= 0. If the denominator on the right-hand side of Eq. (28) is not zero for all values of x, then s(x) is a continuous function. Figure 3(a) illustrates the qualitative shape of the boundary curves ˆIwav = 0 and ˆIst = 0 in such a situation. In this case, the line touches the boundary curve ˆIwav = 0 without intersecting ˆIst = 0 as Dw is increased. As a consequence, the wave instability takes place. The condition is given by (A2 −1)2B2 + 4A(A + 1)2 αβ(A + 1)2 + B2 < 0, (29) (29) which implies fugv −fvgu > ( fu Dv + gv Du)2 4Du Dv . (30) (30) If the denominator on the right-hand side of Eq. (28) vanishes at x−and x+, the boundary curve ˆIst = 0 diverges in two ways depending on the values of n, q, r, α, β, and γ [Fig. 3(b) and (c)]. 7/17 S. Hata et al. PTEP 2014, 013A01 (c) s (x, s; A, B) = 0 ˆIwav x x0 (x, s; A, B) = 0 ˆIST increasing w D x- x+ P0 G (a) (x, s; A, B) = 0 ˆIwav s x x0 (x, s; A, B) = 0 ˆIST increasing w D (b) s (x, s; A, B) = 0 ˆIwav x x0 (x, s; A, B) = 0 ˆIST increasing w D x- x+ P0 G P0 G Fig. 3. 5.2. Conditions not to satisfy Ist = 0 Boundary curves ˆIwav = 0 (red) and ˆIst = 0 (blue) when (a) the condition (30) and (b, c) the condi- tion (34) are satisﬁed. All species are mobile. The positions of the line deﬁned by (10) are shown for two different values of Dw. The boundary curve ˆIst = 0 can behave as shown in (b) or (c), depending on the model parameters. (b) s (x, s; A, B) = 0 ˆIwav x x0 (x, s; A, B) = 0 ˆIST increasing w D x- x+ P0 G (b) (a) G x (c) s (x, s; A, B) = 0 ˆIwav x x0 (x, s; A, B) = 0 ˆIST increasing w D x- x+ P0 G (c) increasing w D Fig. 3. Boundary curves ˆIwav = 0 (red) and ˆIst = 0 (blue) when (a) the condition (30) and (b, c) the condi- tion (34) are satisﬁed. All species are mobile. The positions of the line deﬁned by (10) are shown for two different values of Dw. The boundary curve ˆIst = 0 can behave as shown in (b) or (c), depending on the model parameters. s (x, s; A, B) = 0 ˆIwav (a) x P0 x0 (x, s; A, B) = 0 ˆIST increasing w D G p0 + p0 (x, s; A, B) = 0 ˆIwav s (b) x x0 (x, s; A, B) = 0 ˆIST p0 + p0 s (x, s; A, B) = 0 ˆIwav (a) x P0 x0 (x, s; A, B) = 0 ˆIST increasing w D increasing w D G P0 G p0 + p0 Fig. 4. Boundary curves ˆIwav = 0 (red) and ˆIst = 0 (blue). The reactant V does not diffuse (Dv = 0). The line deﬁned by (10) is shown as black lines for two different values of Dw. The boundary curve ˆIst = 0 can behave as shown in (a) or (b), depending on the parameters n, q, r, α, β, and γ . (x, s; A, B) = 0 ˆIwav s (b) x x0 (x, s; A, B) = 0 ˆIST p0 + p0 increasing w D P0 G (b) (a) increasing w D Fig. 4. Boundary curves ˆIwav = 0 (red) and ˆIst = 0 (blue). The reactant V does not diffuse (Dv = 0). The line deﬁned by (10) is shown as black lines for two different values of Dw. 5.2. Conditions not to satisfy Ist = 0 The boundary curve ˆIst = 0 can behave as shown in (a) or (b), depending on the parameters n, q, r, α, β, and γ . 8/17 PTEP 2014, 013A01 S. Hata et al. In both cases, if x0 ≤x−, the line touches ˆIwav = 0 ﬁrst. Thus, if the condition In both cases, if x0 ≤x−, the line touches ˆIwav = 0 ﬁrst. Thus, if the condition x0 ≤x−= (A2 −1)B −(A + 1)2 B2 + 4αβ A 2A(A + 1)2 , (31) (31) which is equivalent to which is equivalent to fugv −fvgu > 0 (32) fugv −fvgu > 0 (32) and and fu Dv + gv Du ≤0 (33) fu Dv + gv Du ≤0 (33) is satisﬁed, the wave instability takes place as Dw is increased. is satisﬁed, the wave instability takes place as Dw is increased. Next, we assume that the reactant V does not diffuse, that is Dv = 0, leading to A = 0. In this case, the boundary curve ˆIst = 0 is given by s(x)\|A=0 = −γ qx + βγ n + αqr + B(nr −γ q) Bx −αβ −B2 , (34) (34) which diverges at x = B + αβ/B = p0 + αβ/p0 [Figs. 4(a) and (b)]. In this case, if x0 ≤p0 + αβ/p0, the line touches ˆIwav = 0 without intersecting ˆIst = 0 when Dw is increased. Thus, the instability condition for a system with two diffusible reactants is given by which diverges at x = B + αβ/B = p0 + αβ/p0 [Figs. 4(a) and (b)]. In this case, if x0 ≤p0 + αβ/p0, the line touches ˆIwav = 0 without intersecting ˆIst = 0 when Dw is increased. Thus, the instability condition for a system with two diffusible reactants is given by x0 ≤p0 + αβ p0 , (35) (35) which implies gv( fugv −fvgu) ≤0. (36) (36) On the other hand, if the reactant U does not diffuse, Du = 0, one can choose a plane m = A′ p + B′ where A′ = Du/Dv and B′ = m0 −p0Du/Dv, and derive the condition fu( fugv −fvgu) ≤0. (37) (37) Thus, sufﬁcient conditions for the wave instability under increasing diffusion constant Dw of the reactant W has been derived for the four cases depending on the diffusion constants. The ﬁrst two con- ditions (26) and (27) are common to all cases. 5.2. Conditions not to satisfy Ist = 0 The last condition depends on the diffusion constants, i.e. we have ⎧ ⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎩ fugv −fvgu > ( fu Dv + gv Du)2 4Du Dv (Du ̸= 0, Dv ̸= 0), (38) fugv −fvgu > 0 and fu Dv + gv Du ≤0 (Du ̸= 0, Dv ̸= 0), (39) fu( fugv −fvgu) < 0 (Du = 0, Dv ̸= 0), (40) gv( fugv −fvgu) < 0 (Du ̸= 0, Dv = 0). (41) (38) (41) erent requirements can, however, be expressed in a single equation: These different requirements can, however, be expressed in a single equation: det fu + Du fv gu gv + Dv ̸= 0 for any < 0, (42) (42) when at least two diffusion constants are non-vanishing. when at least two diffusion constants are non-vanishing. The wave instability may take place also under the variation of the other two diffusion constants Du or Dv. The corresponding sufﬁcient conditions for each case can be obtained by permutating the three variables u, v, and w. 9/17 PTEP 2014, 013A01 S. Hata et al. In summary, the wave instability occurs under increasing diffusion constants Du, Dv, or Dw if the conditions gv + hw > 0 and gu fv + hu fw < 0 and det gv + Dv gw hv hw + Dw ̸= 0 for any < 0, (43) and det gv + Dv gw hv hw + Dw ̸= 0 for any < 0, (43) (43) hw + fu > 0 and hvgw + fvgu < 0 and det hw + Dw hu fw fu + Du ̸= 0 for any < 0, (44) and det hw + Dw hu fw fu + Du ̸= 0 for any < 0, (44) (44) or or fu + gv > 0 and fwhu + gwhv < 0 and det fu + Du fv gu gv + Dv ̸= 0 for any < 0, (45) and det fu + Du fv gu gv + Dv ̸= 0 for any < 0, (45) (45) are respectively satisﬁed. are respectively satisﬁed. are respectively satisﬁed. are respectively satisﬁed. 6. Numerical examples In order to illustrate the results, we examine several three-component reaction-diffusion systems. The ﬁrst example is a chemical reaction-diffusion system introduced by Meinhardt [16], where self- enhancement of an activator U is antagonized by two inhibitors V and W. The system is described by the equations ⎧ ⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎩ du dt = −ruu + s u2 + bu v 1 + suu2 (1 + sww) + Du∇2u, dv dt = −rvv + su2 + bv + Dv∇2v, dw dt = −rww + rwu + Dw∇2w. (46) (46) We ﬁx the parameters at ru = 1.0, rv = 1.0, rw = 0.01, bu = 0.2, bv = 0, su = 0, sw = 2.0, and s = 1.0, yielding the uniform steady state (¯u, ¯v, ¯w) ≃(0.646, 0.417, 0.646). The Jacobian matrix at the uniform steady state is J ≃ ⎛ ⎜⎝ 0.352 −1.548 −0.564 1, 291 −1 0 0.01 0 −0.01 ⎞ ⎟⎠. (47) (47) This satisﬁes the conditions (17), so that the uniform steady state is stable when diffusion is absent. Below, we demonstrate numerically wave instability in the system (46) to check the validity of the derived conditions (43)–(45). Suppose that the inhibitor W does not diffuse, i.e. Dw = 0. In this case, the conditions (44) are always satisﬁed irrespective of the diffusion mobility Du of the activator U. Thus, the wave instability will be found when Dv is gradually increased. Figure 5 illustrates the linear stability analysis in this case, and also presents the results of numerical simulations which were performed for a one-dimensional system under periodic boundary condi- tions. The boundary surfaces Iwav = 0 and Ist = 0 and the x-p plane in the m-p-s space are shown 10/17 PTEP 2014, 013A01 S. Hata et al. space time 0 0.2 0.4 0.6 0.8 3000 3100 3200 3300 0.58 0.6 0.62 0.64 0.66 0.68 0.7 0.72 3400 1 0 5 10 15 20 25 −0.1 −0.05 0 0.05 0.1 k / 2p Re , Im  Re  Im  −2 −1 0 1 2 −15 −10 −5 0 x p (a) (b) (c) (d) P P0 kc Fig. 5. Wave instability in the model (46). (a) Boundary surfaces Iwav(m, p, s) = 0 (red surface), Ist(m, p, s) = 0 (blue surface), and the x-p plane (gray transparent plane) in the three-dimensional m-p-s space. 6. Numerical examples (b) Boundaries ˆIwav(x, p) = 0 (red curve) and ˆIst(x, p) = 0 (blue curve) on the x-p plane. The lines for two different diffusion constants, Dv = 2.5 (black dashed line) and Dv = 156 (black solid line), are shown. (c) The real part (red curve) and the imaginary part (blue curve) of the linear growth rate λ as functions of wave number k. (d) The space-time plot of the ﬁnal established patterns at Dv = 157.5. The diffusion constants of two species U and W are ﬁxed at Du = 2.5 and Dw = 0 in this ﬁgure. (a) (c) −2 −1 0 1 2 −15 −10 −5 0 x p (b) P P0 (b) (a) (d) 0 5 10 15 20 25 −0.1 −0.05 0 0.05 0.1 k / 2p Re , Im  Re  Im  (c) kc (c) space time 0 0.2 0.4 0.6 0.8 3000 3100 3200 3300 0.58 0.6 0.62 0.64 0.66 0.68 0.7 0.72 3400 1 (d) Fig. 5. Wave instability in the model (46). (a) Boundary surfaces Iwav(m, p, s) = 0 (red surface), Ist(m, p, s) = 0 (blue surface), and the x-p plane (gray transparent plane) in the three-dimensional m-p-s space. (b) Boundaries ˆIwav(x, p) = 0 (red curve) and ˆIst(x, p) = 0 (blue curve) on the x-p plane. The lines for two different diffusion constants, Dv = 2.5 (black dashed line) and Dv = 156 (black solid line), are shown. (c) The real part (red curve) and the imaginary part (blue curve) of the linear growth rate λ as functions of wave number k. (d) The space-time plot of the ﬁnal established patterns at Dv = 157.5. The diffusion constants of two species U and W are ﬁxed at Du = 2.5 and Dw = 0 in this ﬁgure. in Fig. 5(a). Note that, while the variable x was introduced by using the variables m and p in Eq. (18), here it depends on the variables s and m, so that the line lies on the x-p plane irrespective of the value of Dv. The boundary curves ˆIwav = 0 and ˆIst = 0 in the x-p plane are shown in Fig. 5(b). Figure 5(c) plots the linear growth rate λ for the instability threshold. The emerging wave pattern is shown in Fig. 5(d). PTEP 2014, 013A01 PTEP 2014, 013A01 S. Hata et al. As another example, we apply our theory to an ecological reaction-diffusion (dispersal predator- prey) system with three species U, V , and W, where the top predator W feeds on intermediate species V , which is in turn a predator for prey U. All species are able to diffuse, and their diffusion constants are different. Such a system can be modelled by the equations ⎧ ⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎩ du dt = au −buu −cu w u + μ u + Du∇2u, dv dt = av 1 −cv v u −dv w v + ν v + Dv∇2v, dw dt = aw 1 −dw w v w + Dw∇2w, (48) (48) where the Holling type II dependence and a linear function are employed to describe the predator- prey interactions [5,17]. The parameters are ﬁxed at au = 3, bu = 1, cu = 1, av = 6, cv = 1, dv = 1, aw = 4, dw = 0.25, and μ = ν = 0.25, which give a uniform steady state (¯u, ¯v, ¯w) ≃ (1.084, 2.557, 10.23). If all species have the same mobilities, Du = Dv = Dw, the steady state is stable. The Jacobian matrix at the steady state is J ≃ ⎛ ⎜⎝ 0.471 −0.813 0 5.552 0.962 −0.911 0 16 −4 ⎞ ⎟⎠. (49) (49) By substituting this into the inequalities (45), one can directly verify that the system satisﬁes the sufﬁcient conditions. Then, our theory tells us that the wave instability will take place when the diffusion constant Dw of the predator W is increased. Figure 6(a) shows the boundary surfaces Iwav = 0 and Ist = 0 and the x-s plane in m-p-s space. The boundary curves ˆIwav = 0 and ˆIst = 0 in the x-s plane are displayed in Fig. 6(b). Figure 6(c) gives the linear growth rate λ as a function of the wave number k at the instability threshold. The developed wave pattern is shown in Fig. 6(d). Starting from equal mobilities of all three species, Du = Dv = Dw = 20, the mobility Dw of the top predator W is gradually increased. When it comes up to a threshold Dw = 892, the line touches the boundary curve ˆIwav = 0 without having intersections with ˆIst = 0 [Fig. 6(b)]. 6. Numerical examples Starting from equal mobilities of the species U and V , Du = Dv = 2.5, we gradually increase the mobility Dv. When it reaches a certain threshold, the line touches the boundary curve ˆIwav = 0 at P0 without having intersections with ˆIst = 0, as shown in Fig. 5(b). At the instability threshold, the real part of the linear growth rate vanishes, Reλ(k) = 0, for the critical wave number kc ≃8, whereas the imaginary part remains ﬁnite [Fig. 5(c)]. Thus, the wave instability takes place and the uniform steady state becomes unstable. When the instability occurs, traveling waves with wave number k = 8 spontaneously develop, as can be seen in Fig. 5(d). 11/17 PTEP 2014, 013A01 Correspondingly, the real part of the linear growth rate vanishes at the threshold, Reλ(k) = 0, for the critical wave number kc ≃4, whereas the imaginary part remains ﬁnite [Fig. 6(c)]. As a result, the wave insta- bility takes place and the uniform steady state becomes unstable. After the instability, spontaneous development of traveling wave patterns with wave number k = 4 is observed [Fig. 6(d)]. Thus, as illustrated by the above examples, the derived sufﬁcient conditions (43)–(45) indeed provide a useful criterion when searching for the wave instabilities in speciﬁc reaction-diffusion systems. It should be emphasized that the derived conditions (43)–(45) are sufﬁcient, but not necessary. This means that systems may exist where such conditions do not hold, but the wave instability is observed nonetheless. An example of such a situation is provided by the extended Brusselator system [12]: ⎧ ⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎩ du dt = a −(1 + b)u + u2v −cu + dw + Du∇2u, dv dt = bu −u2v + Dv∇2v, dw dt = cu −dw + Dw∇2w, (50) (50) 12/17 PTEP 2014, 013A01 S. Hata et al. space time 0.2 0.4 0.6 0.8 1 3000 3005 3010 3015 1 1.02 1.04 1.06 1.08 1.1 1.12 1.14 1.16 −20 −15 −10 −5 0 s −3 −2 −1 0 1 2 k / 2 0 2 4 6 8 10 −1.5 −1 −0.5 0 0.5 1 1.5 Re λ Re λ Im λ kc 0.5 1 1.5 2 2.5 3 3.5 Im λ 0 3020 π x (a) (b) (c) (d) P0 P Fig. 6. Wave instability in the ecological model (48). (a) Boundary surfaces Iwav(m, p, s) = 0 (red surface), Ist(m, p, s) = 0 (blue surface), and the x-s plane (gray transparent plane) in the three-dimensional m-p-s space. (b) Boundaries ˆIwav(x, s) = 0 (red curve) and ˆIst(x, s) = 0 (blue curve) on the x-s plane. The lines for two different diffusion constants, Dw = 20 (black dashed line) and Dw = 892 (black solid line), are shown. (c) The real part (red curve) and the imaginary part (blue curve) of the linear growth rate λ as functions of the wave number k. (d) The space-time plot of the ﬁnal established patterns at Dw = 900. The diffusion constants of the two species U and V are ﬁxed at Du = Dv = 20 in this ﬁgure. PTEP 2014, 013A01 (a) ( ) −20 −15 −10 −5 0 s −3 −2 −1 0 1 2 x (b) (d) P0 P (a) (b) k / 2 0 2 4 6 8 10 −1.5 −1 −0.5 0 0.5 1 1.5 Re λ Re λ Im λ kc 0.5 1 1.5 2 2.5 3 3.5 Im λ π (c) (c) (d) space time 0.2 0.4 0.6 0.8 1 3000 3005 3010 3015 0 3020 ( ) Fig. 6. Wave instability in the ecological model (48). (a) Boundary surfaces Iwav(m, p, s) = 0 (red surface), Ist(m, p, s) = 0 (blue surface), and the x-s plane (gray transparent plane) in the three-dimensional m-p-s space. (b) Boundaries ˆIwav(x, s) = 0 (red curve) and ˆIst(x, s) = 0 (blue curve) on the x-s plane. The lines for two different diffusion constants, Dw = 20 (black dashed line) and Dw = 892 (black solid line), are shown. (c) The real part (red curve) and the imaginary part (blue curve) of the linear growth rate λ as functions of the wave number k. (d) The space-time plot of the ﬁnal established patterns at Dw = 900. The diffusion constants of the two species U and V are ﬁxed at Du = Dv = 20 in this ﬁgure. where activator U and inhibitor V react chemically, and the activator U reversibly transforms into non-reacting species W. The parameters are ﬁxed at a = 1, b = 2.9, c = 1, and d = 1, which yield the uniform steady state (¯u, ¯v, ¯w) = (1, 2.9, 1). The steady state is stable if all three species have the same mobility, Du = Dv = Dw. The Jacobian matrix at the steady state is The Jacobian matrix at the steady state is J ≃ ⎛ ⎜⎝ 0.9 1 1 −2.9 −1 0 1 0 −1 ⎞ ⎟⎠, (51) (51) which gives fu + gv < 0 and fwhu + gwhg > 0. Thus, the system does not satisfy the condi- tions (45). which gives fu + gv < 0 and fwhu + gwhg > 0. Thus, the system does not satisfy the condi- tions (45). However, numerical simulations show that the wave instability takes place in this model when the nonreactive species W diffuses sufﬁciently faster than the other two species (Fig. 7). The linear 13/17 PTEP 2014, 013A01 S. Hata et al. PTEP 2014, 013A01 space time 0.2 0.4 0.6 0.8 1 3000 3020 3040 3060 3080 0.98 0.985 0.99 0.995 1 1.005 1.01 1.015 1.02 k / 2 Im λ 0 3100 π (a) (b) 0 5 10 15 20 25 −0.2 −0.1 0 0.1 0.2 Re λ 0.5 0.7 0.9 1.1 1.3 Re λ Im λ kc Fig. 7. Wave instability in the extended Brusselator system (50). (a) The real part (red curve) and the imaginary part (blue curve) of the linear growth rate λ at the instability threshold as functions of wave number k. (b) The space-time plot of the ﬁnal established pattern at Du = Dv = 0.3 and Dw = 6.95. k / 2 Im λ π (a) 0 5 10 15 20 25 −0.2 −0.1 0 0.1 0.2 Re λ 0.5 0.7 0.9 1.1 1.3 Re λ Im λ kc (a) (b) space time 0.2 0.4 0.6 0.8 1 3000 3020 3040 3060 3080 0.98 0.985 0.99 0.995 1 1.005 1.01 1.015 1.02 0 3100 (b) Fig. 7. Wave instability in the extended Brusselator system (50). (a) The real part (red curve) and the imaginary part (blue curve) of the linear growth rate λ at the instability threshold as functions of wave number k. (b) The space-time plot of the ﬁnal established pattern at Du = Dv = 0.3 and Dw = 6.95. growth rate as a function of wave number is shown in Fig. 7(a). The established traveling wave pattern is displayed in Fig. 7(b). growth rate as a function of wave number is shown in Fig. 7(a). The established traveling wave pattern is displayed in Fig. 7(b). 7. Discussion and conclusions We have constructed alternative sufﬁcient conditions for the wave instability in general three- component reaction-diffusion systems. The conditions are formulated in terms of the Jacobian matrix elements at a steady state and of the diffusion constants. They do not depend on model details. Once these conditions are satisﬁed, wave instability occurs as we increase the diffusion mobility of one of the reacting species. Our general results are applicable for systems of various origins, including biological, chemical, physical, and ecological systems. Our analysis has revealed that the wave instability may occur even if one of three reactants is immobile. This result can be important in a variety of applications involving both diffusible and non-diffusible reactants. In contrast to necessary and sufﬁcient conditions, different sets of merely sufﬁcient conditions may be derived and hold for the same system. In fact, another set of sufﬁcient conditions for the wave instability in general three-species reaction-diffusion systems has recently been constructed in Ref. [15]. In Ref. [15], the authors focused particularly on how the stability of a three-species system is affected by instabilities of its two-species subsystems. They classiﬁed the instabilities of the subsys- tems into several types and analyzed their effect on the complete three-species system in detail. Their investigation was based on linear stability analysis, and the sufﬁcient conditions were formulated in terms of the elements of the Jacobian matrix and the diffusion constants of the three species. They also derived sufﬁcient conditions for the classical (stationary) Turing instability in three-component reaction-diffusion systems. In our study, we have constructed the sufﬁcient conditions using a different method, based on linear stability analysis and inspection of the instability threshold in the parameter space. Once the Jacobian matrix is obtained, the boundary surfaces of the instabilities and the line whose slope corresponds to the diffusion constants of the three species can be drawn as in Figs. 1 and 2. If the line touches the boundary surface as its slope is varied, the instability occurs. By considering the situation where 14/17 PTEP 2014, 013A01 S. Hata et al. the line always touches the boundary surfaces of the wave instability when one of the three diffusion constants is increased, we have derived sufﬁcient conditions that are different from those in Ref. [15]. We stress that these two sets of sufﬁcient conditions are complementary. Acknowledgements The authors acknowledge the ﬁnancial support through the DFG SFB 910 program “Control of Self-Organizing Nonlinear Systems” in Germany, through the Fellowship for Research Abroad, KAKENHI and the FIRST Aihara Project (JSPS), and the CREST Kokubu Project (JST) in Japan. 7. Discussion and conclusions There is an overlap between them, but neither set of sufﬁcient conditions includes the other one. For example, our con- ditions can predict the wave instability in the ecological model (48); however, this model does not satisfy the conditions [(iv)–(1) and (2) in Corollary 1.1] derived in Ref. [15]. On the other hand, our conditions cannot predict the wave instability in the extended Brusselator system (50), while the conditions in Ref. [15, Sect. 3.2] can do this. Although our conditions have been derived for continuous media, they are also applicable for reaction-diffusion networks where reactants diffuse over links and undergo local reactions on each node [18–21]. In such systems, diffusion processes are described by the Laplacian matrix instead of the Laplacian differential operator in Eqs. (1) and (2). Expanding small perturbations over eigen- vectors of the Laplacian matrix, we obtain the same characteristic equation as that for continuous media (9), where (m, p, s) are parameterized by the Laplacian eigenvalues instead of −k2 in Eq. (10)—see Appendix A. Thus, the critical conditions Iwav = 0 and Ist = 0 are valid, so that we also obtain the sufﬁcient conditions (43)–(45) in networks. This may be useful for ﬁnding the oscillatory Turing bifurcation in networks of coupled reactors or biological cells [22,23]. Appendix A. Linear stability analysis on reaction–diffusion networks We consider three reactant species U, V , and W on a network of size N. The local densities of the reactants on the network node i are denoted as ui = [U]i, vi = [V ]i, and wi = [W]i. The network architecture is determined by the adjacent matrix A whose elements Ai j are 1 if the nodes i and j are connected, and 0 otherwise. All reactants diffuse over network links and undergo local reactions in each node. Such dynamics are described by the equations ⎧ ⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩ ˙ui = f (ui, vi, wi) + Du N ! j=1 Li ju j, ˙vi = g(ui, vi, wi) + Dv N ! j=1 Li jv j, ˙wi = h(ui, vi, wi) + Dw N ! j=1 Li jw j, (A1) (A1) for i = 1, . . . , N, where the functions f , g, and h specify the local reactions, Du, Dv, and Dw are the diffusion constants of the three reactants, and Li j = Ai j −δi j " j Ai j is the Laplacian matrix. for i = 1, . . . , N, where the functions f , g, and h specify the local reactions, Du, Dv, and Dw are the diffusion constants of the three reactants, and Li j = Ai j −δi j " j Ai j is the Laplacian matrix. j The linear stability analysis is performed on the system in an analogous way to continuous media (See Sect. 3). The system is assumed to have a uniform steady state (¯u, ¯v, ¯w) which is determined by f (¯u, ¯v, ¯w) = g(¯u, ¯v, ¯w) = h(¯u, ¯v, ¯w) = 0. Small perturbations are introduced on the steady state as (ui, vi, wi) = (¯u, ¯v, ¯w) + (δui, δvi, δwi). We substitute this into Eq. (A1) to obtain linearized differential equations. Introducing the rescaled variables in the same way as in Eq. (3), the linearized 15/17 PTEP 2014, 013A01 S. Hata et al. PTEP 2014, 013A01 S. Hata et al. are rewritten as differential equations are rewritten as differential equations are rewritten as are rewritten as ⎧ ⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎪⎩ d d˜t δ ˜ui = m0δ ˜ui + αδ ˜vi + nδ ˜wi + Duμ N ! j=1 Li jδ ˜u j, d d˜t δ ˜vi = βδ ˜ui + p0δ ˜vi + qδ ˜wi + Dvμ N ! Appendix A. Linear stability analysis on reaction–diffusion networks j=1 Li jδ ˜v j, d d˜t δ ˜wi = rδ ˜ui + γ δ ˜vi + s0δ ˜wi + Dwμ N ! j=1 Li jδ ˜w j, (A2) (A2) where m0, p0, s0, n, q, r, and μ are given in Eq. (5). Perturbations are expanded over a set of Laplacian eigenvectors {φ(κ) i } as where m0, p0, s0, n, q, r, and μ are given in Eq. (5). Perturbations are expanded over a set of Laplacian eigenvectors {φ(κ) i } as δ ˜ui(t) = N ! κ=1 ˜U (κ) exp[λ(κ)t]φ(κ) i , δ ˜vi(t) = N ! κ=1 ˜V (κ) exp[λ(κ)t]φ(κ) i , (A3) δ ˜wi(t) = N ! κ=1 ˜W (κ) exp[λ(κ)t]φ(κ) i , (A3) where λ(κ) is the linear growth rate of the κth eigenmode. Substituting into Eq. (A2), we obtain the following matrix equation for each eigenmode: where λ(κ) is the linear growth rate of the κth eigenmode. Substituting into Eq. (A2), we obtain the following matrix equation for each eigenmode: λ(κ) ⎛ ⎜⎝ ˜U (κ) ˜V (κ) ˜W (κ) ⎞ ⎟⎠= ⎛ ⎜⎝ m0 + Duμ(κ) α n β p0 + Dvμ(κ) q r γ s0 + Dwμ(κ) ⎞ ⎟⎠ ⎛ ⎜⎝ ˜U (κ) ˜V (κ) ˜W (κ) ⎞ ⎟⎠, (A4) (A4) where (κ) is the Laplacian eigenvalue, which is deﬁned by " j Li jφ(κ) j = (κ)φ(κ) i . Assuming that the equations (A4) have non-trivial solutions, we obtain the same characteristic equation as Eq. (9) with where (κ) is the Laplacian eigenvalue, which is deﬁned by " j Li jφ(κ) j = (κ)φ(κ) i . Assuming that the equations (A4) have non-trivial solutions, we obtain the same characteristic equation as Eq. (9) with m = m(κ) = m0 + (κ)μDu, p = p(κ) = p0 + (κ)μDv, (A5) s = s(κ) = s0 + (κ)μDw. (A5) Thus, the critical conditions for the instabilities (15) and (16) are valid, and therefore the sufﬁcient conditions (43)–(45) are also applicable for reaction-diffusion networks. [6] S. Sick, S. Reiniker, J. Timmer, and T. Schlake, Science, 314, 1447 (2006). [7] S. Kondo and R. Asai, Nature, 376, 765 (1995). y ( ) [5] J. Murray, Mathematical Biology (Springer, New York, 2003). ( p g ) [4] H. Meinhardt and A. Gierer, Bioessays, 22, 753 (2000). D. Walgraef, Spatio-Temporal Pattern Formation, with Examples in Physics, Chemistry and Materia Science (Springer, New York, 1997). [1] A. M. Turing, Philos. Trans. R. Soc. London B, 237, 37 (1952). ] A. M. Turing, Philos. Trans. R. Soc. London B, 237, 37 (1952). References [1] A. M. Turing, Philos. Trans. R. Soc. London B, 237, 37 (1952). [1] A. M. Turing, Philos. Trans. R. Soc. London B, 237, 37 (1952). [2] E. Knobloch and J. De Luca, Nonlinearity, 3, 975 (1990). [3] D. Walgraef, Spatio-Temporal Pattern Formation, with Examples in Physics, Chemistry and Materials Science (Springer, New York, 1997). [3] D. Walgraef, Spatio-Temporal Pattern Formation, with Examples in Physics, Chemistry and Materials Science (Springer, New York, 1997). [4] H. Meinhardt and A. Gierer, Bioessays, 22, 753 (2000). [5] J. Murray, Mathematical Biology (Springer, New York, 2003). 16/17 PTEP 2014, 013A01 S. Hata et al. [8] A. Nakamasu, G. Takahashi, A. Kanbe, and S. Kondo, Proc. Natl. Acad. Sci. U.S.A., 106, 8429 (2009). [9] V. Castets, E. Dulos, J. Boissonade, and P. De Kepper, Phys. Rev. Lett., 64, 2953 (1990). [8] A. Nakamasu, G. Takahashi, A. Kanbe, and S. Kondo, Proc. Natl. Acad. Sci. U.S.A., 106, 8429 (2009). y ( ) [10] Q. Ouyang and H. L. Swinney, Nature, 352, 610 (1991). [10] Q. Ouyang and H. L. Swinney, Nature, 352, 610 (1991). yang and H. L. Swinney, Nature, 352, 610 (1991). [12] L. Yang, M. Dolnik, A. M. Zhabotinsky, and I. R. Epstein, J. Chem. Phys., 117, 7259 ] L. Yang, M. Dolnik, A. M. Zhabotinsky, and I. R. Epstein, J. Chem. Phys., 117, 7259 (2002). [13] V. K. Vanag and I. R. Epstein, Phys. Rev. Lett., 87, 228301 (2001). ] V. K. Vanag and I. R. Epstein, Phys. Rev. Lett., 87, 228301 (2001). ] K. A. J. White and C. A. Gilligan, Philos. Trans. R. Soc. London B, 353, 543 (1998). [14] K. A. J. White and C. A. Gilligan, Philos. Trans. R. Soc. London B, 3 ] A. Anma, K. Sakamoto, and T. Yoneda, Kodai Math. J., 35, 215 (2012). [15] A. Anma, K. Sakamoto, and T. Yoneda, Kodai Math. J., 35, 215 (2012) [16] H. Meinhardt, Physica D, 199, 264 (2004). [ ] y ( ) [17] C. S. Holling, The Canadian Entomologist, 91, 385 (1959). [17] C. S. Holling, The Canadian Entomologist, 91, 385 (1959). 17] C. S. Holling, The Canadian Entomologist, 9 [18] H. Nakao and A. S. Mikhailov, Nat. Phys., 6, 544 (2010). [18] H. Nakao and A. S. Mikhailov, Nat. Phys., 6, 544 (2010). [19] S. Hata, H. Nakao, and A. S. Mikhailov, Europhys. Lett., 98, 64004 (2012). [19] S. Hata, H. Nakao, and A. S. References Mikhailov, Europhys. Lett., 98, 64004 (2012). [20] N. E. Kouvaris, H. Kori, and A. S. Mikhailov, PLoS ONE, 7, 45029 (2012). [20] N. E. Kouvaris, H. Kori, and A. S. Mikhailov, PLoS ONE, 7, 45029 (2012). [21] M. Wolfrum, Physica D, 241, 1351 (2012). y ( ) [22] H. G. Othmer and L. E. Scriven, J. Theor. Biol., 32, 507 (1971). 22] H. G. Othmer and L. E. Scriven, J. Theor. Bio [23] W. Horsthemke, K. Lam, and P. K. Moore, Phys. Lett. A, 328, 444 (2004) [23] W. Horsthemke, K. Lam, and P. K. Moore, Phys. Lett. A, 328, 444 (2004). 17/17
https://openalex.org/W4312235393	https://ieeexplore.ieee.org/ielx7/6287639/9668973/09919224.pdf	English	null	Corrections to “Multicriteria and Statistical Approach to Support the Outranking Analysis of the OECD Countries”	IEEE access	2,022	cc-by	457	DANIEL AUGUSTO DE MOURA PEREIRA 1, MARCOS DOS SANTOS 2, IGOR PINHEIRO DE ARAÚJO COSTA 3,4, MIGUEL ÂNGELO LELLIS MOREIRA 3,4, ADILSON VILARINHO TERRA3,4, CLAUDIO DE SOUZA ROCHA JUNIOR 4, AND CARLOS FRANCISCO SIMÕES GOMES 4 1Production Engineering Department, Federal University of Campina Grande (UFCG), Campina Grande 58428-830, Brazil 2Systems and Computing Department, Military Institute of Engineering (IME), Rio de Janeiro 22290-270, Brazil 3Naval Systems Analysis Centre, Rio de Janeiro 20180-001, Brazil 4Production Engineering Department, Federal Fluminense University, Rio de Janeiro 24220-008, Brazil Corresponding author: Igor Pinheiro De Araújo Costa (costa_igor@id.uff.br) In the above article [1], two references [2] and [3] were missing. In the above article [1], two references [2] and [3] were missing. article [1] has a similar part in statistical approach with the two references [2] and [3]. The ﬁrst sentence of Section III should read ‘‘This section aims to present an extensive methodological anal- ysis concerning [2] and [3]. Comparing [1] to [2] and [3], the study presents a multimethodological approach, integrating methodological concepts based on MCDA, through the PROMETHEE-SAPEVO-M1, ELECTRE-MOr, and VIKOR methods, building the ranking of countries from OECD.’’ [3] E. P. Eufrazio, ‘‘Proposition of a synthetic quality of life index: An appli- cation of multivariate analysis and clustering using the OECD bet- ter life index,’’ Federal Fluminense Univ., Niterói, Brazil, Tech. Rep., 2019. eived 30 September 2022, accepted 30 September 2022, date Received 30 September 2022, accepted 30 September 2022, date of current version 13 October 2022. Digital Object Identifier 10.1109/ACCESS.2022.3211691 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ VOLUME 10, 2022 REFERENCES [1] D. A. D. M. Pereira, M. D. Santos, I. P. De Araujo Costa, M. A. L. Moreira, A. V. Terra, C. D. S. R. Junior, and C. F. S. Gomes, ‘‘Multicrite- ria and statistical approach to support the outranking analysis of the OECD countries,’’ IEEE Access, vol. 10, pp. 69714–69726, 2022, doi: 10.1109/ACCESS.2022.3187001. [2] E. P. Eufrazio and H. G. Costa, ‘‘A proposed synthetic index of quality life for OECD countries,’’ in Innovation for Systems Infor- mation and Decision Meeting. Cham, Switzerland: Springer, 2020, pp. 62–76. One sentence is appended as acknowledgment: ‘‘The authors would also like to express their sincere gratitude to Prof. H. G. Costa and Prof. E. P. Eufrazio for their discussion in the statistical analysis.’’ [3] E. P. Eufrazio, ‘‘Proposition of a synthetic quality of life index: An appli- cation of multivariate analysis and clustering using the OECD bet- ter life index,’’ Federal Fluminense Univ., Niterói, Brazil, Tech. Rep., 2019. The authors requested this correction for complemen- ting missing references as well as explaining why the above is work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ 107630
https://openalex.org/W2808768240	https://ri.diva-portal.org/smash/get/diva2:1115750/FULLTEXT01	English	null	Patient-centred outcome metrology for healthcare decision-making	Journal of physics. Conference series	2,018	cc-by	4,365	Related content Related content The problem of patient-centred outcome measurement in psychiatry: why metrology hasn’t mattered and why it should S P Barbic, S J Cano and S Mathias - Metrology in Pharmaceutical Industry - A Case Study Priscila D. Yuvamoto, Ricardo K. S. Fermam and Elizabeth S. Nascimento - Forensic Metrology: Its Importance and Evolution in the United States JD Ted Vosk - The problem of patient-centred outcome measurement in psychiatry: why metrology hasn’t mattered and why it should S P Barbic, S J Cano and S Mathias - Metrology in Pharmaceutical Industry - A Case Study Priscila D. Yuvamoto, Ricardo K. S. Fermam and Elizabeth S. Nascimento - Forensic Metrology: Its Importance and Evolution in the United States JD Ted Vosk - This content was downloaded from IP address 31.208.188.118 on 20/06/2018 at 06:43 Journal of Physics: Conference Series Journal of Physics: Conference Series The problem of patient-centred outcome measurement in psychiatry: why metrology hasn’t mattered and why it should S P Barbic, S J Cano and S Mathias - To cite this article: S J Cano et al 2018 J. Phys.: Conf. Ser. 1044 012057 Forensic Metrology: Its Importance and Evolution in the United States JD Ted Vosk - View the article online for updates and enhancements. 1234567890 ‘’“” 2017 IMEKO TC1-TC7-TC13 Joint Symposium IOP Publishing IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 1 Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. Published under licence by IOP Publishing Ltd 5 To whom any correspondence should be addressed. E-mail: stefan.cano@modusoutcomes.com Abstract. Valid and precise quantification of clinical variables is essential for appropriate interpretation to inform healthcare decision making. The outcomes produced from different measurement procedures and instruments, purporting to quantify the same measurand, should be directly comparable. This ensures the appropriate application and widespread adoption of clinical research findings. Metrology provides a framework for the development of a common language of reference measurement systems, which have the potential to improve the accuracy and comparability of patients’ results. However, the practices, procedures and instruments used in social measurement are currently excluded from any formal metrological framework. In this paper, we build on previous arguments, and propose a new international body to bring together metrology, psychometrics, philosophy, and clinical management to support the global comparability and equivalence of measurement results in patient centred outcome measurement to improve healthcare. Patient-centred outcome metrology for healthcare decision- making S J Cano1,5, L R Pendrill2, S P Barbic3 and W P Fisher Jr4 S J Cano1,5, L R Pendrill2, S P Barbic3 and W P Fisher Jr4 1 Modus Outcomes, Spirella Building, Letchworth Garden City, SG6 4ET, UK 2 RI.SE, Eklandagatan 86, S-412 61, Göteborg, Sweden 3 Faculty of Medicine, Department of Occupational Science and Occupational Therapy, the University of British Columbia, Vancouver, BC V6T 2A1, Canada 4 BEAR Center, Graduate School of Education, University of California, Berkeley, CA 92720 USA S J Cano1,5, L R Pendrill2, S P Barbic3 and W P Fisher Jr4 1 Modus Outcomes, Spirella Building, Letchworth Garden City, SG6 4ET, UK 2 RI.SE, Eklandagatan 86, S-412 61, Göteborg, Sweden 3 Faculty of Medicine, Department of Occupational Science and Occupational Therapy, the University of British Columbia, Vancouver, BC V6T 2A1, Canada 4 BEAR Center, Graduate School of Education, University of California, Berkeley, CA 92720 USA 1 Modus Outcomes, Spirella Building, Letchworth Garden City, SG6 4ET, UK 2 RI.SE, Eklandagatan 86, S-412 61, Göteborg, Sweden 3 Faculty of Medicine, Department of Occupational Science and Occupational Therapy, the University of British Columbia, Vancouver, BC V6T 2A1, Canada 4 BEAR Center, Graduate School of Education, University of California, Berkeley, CA 92720 USA 2 RI.SE, Eklandagatan 86, S-412 61, Göteborg, Sweden 3 Faculty of Medicine, Department of Occupational Science and Occupational Therapy, the University of British Columbia, Vancouver, BC V6T 2A1, Canada 4 BEAR Center, Graduate School of Education, University of California, Berkeley, CA 92720 USA E-mail: stefan.cano@modusoutcomes.com 1. Introduction Previously, we put social measurement ‘on trial’, and provided two perspectives arguing why measurement in the social and in the physical sciences are incompatible and countered with two perspectives supporting compatibility [1]. We concluded that measurement, whether physical or social, should have the same definition (i.e., the ratio of two magnitudes of the same thing in which the denominator is the unit, providing for invariant comparisons) and the same broad goal (i.e., quantification of meaningful variables). The role of social measurement methods in high-stakes decision making for education, psychology, and health more generally, highlights the need for the field to advance. We offered a way forward potentially applicable to both physical and social measurement (see section 3 below). In this paper, we proffer potential next steps, focussing on the area of healthcare. We 2017 IMEKO TC1-TC7-TC13 Joint Symposium 1234567890 ‘’“” IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 propose establishing a joint international effort aimed at delivering patient-centred outcome measures (PCOMs) in support of healthcare policy decision-making. propose establishing a joint international effort aimed at delivering patient-centred outcome measures (PCOMs) in support of healthcare policy decision-making. 2. The quest for better healthcare Patient experience affects patient safety, clinical effectiveness, allocation of funding [2], and is associated with: health outcomes; adherence to therapeutic regimens; preventative care; healthcare resources; quality of care delivery; and adverse events [3]. It is unsurprising, then, that patients who are more engaged in decision-making have better outcomes [4]. Therefore, decision makers within health systems increasingly rely on timely and relevant information about patient outcomes to improve policy and practice. PCOMs are instruments (e.g., questionnaires, biometric equipment, wearables) that quantify patients’ health, health-related quality of life and other latent health constructs, such as pain, mood, and function. They can also be used to quantify patient experience and satisfaction of healthcare, which play an increasingly central role in evidence-based medicine [5]. PCOMs have the potential to ‘speak’ to patients, carers and clinicians. Used alone, or in tandem with surrogate data (e.g., analysed in the laboratory), PCOMs offer the opportunity for more meaningful and interpretable measurement of patient outcomes. This helps to improve our understanding of the natural course of disease and guiding treatment decisions. There is increasing interest in embedding PCOMs within health information systems [6]. It has been suggested that they have the potential to transform healthcare into a more patient-centred model [7]. In 2017, the Organisation for Economic Cooperation and Development (OECD) went so far as to claim that a specific type of PCOM (viz., patient reported outcome measures; PROMS) have a role in benchmarking the performance of whole healthcare systems [8]. The routine collection of PROMs has already been implemented internationally, including United Kingdom, Australia, United States of America, Sweden, Canada, and the Netherlands [8]. Thus, PCOMs have a potentially significant role in modern healthcare policy decision-making and evidence-informed practice. But are we able to deliver high-stakes measurement of this kind? Some suggest not. For example, it has recently been estimated that only 10% of existing PROMs have been developed based on patient-identified priorities [9]. In addition, reflecting upon the current OECD initiative, Angela Coulter suggests that using PCOMs for high stakes decision making may be premature [5]: “Multi-purpose applications of [PCOMs] - using them in individual clinical care and aggregating the data for performance assessment - remains largely aspirational at present...” Elizabeth Teisberg provides a potential reason as to why [10]: Elizabeth Teisberg provides a potential reason as to why [10]: 4. International co-operation, innovation and communication 4. International co operation, innovation and communication In laboratory medicine, the standardization of measurements is a high priority, with the goal of comparability of results obtained using routine procedures [16, 19]. Recommendations of the IFCC Task Force on the Impact of Laboratory Medicine on Clinical Management and Outcomes (TF-ICO) include: “Effective collaboration with clinicians, and a determination to accept patient outcome and patient experience as the primary measure of laboratory effectiveness.” [14]. The benefits of this are clear. For example, the standardization of cholesterol testing in the evaluation of cardiovascular risk has led to savings in the region of 100 million USD/year in treatment costs with a parallel significant reduction of untreated “false-negative” individuals who are at increased risk [20]. But international cooperation for developingreference measurement for PCOMs is in its infancy But, international cooperation for developing reference measurement for PCOMs is in its infancy. Its aims could include traceability, comparability, and the inclusion (and possible reduction) of measurement uncertainty. This would require the consideration of five key components: 1) key stakeholder involvement (including patients, clinicians, caregivers, policy makers); 2) a clear definition of the measurand in regards to the intended clinical use [21]; 3) a clear definition of the clinically allowable error of measurements; 4) international cooperation and consensus to navigate the complexities of the development of metrologically sound reference measurement systems; and 5) continued clinical validation of newly calibrated instruments. y International cooperation would require the involvement of international metrology treaty organizations, professional societies and federations devoted to improving measurement quality in health sciences. One of its central functions would be an online database listing: reference materials; reference methods; and reference measurement services meeting appropriate international standards. This database would allow PCOM providers to select references for calibration traceability and provide support for suppliers of these services. In addition, it would be important to consider three key components to navigate the complexity of the scientific and political landscapes. First, the development and evaluation of new PCOMs is time consuming and resource intensive. As such, we should move towards a non-competitive exercise where expertise and resources are pooled to avoid the development of multiple instruments for the same purpose, thus engendering wider reaching collaboration. Second, as our understanding of the etiology, sequalae, and natural history of many diseases expands, we should be open to innovative ways to capture impact and patient benefit. Elizabeth Teisberg provides a potential reason as to why [10]: “The quest for better health care, driven by measuring safety and quality, is well intentioned and has notable achievements. But like the Biblical story about building a better city, the measurement effort has become a cacophonous muddle that is distracting clinicians, raising the cost-of-care delivery, and not helping consumers make better health care choices…The lack of measurement standards [has resulted]...in contradictory conclusions…The problem isn’t requiring measurement; health care needs meaningful measures…But…despite their vast number, the [existing] metrics don’t measure the things that define health for patients and success for clinicians.” “The quest for better health care, driven by measuring safety and quality, is well intentioned and has notable achievements. But like the Biblical story about building a better city, the measurement effort has become a cacophonous muddle that is distracting clinicians, raising the cost-of-care delivery, and not helping consumers make better health care choices…The lack of measurement standards [has resulted]...in contradictory conclusions…The problem isn’t requiring measurement; health care needs meaningful measures…But…despite their vast number, the [existing] metrics don’t measure the things that define health for patients and success for clinicians.” The absence of standardization of PCOMs, and lack of recourse to formal metrological frameworks, can be ascribed to historical, ideological, and political differences [11]. But the deleterious sequalae of this situation is profound [12-14]. PCOMs are currently in a position not dissimilar to the measurement of temperature in the 17th Century [15]. As in other areas of clinical sciences [16], recourse to a metrological framework can help to provide a resolution. 2 1234567890 ‘’“” 2017 IMEKO TC1-TC7-TC13 Joint Symposium IOP Publishing IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 2017 IMEKO TC1-TC7-TC13 Joint Symposium 1234567890 ‘’“” 2017 IMEKO TC1-TC7-TC13 Joint Symposium IOP Publishing IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 1234567890 ‘’“” IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 3. The promise of the synthesis of metrology, psychometrics and philosophy To ensure decision makers within health systems have access to the necessary objective evide need the same kind of quality-assurance of PCOMs as is established in physics and engineeri q y p y g g [ ] This translates into assurance with objective metrological comparability (‘traceability’) and declared measurement uncertainty. A possible way forward involves the synthesis of metrology, psychometrics and philosophy [1]. Essentially, evidence-based measurement for evidence-based clinical management would involve: • extending traditional metrological concepts via an operational model of a specific measurement system, in which the output of the instrument (which could be human) in response to probing the object (‘entity’) is a performance metric, i.e., how well the set-up performs an assessment [17]; • exploiting the unique properties of Rasch Measurement Theory [18], which enables key metrological components (e.g., references for traceability and means of evaluating measurement uncertainty) to be established; • exploiting the unique properties of Rasch Measurement Theory [18], which enables key metrological components (e.g., references for traceability and means of evaluating measurement uncertainty) to be established; • establishing metrological references founded on objective and clinically meaningful measurement against the backdrop of the specific challenges of social measurement methods. 5. Towards a common language for patient-centred outcome metrology 5. Towards a common language for patient-centred outcome metrology The comparability of findings among measurement procedures and instruments will result in harmonization; a common language to enable interchangeability geographically and temporally. However, as described above, the situation is complex. For example, as Angela Coulter [5] flags, different contexts will require different information needs. It is unlikely, that these needs are amenable to perfect alignment. But with several individual and separate global initiatives currently happening (e.g., ICHOM, COMET [22-25]) this will require more coherent coordination and alignment than is currently the case. There are two components to this situation. First, universality is not an a priori fact of objective existence for some measured constructs. This is typically assumed for physical variables, but not for others, such as psychological and social variables. However, contrary to common assumptions, universality is not an inherent property, even in the domains of physical variables, apart from language and culture. ‘Things’ in themselves do not retain constant and invariant forms over time, space, and levels of information and organizational complexity. Instead, universality is a social construct that emerges as a product of a particular way of organizing experience. It arises via complex processes in which signal-noise ratios are lower than would be expected if physical variables were objectively constituted in and of themselves [26]. The counter-intuitive conundrum is that: “In order for a statistical logistics to enhance precise decision making, it has to incorporate imprecision; in order to be universal, it has to carefully select its locales.” [26], p. 55 This kind of psychosocial stochastic resonance (interdependency of noise and signal) has been posited in other studies of scientific standards and universals [27], pp. 843-844; [28], pp. 69-71. Second, the incorporation of standards in information infrastructures must be sensitive to the discontinuous shifts in context that occur across applications: “With the rise of decentralized technologies used across wide geographical distance, both the need for common standards and the need for situated, tailorable and flexible technologies grow stronger. A lowest common denominator will not solve the demand for customized possibilities; neither will rigid standards resolve the issue.” [29], p.111 In relation to PCOMs, the general assumption is that common standards require common content; the same questions asked of every patient in every health care context and no matter what the individual patient's needs and conditions might be. 4. International co-operation, innovation and communication One size does not fit all, and the most appropriate fit-for-purpose instrumentation should be established early on (e.g., de novo instruments, adaptation of existing instruments, exploitation of information technology resources [e.g., computer adaptive testing], or mobile health technologies). Third, open communication is key to establish wider dissemination of important patient research, further avoiding the duplication of efforts, and involving a wider range of audiences. Ultimately, PCOMs will lead to 3 2017 IMEKO TC1-TC7-TC13 Joint Symposium 1234567890 ‘’“” IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 patient-based evidence and the realization of patient-centered care. Patients should always be involved as key stakeholders in the process of identifying and interpreting meaningful clinical outcomes that resonate with their experience, preferences, expectations and values. patient-based evidence and the realization of patient-centered care. Patients should always be involved as key stakeholders in the process of identifying and interpreting meaningful clinical outcomes that resonate with their experience, preferences, expectations and values. 5. Towards a common language for patient-centred outcome metrology indicates your health is good”) are not automatically commensurable with metalinguistic statements (e.g., “your patients’ PCOM survey results indicate your care is producing satisfactory outcomes relative to the outcomes obtained by your peers”) and neither are automatically commensurable with metacommunicative statements (e.g., “if I adopt that new treatment modality perhaps my patients’ PCOMs will be superior to my colleagues’ PCOMs”). An individual’s array of clinical indicators and responses to a health status survey may correspond well with expectations formed regarding the natural history of a chronic disease or the developmental effects of aging and growth. The information relevant and applicable at that level, however, is not the same as the information needed at the metalinguistic or metacommunicative levels. The consequences of not attending to the specific demands that must be met to coordinate common standards for expressing the information needed across these levels is well stated by Star and Ruhleder: “If we, in large-scale information systems implementation, design messaging systems blind to the discontinuous nature of the different levels of context, we end up with organizations which are split and confused, systems which are unused or circumvented, and a set of circumstances of our own creation which more deeply impress disparities on the organizational landscape.” [29], p. 118 “If we, in large-scale information systems implementation, design messaging systems blind to the discontinuous nature of the different levels of context, we end up with organizations which are split and confused, systems which are unused or circumvented, and a set of circumstances of our own creation which more deeply impress disparities on the organizational landscape.” [29], p. 118 Suggestions for coherently coordinating information across these levels are offered elsewhere [32, 33]. But, in the end, a common language for patient-centered outcome metrology would significantly contribute to improvements in patient-centered care by allowing results of clinical studies undertaken in different locations or times to be universally applied. The ‘semantic interoperability’ has to be good i.e. the “the ability to ensure that the precise meaning of exchanged information is unambiguously interpretable by any other system, service or user”, where we borrow the term from the context of digital solutions that enable public administrations, businesses and citizens to benefit from interoperable cross-border and cross-sector public services [34]. 5. Towards a common language for patient-centred outcome metrology The unworkable rigidity of this approach has prevented the development of information infrastructure standards in health care that would be of great value in patient care, quality improvement initiatives, and accountability. With the increasing availability of computerized adaptive administrations of PCOM items from calibrated banks [30], this lowest common denominator approach is being supplanted by methods better able to meet the demand for customized information relevant to the needs of individual patients. p But this shift from fixed forms to dynamically configured PCOMs is insufficient for the need of a health care information infrastructure integrating local sensitivities with universal standards across applications. In the same way public health has adopted an ecological frame of reference distinguishing micro-, meso-, and macro-levels of complexity in its policy and practice implications [31], so too, must PCOMs be used to similarly attend to the relational properties obtained in applications requiring different forms of information and organization [29]. For example, spanning the gulf of complexity from individual interpretation of PCOMs (micro-), aggregating outcomes for quality improvement purposes (meso-), and producing theories of health status and the impact of care (macro-). Each level demands a learning and subsequent relearning about patients' health status. g q g p In other words, denotative statements of fact at the local, individual level (e.g., “your PCO 4 2017 IMEKO TC1-TC7-TC13 Joint Symposium 1234567890 ‘’“” 2017 IMEKO TC1-TC7-TC13 Joint Symposium IOP Publishing IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 1234567890 ‘’“” 2017 IMEKO TC1-TC7-TC13 Joint Symposium IOP Publishing IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 1234567890 ‘’“” IOP Conf. Series: Journal of Physics: Conf. Series 1044 (2018) 012057 doi :10.1088/1742-6596/1044/1/012057 indicates your health is good”) are not automatically commensurable with metalinguistic statements (e.g., “your patients’ PCOM survey results indicate your care is producing satisfactory outcomes relative to the outcomes obtained by your peers”) and neither are automatically commensurable with metacommunicative statements (e.g., “if I adopt that new treatment modality perhaps my patients’ PCOMs will be superior to my colleagues’ PCOMs”). An individual’s array of clinical indicators and responses to a health status survey may correspond well with expectations formed regarding the natural history of a chronic disease or the developmental effects of aging and growth. The information relevant and applicable at that level, however, is not the same as the information needed at the metalinguistic or metacommunicative levels. 5. Towards a common language for patient-centred outcome metrology It would also support an effective application of evidence-based medicine and use of guidelines (established by scientific or professional bodies) to define specific decision limits for diagnosis and therapeutic interventions. Thus, a joint international effort aimed at delivering PCOMs in support of modern healthcare policy decision-making could lead to the appropriate embedding of PCOMs in health systems. This would allow us to evaluate the effectiveness and cost-effectiveness of major policy initiatives, to identify conditions or interventions where substantial variation in patient outcome exists across providers, and improve the choice of treatment plans for individual patients. References [1] Cano S, Vosk T, Pendrill L and Stenner A J 2016 J. Phys. Conf. Series 772 012025 [2] Basch E 2017 N. Engl. J. Med. 376 105-8 [3] Doyle C, Lennox L and Bell D 2013 BMJ Open 3 e001570 [4] Greenfield S, Kaplan S and Ware J 1985 Ann. Int. Med. 102 520-8 [5] Coulter A 2017 BMJ 356 j816 [6] Devlin N and Appleby J 2010 Getting the most out of PROMs: putting health outcomes at the heart of NHS decision-making (London, UK: Kings Fund) [7] Black N 2013 BMJ 346 f167 [8] OECD 2017 Recommendations to OECD Ministers of Health from the high-level reflection group on the future of health statistics (Paris: OECD) [9] Wiering B, de Boer D and Delnoij D 2017 Health Expectations 20 11-23 [10] Teisberg E 2015 The Quality Tower Of Babel http://healthaffairs.org/blog [1] Cano S, Vosk T, Pendrill L and Stenner A J 2016 J. Phys. Conf. Series 772 012025 [2] Basch E 2017 N. Engl. J. Med. 376 105-8 [3] Doyle C, Lennox L and Bell D 2013 BMJ Open 3 e001570 [4] Greenfield S, Kaplan S and Ware J 1985 Ann. Int. Med. 102 520-8 [5] Coulter A 2017 BMJ 356 j816 [6] Devlin N and Appleby J 2010 Getting the most out of PROMs: putting health outcomes at the heart of NHS decision-making (London, UK: Kings Fund) [7] Black N 2013 BMJ 346 f167 [8] OECD 2017 Recommendations to OECD Ministers of Health from the high-level reflection group on the future of health statistics (Paris: OECD) [8] OECD 2017 Recommendations to OECD Ministers of Health from the high-level reflection group on the future of health statistics (Paris: OECD) ] Wiering B, de Boer D and Delnoij D 2017 Health Expectations 20 11-23 [9] Wiering B, de Boer D and Delnoij D 2017 Health Expectations 20 11-23 [10] Teisberg E 2015 The Quality Tower Of Babel http://healthaffairs.org/blog 5 2017 IMEKO TC1-TC7-TC13 Joint Symposium [11] Cano S and Hobart J 2011 Pat. Pref. Adherence 5 279-90 [12] Hunt S 1997 Qual. Life Res. 6 205-12 [12] Hunt S 1997 Qual. Life Res. 6 205-12 [13] Boyce M and Browne J 2013 Qual. Life Res. 22 2265-78 [13] Boyce M and Browne J 2013 Qual. Life Res. 22 2265-78 [14] Hobart J, Cano S, Zajicek J and Thompson A 2007 Lancet Neurol. 6 1094-105 [15] Chang H 2007 Inventing Temperature: Measurement and Scientific Progress (Oxford: Oxford University Press) [16] Jones G and Jackson C 2016 Clin Chim Acta 453 86-94 [17] Pendrill L 2014 NCSLi Measure: The Journal of Measurement Science 9 22-33; Pendrill L R and Petersson N 2016 Measurement Science & Technology, 27, 094003 [18] Andrich D 2017 J. Phys. Conf. Series in press [18] Andrich D 2017 J. Phys. Conf. Series in press [19] Hallworth M, et al 2015 Med Clin Chem 61 589–99 [20] Panteghini M 2007 Clin. Biochem. Rev. 28 97-104 [21] FDA 2013 Roadmap to Patient-focused Outcome Measurement in Clinical Trials http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/DrugDevelopmentToolsQ ualificationProgram/UCM370174.pdf [22] Gargon E et al 2015 Trials 18 54 [23] International Consortium for Health Outcomes Measurement (ICHOM) 2017. ICHOM and OECD sign Letter of Intent to collaborate http://www.ichom.org/news/ichom-and-oecd-sign- letter-of-intent-to-collaborate-on-the-collection-analysis-and-publishing-of-patient-reported- outcomes/. [24] Williamson P R, et al 2012 Trials 13 132 [25] Porter M E, Larsson S and Lee T H 2016 N. Engl. J. Med. 374 504-6 [26] Berg M and Timmermans S 2000 Configurations 8 31-61 [26] Berg M and Timmermans S 2000 Configurations 8 31-61 [27] Galison P 1997 Image and logic (Chicago: University of Chicago Press) [28] Fisher W P and Wilson M 2015 Pensamiento Educativo: Revista de Investigacion Educacional Latinoamericana 52 55-78 [29] Star S and Ruhleder K 1996 Info. Sys. Res. 7 111-34 [29] Star S and Ruhleder K 1996 Info. Sys. Res. 7 111-34 [30] Cook K, O'Malley K and Roddey T 2005 Health Serv. Res. 40 1694-711 [31] Susser M and Susser E 1996 Am. J. Public Health 86 674-7 [32] Fisher W P, Oon E and Benson S 2017 J. Phys. Conf. Series in press [33] Fisher W P and Stenner A J 2017 J. Phys. Conf. Series in press [34] ISA 2017 Interoperability Solution for European Public Administrations (EU Commission: https://ec.europa.eu/isa2/isa2_en) [34] ISA 2017 Interoperability Solution for European Public Administrations (EU Commission: https://ec.europa.eu/isa2/isa2_en) 6 6
https://openalex.org/W3165518535	https://www.researchsquare.com/article/rs-92597/latest.pdf	English	null	Influence of grass lawns on the summer thermal environment and microclimate of heritage sites: a case study of Fuling mausoleum, China	Heritage science	2,021	cc-by	8,583	In§uence of Lawns on the Summer Thermal Environment and Microclimate of Heritage Sites: A Case Study of Fuling Mausoleum, China Xiaoyu Wang (  704216608@qq.com ) oyu Wang (  704216608@qq.com ) enyang University of Technology https://orcid.org/0000-0003-3223-6811 Keywords: Thermal environment; Microclimate; Fuling mausoleum; Heritage site; CFD simulation Keywords: Thermal environment; Microclimate; Fuling mausoleum; Heritage site; CFD simulation Research article Keywords: Thermal environment, Microclimate, Fuling mausoleum, Heritage site, CFD simulation Posted Date: October 27th, 2020 DOI: https://doi.org/10.21203/rs.3.rs-92597/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published on January 13th, 2021. See the published version at https://doi.org/10.1186/s40494-020-00479-9. Influence of lawns on the summer thermal environment and microclimate of heritage sites: A case study of Fuling mausoleum, China Xiaoyu Wang1, Peng Liu1, Gongwen Xu1 1 School of Architecture & Civil Engineering, Shenyang University of Technology, Shenyang 110870, China Corresponding author. E-mail address: 704216608@qq.com (X.Y. Wang) # Equal Contribution Abstract: The thermal environment and microclimate of heritage sites has been severely impacted by rapid urbanization. This study collected various meteorological measurement data as a reference for computational fluid dynamics (CFD) simulation settings. Then CFD was applied to simulate the impact of lawns on the thermal environment and microclimate of Fuling Mausoleum. We found that lawns and soil can cool the air through evaporation, and thus have a specific cooling effect on the bricked ground. After lawns were planted, the bricked ground temperature decreased by 1.56– 17.54°C than that before lawns were planted at 14:00, a decrease of 2.68%–24.20%. Under normal circumstances, when the wind speed or relative humidity increased, the ground temperature dropped. Greenbelt vegetation can adjust the microclimate and human thermal comfort indicators. The consistency of the difference between the actual measurement and the CFD simulation results shows that CFD simulation can thus accurately reflect the internal temperature field distribution if the selection of simulation parameters is reasonable. Theoretical calculation and analysis, experimental measurement research, and modern computer simulation analysis methods applied together constitute a complete system for studying modern physical environmental problems and can provide reliable and economic results. 1. Introduction As a result of the rapid expansion of urban areas, heritage sites that were initially located in the suburbs or rural areas of many cities have gradually become important inner city areas. The phenomenon of urbanization has caused severe impacts and changes to the natural environment, and the environmental landscape is also deteriorating. The natural vegetation cover has been replaced by buildings and urban infrastructure, which results in the well-known urban heat island effect [1]. Several adaptation measures have been put into practice for their potential to affect the urban microclimate, mainly to reduce high temperatures [2,3,4,5,6]. The greening and vegetation of urban areas are the most investigated adaptation approaches as they can reduce the surface temperature of the surrounding environment [3,7,8]. The ambient temperature can also be decreased by evaporation and transpiration of plants. The main methods to study the effectiveness of vegetation in mitigating the heat island effect are field measurements [7,9,10], ground control experiments [11], and computational fluid dynamics (CFD) simulation [12,13,14,15,16,17,18]. Of these, CFD evaluates the effectiveness of vegetation coverage in detail on the outdoor urban microclimate, which is an economical and practical method of achieving accurate predictions. Hong and Lin [19] used the outdoor thermal simulation platform to discover the influence of the green layout on the outdoor environment and pedestrian thermal comfort. Bruce and Fleer [12] used CFD and ENVI-met models to simulate the microclimate and studied the impact of ground greening on the thermal stress coverage of small parks. Dimoudi and Nikolopoulou [13] used CFD model to compare the cooling effect of different methods and the vegetation arrangement in general building facilities. Li and Yu [20] used CFD to simulate the thermal environment and residential microclimate around a single building. They verified their simulation by comparing between the measured data with simulated temperature of the outdoor thermal environment around buildings. Bowler et al. [7] conducted an empirical study on the effect of urban greening for cooling effect. The main emphasis of previous research is on the evaluation of the indoor microclimate of heritage buildings and how to improve the comfort of users. Corgnati et al. [21] also described a methodology for evaluating the microclimate in museums between traditional convenience and thermal comfort. According to European standards such as EN 15251 [22], La Gennusa et al. [23] analyzed the indoor climate and indoor environment in historical Italian buildings, including temperature and humidity. 1. Introduction CFD technology involves the solving of various conservation control partial differential equations describing fluid flow, heat transfer, and mass transfer and the visualization of the different solved flow or heat transfer phenomena. CFD technology, theoretical analysis and experimental research have become three essential fluid mechanics research methods. These three methods complement each other [24,25]. To our knowledge, few studies focus on the outdoor thermal environment and microclimate for heritage sites. Because of urbanization around heritage sites, the summer temperatures are increasing. This study focused on the cooling effect of grassland and vegetation on the environment in summer. On the basis of actual measurement, the CFD simulation experiment was applied to analyze the cooling effect of the vegetation. First, infrared images and precise measurements are taken with an infrared thermal imaging camera to collect various physical data in meteorology as a reference for CFD simulation settings. Then, CFD was used to simulate the impact of lawns on the thermal environment and microclimate of the heritage site. The research includes: 1) Analysis and comparison of the actual measurement with simulation results; 2) Analysis of the simulation results before and after lawns were planted; 3) Analysis of the simulation results at different wind speeds and relative humidity; and 4) Analysis and comparison of the actual measurement of microclimate with the human thermal comfort index value. 2.1 Study site Shenyang is located in central Liaoning Province, southern northeast China (41°12N–43°02N, 122°25E–123°48′E). Fuling Mausoleum is located in the eastern suburbs of Shenyang. It is the tomb of the Qing patriarch Nurhachi and his queen Yehenala and is one of the three mausoleums of Shengjing. Fuling Mausoleum was built in 1629 and was completed in 1651. After a great deal of construction works during the Shunzhi, Kangxi, and Qianlong periods of the Qing Dynasty, a large-scale and well-equipped ancient emperor tomb complex was formed. When it was first built, it was called “Xianhan Mausoleum” or “Taizu Mausoleum”. In the first year of Chongde (1636), it was named “Fuling Mausoleum”, which implies the long-lasting fortune of the Qing Dynasty [26]. In 2004, The Three Mausoleums of Shengjing (referring to Fuling, Zhaoling, and Yongling Mausoleums), were included on the UNESCO World Heritage Site List [27]. Fig.1 The environment surrounding Fuling Mausoleum (white is the CFD simulation model). Fig.1 The environment surrounding Fuling Mausoleum (white is the CFD simulation model). Fuling Mausoleum has a detailed layout and distinct levels and has a total area of approximately 194,800m2. It is shaped like an inner city with an outer wall. It comprises three parts—the front yard, the square city and the treasure city—and it gradually increases in elevation from south to north. The mausoleum building complex remains relatively intact [28]. There are 32 existing ancient buildings, which are symmetrically distributed with the Shinto as the central axis. It is a complex of imperial tombs that integrates the characteristics of the Manchu and Han ethnicities, which differs from the tombs of the Ming Dynasty and those built after the Qing Dynasty [26]. This article focuses on the three main buildings of Fuling Mausoleum: Stele Pavilion, Long’en Gate and Long’en Hall (Fig. 1). 2.2 On-site measurements July is the hottest month of the year in Shenyang. The daily average temperature is between 28 and 29°C, and the weather is usually in the range 24–32°C. The maximum ground temperature can reach 63°C. In July, the average wind speed is 3.64–3.86 m/s and the wind direction is mainly southerly [29]. On July 12, 2020, when field measurements were taken at Fuling Mausoleum, the weather was clear and the wind speed was relatively low. To study the impact of lawns on the summer thermal environment and microclimate of Fuling Mausoleum, the project team members selected five-time points—10:00, 11:00, 12:00, 13:00, and 14:00—and conducted the field measurements at these times. The measuring instruments (Table 1, Fig.2) and measurement groups taken were as follows: Table 1 Measuring instruments and measuring parameters Measuring instruments Measuring parameters Instrument accuracy Black bulb thermometer Temperature and humidity tester WBGT Tester Air temperature ±0.1℃ Black bulb temperature ±0.1℃ WBGT Heat index ±0.1℃ Relative humidity ±5％RH Anemometer Wind speed ±3％或±0.1dgt Infrared thermal imaging thermometer Object temperature imaging ±2％或±2℃ Heat flux sensor Soil heat flux density、Heat flux density of bricked ground 20-100 ㎶·w-1·m2 Table 1 Measuring instruments and measuring parameters Table 1 Fig. 2 Measuring instruments: (a) black bulb thermometer, (b) anemometer, (c) heat flux sensor (photos by Xiaoyu Wang). Fig. 2 Measuring instruments: (a) black bulb thermometer, (b) anemometer, (c) heat flux sensor (photos by Xiaoyu Wang). (1) Outdoor thermal environment. The primary measurement data were air temperature, black bulb temperature, relative humidity, wind speed, and the wet bulb globe temperature (WBGT) value. Measurement were taken 1.6 m from the ground in the open area in front of Long’en Gate to obtain biological data to describe the environment, and to measure the thermal environment and microclimate physical data of the surrounding lawns and the bricked ground. (2) Infrared thermal imaging. The primary measurement contents were the real- time temperature of the lawns and the bricked ground. The measurement locations were selected in front of the Stele Pavilion, Long’en Gate, and Long’en Hall and included the bricked ground and lawns to the left and right of these structures. Infrared thermal imaging photos were taken to determine the real-time temperature at each measurement point. (3) Surface heat flux measurement. The primary measurement contents were the heat flux data for the bricked ground and the green soil. The heat flow measurement points were placed on the floor of the bricked ground and at 10 cm soil depth in the lawns to measure the heat flux data of the different ground surfaces. The surface of the lawns was in an endothermic state and the measured heat flux was between −16.9 and −16.8 W/m2, whereas the bricked ground was in an exothermic state, and its heat flux was between 90.7 and 137.9 W/m2 (Table 2). Table 2 Actual measurement data Time (Hour) Air temperature (℃) Wind speed (m/s) Relative humidity(%) Heat flux of lawns (W/ m2) Heat flux of bricked ground (W/ m2) 10 26.1 3.6 68 -16 90.7 12 31.4 4 46.9 -16.8 108.9 14 33.7 5 40.1 -16.8 137.9 Table 2 2.3 CFD approach (2) Mesh settings The mesh system setting was a numerical calculation method based on the finite volume method. The calculation area was divided into the appropriate mesh, and calculations were performed to determine the mass, momentum, and energy conditions in each mesh. In this study, the mesh was generated using ANSYS-ICEM CFD software using tetrahedral elements in the fluid domain. The mesh contains between 2.35 and 2.4 million tetrahedral elements (Fig. 3). Fig.3 (a) General views of the computational domain (geometry and mesh) for Fuling Mausoleum. (b) Detail of the mesh refinement with tetrahedral elements. Fig.3 (a) General views of the computational domain (geometry and mesh) for Fuling Mausoleum. (b) Detail of the mesh refinement with tetrahedral elements. 2.3 CFD approach Based on the actual measurement and theoretical research, the heat flux determined using ANSYS/Fluent was used to set two types of lawns and bricked ground to simulate the thermal environment and microclimate [30]. The heat flux of lawns and the bricked ground was measured as a reference. The grass material was found to absorb heat, thus the heat flux was set to a negative value for these areas. Conversely, the brick material emit heat, thus the heat flux was set to a positive value for these areas. (1) Calculation area and boundary condition settings (1) Calculation area and boundary condition settings The calculation area was set to a rectangular parallelepiped area (1200 m long, 400 m wide, and 70 m high). The calculation area was mainly based on the ancient buildings from Stele Pavilion to Ming building and the surrounding environment. The inlet of the simulation calculation range was set to the velocity inlet, the outlet was set to the pressure outlet, the boundary condition of the ground was set to the wall, and the external calculation domain was set to the symmetry to assume actual atmospheric environmental conditions. (3) RNG k–ε Model We used the RNG k–ε model for viscous modeling. The RNG k–ε model was proposed by Orszag and Yakhot [31]. Although similar in form to the standard k–ε model, the RNG k–ε model is more accurate and reliable for a wider class of flows [32]. The turbulence kinetic energy, and its rate of dissipation, ε, are obtained from the following transport equations [30]:   ( ) i k eff k b M k i j j k k ku G G Y S t x x x                            (1) (1) and     2 1 3 2 ( ) i eff k b i j j u C G C G C R S t x x x k k                                     (2) (2) where Gk and Gb represent the turbulence kinetic energy generated from the mean velocity gradient and buoyancy, respectively; YM represents the contribution of the fluctuating dilatation in compressible turbulence to the overall dissipation rate; αk and αε denote the inverse effective Prandtl numbers for k and ε, respectively; and Sk and Sε are user-defined source terms. (4) Wind profile (4) Wind profile Regarding the wind boundary layer over the terrain, the velocity profile specifying the inlet boundary condition for each simulation was defined according to this well- established equation [33], 10 10 p n z z U U        (3) (3) Where Uz denotes the air velocity which depends on height z; U10 is the air velocity at the height of 10 m (assumed to be 4 or 10 m/s); and np is the Hellman exponent, which depends on the atmospheric stability and the nature of the terrain. np near the suburbs of the city is assumed to be 0.22 [34]. (5) Settings of the numerical simulation scheme The real values collected according to actual measurement served as the basis for the verification of the numerical simulations. The impact of lawns and bricked ground on the thermal environment of ancient buildings was considered through the relative evaluation of the various physical values. The basic parameters of each material are shown in Table 3 [35]. The settings of the specific numerical simulation scheme are shown in Fig. 4. Table 3 Basic parameters of each material Material Density (kg·m-3) Specific heat capacity (J·kg-1·k-1) Thermal Conductivity (W·m-1·k-1) Absorption coefficient Refractive coefficient Bricked ground 2344 750 1.00 0.87 0.90 Wall 1800 840 1.00 0.73 0.92 Lawns 700 2310 1.20 0.00 0.20 Table 3 Basic parameters of each material Table 3 Figure 4. Settings of the specific numerical simulation scheme. Figure 4. Settings of the specific numerical simulation scheme. 3. Results and discussion The heat exchange and distribution in the thermal environment of this site constitute a complex system, including various heat transfer processes such as convection, radiation, and heat conduction among mountains, rivers, air, and the earth. We focus on the influence of lawns on the summer thermal environment and microclimate of the architectural heritage site, which is mainly reflected in how wind Fig.5 Photo and infrared images of the site in front of Stele Pavilion: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Fig.5 Photo and infrared images of the site in front of Stele Pavilion: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). speed, temperature, and humidity affect the cooling effect of vegetation. 3.1 Analysis of the actual measurement The project team took photos and infrared images of the bricked ground and lawns to the left and right in front of Stele Pavilion, Long’en Gate, and Long’en Hall at the five-time points and obtained the data for each measuring point. To accommodate space limitations, we focused on the most explicit infrared image at 14:00 to study the role of local lawns on the thermal environment and microclimate regulation. Figs. 5–7 present photos and infrared images of the bricked ground in front of Stele Pavilion, Long’en Gate, and Long’en Hall and lawns to the left and right of these sites taken at 14:00 using TESTO thermal imaging camera. The air temperature is affected by solar radiation and ground radiation. The air temperature peaked at around 14:00. After the ground absorbed solar radiation, the temperature also increased. The minimum temperature of green soil was still low, which indicates that the green soil could cool air through evaporation and cause a relatively low surface temperature. However, at the same time, the massive radiation absorption coefficient of the soil allowed it to absorb more heat radiation, which resulted in higher local heat production. Fig.6 Photo and infrared images of the site in front of Long’en Gate: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Fig.6 Photo and infrared images of the site in front of Long’en Gate: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Fig.7 Photo and infrared images of the site in front of Long’en Hall: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Fig.7 Photo and infrared images of the site in front of Long’en Hall: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). As the temperature increased, the temperature change increased. The bricked ground did not have the evapotranspiration effect of green soil. The coefficient of absorbing radiation was small; therefore, the minimum and maximum temperature did not change as much as the soil of lawns. 3.1 Analysis of the actual measurement However, the greening of the surrounding environment affected on the bricked ground and had a specific cooling effect. As the temperature increased, the temperature change increased. The bricked ground did not have the evapotranspiration effect of green soil. The coefficient of absorbing radiation was small; therefore, the minimum and maximum temperature did not change as much as the soil of lawns. However, the greening of the surrounding environment affected on the bricked ground and had a specific cooling effect. As shown in Figs.5–7, the lowest and highest temperatures of lawns to the left in front of Stele Pavilion were 29.7 and 64.3°C, respectively, those of lawns to the right were 33.1 and 63.2°C; and of the bricked ground were 45.4 and 61.8°C. The lowest and highest temperatures of lawns to the left in front of Long’en Gate were 31.3 and 63.0°C, respectively, those of lawns to the right were 33.4 and 72.7°C; and of the bricked ground were 45.4 and 63.6°C. The lowest and highest temperatures of lawns to the left in front of Long’en Hall were 39.8 and 59.7°C, respectively, those of lawns to the right were 33.4 and 62.0°C; and of the bricked ground were 46.6 and 66.4°C. 3.2 Analysis and comparison of the actual measurement with simulation results The following relative error formula was used to assess the validity of the calculation results [36]: = 100% L   (4) (4) where  is the relative error between the measured value and the simulated value at a given time,  is the absolute error at that time (the absolute value of the difference between the measured value and the simulated value), and L is the measured value at the time (°C). The error results of the measured and simulated temperature values of each measuring point on a typical day in each season are given in Table 4. Table 4 Relative error of the measured and simulated temperature values [31] Time (Hour) Relative error  ratio in each interval (%) 0%≤≤5% 5%＜≤10% 10%＜≤15% ＞15% 10 26.7% 33.3% 33.3% 6.7% 12 40% 30% 26.7% 3.3% 14 53.3% 36.7% 6.7% 3.3% Table 4 Relative error of the measured and simulated temperature values [31] The consistency of the difference between the actual measurement and the CFD simulation results indicates that CFD could accurately reflect the internal temperature field distribution if the selection of simulation parameters was reasonable (Figs.8-10). 3.1 Analysis of the actual measurement In particular, the difference between the actual measurement and the simulation was smallest at 14:00 and the simulated data were the closest to the actual measured data (see Table 4 and Fig. 10). Fig.8 Temperature cloud diagrams and measurement point locations of the numerical simulation at 10:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 10:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig.8 Temperature cloud diagrams and measurem 10:00 (a) after lawns were planted and (b) before measurement with simulation results in front Pavilion, (d) in front of Long’en Ga Fig.8 Temperature cloud diagrams and measurement point locations of the numerical simulation at 10:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 10:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig.8 Temperature cloud diagrams and measurement point locations of the numerical simulation at 10:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 10:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig. 8(a) and (b) show the temperature cloud diagrams and the measurement point locations of the numerical simulation in front of Stele Pavilion, Long’en Gate, and Long’en Hall at 10:00. Fig. 8(c)-(e) show the temperature comparison between the measured and simulated thermal environments in front of Stele Pavilion, Long’en Gate, and Long’en Hall at 10:00. Comparing the measured and simulated temperature curves, the difference between the measured and simulated values at 10 points was 0.15–7.16°C and the proportion was 0.51%–18.8%. Fig. 9(a)-(b) show the temperature cloud diagrams and the measurement point locations of the numerical simulation in front of Stele Pavilion, Long’en Gate, and Long’en Hall at 12:00. Fig. 9(c)-(e) show the temperature comparison between the measured and simulated thermal environments in front of Stele Pavilion, Long’en Gate, and Long’en Hall at 12:00. Comparing the Fig.9 Temperature cloud diagrams and measurement point locations of the numerical simulation at 12:00 (a) after lawns were planted and (b) before lawns were planted. 3.3 Analysis of the simulation results before and after lawns were planted Typically, after lawns were planted, the temperature of the bricked ground in Fuling Mausoleum was lower than that before lawns were planted at 10:00, 12:00 and 14:00. Comparison of the simulated temperature curves before and after lawns were planted showed that the temperature of the bricked ground in front of Stele Pavilion at 10:00 decreased by 5.30–6.67°C after lawns were planted, a decrease of 9.17%–11.24% (Fig. 8). The temperature of the bricked ground in front of Long’en Gate decreased by 6.0°C after lawns were planted, a decrease of 11.85%. The bricked ground temperature in front of Long’en Hall decreased by 9.77–18.20°C after lawns were planted, a decrease of 16.86%–26.45%. Comparison of the simulated temperature curves before and after lawns were planted showed that the temperature of the bricked ground in front of Stele Pavilion at 12:00 decreased by 2.62–4.19°C after lawns were planted, a decrease of 4.08%–6.40% (Fig. 9). The temperature of the bricked ground in front of Long’en Gate decreased by 0.02–1.55°C after lawns were planted, a decrease of 0.04%–2.57%. The bricked ground temperature in front of Long’en Hall decreased by 0.90–3.38°C after lawns were planted, a decrease of 1.60%–5.93%. Comparison of the simulated temperature curves before and after lawns were planted showed that the temperature of the bricked ground in front of Stele Pavilion at 14:00 decreased by 2.69– 4.26°C after lawns were planted, a decrease of 3.91%–6.09% (Fig. 10). The temperature of the bricked ground in front of Long’en Gate decreased by1.56–9.40°C after lawns were planted, a decrease of 2.68%–14.75%. The bricked ground temperature in front of Long’en Hall decreased by 11.47–17.54°C after lawns were planted, a decrease of 17.56%–24.20%. The temperature of some measured points had increased because of the effect of airflow vortexes. In air vortexes, the heat could not be taken away in time, which had increased the bricked ground temperature. 3.1 Analysis of the actual measurement (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 12:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig.9 Temperature cloud diagrams and measurement point locations of the numerical simulation at 12:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 12:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig.10 Temperature cloud diagrams and measurement point locations of the numerical simulation at 14:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 14:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig.10 Temperature cloud diagrams and measurement point locations of the numerical simulation at 14:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 14:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Fig.10 Temperature cloud diagrams and measurement point locations of the numerical simulation at 14:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 14:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. measured and simulated temperature curves, the difference between the measured and simulated values at 10 points was 0.22-8.74°C and the proportion was 0.48-14.2%. Fig. 10(a)-(b) show the temperature cloud diagrams and the measurement point locations of the numerical simulation in front of Stele Pavilion, Long’en Gate, and Long’en Hall at 14:00. Fig. 10(c)-(e) show the temperature comparison between the measured and simulated thermal environments in front of Stele Pavilion, Long’en Gate, and Long’en Hall at 14:00. Comparing the measured and simulated temperature curves, the difference between the measured and simulated values at 10 points was 0.02-7.55°C and the proportion was 0.04-16.6%. 3.4 Comparison and analysis of different wind speeds and relative humidity According to the comparison of actual measurement and simulation data, the difference between the actual measurement and the simulation at Fuling Mausoleum was smallest at 14:00 and the simulation data were the closest to the actual measurement data. In this study, the simulation parameters of wind speed and relative humidity at 14:00 were changed to study the influence of different results on the thermal environment and microclimate. Fig. 11(a)–(c) present the wind speed cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at 10:00, 12:00, and 14:00. Fig. 11(d)–(f) present the relative humidity cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at 10:00, 12:00, and 14:00. Fig.11 (a-c) Wind speed cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (b) 12:00, and (c) 14:00. (d-f) Relative humidity cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (e) 12:00, and (f) 14:00. Fig.11 (a-c) Wind speed cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (b) 12:00, and (c) 14:00. (d-f) Relative humidity cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (e) 12:00, and (f) 14:00. Fig.12 Comparison of simulation results at different wind speeds at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Fig.12 Comparison of simulation results at different wind speeds at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Fig.13 Comparison of simulation results at different relative humidity at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Fig.13 Comparison of simulation results at different relative humidity at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Typically, as the wind speed increased, the ground temperature dropped. If the building was blocked by other walls, some of the measured points may be affected by airflow vortexes. Fig. 12(a)–(c) show temperature comparison diagrams of Stele Pavilion, Long’en Gate, and Long’en Hall with different wind speed at 14:00. 3.4 Comparison and analysis of different wind speeds and relative humidity We found that after lawns were planted, with a wind speed of 5m/s compared with 4m/s, the temperature of the bricked ground in front of Stele Pavilion decreased by 5.72–6.24°C, a decrease of 8.21%–8.67%; the temperature of the bricked ground in front of Long’en Gate increased by 2.37–5.83°C, an increase of 4.03%–9.70%; and the temperature of the bricked ground in front of Long’en Hall decreased by 14.17–19.27°C, a decrease of 20.50%–26.34%. Compared with a wind speed of 5m/s, at a wind speed of 6m/s, the temperature of the bricked ground in front of Stele Pavilion decreased by 4.06–4.44°C, a decrease of 6.39%–6.71%; the bricked ground temperature in front of Long’en Gate decreased by 2.35–2.65°C, a decrease of 4.24%–4.70%; and the temperature of bricked ground in front of Long’en Hall increased by 6.10–9.53°C, an increase of 11.37%– 17.69%. Similar to the relative humidity, as the relative humidity increased, the ground temperature dropped. If the building was blocked by other walls, the wind speed will decrease and the temperature of the ground will increase when the relative humidity increases. Fig. 13(a)–(c) show the temperature comparison diagrams of Stele Pavilion, Long’en Gate, and Long’en Hall with different relative humidity at 14:00. We found that compared with lawns planted at 10% relative humidity, after lawns were planted at 40% relative humidity, the temperature of the tiled ground in front of Stele Pavilion decreased by 0.08–0.11°C, a decrease of 0.12%–0.17%; the temperature of the bricked ground in front of Long’en Gate increased by 2.55–5.25°C, an increase of 4.83%– 10.69%; and the temperature of the bricked ground in front of Long’en Hall decreased by 2.57–6.82°C, a decrease of 4.47%–11.23%. At 70% relative humidity instead of 40% relative humidity, the temperature of the bricked ground in front of Stele Pavilion decreased by 0.21–0.28°C, a decrease of 0.34%–0.43%; the temperature of the bricked ground in front of Long’en Gate increased by 3.27–8.62°C, an increase of 5.91%– 15.86%; and the temperature of the bricked ground in front of Long’en Hall increased by 0.33–4.40°C, an increase of 0.59%–8.17%. 3.5 Actual measurement of microclimate with the human thermal comfort index value. Temperature is an important factor affecting human comfort and physical health. In the process of environment creation, it is essential that using vegetation to reduce air temperature and provide an excellent and suitable temperature environment for humans and the protection of architectural heritage (Fig. 14) [29]. At present, the most popular outdoor environmental evaluation indicators in China are the standard sufficient temperature and WBGT. The standard effective temperature index is used to evaluate the comfort of the environment (Table 5) [37] and the WBGT index is used to measure the safety of the thermal environment (Table 6) [38]. The project team members measured the air temperature, black bulb temperature, and WBGT values at five-time points (Table 7). The human thermal comfort indexes in Table 5 and 6 and the measured results in Table 7 show that at 10:00 as solar radiation illuminance increased, the air temperature gradually increased. At this time, the black bulb temperature value of the environment Fig.14 Psychrometric chart. Fig.14 Psychrometric chart. was 27.7°C, between 25.6 and 30°C, and the WBGT index was 23.3°C, between 18 and 24°C, indicating a “warm” thermal sensation range. At 12:00, as a result of intense solar radiation, the ground accumulated a considerable amount of heat. The black bulb temperature measured was 36.7℃ and WBGT measured was 27.8°C. The thermal comfort index was in the “hot” thermal sensation zone. At 14:00, solar radiation was the strongest. The black bulb temperature measured was 39.6°C and WBGT measured was 29.7°C. In these cases, the environment was “hot” or “very hot”. The main influences of outdoor environmental parameters on human thermal perception are temperature, humidity, solar radiation, and wind speed. Comparing the actual measurement with numerical simulation results, the thermal comfort of the investigation environment was calculated. The numerical simulation results were very close to the actual measurement data (Fig. 15). Table 5 Standard effective temperature index [37] Thermal comfort index Comfortable Slightly warmer Slightly hotter Hot Very hot Standard effective temperature (℃) 22.2-25.6 25.6-30 30-34.5 34.5-37.5 ≥37.5 Table 6 WBGT thermal comfort index [38] WBGT Index Comfortable warm Hot Very hot Extremely hot WBGT (℃) <18 18-24 24-28 28-30 >30 Table 5 Standard effective temperature index [37] Table 5 Table 7 Real-time measurement of WBGT and black bulb temperature Time (Hour) Air temperature (℃) black bulb temperature (℃) WBGT temperature (℃) 10 26.1 27.7 23.3 11 29 34.8 25.2 12 31.4 36.7 27.8 13 31.6 37.7 28.5 14 33.7 39.6 29.7 Fig.15 Numerical simulation results were very close to the actual measurement data. Real-time measurement of WBGT and black bulb temperature Fig.15 Numerical simulation results were very close to the actual measurement data. Fig.15 Numerical simulation results were very close to the actual measurement data. Acknowledgements g Not applicable. 4. Conclusions The aims of this study were to compare measurement data and simulation results before and after planting grassland, compare the simulation results based on different wind speeds and relative humidity, and compare the actual measurements of the microclimate using the human thermal comfort index value, to reduce the damage to heritage sites caused by the heat island effect. The results of this research can be applied to the thermal environment and thermal comfort research of other heritage sites. The following conclusions were drawn from this study. (1) Because lawns and soil can cool the air through evaporation and the soil has a significant radiation absorption coefficient, it can also absorb more heat radiation and generate higher temperatures locally. The greening of the surrounding environment can affect the bricked ground and lead to a specific cooling effect. (2) The consistency of the difference between the actual measurement and the CFD simulation results indicates that the selection of simulation parameters was reasonable and could accurately reflect the internal temperature field distribution at Fuling (2) The consistency of the difference between the actual measurement and the CFD simulation results indicates that the selection of simulation parameters was reasonable and could accurately reflect the internal temperature field distribution at Fuling Mausoleum. The CFD simulation was able to predict the temperature of the surrounding environment at heritage sites. Mausoleum. The CFD simulation was able to predict the temperature of the surrounding environment at heritage sites. Mausoleum. The CFD simulation was able to predict the temperature of the surrounding environment at heritage sites. (3) The temperature of the bricked ground at Fuling Mausoleum after lawns were planted was lower than that before lawns were planted. As the wind speed and relative humidity increased, the ground temperature dropped. If the building was blocked by other walls, the temperature of some measured points may be affected by airflow vortexes and the temperature of the bricked ground will increase when the wind speed increases. (4) The main influences of outdoor environmental parameters on human thermal perception are temperature, humidity, solar radiation, and wind speed. The actual measurement data were very close to the numerical simulation results. (4) The main influences of outdoor environmental parameters on human thermal perception are temperature, humidity, solar radiation, and wind speed. The actual measurement data were very close to the numerical simulation results. Availability of data and materials The data is available within the article The data is available within the article. Abbreviations CFD: computational fluid dynamics; WBGT: wet bulb globe temperature. Ethics approval and consent to participate Not applicable. Competing interests The authors declare that they have no competing interests. Ethics approval and consent to participate Not applicable. pp Not applicable. Author Contributions X.W., P.L. and G.X. contributed equally to this work. Funding g This work was supported by the National Natural Science Foundation of China (Project No: 51978417), Natural Science Foundation of Liaoning Province (Project No: 2019- ZD-0207), Scientific Research Project of the Educational Department of Liaoning Province (Project No: LJGD2019012) and the Education and Teaching Reform Project of Shenyang University of Technology, China. Consent for publication Written informed consent for publication was obtained from all participants. Author details Author details 1 School of Architecture & Civil Engineering, Shenyang University of Technology, Shenyang 110870, China References 1. Ng, E., Chen, L., Wang, Y., Yuan, C., 2012. A study on the cooling effects of greening in a high- density city: an experience from Hong Kong. Build Environ 47, 256-271. https://doi.org/10.1016/j.buildenv.2011.07.014. 2. Davies, M., Steadman, P., Oreszczyn, T., 2008. Strategies for the modification of the urban climate and the consequent impact on building energy use. Energy Policy 36, 4548–4551. https://doi.org/10.1016/j.enpol.2008.09.013. 3. Gago, E.J., Roldan, J., Pacheco-Torres, R., Ordóñez, J., 2013. The city and urban heat islands: a review of strategies to mitigate adverse effects Renew. Sustain. Energy Rev 25, 749-758. https://doi.org/10.1016/j .rser.2013.05.057. 4. Revi, A., Satterthwaite, D.E., Aragón-Durand, F., Corfee-Morlot, J., Kiunsi, R., Pelling, M., Roberts, D., Solecki, W., 2014. Urban Areas. Climate Change 2014: Impacts, Adaptation, and Vulnerability. Part A. Cambridge University Pres, Cambridge, UK. 5. Rizwan, A.M., Dennis, Y.C.L., Liu, C., 2008. A review on the generation, determination and mitigation of urban heat island. Environ. Sci 20, 120-128. https://doi.org/10.1016/S1001- 0742(08)60019-4. 6. Rosenfeld, A.H., Akbari, H., Bretz, S.E., Fishman, B.L., Kurn, D.M., Sailor, D.J., Taha, H., 1995. Mitigation of urban heat islands: materials, utility programs, updates. Energy Build 22, 255-265. https://doi.org/10.1016/0378-7788(95)00927-P. 7. Bowler, D.E., Buyung-Ali, L., Knight, TM., Pullin, A.S., 2010. Urban greening to cool towns and cities: a systematic review of the empirical evidence. Landsc. Urban Plan 97, 147–155. https://doi.org/10.1016/ j.landurbplan.2010.05.006. 8. Gill, S., Handley, .J, Ennos, A.R., Pauleit, S., 2007. Adapting cities for climate change: the role of the green infrastructure. Built Environ 33, 115–133. 9. Chan, A.L.S., 2011. Developing a modified typical meteorological year weather file for Hong Kong taking into account the urban heat island effect. Build Environ 46, 2434-2441. https://doi.org/10.1016/j.buildenv.2011.04.038. 10. Feyisa, G.L., Dons, K., Meilby, H., 2014. Efficiency of parks in mitigating urban heat island effect: an example from Addis Ababa. Landscape Urb Plann 123, 87-95. https://doi.org/10.1016/j.landurbplan.2013.12.008. 11. Park, M., Hagishima, A., Tanimoto, J., Narita, K-i., 2012. Effect of urban vegetation on outdoor thermal environment: field measurement at a scale model site. Build Environ 56, 38-46. https://doi.org/10.1016/j.buildenv.2012.02.015. 12. Bruse, M., Fleer, H., 1998. Simulating surface–plant–air interactions inside urban environments with a three dimensional numerical model. Environ Model Softw 13, 373-384. https://doi.org/10.1016/S1364-8152(98)00042-5. 13. Dimoudi, A., Nikolopoulou, M., 2003. Vegetation in the urban environment: microclimatic analysis and benefits. Energy Build 35, 69-76. 14. Fröhlich, D., Matzarakis, A., 2013. Modeling of changes in thermal bioclimate: examples based on urban spaces in Freiburg, Germany. Theor Appl Climatol 111, 547-558. https://doi.org/10.1007/s00704-012-0678-y. 14. References Fröhlich, D., Matzarakis, A., 2013. Modeling of changes in thermal bioclimate: examples based on urban spaces in Freiburg, Germany. Theor Appl Climatol 111, 547-558. https://doi.org/10.1007/s00704-012-0678-y. 15. Srivanit, M., Hokao, K., 2013. Evaluating the cooling effects of greening for improving the outdoor thermal environment at an institutional campus in the summer. Build Environ 66, 158-172. 15. Srivanit, M., Hokao, K., 2013. Evaluating the cooling effects of greening for improving the outdoor thermal environment at an institutional campus in the summer. Build Environ 66, 158-172. https://doi.org/10.1016/j.buildenv.2013.04.012. 16. Vidrih, B., Medved, S., 2013. Multiparametric model of urban park cooling island. Urb Forest Urb Greening 12, 220-229. https://doi.org/10.1016/j.ufug.2013.01.002 17. Wen, C.Y., Juan, Y.H., Yang, A.S., 2017. Enhancement of city breathability with half open spaces in ideal urban street canyons. Build Environ 112, 322-336. https://doi.org/10.1016/j.buildenv.2016.11.048. 18. Yang AS, Juan YH, Wen CY, Su YM, Wu YC, 2016. Investigation on wind environments of surrounding open spaces around a public building. J Mech, 1-13. https://doi.org/10.1017/jmech.2016.47. 19. Hong, B., Lin, B.R., 2014. Numerical studies of the outdoor wind environment and thermal comfort at pedestrian level in housing blocks with different building layout patterns and trees arrangement. Renewable Energy 5, 1-10. 20. Li, X.T., Yu, Z., 2005. Numerical analysis of outdoor thermal environment around buildings. Building and Environment 40, 853-866. https://doi.org/10.1016/j.buildenv.2004.08.022 21. Corgnati, S.P., Fabi, V., Filippi, M., 2009. A methodology for microclimatic evaluation in museum: Application to a temporary exhibit. Building and Environment 44, 1253–1260. https://doi.org/10.1016/j.buildenv.2008.09.012. 22. EN15251, 2008. Indoor environmental input parameters for design and assess-ment of energy performance of buildings addressing indoor air quality, thermalenvironment, lighting and acoustics. 23. La Gennusa, N., Lascari, G., Rizzo, G., Scaccianoce, G., 2008. Conflicting needs of thethermal indoor environment of museums: in search of a practical com-promise. Journal of Cultural Heritage , 125–134. http://www.sciencedirect.com/science/article/pii/S1296207408000186. 24. Shuzo, 2007. CFD and building environment design. China Architecture Publishing, Beijing. 25. Yuan, J., Liu, J.P., 2007. Numerical Simulation of Urban Street Valley Thermal Environment and Planning and Design Counter measures. Architecture Journal 3, 37-39. 26. Wang, X.Y., 2016. Research on the overall protection model of the architectural cultural heritage area of the Qing Dynasty in Liaoning. Science Press, Beijing. 27. Wang, Y.C., 2004. Fuling mausoleum in Shengjing (Qing Culture Series). Shenyang press, Shenyang. 28. Liang, Y., 2010. The Cultural Value of the Three Tombs in Shengjing. Vicissitudes 4, 145-147. 29. China Meteorological Information centre, 2007. 37. McCullough, E.A., Jones, B.W., Zbikowski, P.J., 1983. The effect of garment design on the thermal insulation values of clothing. ASHRAE Transactions 89, 327-352. References Special meteorological data set for China's building thermal environment analysis. China Architecture Publishing, Beijing. 30. ANSYS, 2016. ANSYS FLUENT 16.0 Theory Guide. Canonsburg, Pennsylvania, U.S.A. 31. Orszag, S.A., Yakhot, V., 1986. Renormalization group analysis of turbulence. Proceedings of the International Congress of Mathematicians, Berkeley, 1395-1399. 32. Orszag, S.A., Yakhot, V., Flannery, W.S., Boysan, F., Choudhury, D., Maruzewski, J., Patel, B., 1993. Renormalization Group Modeling and Turbulence Simulations. International Conference on Near-Wall Turbulent Flows, Tempe, Arizona. 33. Hanna, S.R., G. A. B., Hosker, R.P., 1982. Handbook on Atmospheric Diffusion. DOE/TIC 11223. Technical Information Center, U.S. Department of Energy, Washington, DC. 34. Editorial department of Architectural Design Information Collection, 1994. Architectural Design Information Collection (Second Edition). China Architecture Publishing, Beijing. 35. Launder, B. E., D. B. S. 1974. The numerical computation of turbulent flows. Computer Methods in Applied Mechanics and Engineering 3 (2):269-289. 36. Sheng, J.S., Zhang, W., Gao, Z., 2019. Simulation Research of Effects of Different Underlying Surface Types on Outdoor Thermal Environment. Journal of Tianjin Chengjian University 25, 437- 442. 37. McCullough, E.A., Jones, B.W., Zbikowski, P.J., 1983. The effect of garment design on the thermal insulation values of clothing. ASHRAE Transactions 89, 327-352. 38. Zhang, W., Gao, Z., 2015. Research Progress of Outdoor Thermal Comfort Index. Journal of Environment and Health 32, 836-840. Figures Figure 3 (a) General views of the computational domain (geometry and mesh) for Fuling Mausoleum. (b) Detail of the mesh re¦nement with tetrahedral elements. (a) General views of the computational domain (geometry and mesh) for Fuling Mausoleum. (b) Detail of the mesh re¦nement with tetrahedral elements. (a) General views of the computational domain (geometry and mesh) for Fuling Mausoleum. (b) Detail of the mesh re¦nement with tetrahedral elements. Figure 4 Settings of the speci¦c numerical simulation scheme. Figure 1 The environment surrounding Fuling Mausoleum (white is the CFD simulation model). The environment surrounding Fuling Mausoleum (white is the CFD simulation model). The environment surrounding Fuling Mausoleum (white is the CFD simulation model). Figure 2 Figure 2 Figure 2 Measuring instruments: (a) black bulb thermometer, (b) anemometer, (c) heat §ux sensor (photos by Xiaoyu Wang). Figure 3 (a) General views of the computational domain (geometry and mesh) for Fuling Mausoleum. (b) Detail of the mesh re¦nement with tetrahedral elements. Figure 3 Figure 4 Settings of the speci¦c numerical simulation scheme. Figure 5 Photo and infrared images of the site in front of Stele Pavilion: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 5 Photo and infrared images of the site in front of Stele Pavilion: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 5 Photo and infrared images of the site in front of Stele Pavilion: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 6 Photo and infrared images of the site in front of Long’en Gate: (a) photo of the site and overhead of (b) awns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 6 Photo and infrared images of the site in front of Long’en Gate: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 6 Photo and infrared images of the site in front of Long’en Gate: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Photo and infrared images of the site in front of Long’en Gate: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 7 Photo and infrared images of the site in front of Long’en Hall: (a) photo of the site and overhead of (b) awns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). Figure 7 Photo and infrared images of the site in front of Long’en Hall: (a) photo of the site and overhead of (b) lawns to the left, (c) bricked ground, (d) lawns to the right (photo by Xiaoyu Wang). ment point locations of the numerical simula wns were planted (c-e) Comparison of actual Figure 8 Temperature cloud diagrams and measurement point locations of the numerical simulatio after lawns were planted and (b) before lawns were planted. Figure 4 (c-e) Comparison of actual m with simulation results in front of Stele Pavilion at 10:00 (c) in front of Stele Pavilion, (d) in Long’en Gate, and (e) in front of Long’en Hall. Figure 8 Temperature cloud diagrams and measurement point locations of the numerical simulation at 10:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 10:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Figure 9 Temperature cloud diagrams and measurement point locations of the numerical simulatio after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual m with simulation results in front of Stele Pavilion at 12:00 (c) in front of Stele Pavilion, (d) i Long’en Gate, and (e) in front of Long’en Hall. Figure 9 Figure 9 Temperature cloud diagrams and measurement point locations of the numerical simulation at 12:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 12:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Figure 10 Temperature cloud diagrams and measurement point locations of the numerical simulatio after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual m with simulation results in front of Stele Pavilion at 14:00 (c) in front of Stele Pavilion, (d) i Figure 10 Temperature cloud diagrams and measurement point locations of the numerical simulation at 14:00 (a after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 14:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Figure 10 Temperature cloud diagrams and measurement point locations of the numerical simulation at 14:00 (a) after lawns were planted and (b) before lawns were planted. (c-e) Comparison of actual measurement with simulation results in front of Stele Pavilion at 14:00 (c) in front of Stele Pavilion, (d) in front of Long’en Gate, and (e) in front of Long’en Hall. Figure 11 (a-c) Wind speed cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (b) 12:00, and (c) 14:00. (d-f) Relative humidity cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (e) 12:00, and (f) 14:00. Figure 11 (a-c) Wind speed cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (b) 12:00, and (c) 14:00. (d-f) Relative humidity cloud diagrams in the vertical direction of the central axis of Fuling Mausoleum at (a) 10:00, (e) 12:00, and (f) 14:00. Figure 12 Comparison of simulation results at different wind speeds at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Figure 12 Comparison of simulation results at different wind speeds at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall Comparison of simulation results at different wind speeds at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Figure 13 Comparison of simulation results at different relative humidity at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Figure 13 Figure 13 Comparison of simulation results at different relative humidity at (a) Stele Pavilion, (b) Long’en Gate, and (c) Long’en Hall. Comparison of simulation results at different relative humidity at (a) Stele Pavilion, (b) Long’en Gate, and Figure 14 Psychrometric chart. Psychrometric chart. Psychrometric chart. Figure 15 Numerical simulation results were very close to the actual measurement data. Figure 15 Figure 15 Numerical simulation results were very close to the actual measurement data.
https://openalex.org/W4229065747	https://discovery.ucl.ac.uk/id/eprint/10151703/1/xk5092.pdf	English	null	<i>Ab initio</i> simulations of α- and β-ammonium carbamate (NH<sub>4</sub>·NH<sub>2</sub>CO<sub>2</sub>), and the thermal expansivity of deuterated α-ammonium carbamate from 4.2 to 180 K by neutron powder diffraction	Acta crystallographica. Section B, Structural science, crystal engineering and materials./Acta crystallographica. Section B, Structural science, crystal engineering and materials	2,022	cc-by	12,354	research papers Ab initio simulations of a- and b-ammonium carbamate (NH4NH2CO2), and the thermal expansivity of deuterated a-ammonium carbamate from 4.2 to 180 K by neutron powder diffraction ISSN 2052-5206 ISSN 2052-5206 Christopher M. Howard,a,b* Ian G. Wood,b Kevin S. Knightb and A. Dominic Fortesc,b Received 31 January 2022 Accepted 8 March 2022 aBayerisches Geoinstitut (BGI), University of Bayreuth, 95447 Bayreuth, Germany, bDepartment of Earth Sciences, University College London, Gower Street, London WC1E 6BT, United Kingdom, and cISIS Facility, Rutherford Appleton Laboratory, Harwell Science and Innovation Campus, Didcot, Oxfordshire OX11 0QX, United Kingdom. Correspon- dence e-mail: christopher.howard@frm2.tum.de Edited by A. Katrusiak, Adam Mickiewicz University, Poland Edited by A. Katrusiak, Adam Mickiewicz University, Poland Edited by A. Katrusiak, Adam Mickiewicz University, Poland Keywords: powder diffraction; thermal expan- sion; DFT; neutron diffraction; negative linear compressibility; ammonium carbamate. CCDC references: 2162678; 2162679 Supporting information: this article has supporting information at journals.iucr.org/b Keywords: powder diffraction; thermal expan- sion; DFT; neutron diffraction; negative linear compressibility; ammonium carbamate. Experimental and computational studies of ammonium carbamate have been carried out, with the objective of studying the elastic anisotropy of the framework manifested in (i) the thermal expansion and (ii) the compressibility; furthermore, the relative thermodynamic stability of the two known polymorphs has been evaluated computationally. Using high-resolution neutron powder diffraction data, the crystal structure of -ammonium carbamate (ND4ND2CO2) has been reﬁned [space group Pbca, Z = 8, with a = 17.05189 (15), b = 6.43531 (7), c = 6.68093 (7) A˚ and V = 733.126 (9) A˚ 3 at 4.2 K] and the thermal expansivity of -ammonium carbamate has been measured over the temperature range 4.2–180 K. The expansivity shows a high degree of anisotropy, with the b axis most expandable. The ab initio computational studies were carried out on the - and -polymorphs of ammonium carbamate using density functional theory. Fitting equations of state to the P(V) points of the simulations (run athermally) gave the following values: V0 = 744 (2) A˚ 3 and bulk modulus K0 = 16.5 (4) GPa for the -polymorph, and V0 = 713.6 (5) A˚ 3 and K0 = 24.4 (4) GPa for the -polymorph. The simulations show good agreement with the thermoelastic behaviour of -ammonium carbamate. Both phases show a high-degree of anisotropy; in particular, - ammonium carbamate shows unusual compressive behaviour, being determined to have negative linear compressibility (NLC) along its a axis above 5 GPa. The thermodynamically stable phase at ambient pressure is the -polymorph, with a calculated enthalpy difference with respect to the -polymorph of 0.399 kJ mol1; a transition to the -polymorph could occur at 0.4 GPa. Supporting information: this article has supporting information at journals.iucr.org/b research papers Pluto has extensive deposits of N2, CH4 and CO ices across its surface in various mixtures (Protopapa et al., 2017; Schmitt et al., 2017; Moore et al., 2016) and a thin N2-dominated atmosphere (Gladstone & Young, 2019), with both Pluto and its moon Charon manifesting the spectroscopic signature of ammoniated material (Dalle Ore et al., 2018, 2019; Grundy et al., 2016). The dwarf planet Ceres also manifests spectroscopic signatures of ammoniated species, in the form of phyllo- silicates (Ammannito et al., 2016; Raponi et al., 2019), which are seen in the regions with high albedo, in particular in the complex crater Occator (De Sanctis et al., 2015, 2016), suggesting an NH4-rich interior. The Kuiper belt object Arrokoth has abundant CH3OH on the surface (Stern, 2019), with other potential ices, including H2O, NH3, H2CO and CO2, in trace amounts (Grundy et al., 2020). In other words, all the constituents necessary to make ammonium carbonates (and related substances) are present and so are likely to form these materials wherever partial melts of ammonia–water are exposed to solid or gaseous CO2. solid organics (Artemieva & Lunine, 2003) and both urea and ammonium carbamate/carbonate may be substantial by- products, persisting on geological timescales at the surface temperature of 95 K. Not only are ammonium carbonates relevant to the outer solar system, but they also play a role in industry here on Earth; ammonia–water solutions may be used to capture CO2 from ﬂue gases, deposited in solid form as ammonium carbonates (Han et al., 2013). Furthermore, there has been some interest in using ammonium carbonates for hydrogen storage in fuel cells (Lan et al., 2012). There has nevertheless been comparatively little interest in studying the structures and properties of these compounds; indeed, the structure of ammonium carbonate monohydrate was determined comparatively recently (Fortes et al., 2014), and the structure of the sesquicarbonate was determined in 2003 (Margraf et al., 2003). Furthermore, the identity of a second () phase of ammonium carbamate was ﬁrst reported in 2007 (Kuhn et al., 2007), while the only published crystal structure of -ammonium carbamate was that determined in 1973 by single-crystal X-ray diffraction (Adams & Small, 1973). Both ammonium carbamate phases are orthorhombic: -ammonium carbamate has the space group Pbca, with Z = 8, while -ammonium carbamate has the space group Ibam, with Z = 8. research papers In the same context, ammonium carbonates may have some astrobiological relevance since removal of water leads – via ammonium carbamate – to urea. It is known from laboratory analogue experiments that the organic molecules produced photochemically in the atmosphere of Saturn’s giant satellite Titan may be hydrolysed in aqueous ammonia to form both urea and amino acids (Poch et al., 2012). Further hydrolysis of urea would be expected to form ammonium carbonates; on Earth, this process is mediated in soils with the aid of bacterial urease, whereafter the carbonate breaks down to ammonia and water. Meteorite impacts into Titan’s surface would provide the requisite liquid (as impact melt) to hydrolyse any It is surprising that ammonium carbamate has apparently been of little interest, given the similarity of the carbamate ion to the primary amide group; previous work on ammonium carbamate has mainly focused on the thermodynamics and kinetics, typically related to the formation of urea, rather than crystallographic studies. Whilst urea has been extensively studied, the same is not true of the carbonate/carbamate family of compounds. Figure 1 The ternary system of NH3 + CO2 H2O. The compositions to form in this system are: (1) ammonium carbonate monohydrate; (2) ammonium sesquicarbonate monohydrate; (3) ammonium bicarbonate; (4) - and - ammonium carbamate (open circle); (5) urea; (6) CO2 clathrate hydrate; (7) solid carbonic acid; (8) ammonia dihydrate; (9) ammonia mono- hydrate; (10) ammonia hemihydrate. The goals of this study were to calculate the bulk properties of both phases of ammonium carbamate as a function of pressure, preparatory to the development of a planetary model. Without knowing how something as fundamental as the density varies with pressure and temperature, as well as high-pressure polymorphism, the development of accurate structural and evolutionary models of icy planetary interiors is severely limited. To determine the necessary physical prop- erties includes calculating the bulk elastic properties as a function of pressure using density functional theory (DFT); in particular, how the different arrangement of structural units and hydrogen bonds affects the bulk modulus of the two polymorphs. An opportunity to test the anisotropy of the -polymorph arose during an attempt to measure the thermal expansion of ammonium carbonate monohydrate. During a synthesis of that product, ammonium carbamate was formed instead, and neutron powder diffraction data were collected as a function of temperature. 1. Introduction With recent robotic missions to icy planetary bodies in our solar system, most notably the ﬂybys of Pluto, Charon and Arrokoth by the New Horizons spacecraft, and of Ceres by the Dawn spacecraft, the importance of characterizing ‘ices’ composed of primordial mixtures of ammonia, carbon dioxide and water has become more urgent. Although long suspected to be present in the outer solar system (Kargel, 1992), solids in the ternary NH3 + CO2 H2O system (Fig. 1), which includes two phases of ammonium carbamate (NH4NH2CO2), am- monium carbonate monohydrate [(NH4)2CO3H2O], ammo- nium sesquicarbonate monohydrate {(NH4)4[H2(CO3)3]H2O}, ammonium bicarbonate [(NH4)2HCO3] and the water-deﬁ- cient compound urea [CO(NH2)2], appear increasingly likely to occur as ‘rock-forming’ minerals. Published under a CC BY 4.0 licence 459 https://doi.org/10.1107/S2052520622002645 Acta Cryst. (2022). B78, 459–475 460 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 2.3. Structural refinement Ammonium carbamate was formed serendipitously in the process of attempting to synthesize deuterated ammonium carbonate monohydrate, during which a 28 wt% ND3 solution was exposed to CO2 from sublimating dry ice pellets, forming a slurry of white material. Consequently, the specimen contained excess water in the form of water ice, which also acted as a reservoir for small amounts of excess ammonium. The resulting white compound was stored in a freezer. The sample was taken out of the freezer and allowed to warm and soften, after which the partially molten slurry was transferred into a cryomortar under liquid nitrogen and ground to a ﬁne powder using a similarly cooled pestle. Once fully ground, the material was loaded into an aluminium-framed slab-geometry can of inner dimensions 23 18 15 mm (h w d) relative to the incident neutron beam, also precooled in liquid N2. Vanadium foil windows were indium-sealed to the front and back faces of the slab can, with the exposed components on the incident-beam side being masked with Gd foil. The sample holder was wired with an RhFe resistance thermometer and a 30 mm cartridge heater, after which, it was immersed in liquid nitrogen for transfer into a He cryostat operating at 100 K. Once oriented in the beam, the cryostat temperature was reduced to 4.2 K for measurements to commence. In the 4.2 K neutron powder diffraction dataset, two phases were identiﬁed by inspection of the peak positions – deuter- ated water ice and -ammonium carbamate. The Rietveld method was used to reﬁne the structure of D2O ice Ih and -ammonium carbamate, starting from the previously published atomic coordinates based upon the single-crystal X-ray investigation of Adams & Small (1973). In this reﬁne- ment, the unit-cell parameters, scale factor, peak-proﬁle coefﬁcients (GSAS proﬁle function #3) and a twelve-term Chebyshev polynomial background were allowed to vary and the atomic coordinates and isotropic displacement parameters (Uiso) were reﬁned independently, without restraints. It was found necessary to reﬁne a wavelength-dependent absorption/ reﬂectivity correction (GSAS model #1). The ﬁnal ﬁt to the data is excellent, the Rwp value being better than 2%; a Figure 2 Neutron powder diffraction pattern acquired at 4.2 K in the highest resolution backscattering banks of HRPD. research papers The thermal expansion derived from these data provides a useful measurement of the structural aniso- tropy for comparison with the calculations. Figure 1 g The ternary system of NH3 + CO2 H2O. The compositions to form in this system are: (1) ammonium carbonate monohydrate; (2) ammonium sesquicarbonate monohydrate; (3) ammonium bicarbonate; (4) - and - ammonium carbamate (open circle); (5) urea; (6) CO2 clathrate hydrate; (7) solid carbonic acid; (8) ammonia dihydrate; (9) ammonia mono- hydrate; (10) ammonia hemihydrate. Here, therefore, we present the ﬁrst ab initio calculations to determine the high-pressure behaviour of both - and -ammonium carbamate; the ﬁrst neutron powder diffraction Acta Cryst. (2022). B78, 459–475 research papers parameters from data measured only in the 100–200 ms win- dow whilst avoiding peak overlap from the additional peaks at shorter d-spacings. For accurate structure reﬁnement, how- ever, it is more desirable to measure the shorter d-spacing peaks in the 30–130 ms window. data of -ammonium carbamate, thereby providing the ﬁrst accurate H-atom positions to allow characterization of the hydrogen bonding in the crystal; and thermal expansivity measurements from 4.2 to 180 K. These results begin to bridge the current gap in our knowledge of the behaviour of ammonium carbonates under conditions related to the icy moons of the outer solar system. Consequently, we collected a low-noise structural data set at 4.2 K in the 30–130 ms window, counting for 6.5 h (corre- sponding to an integrated proton beam current of 250 mA h) and then measured on warming using only the 100–200 ms window. These thermal expansion data were collected in 5 K increments from 10–150 K and (due to time constraints) in 10 K increments up to 180 K. Counting times were 15 min, corresponding to an integrated proton beam current of 15 mA h per step. The diffraction data were focussed, nor- malized to the incident beam spectrum and corrected for detector efﬁciency by reference to a vanadium standard using Mantid (Arnold et al., 2014) and then exported in a format suitable for analysis with GSAS/Expgui (Larson & Von Dreele, 2004; Toby, 2001). 2.3. Structural refinement Red symbols represent the measured data, green lines the result of Rietveld reﬁnement, pink lines the difference proﬁle and the tick marks the expected positions of each Bragg reﬂection for ND4ND2CO2 (black) and D2O ice Ih (red). 2. Experimental Time-of-ﬂight neutron powder diffraction data were collected using the ‘High Resolution Powder Diffractometer’ (HRPD) at the STFC ISIS neutron spallation source of the Rutherford– Appleton Laboratory, UK. The HRPD instrument was chosen due to its high resolution (d/d ’ 8 104 in the back- scattering bank), allowing the highest precision in reﬁned unit- cell parameters and accurate partitioning of intensity in well- resolved high-Q Bragg peaks. C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 461 2.2. Data acquisition Time-of-ﬂight data on HRPD are measured routinely in a range of 100 ms bandwidth ‘windows’ where disc choppers act as bandwidth selectors, these typically being 30–130 or 100– 200 ms. In the instrument’s backscattering detector bank (2 = 154–176), these time-of-ﬂight ranges provide access to d- spacings in the ranges 0.65–2.60 (30–130 ms) and 2.15–4.00 A˚ (100–200 ms) at the highest resolution. Intermediate-resolu- tion data are simultaneously measured in the 90 detector banks (0.85–3.90 A˚ in 30–130 ms and 2.82–6.00 A˚ in 100– 200 ms). Since HRPD has an essentially Q-independent resolution function, we can obtain high-precision unit-cell Figure 2 Neutron powder diffraction pattern acquired at 4.2 K in the highest resolution backscattering banks of HRPD. Red symbols represent the measured data, green lines the result of Rietveld reﬁnement, pink lines the difference proﬁle and the tick marks the expected positions of each Bragg reﬂection for ND4ND2CO2 (black) and D2O ice Ih (red). C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 4 461 Acta Cryst. (2022). B78, 459–475 research papers carbamate were input into CASTEP (Clark et al., 2005). Static density functional theory (DFT) calculations were then carried out using the PBE pseudopotential (Perdew et al., 1996) with the Tkatchenko–Schefﬂer dispersion correction (Tkatchenko & Schefﬂer, 2009). The importance of dispersion corrections for the study of elastic properties and anisotropy in planetary ices was demonstrated recently by Meusburger et al. (2021). For the work reported here, we had found that geometry optimizations completed without dispersion cor- rections led to disagreements of 10% in the unit-cell para- meters, these being reduced to 1–3% with the use of dispersion corrections. Similarly, the estimated phase-transi- tion pressures reported later were reduced from 2 to 0.4 GPa. graphical depiction of the ﬁt to the neutron powder data is given in Fig. 2. The reﬁned structural parameters are given in the crystallographic information ﬁle (CIF) in the supporting information. At each temperature step, the unit-cell parameters of both phases were determined using the Le Bail method (with initial intensities based on the GSAS Fcalc method); in addition to the unit-cell parameters, the peak proﬁle parameters and the background parameters were allowed to vary. Consideration of the axial ratio of the D2O ice Ih present in the sample (see Fig. 3 and x4.1) indicated that, at 180 K, it was sufﬁciently pure to allow its use as an internal standard. 2.2. Data acquisition To ensure uniformity with previous work, the unit-cell parameters of D2O at 180 K were therefore ﬁxed to those measured in a previous experi- ment with a silicon standard (Fortes, 2018), and the diffrac- tometer constants subsequently reﬁned. Convergence tests were carried out to optimize the k-point sampling of the Brillouin zone within the Monkhorst–Pack scheme (Monkhorst & Pack, 1976) and the kinetic energy cut- off of the plane-wave basis set. For -ammonium carbamate, a converged grid of 2 6 5 (0.03 A˚ 1 reciprocal lattice spacing), with an energy cut-off of 1000 eV, yielded a total energy convergence better than 103 eV. For -ammonium carbamate, a k-point mesh of 2 2 3 (0.05 A˚ 1 reciprocal 462 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 4.1. Behaviour of D2O ice Ih upon warming ( ) ( ) ( ) ( ) N1—H(D)4 1.055 1.052 (5) 0.93 (3) 0.85 (6) N1—H(D)5 1.051 1.076 (5) 0.93 (3) 0.91 (6) N1—H(D)6 1.064 1.049 (5) 0.93 (3) 0.75 (6) NH(D)4 volume 0.5987 0.6106 0.4166 0.3205 Distortion index 0.0057 0.0098 0.0028 0.08704 Quadratic elongation 1.001 1.0026 1.0026 1.0615 Bond-angle variance 3.4483 10.2483 9.7185 114.3099 H(D)1 O1 1.863 1.949 (5) 2.16 (3) 2.12 (5) H(D)2 O1† 2.516 2.405 (5) 2.70 (3) 2.64 (6) H(D)3 O2 1.674 1.690 (5) 1.82 (3) 1.77 (5) H(D)4 O2 1.762 1.792 (5) 1.93 (3) 2.13 (5) H(D)5 O1 1.767 1.779 (5) 1.93 (3) 1.93 (6) H(D)6 O1 1.862 1.914 (6) 2.03 (3) 2.19 (6) N2—H(D)1 O1 177.1 171.2 (4) 174.7 (8) 174 (5) N2—H(D)2 O1† 148.0 150.4 (4) 149.9 (8) 158 (1) N1—H(D)3 O2 170.2 170.1 (4) 171.4 (8) 166 (4) N1—H(D)4 O2 164.3 166.2 (4) 162.4 (8) 138 (5) N1—H(D)5 O1 170.3 170.5 (4) 168.5 (8) 173 (4) N1—H(D)6 O1 157.5 158.4 (4) 155.4 (7) 158 (5) Notes: (†) in Adams & Small (1973), this is suggested to not be a hydrogen bond. Figure 4 Neutron powder diffraction patterns acquired at 160 K (top) and 170 K (bottom) in the highest resolution backscattering banks of HRPD. Red symbols represent the measured data, green lines the result of Rietveld reﬁnement, pink lines the difference proﬁle and the tick marks the expected positions of each Bragg reﬂection for: (a) ND4ND2CO2 (black); (b) D2O ice Ih (red); (c) ND3D2O (blue); (d) ND32D2O (green). At 160 K, peaks of ND3D2O appear, and at 170 K, peaks of ND32D2O also appear; by 180 K, these peaks have disappeared. Figure 4 thermal expansion is shifted to a lower temperature. Fur- thermore, the deviation is associated almost entirely with the behaviour of the crystal’s a axis; this is most apparent from a plot of the c/a ratio against temperature (Fig. 3). From 180 to 135 K, the c/a ratio is in excellent agreement with both the absolute values and the temperature dependence found in pure D2O (Fortes, 2018); below 135 K, the c/a ratio increases systematically. The dependence of the c/a ratio in ice Ih on both thermal history and chemical doping was described by Fortes (2019), who attributed the observed effects to changes in partial order in the crystal. C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 46 3. Computational methods To provide further information for planetary models, such as high-pressure behaviour, and to provide an accurate structure for -ammonium carbamate, both structures of ammonium Figure 3 (a)/(b) Reﬁned unit-cell parameters, (c) unit-cell volume and (d) c/a ratio of D2O ice Ih compared to literature values. Solid symbols are from Le Bail reﬁnements at each temperature step and black symbols are from Fortes (2018). There is a clear change in thermal expansion behaviour – principally of the a axis – below ca 135 K. Figure 3 (a)/(b) Reﬁned unit-cell parameters, (c) unit-cell volume and (d) c/a ratio of D2O ice Ih compared to literature values. Solid symbols are from Le Bail reﬁnements at each temperature step and black symbols are from Fortes (2018). There is a clear change in thermal expansion behaviour – principally of the a axis – below ca 135 K. 462 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 462 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate Acta Cryst. (2022). B78, 459–475 research papers research papers lattice spacing), with an energy cut-off of 980 eV, yielded a total energy convergence better than 103 eV. 4.1. Behaviour of D2O ice Ih upon warming 4.1. Behaviour of D2O ice Ih upon warming DFT athermal HRPD neutron 4.2 K X-ray 293 (2) K (Baisch et al., 2006) X-ray 295 K (Adams & Small, 1973) C1—O1 1.300 1.247 (5) 1.269 (3) 1.270 (5) C1—O2 1.285 1.272 (5) 1.266 (2) 1.279 (4) C1—N2 1.361 1.362 (4) 1.340 (3) 1.348 (5) N2—H(D)1 1.034 1.018 (5) 0.84 (3) 0.85 (6) N2—H(D)2 1.019 1.002 (5) 0.81 (3) 0.83 (7) N1—H(D)3 1.044 1.067 (4) 0.94 (4) 1.00 (5) N1—H(D)4 1.055 1.052 (5) 0.93 (3) 0.85 (6) N1—H(D)5 1.051 1.076 (5) 0.93 (3) 0.91 (6) N1—H(D)6 1.064 1.049 (5) 0.93 (3) 0.75 (6) NH(D)4 volume 0.5987 0.6106 0.4166 0.3205 Distortion index 0.0057 0.0098 0.0028 0.08704 Quadratic elongation 1.001 1.0026 1.0026 1.0615 Bond-angle variance 3.4483 10.2483 9.7185 114.3099 H(D)1 O1 1.863 1.949 (5) 2.16 (3) 2.12 (5) H(D)2 O1† 2.516 2.405 (5) 2.70 (3) 2.64 (6) H(D)3 O2 1.674 1.690 (5) 1.82 (3) 1.77 (5) H(D)4 O2 1.762 1.792 (5) 1.93 (3) 2.13 (5) H(D)5 O1 1.767 1.779 (5) 1.93 (3) 1.93 (6) H(D)6 O1 1.862 1.914 (6) 2.03 (3) 2.19 (6) N2—H(D)1 O1 177.1 171.2 (4) 174.7 (8) 174 (5) N2—H(D)2 O1† 148.0 150.4 (4) 149.9 (8) 158 (1) N1—H(D)3 O2 170.2 170.1 (4) 171.4 (8) 166 (4) N1—H(D)4 O2 164.3 166.2 (4) 162.4 (8) 138 (5) N1—H(D)5 O1 170.3 170.5 (4) 168.5 (8) 173 (4) N1—H(D)6 O1 157.5 158.4 (4) 155.4 (7) 158 (5) Notes: (†) in Adams & Small (1973), this is suggested to not be a hydrogen bond. When comparing D2O ice Ih unit-cell parameters to literature values, there is a signiﬁcant difference in the ice unit- cell parameters compared with pure ice at low temperatures (Fig. 3). Although the unit-cell volumes at 180 K agree well, there is a growing divergence at lower temperatures, such that the unit-cell volume is smaller than that of pure D2O ice Ih at 10 K by 0.11% and the transition to negative volumetric thermal expansion is shifted to a lower temperature. Fur- thermore, the deviation is associated almost entirely with the behaviour of the crystal’s a axis; this is most apparent from a plot of the c/a ratio against temperature (Fig. 3). 4.1. Behaviour of D2O ice Ih upon warming From 180 to 135 K, the c/a ratio is in excellent agreement with both the absolute values and the temperature dependence found in pure D2O (Fortes, 2018); below 135 K, the c/a ratio increases systematically. The dependence of the c/a ratio in ice Ih on both thermal history and chemical doping was described by Fortes (2019), who attributed the observed effects to changes in partial order in the crystal. We therefore propose that small concentrations of ammonium ions present in our D2O ice Ih sample effects similar very subtle structural changes that drive the c/a ratio to adopt a nearly ideal hexagonal close-packed value (1.630). In addition to the likely very small quantities of ammonium ions trapped in the crystal structure of ice Ih, it appears that Figure 4 Neutron powder diffraction patterns acquired at 160 K (top) and 170 K (bottom) in the highest resolution backscattering banks of HRPD. Red symbols represent the measured data, green lines the result of Rietveld reﬁnement, pink lines the difference proﬁle and the tick marks the expected positions of each Bragg reﬂection for: (a) ND4ND2CO2 (black); (b) D2O ice Ih (red); (c) ND3D2O (blue); (d) ND32D2O (green). At 160 K, peaks of ND3D2O appear, and at 170 K, peaks of ND32D2O also appear; by 180 K, these peaks have disappeared. Table 1 Table 1 Table 1 Comparison of bond lengths, polyhedral volumes, various distortion metrics and hydrogen-bond parameters (A˚ , ) in -ammonium carba- mate. Comparison of bond lengths, polyhedral volumes, various distortion metrics and hydrogen-bond parameters (A˚ , ) in -ammonium carba- mate. A series of calculations were then performed in which the ions were allowed to move according to the Hellman– Feynman forces with the symmetry maintained, and the unit- cell shape was allowed to vary; the initial structural para- meters of -ammonium carbamate were those obtained from our neutron powder diffraction data at 4.2 K and of -ammonium carbamate were taken from Kuhn et al. (2007). The values are from the outputs of CASTEP and measurements on HRPD, a powder X-ray diffraction experiment (Baisch et al., 2006) and a single-crystal X-ray diffraction experiment (Adams & Small, 1973). Note that the HRPD sample was deuterated. Atomic labels in Baisch et al. (2006) correspond to: H11 = H3, H12 = H2, H22 = H5, H23 = H6, H24 = H4 and N2 = N1, according to the labelling of Adams & Small (1973). The distortion index is based on bond lengths in the polyhedra and was deﬁned by Baur (1974). ‘Quadratic elongation’ gives a quantative measure of polyhedral distortion, which is independent of the effective polyhedron size (Robinson et al., 1971), and ‘bond-angle variance’ gives a quantative measure of how the bond angles in the tetrahedra deviate from the ideal value of 109.28 – lower values equate to bond angles closer to the ideal (Robinson et al., 1971). 4. Results and discussion 4. Results and discussion 4.1. Behaviour of D2O ice Ih upon warming We therefore propose that small concentrations of ammonium ions present in our D2O ice Ih sample effects similar very subtle structural changes that drive the c/a ratio to adopt a nearly ideal hexagonal close-packed value (1.630). Figure 4 g Neutron powder diffraction patterns acquired at 160 K (top) and 170 K (bottom) in the highest resolution backscattering banks of HRPD. Red symbols represent the measured data, green lines the result of Rietveld reﬁnement, pink lines the difference proﬁle and the tick marks the expected positions of each Bragg reﬂection for: (a) ND4ND2CO2 (black); (b) D2O ice Ih (red); (c) ND3D2O (blue); (d) ND32D2O (green). At 160 K, peaks of ND3D2O appear, and at 170 K, peaks of ND32D2O also appear; by 180 K, these peaks have disappeared. In addition to the likely very small quantities of ammonium ions trapped in the crystal structure of ice Ih, it appears that 463 Acta Cryst. (2022). B78, 459–475 research papers Figure 6 The R2 2(8) carbamate ionic dimers seen in the structure of -ammonium carbamate (ND4ND2CO2). Faded tetrahedra are below the plane of the centrosymmetric ring structure and the dashed lines represent H(D) bonds. there was also sufﬁcient excess ammonia present – likely as an amorphous aqueous solid in pores and on grain boundaries – to form ammonia hydrates. These were seen in the diffraction patterns during warming of the sample, with the ﬁrst impurity phase appearing at 160 K. Peaks from this phase became larger at 170 K and were joined by peaks from another accessory phase (Fig. 4). During the ﬁnal measurement at 180 K, all impurity peaks had disappeared completely. We matched the peaks from both accessory phases; the ﬁrst to appear was ammonia monohydrate (ND3D2O) and the second was ammonia dihydrate (ND32D2O). Since these are both known to melt between 170 and 180 K, it is not surprising that they are absent from the 180 K diffraction data. The presence of excess water in the form of ice meant we were obliged to keep the sample below the melting point of water. The addition of even small amounts of water from moisture in the atmosphere to ammonium carbamate is enough to slowly form ammonium bicarbonate. 4.1. Behaviour of D2O ice Ih upon warming With the presence of ammonium ions in the sample, eventually melting and forming ammonium hydrates, the melting point of water would be signiﬁcantly reduced, and we could not be certain as to when the sample would start to decompose, so it was decided to halt the experiment at 180 K to be certain the sample did not melt. Figure 6 g The R2 2(8) carbamate ionic dimers seen in the structure of -ammonium carbamate (ND4ND2CO2). Faded tetrahedra are below the plane of the centrosymmetric ring structure and the dashed lines represent H(D) bonds. There are many different possible arrangements of these rings within a crystal structure, including various ring, tape, and sheet-like motifs (Leiserowitz & Hagler, 1983). In ammonium carbamate, layers of these hexagonal rings are alternately arranged perpendicular to one another, known as a herringbone motif arrangement. Indeed, the space group of ammonium carbamate, Pbca, can generate an interlayer herringbone motif between the terminal atoms of the amide residue groups, R, which lie on the outside of the bilayer, due to the b-glide plane (Leiserowitz & Hagler, 1983). The form of the ammonium carbamate motifs in -ammonium carbamate is therefore similar to the shallow glide motif in primary amide structures. 464 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 4.2. a-Ammonium carbamate 4.2.1. Structure and bonding. The structure consists of ND4 + tetrahedra and planar ND2CO2 anions linked by moderately strong hydrogen bonds [H(D) O in the range 1.690–1.949 A˚ ]. Pairs of carbamate ions run parallel to the b axis, which are linked via ammonium tetrahedra (Fig. 5). Hydrogen bonding between centrosymmetric pairs of carba- mate ions creates an R2 2(8) ring motif, measuring 4.286 A˚ , between centrosymmetric pairs of C atoms; these rings are typically seen in primary amide structures, with amide mol- ecules, R-CONH2, forming the rings (Fig. 6). In the shallow glide motif, the spacing between layers of the same orientation is typically 3.4 A˚ ; in ammonium carbamate, this spacing is almost doubled in length, at 6.436 or 6.681 A˚ (equivalent to the b and c axes, respectively), due to the interlayer motif acting as an additional layer. Three C atoms of Figure 5 The structure of -ammonium carbamate, viewed parallel to the b axis. The grey dashed lines are H(D) O hydrogen bonds, and the solid black line is the unit cell. Figure 5 Figure 5 The structure of -ammonium carbamate, viewed parallel to the b axis. The grey dashed lines are H(D) O hydrogen bonds, and the solid black line is the unit cell amate, viewed parallel to the b axis. The grey dashed lines are H(D) O hydrogen bonds, and the solid black line is 464 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate Acta Cryst. (2022). B78, 459–475 research papers an angle of 150.4 (4). This work shows that the D1 O1 and D6 O1 hydrogen bonds are slightly longer than the remaining two bonds, agreeing with the work of Adams & Small (1973). an angle of 150.4 (4). This work shows that the D1 O1 and D6 O1 hydrogen bonds are slightly longer than the remaining two bonds, agreeing with the work of Adams & Small (1973). Parameters obtained by ﬁtting a second-order Gru¨neisen–Debye model with a single Debye temperature [equation (S2)] to the unit-cell volume of -ammonium carbamate between 4.2 and 180 K. V0 (A˚ ) K0/ (GPa) D (K) Q (J) b Adjusted R2 Second-order 733.15 (1) 49 (1) 208 (5) 3.60 (9) 1017 10 (1) 0.99991 4.2.2. Thermal expansion of a-ammonium carbamate. 4.2. a-Ammonium carbamate The thermal expansion was determined for deuterated -ammonium carbamate from the unit-cell parameters collected on HRPD at 4.2 K, in 5 K increments from 10 to 150 K, and in 10 K increments from 150 to 180 K. In order to model the thermal expansion to obtain useful thermoelastic parameters, a second-order Gru¨neisen–Debye model, with a single Debye temperature [equation (S2) in the supporting information], was used in least-squares ﬁtting to the unit-cell volume, with the values V0, D, Q and b freely reﬁned (further information regarding Gru¨neisen–Debye models can be found in the supporting information). The results of this ﬁt are given in Table 2, and shown in Fig. 7. The calculated thermal expansion coefﬁcients are shown in Fig. 8. The thermal expansion appears to be dominated by low- frequency vibrational modes with characteristic temperatures in the range 200–230 K; this suggests that the phonon density of states has a cut-off at 17.2–19.8 meV (equivalent to In order to model the thermal expansion to obtain useful thermoelastic parameters, a second-order Gru¨neisen–Debye model, with a single Debye temperature [equation (S2) in the supporting information], was used in least-squares ﬁtting to the unit-cell volume, with the values V0, D, Q and b freely reﬁned (further information regarding Gru¨neisen–Debye models can be found in the supporting information). The results of this ﬁt are given in Table 2, and shown in Fig. 7. The calculated thermal expansion coefﬁcients are shown in Fig. 8. separate hexagonal ring structures usually have a separation of 10 A˚ , but here the distance is 9.27 A˚ ; this is accom- modated by the tilt angle of the ring structures being 68 rather than a typical 61 (Leiserowitz & Hagler, 1983). In terms of bonds and bond lengths (see Table 1), the length of the N2—D1 bond is longer than the N2—D2 bond, due to the D1 atom donating a hydrogen bond [forming part of the R2 2(8) H(D)-bonded motif]. 4.2. a-Ammonium carbamate The white circles show values from the numerical derivative of the reﬁned unit-cell data, the solid red lines indicate the expansivities computed from the exponential model [equation (1)] and the solid blue line is from a power-law/exponential model (see Fortes, 2018) ﬁtted to the D2O ice Ih present as a contaminant in the sample. volume. Therefore, the thermal expansion data were also determined by ﬁtting the unit-cell parameters using a linear combination of a power law and an exponential function: wavenumbers 140–160 cm1 or 4.2–4.8 THz, respectively). The ﬁt of the Gru¨neisen–Debye model with a single Debye temperature also suggests that the heat capacity of -ammonium carbamate does not approach the classical Dulong and Petit high-temperature limit (3nkB), but instead tends towards a high-temperature limit of 2nkB. T ð Þ ¼ h p Tðq=TÞi þ r exp s T ð1Þ ð1Þ The value of K0/ found from this Gru¨neisen–Debye ﬁt to the unit-cell volume is 49 (1) GPa. This value is much greater than that calculated from ﬁtting an equation of state to the outputs of the DFT calculations (see x4.2.4). The reason for this is most likely that a Debye model with a single Debye temperature does not properly describe the internal energy of a molecular solid such as -ammonium carbamate (see supporting information). An alternative explanation is that 6¼ 1; if a value of K0 is assumed from the high-pressure DFT simulations, where K0 = 16.5 GPa, equation (S2) gives 0.33 (1). It could also be that -ammonium carbamate shows a temperature dependence of , which is assumed to be constant in this approximation. where p and r are scaling terms, and q and s are a ‘char- acteristic temperature’ indicative of the saturation behaviour of the function. The temperature dependence of the unit-cell parameters, or unit-cell volume, is then given by: X T ð Þ ¼ X0 exp Z T 0 T ð ÞdT ð2Þ ð2Þ where X(T) is either an axial length or the unit-cell volume. The results (Table 3) have small uncertainties; furthermore, they lack the oscillatory behaviour of polynomial functions and give the correct behaviour at temperatures close to 0 K. As can be seen in Fig. 4.2. a-Ammonium carbamate The N2—D2 bond was suggested not to be involved in a hydrogen bond by Adams & Small (1973); it appears that this H(D) atom may be sharing an interaction with the O1 atom, although the (H)D2 O1 distance is longer and more distorted than the remaining H(D) bonds – the D2 O1 bond length measures 2.405 (5) A˚ , with The thermal expansion appears to be dominated by low- frequency vibrational modes with characteristic temperatures in the range 200–230 K; this suggests that the phonon density of states has a cut-off at 17.2–19.8 meV (equivalent to Figure 7 Unit-cell parameters (a)–(c) and unit-cell volume (d) of deuterated -ammonium carbamate from 4.2 to 180 K. The solid symbols are from Le Bail reﬁnements, with the solid red lines indicating the exponential model [equations (1) and (2)] ﬁtted to the unit-cell parameters. A second-order (black line) Gru¨neisen–Debye model with a single Debye temperature [equation (S2)] is also ﬁtted to the unit-cell volume. The second-order Gru¨neisen–Debye model is not distinguishable from the exponential model at this scale. 7 Figure 7 Unit-cell parameters (a)–(c) and unit-cell volume (d) of deuterated -ammonium carbamate from 4.2 to 180 K. The solid symbols are from Le Bail reﬁnements, with the solid red lines indicating the exponential model [equations (1) and (2)] ﬁtted to the unit-cell parameters. A second-order (black line) Gru¨neisen–Debye model with a single Debye temperature [equation (S2)] is also ﬁtted to the unit-cell volume. The second-order Gru¨neisen–Debye model is not distinguishable from the exponential model at this scale. C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 465 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 465 465 Acta Cryst. (2022). B78, 459–475 research papers Figure 8 (a)–(c) Linear and (d) volume thermal expansion coefﬁcients of deuterated -ammonium carbamate with D2O ice Ih. The white circles show values from the numerical derivative of the reﬁned unit-cell data, the solid red lines indicate the expansivities computed from the exponential model [equation (1)] and the solid blue line is from a power-law/exponential model (see Fortes, 2018) ﬁtted to the D2O ice Ih present as a contaminant in the sample. Figure 8 (a)–(c) Linear and (d) volume thermal expansion coefﬁcients of deuterated -ammonium carbamate with D2O ice Ih. 466 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 4.2. a-Ammonium carbamate 7, despite its ad hoc nature, the expo- nential model does an excellent job of ﬁtting the experimental results, showing only extremely small differences from the Gru¨neisen–Debye model (nevertheless, we present the ﬁtting of different Debye models in the supporting information). Although Gru¨neisen–Debye models can provide useful thermoelastic parameters, they can be difﬁcult to evaluate (due to their complexity and, in this case, lack of heat capacity data) and are only dimensionally correct for the unit-cell Acta Cryst. (2022). B78, 459–475 research papers The linear expansivities and the volumetric expansivity, derived from equation (1), are shown in Fig. 8. The symbols are obtained as point-by-point derivatives of the experimental unit-cell parameters, providing a useful visual indicator of the uncertainties due to their scatter. There is a large degree of anisotropy in the structure upon warming; the greatest expansion is along the b axis, with the a and c axes having similar expansivities. The values of the volumetric thermal expansion coefﬁcient are comparable with water ice and ammonium carbonate monohydrate at the same due to their scatter. water ice and ammonium carbonate monohyd ). B78, 459–475 C. M. Howard et al. Ab initio simulations of - and -ammonium y–volume [E(V)] and (b) pressure–volume [P(V)] curves of -ammonium carbamate, calculated using CASTEP. T order Birch–Murnaghan equation of state [equations (S5) and (S6), BMEOS3]. (c)–(f) The relative change of the le unction of pressure of -ammonium carbamate, showing the a axis (green), b axis (blue), c axis (red) and unit-cell v how the different compressive regimes explained in the text that are not obvious in the E(V) and P(V) curves. T to the points to guide the eye, where an equation of state cannot be ﬁtted. Figure 9 Plots of (a) energy–volume [E(V)] and (b) pressure–volume [P(V)] curves of -ammonium carbamate, calculated using CASTEP. The solid lines show the best ﬁt third-order Birch–Murnaghan equation of state [equations (S5) and (S6), BMEOS3]. (c)–(f) The relative change of the length of the unit-cell dimensions as a function of pressure of -ammonium carbamate, showing the a axis (green), b axis (blue), c axis (red) and unit-cell volume (black). The shaded regions show the different compressive regimes explained in the text that are not obvious in the E(V) and P(V) curves. The dashed lines are simple linear ﬁts to the points to guide the eye, where an equation of state cannot be ﬁtted. C. M. C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate research papers To provide information relevant to simulating -ammonium carbamate in models of icy planetary interiors, the high-pressure behaviour was calculated by a series of relaxations over a pressure range between 0.9981 and +14.9952 GPa. ð3Þ with the same values of r and s as determined from the ﬁt to the unit-cell volume, and 0 = 1523.87 (2) kg m3. 4.2.3. DFT simulations of the structure of a-ammonium carbamate. The unit-cell volume at ambient pressure and 4.2 K, as measured on HRPD, is 733.126 (9) A˚ 3. The zero- pressure athermal relaxation of -ammonium carbamate from CASTEP yielded a unit-cell volume of 743.772 A˚ 3, which equates to an approximately 1.5% overestimation of the measured value (Table 4). Although the simulated -ammo- nium carbamate is protiated, the differences from the experimental structure, which is deuterated, are likely to be minor. To quantify elastic parameters, an equation of state was ﬁtted to the total energy against unit-cell volume [E(V)] and pressure against unit-cell volume [P(V)] curves (BMEOS3; Birch, 1952). These plots are given in Fig. 9, with the para- meters from ﬁtting to the calculated unit-cell parameters of -ammonium carbamate given in Table 5. The values of X0, K0, K0 0 and E0 were allowed to freely reﬁne. A parametric form of the Birch–Murnaghan equation of state was used to ﬁt the compressibility along each of the unit-cell axes (for b3 and c3; the a axis is not ﬁtted, which is explained next). This difference in unit-cell volume can largely be attributed to the angle between the two herringbone-motif carbamate ions – the simulated value has this close to 90, whereas the reﬁned structure at 4.2 K has an angle greater than 90. This means that the b axis has become longer and the c axis shorter in the simulated unit-cell parameters, resulting in the over- and underestimation of the individual unit-cell parameters. The result of relaxing -ammonium carbamate made no signiﬁcant changes to the structure. The bond lengths in the structure are only marginally different between the simulated and measured structures, i.e. the length of the carbamate ion is very similar, and the distance between centrosymmetric pairs of ions is 4.254 A˚ for the simulated and 4.304 (5) A˚ for the measured structures. 4.2. a-Ammonium carbamate Howard et al. Ab initio simulations of - and -ammonium carbamate 46 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 46 467 Acta Cryst. (2022). B78, 459–475 468 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate research papers Table 3 ab e 3 Parameters obtained by ﬁtting of equations (1) and (2) to the unit-cell parameters and unit-cell volume of -ammonium carbamate between 4.2 and 180 K. A comparison of the unit-cell parameters of -ammonium carbamate from the DFTathermal zero-pressure relaxation using CASTEP, from an X-ray single-crystal diffraction (SXD) experiment at room temperature (Adams & Small, 1973) and from a powder X-ray diffraction experiment (PXRD) at room temperature (Baisch et al., 2006). A comparison of the unit-cell parameters of -ammonium carbamate from the DFTathermal zero-pressure relaxation using CASTEP, from an X-ray single-crystal diffraction (SXD) experiment at room temperature (Adams & Small, 1973) and from a powder X-ray diffraction experiment (PXRD) at room temperature (Baisch et al., 2006). X0 (A˚ , A˚ 3) P (104) q r (104) s Adjusted R2 a axis 17.0517 (2) – – 0.268 (8) 51 (2) 0.99827 b axis 6.43539 (9) 0.95 (1) 14.5 (3) – – 0.99973 c axis 6.68077 (6) – – 0.74 (2) 113 (3) 0.99903 Volume (A˚ 3) 733.148 (9) – – 1.65 (1) 60.9 (6) 0.99993 Formula (NH4)+(NH2CO2) Crystal system Orthorhombic Space group Pbca Z 8 Reference This work This work This work Adams & Small (1973) Baisch et al. (2006) Method DFT Neutron Neutron X-ray X-ray T (K) Athermal 4.2 180 295 293 (2) a axis (A˚ ) 17.0796 17.0519 (2) 17.0908 (2) 17.121 (6) 17.119 (4) b axis (A˚ ) 6.61022 6.43518 (7) 6.48215 (8) 6.531 (2) 6.535 (2) c axis (A˚ ) 6.58788 6.68093 (8) 6.7040 (1) 6.742 (3) 6.754 (2) Volume (A˚ 3) 743.772 733.122 (9) 742.71 (1) 753.9 (5) 755.73 (3) temperature (Fortes, 2018; Fortes et al., 2014). We ﬁnd no evidence for negative thermal expansion in the structure. temperature (Fortes, 2018; Fortes et al., 2014). We ﬁnd no evidence for negative thermal expansion in the structure. For making a simple density calculation as a function of temperature for planetary modelling, the density can be described using only the exponential function of equation (1) (as was sufﬁcient to describe the volumetric expansion) and using: slightly longer than the remaining three bonds [excluding H(D)2 O1], and also agrees well with the work of Adams & Small (1973). ðTÞ ¼ 0 exp Z T 0 ðTÞdT ð3Þ 4.2.4. Simulations of the high-pressure behaviour of a-am- monium carbamate. research papers bond lengthens slightly, before shortening again. After this initial reconﬁguration, the changes are mainly due to the H O hydrogen bonds in the rings changing their behaviour, initially lengthening before shortening. Interestingly, these changes only manifest themselves in the lengths of the a and b axes, with barely any effect on the c axis. bond lengthens slightly, before shortening again. After this initial reconﬁguration, the changes are mainly due to the H O hydrogen bonds in the rings changing their behaviour, initially lengthening before shortening. Interestingly, these changes only manifest themselves in the lengths of the a and b axes, with barely any effect on the c axis. The largest NLC region in the a axis, however, occurs over the range 8–11 GPa (region III), continuing to a lesser extent up to 15 GPa (region IV), but this does not seem to translate into a signiﬁcant change in the unit-cell volume, as can be observed in Fig. 10. The NLC apparent in the a axis manifests itself in the b and c axes in different ways: between 9.0 and 10.5 GPa, there is a marked change in slope of the b axis, which occurs at the same point as the end of the major part of the NLC; however, along the c axis, there is no obvious change in slope. This compressive behaviour continues up to a pressure of 5 GPa. However, along the a axis, between 5 and 8 GPa (region II), there is a small degree of negative linear com- pressibility (NLC), followed by an inversion to normal posi- tive linear compressibility (PLC), averaging over this pressure range a small degree contraction along the a axis. The cause of this behaviour is a small rotation of the NH4 + tetrahedra to accommodate the closing of the carbamate rings, causing the carbamate ions to move outwards along the a axis. The cause of this large NLC is once again from the rotation of the NH4 + tetrahedra, forcing the carbamate ions to move apart along the a axis. Alternating chains of ammonium tetrahedra and carbamate ions along b likely reach a critical point, in that they cannot move closer together along either Figure 10 Structural features of -ammonium carbamate as a function of pressure. research papers For making a simple density calculation as a function of pressure, the density was ﬁtted using the equation of state used to ﬁt P(V) points, where X = /0 (see supporting information). The ﬁtted parameters are: 0 = 1394.2 kg m3 (value ﬁxed), K0 = 15.4 (3) GPa and K0 0 = 5.2 (1). There is a high degree of anisotropy in the structure when compressed and, although not immediately apparent from examination of the E(V) and P(V) curves, there are some interesting and complex features in how the structure deforms in response to an external pressure, in particular along the a axis. The initial loading shows subtle changes to the carbamate ions, manifesting in changes mostly along the a axis up to pressures of 5 GPa. During the initial loading, the carbamate ions reconﬁgure, particularly one of the H atoms on the N2 amino group. The N2—H1 and N2—H2 bonds shorten and lengthen, respectively, but in a rather smooth fashion, with the angles C1—N2—H1 and C1—N2—H2 also changing smoothly. This results in the carbamate ions themselves becoming more planar up to a pressure of 2 GPa before transitioning to a decrease in planarity (Fig. 10). The C1—O1 One thing to note is the distortion index of the ammonium tetrahedra in the HRPD sample measured at 4.2 K, which is much larger than the other values in Table 1, but remains an order of magnitude smaller than that of Adams & Small (1973). Apart from the distortion of the tetrahedra, the work on HRPD and the athermal zero-pressure DFT relaxation shows a very good agreement with the bonding geometry, such as the H(D)1 O1 and H(D)6 O1 hydrogen bonds being 468 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate Acta Cryst. (2022). B78, 459–475 research papers research papers This is evident from the transfer of the H O hydrogen bonds from the ammonium tetrahedra to the carbamate ion (Fig. 10). As the NLC begins, the N1—H6 O1 hydrogen bond becomes highly strained because of the ammonium tetrahedra rotating, and instead the bond changes to the O2 atom of the carbamate molecule. This also causes the change in slope of the b axis. the a or the b axis. The only way to accommodate the increasing stress is therefore to rotate the ammonium tetra- hedra. As the carbamate ions had rotated at a lower pressure and had been packed closer together by the contraction along b, one of the O atoms was moved closer to the ammonium tetrahedra, leading to the formation of a weak hydrogen bond. This is evident from the transfer of the H O hydrogen bonds from the ammonium tetrahedra to the carbamate ion (Fig. 10). As the NLC begins, the N1—H6 O1 hydrogen bond becomes highly strained because of the ammonium tetrahedra rotating, and instead the bond changes to the O2 atom of the carbamate molecule. This also causes the change in slope of the b axis. The previously published structure of -ammonium carba- mate indicated that it does not form the centrosymmetric dimer ring motifs R2 2(8) seen in -ammonium carbamate; instead, it forms one-dimensional chains along the mirror plane by the hydrogen bonding of a nitrile H atom to one of the carbonyl O atoms (H4 O2) at c/2 intervals. The remaining nitrile H atom (H3), in this instance, is unlikely to form a hydrogen bond, due to the N2—H3 O2 hydrogen bond having a small angle (115) and long length (3 A˚ ). Each carbonyl O atom has two out-of-plane hydrogen bonds to two separate ammonium tetrahedra, which lie either side of the mirror plane, with the tetrahedra midway between the chains. The ﬁnal transition at a critical pressure of 11.25 GPa (regions III and IV) is caused by the system reconﬁguring to the rotated ammonium tetrahedra, with the carbamate ions forced to translate along the a axis. 4.3. b-Ammonium carbamate 4.3.1. DFT structure of b-ammonium carbamate. The zero- pressure athermal geometry optimization of the -phase from CASTEP gave a unit-cell volume of 711.6 (2) A˚ 3; although this equates to an approximately 3% overestimation of the measured value at 298 K (Kuhn et al., 2007) (Table 6), the overestimation of the unit-cell volume is likely much larger, since, intuitively, it is expected the structure will contract upon cooling. An experiment at low temperature is needed to conﬁrm the degree of overestimation of the unit-cell volume. However, since no experimental data are currently available, we present comparisons between the athermal simulation and the room-temperature experiment, where the differences can largely be attributed to the bond lengths; DFT calculated bond lengths for -ammonium carbamate are compared to the experimental values in Table 7. However, in the DFT-relaxed structure (Fig. 11), the carbamate molecules have rotated, resulting in the H atoms changing position. The N2—H3 O2 hydrogen bond has become quite linear, with an angle of 172, and is much shorter, at 2.074 A˚ . The relaxing of the H atoms has altered the bonding geometry so that, now, -ammonium carbamate forms the same hexagonal ring motif, a double two-centre hydrogen-bonded R2 2(8) ring that is seen in the -phase, albeit with a different motif layout to the latter (see Fig. 12); the motifs instead lay on mirror planes, but are offset by the two glide planes in the structure. These planar R2 2(8) rings are joined via ammonium ions at z = 1 4 and z = 3 4. The N—H bond lengths given in Kuhn et al. (2007) are typically short due to the use of X-ray diffraction. The results of the present work, therefore, improve on the accuracy on the H-atom coordinates and are thus a much better determination of the hydrogen-bond lengths in the structure. Once again, this allows for more deﬁnitive statements regarding the structure and bonding to be made. It can also be seen in the experimental values that the ammonium tetrahedra are extremely distorted (Table 8). In the relaxed structure, the NH4 tetrahedra have a smaller distortion, with each of the H atoms donating a hydrogen bond to different carbamate O atoms. The length of the H1 O1 and H2 O2 hydrogen bonds, which form the out- 0 C. M. Howard et al. research papers Each of the panels shows different regions (shaded regions are II and IV, corresponding to 5–8 and >11 GPa, respectively) of the compressive behaviour that manifests along the a and b axes. Panels (a)/(b) show the behaviour of the carbamate molecule changing up to pressures of 2 GPa, (c)–(e) show the changes up to 5 GPa and (f)–(h) show the changing bond geometry as the O2 H6 hydrogen bond is replaced by the O1 H6 hydrogen bond. Figure 10 Figure 10 Structural features of -ammonium carbamate as a function of pressure. Each of the panels shows different regions (shaded regions are II and IV, corresponding to 5–8 and >11 GPa, respectively) of the compressive behaviour that manifests along the a and b axes. Panels (a)/(b) show the behaviour of the carbamate molecule changing up to pressures of 2 GPa, (c)–(e) show the changes up to 5 GPa and (f)–(h) show the changing bond geometry as the O2 H6 hydrogen bond is replaced by the O1 H6 hydrogen bond. C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 46 469 Acta Cryst. (2022). B78, 459–475 470 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate research papers Table 5 Table 6 Comparison of bond lengths (A˚ ), polyhedral volumes, various distortion metrics and hydrogen-bond parameters (A˚ , ) in -ammonium carba- mate. Fitted third-order Birch–Murnaghan equation of state parameters [equations (S5) and (S6)] of -ammonium carbamate. Fits to equation (S5) were achieved using only those pressure values that were greater than 0 GPa. For the b and c axes, the values ﬁtted were b3 and c3. No equation of state could be ﬁtted for the a axis due to the presence of negative linear compression. The -ammonium carbamate values are from the outputs of CASTEP and a powder X-ray diffraction experiment DFT athermal X-ray 298 K (Kuhn et al., 2007) C1—O1 1.300 1.287 (9) C1—O2 1.285 1.320 (9) C1—N2 1.361 1.294 (9) N2—H1 1.034 0.89 (6) N2—H2 1.019 0.83 (6) N1—H1 1.044 0.99 (4) N1—H2 1.055 1.12 (4) N2—H3 1.051 0.89 (7) N2—H4 1.064 0.83 (6) NH4 volume 0.5987 0.5883 Distortion index 0.0057 0.0600 Quadratic elongation 1.001 1.0231 Bond-angle variance 3.4483 65.0215 H1 O1 1.710 1.87 (4) H2 O2 1.804 1.77 (4) H3 O2 2.074 3.01 (6) H4 O2 2.091 2.15 (6) N1—H1 O1 171.31 100.5 (6) N1—H2 O2 165.40 157.2 (8) N2—H3 O2 171.67 115.0 (6) N2—H4 O2 176.67 143 (2) X0 (A˚ , A˚ 3) K0 (GPa) K0 0 K00 0 (GPa1) E0 (eV molecule1) BMEOS3, P(b)† 6.594 (9) 10.3 (4) 3.6 (1) 0.35 (2) BMEOS3, P(c) 6.595 (7) 16.2 (4) 3.08 (5) 0.236 (4)* BMEOS3, P(V) 744 (2) 16.5 (4) 5.0 (1) 0.36 (3)* BMEOS3, E(V) 743.7 (6) 16.0 (3) 5.35 (9) 0.44 (3)* 1669.9149 (4) Notes: (†) ﬁtted to points below 9 GPa due to change of slope; () derived from ﬁtted values of K0 and K0 0. Notes: (†) ﬁtted to points below 9 GPa due to change of slope; () derived from ﬁtted values of K0 and K0 0. the a or the b axis. The only way to accommodate the increasing stress is therefore to rotate the ammonium tetra- hedra. As the carbamate ions had rotated at a lower pressure and had been packed closer together by the contraction along b, one of the O atoms was moved closer to the ammonium tetrahedra, leading to the formation of a weak hydrogen bond. 4.3. b-Ammonium carbamate Ab initio simulations of - and -ammonium carbamate Acta Cryst. (2022). B78, 459–475 research papers Table 8 Fitted equation of state parameters [equations (S5) and (S6)] of -ammonium carbamate. Table 7 A comparison of the unit-cell parameters of -ammonium carbamate from the DFT athermal zero-pressure relaxation using CASTEP and a powder X-ray diffraction (PXRD) experiment at room temperature (Kuhn et al., 2007). Fits to equation (S5) were achieved using only those pressure values that were greater than 0 GPa. For the axial values, the cubes of the axes were ﬁtted. Formula (NH4)+(NH2CO2) Crystal system Orthorhombic Space group Ibam Z 8 Reference This work Kuhn et al. (2007) Method DFT X-ray T (K) Athermal 298 a axis (A˚ ) 9.8718 10.1428 b axis (A˚ ) 9.5064 9.1579 c axis (A˚ ) 7.6071 7.4485 Volume (A˚ 3) 713.89 691.87 X0 (A˚ , A˚ 3) K0 (GPa) K0 0 K00 0 (GPa1) E0 (eV molecule1) P(a) 9.872 (3) 39 (2) 9.3 (6) 0.9 (2) P(b) 9.508 (3) 32 (1) 17 (1) 6 (1) P(c) 7.620 (8) 12.2 (3) 2.76 (5) 0.345 (4) P(V) 713.6 (5) 24.4 (4) 4.09 (9) 0.164 (7) E(V) 711.6 (2) 23.3 (2) 4.51 (8) 0.200 (7) 1669.9084 (3) In -ammonium carbamate, the a and b axes compress at a much slower rate than the c axis (Fig. 13). The compression seen along the a and b axes is caused by the similar hydrogen- bond geometry in both directions from the R2 2(8) ring motifs of the carbamate ions. The structure is much less stiff perpen- dicular to the planar layers, in which the planar layers of carbamate ions are linked by ammonium tetrahedra halfway between. This can be seen in the compression of the medium- range hydrogen bonds that connect the layers of carbamate ions via ammonium tetrahedra; the O1 H1 and O2 H2 hydrogen bonds compress by 5.32 and 6.65%, respectively, over a pressure range of 0 to 10.025 GPa. of-plane bonds between the NH4 tetrahedra and planar NH2CO2 ions, are shorter than the remaining two hydrogen bonds that form parallel to the planes. of-plane bonds between the NH4 tetrahedra and planar NH2CO2 ions, are shorter than the remaining two hydrogen bonds that form parallel to the planes. 4.3.2. Simulations of the high-pressure behaviour of b-am- monium carbamate. C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 471 4.3. b-Ammonium carbamate As with -ammonium carbamate, the high-pressure behaviour of -ammonium carbamate was simulated to provide information relevant for planetary models. A series of static relaxations were performed over a pressure range between 1.9977 and +10.025 GPa. To gain information on the bulk elastic parameters, E(V) and P(V) curves were ﬁtted with a BMEOS3 equation of state, with the plots given in Fig. 13 and values of the ﬁt in Table 8. The values of X0, K0, K0 0 and E0 were allowed to freely reﬁne. Plots of the unit-cell parameters as a function of pressure are given in Fig. 13. A parametric form of the Birch–Murnaghan equation of state was used to ﬁt the compressibility along each of the unit-cell axes (for a3, b3 and c3). The density as a function of pressure was ﬁtted using the same equation of state used to ﬁt P(V) points. The ﬁtted parameters are: 0 = 1452.7 (8) kg m3, K0 = 24.2 (3) GPa and K0 0 = 4.11 (8). Figure 12 Comparison of the hydrogen-bond motif of -ammonium carbamate in the ab plane from (a) an X-ray diffraction experiment (Kuhn et al., 2007) and (b) the zero-pressure structure from CASTEP with symmetry operators overlain. The relaxed structure shows the same hexagonal-ring motif present in the -phase. Figure 11 View of the DFT-simulated structure of -ammonium carbamate parallel to the b axis. The grey dashed lines are H O hydrogen bonds and the solid black line is the unit cell. Figure 12 Figure 12 Figure 12 Comparison of the hydrogen-bond motif of -ammonium carbamate in the ab plane from (a) an X-ray diffraction experiment (Kuhn et al., 2007) and (b) the zero-pressure structure from CASTEP with symmetry operators overlain. The relaxed structure shows the same hexagonal-ring motif present in the -phase. Figure 11 g View of the DFT-simulated structure of -ammonium carbamate parallel to the b axis. The grey dashed lines are H O hydrogen bonds and the solid black line is the unit cell. 471 Acta Cryst. (2022). B78, 459–475 472 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate research papers Figure 14 Calculated enthalpy difference between -ammonium carbamate (normalized to zero) and -ammonium carbamate (red squares). Enthalpies were output from the dispersion-corrected simulations of CASTEP, with the lines a simple polynomial ﬁt (solid) and a linear ﬁt (dashed) to the points. The relative stability of the two phases is assessed by calculating the enthalpy (H) as a function of pressure for each phase, calculated using H = U + PV; since the simulations are athermal the enthalpy is equal to the Gibbs free energy. The enthalpies are compared in Fig. 14. The -polymorph has the lowest enthalpy of the two phases and is hence the stable phase at zero-pressure in the athermal limit, with a difference of 399.33 J mol1 with respect to the -polymorph. When increasing the pressure, the enthalpies of the two phases cross; the transition pressure between the two phases, where H = 0, is 0.4 GPa. The -polymorph has the lowest enthalpy above 0.4 GPa and is then the stable phase in the athermal limit. This trend, of the -polymorph having an increasing enthalpy difference relative to the -polymorph, continues up to pres- sures of 2 GPa. However, as the -phase starts to undergo subtle structural changes above this pressure, the trend reverses and the enthalpy difference between the two phases begins to decrease. Fig Cal (no Ent g Calculated enthalpy difference between -ammonium carbamate (normalized to zero) and -ammonium carbamate (red squares). Enthalpies were output from the dispersion-corrected simulations of CASTEP, with the lines a simple polynomial ﬁt (solid) and a linear ﬁt (dashed) to the points. 5.1. Thermoelastic behaviour The presence of negative linear compression in -ammo- nium carbamate is rather uncommon, since most materials compress in all directions when hydrostatically stressed. It seems rather curious that the negative linear compression behaviour calculated to exist in the -phase is not seen in -ammonium carbamate, since both phases consist of the same centrosymmetric R2 2(8) ring motifs. If it were found in the C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 47 research papers research papers p p 13 (a) energy–volume [E(V)] and (b) pressure–volume [P(V)] curves of -ammonium carbamate, calculated using CASTEP. The solid l -ﬁt third-order Birch–Murnaghan equation of state (BMEOS3) [equations (S5) and (S6)]. (c)–(f) The unit-cell dimensions of -a te as a function of pressure. The solid lines are ﬁtted BMEOS3 [equation (S5)], corresponding to the a axis (green), b axis (blue), c axis volume (black) Figure 13 Plots of (a) energy–volume [E(V)] and (b) pressure–volume [P(V)] curves of -ammonium carbamate, calculated using CASTEP. The solid lines show the best-ﬁt third-order Birch–Murnaghan equation of state (BMEOS3) [equations (S5) and (S6)]. (c)–(f) The unit-cell dimensions of -ammonium carbamate as a function of pressure. The solid lines are ﬁtted BMEOS3 [equation (S5)], corresponding to the a axis (green), b axis (blue), c axis (red) and unit-cell volume (black). 472 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 472 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate Acta Cryst. (2022). B78, 459–475 research papers References Adams, J. M. & Small, R. W. H. (1973). Acta Cryst. B29, 2317–2319. Ammannito, E. et al. (2016). Science, 353, aaf4279. Adams, J. M. & Small, R. W. H. (1973). Acta Cryst. B29, 2317–2319. Ammannito, E. et al. (2016). Science, 353, aaf4279. Arnold, O. et al. (2014). Nucl. Instrum. Methods Phys. Res. A, 764, 156–166. Arnold, O. et al. (2014). Nucl. Instrum. Methods Phys. Res. A, 764, 156–166. Artemieva, N. & Lunine, J. (2003). Icarus, 164, 471–48 Baisch, U., Pagano, S., Zeuner, M. & Schnick, W. (2006). Eur. J. Inorg. Chem. 2006, 3517–3524. Baur, W. H. (1974). Acta Cryst. B30, 1195–1215. research papers -structure, a ‘wine-rack’ motif that forms between the carbamate ions on the mirror planes and the ammonium ions that form halfway between these layers would cause the carbamate ions on the mirror planes to squeeze closer toge- ther, with the ammonium ions moving further apart to accommodate this change. Further experimental work, in particular a high-pressure study, is needed to verify the exis- tence of the calculated NLC seen in -ammonium carbamate, whether a phase transition would occur before the NLC region and to investigate the phase relationship between the two known phases. ammonium carbamate for further ab initio simulations have been obtained, and the thermal expansion coefﬁcients of - ammonium carbamate are now also known. However, further work is still required to understand the properties of ammo- nium carbamate in planetary environments. In particular, vibrational spectra are required for unambiguous identiﬁca- tion on planetary surfaces, as is a study of the P–T phase diagram under conditions applicable to icy bodies, i.e. high pressure and low temperature. 474 C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate Funding information When the two phases of ammonium carbamate are compared against thermal expansion measurements of other compounds in this ternary system, the high degree of aniso- tropy is comparable. The ammonium carbonates are highly anisotropic, typically with one unique axis, and two axes that behave similarly, due to planar layers of interconnected ammonium tetrahedra. Indeed, in ammonium carbonate monohydrate, the c axis has been shown from variable- temperature neutron single-crystal data to exhibit negative linear expansion, whilst the a and b axes exhibit large positive expansivities of similar magnitude (Fortes et al., 2014). Funding for this research was provided by: Science and Technology Facilities Council (grant No. ST/K000934/1). 4.4. Thermodynamics Since the total internal energies of both structures have been simulated using DFT as a function of pressure, it is possible to assess the relative stability of both phases in order to determine where a phase change may occur as a function of pressure. Figure 15 Figure 15 Projections of the thermal expansion coefﬁcient representation surface in the yz (top), xz (middle) and xy (bottom) planes at 60, 120 and 180 K. Solid green lines are positive values. Corresponding projections of the structure of -ammonium carbamate are shown on the right. Figure 15 Projections of the thermal expansion coefﬁcient representation surface in the yz (top), xz (middle) and xy (bottom) planes at 60, 120 and 180 K. Solid green lines are positive values. Corresponding projections of the structure of -ammonium carbamate are shown on the right. 473 Acta Cryst. (2022). B78, 459–475 research papers 6. Conclusion The measured thermal expansion correlates well with the calculated high-pressure behaviour of -ammonium carba- mate. The high degree of anisotropy under pressure in both these polymorphs is due to details in the structure: -ammonium carbamate’s b axis is the most compressible, due to few hydrogen bonds forming between layers of ammonium tetrahedra and carbamate ions, which run parallel to the c axis, and hence provide little resistance to compression; -ammo- nium carbamate is most compressible along c due to planar layers of carbamate ions running in chains in the ab plane, with ammonium tetrahedra forming the interlayer halfway between these parallel chains. The a and b axes have lower but similar compressibility due these to planar layers of carbamate ions. We report dispersion-corrected density functional theory simulations of both - and -ammonium carbamate, the ﬁrst neutron powder diffraction study of -ammonium carbamate, together with the ﬁrst structure reﬁnement at non-ambient temperature, measured at 4.2 K. We also carried out the ﬁrst thermal expansion measurements of -ammonium carbamate (from 4.2 to 180 K) and have shown how the magnitudes of the axial expansivities are related to the crystal structure. We ﬁnd evidence for the unusual property of NLC in the a axis of -ammonium carbamate above 5 GPa, with the largest degree of negative expansivity occurring between 9 and 10.5 GPa. This work has provided important parameters to be incorpo- rated into structural models of icy bodies in the Solar System. Further experiments at high pressure are needed to investi- gate the calculated negative linear compressibility in -am- monium carbamate and the possible phase transition between the two known polymorphs. In terms of thermal expansion, the direction of greatest thermal expansion is perpendicular to the corrugated sheets of alternating ammonium tetrahedra and centrosymmetric carbamate ionic dimers, whereas the minimum thermal expansion is in the plane of these alternating sheets. The intermediate value is parallel to stacks of centrosymmetric carbamate ions running along c, connected by chains of ammonium tetrahedra. The limited bonding between the corrugated sheets running along the b axis provides little resistance to expansion, with a three-dimensional bonding network running in the planes of these sheets resisting expansion (Fig. 15). Based upon the DFT calculations, we would expect that -ammonium carbamate should have a smaller thermal expansivity than -ammonium carbamate, due to its lower compressibility. Acknowledgements The authors wish to thank the STFC for funding and the STFC ISIS neutron spallation source for the provision of beam time (RB1410134; doi: 10.5286/ISIS.E.55377240). CMH thanks STFC for a postgraduate studentship. Open access funding enabled and organized by Projekt DEAL. Funding information C. M. Howard et al. Ab initio simulations of - and -ammonium carbamate 475 De Sanctis, M. C. et al. (2016). Nature, 536, 54–57. 5.2. Planetary modelling The work presented here provides some important values for use in planetary modelling: the density as a function of pressure has been calculated for both structures of ammonium carbamate, accurate crystal structures of both phases of ( ) Dalle Ore, C. M., Protopapa, S., Cook, J. C., Grundy, W. M., Cruikshank, D. P., Verbiscer, A. J., Ennico, K., Olkin, C. B., Stern, S. A., Weaver, H. A. & Young, L. A. (2018). Icarus, 300, 21–32. De Sanctis, M. C. et al. (2015). Nature, 528, 241–244. Acta Cryst. (2022). B78, 459–475 Acta Cryst. (2022). B78, 459–475 475 research papers research papers De Sanctis, M. C. et al. (2016). Nature, 536, 54–57. Margraf, G., Lerner, H.-W., Bolte, M. & Wagner, M. (2003). Z. Naturforsch. B, 58, 511–513. Fortes, A. D. (2018). Acta Cryst. B74, 196–216. Meusburger, J. M., Hudson-Edwards, K. A., Tang, C. C., Crane, R. A. & Fortes, A. D. (2021). Icarus, 368, 114611. Fortes, A. D. (2019). Phys. Chem. Chem. Phys. 21, 8264–8274. Fortes, A. D., Wood, I. G., Alfe`, D., Herna´ndez, E. R., Gutmann, M. J. & Sparkes, H. A. (2014). Acta Cryst. B70, 948–962. Monkhorst, H. J. & Pack, J. D. (1976). Phys. Rev. B, 13, 5188–5192. Monkhorst, H. J. & Pack, J. D. (1976). Phys. Rev. B Moore, J. M. et al. (2016). Science, 351, 1284–1293. Gladstone, G. R. & Young, L. A. (2019). Annu. Rev. Earth Planet. Sci. 47, 119–140. Perdew, J. P., Burke, K. & Ernzerhof, M. (1996). Phys. Rev. Lett. 77, 3865–3868. Grundy, W. M. et al. (2016). Science, 351, aad9189. Poch, O., Coll, P., Buch, A., Ramı´rez, S. I. & Raulin, F. (2012). Planet. Space Sci. 61, 114–123. Grundy, W. M. et al. (2020). Science, 367, eaay3705. Han, K., Ahn, C. K., Lee, M. S., Rhee, C. H., Kim, J. Y. & Chun, H. D. (2013). Int. J. Greenhouse Gas Control, 14, 270–281. p Protopapa, S. et al. (2017). Icarus, 287, 218–228. Raponi, A., De Sanctis, M. C., Carrozzo, F. G., Ciarniello, M., Castillo- Rogez, J. C., Ammannito, E., Frigeri, A., Longobardo, A., Palomba, E., Tosi, F., Zambon, F., Raymond, C. A. & Russell, C. T. (2019). Icarus, 320, 83–96. Kargel, J. S. (1992). Icarus, 100, 556–574. Kuhn, N., Stro¨bele, M. & Meyer, H.-J. (2007). Z. Anorg. Allg. Chem. 633, 653–656. Lan, R., Irvine, J. T. S. & Tao, S. W. (2012). Int. J. Hydrogen Energy, 37, 1482–1494. Robinson, K., Gibbs, G. V. & Ribbe, P. H. (1971). Science, 172, 567– 570. Larson, A. C. & Von Dreele, R. B. (2004). Report 86-748. Los Alamos National Laboratory, NM, USA. Schmitt, B. et al. (2017). Icarus, 287, 229–260. Stern, S. A. et al. (2019). Science, 364, eaaw9771. Leiserowitz, L. & Hagler, A. T. (1983). Proc. Roy. Soc. London Ser. Math. Phys. Eng. Sci. 388, 133–175. Tkatchenko, A. & Schefﬂer, M. (2009). Phys. Rev. Lett. 102, 073005. Toby, B. H. (2001). J. Appl. Cryst. 34, 210–213. Toby, B. H. (2001). J. Appl. Cryst. 34, 210–213. 475 Acta Cryst. (2022). B78, 459–475
https://openalex.org/W2798014810	https://www.frontiersin.org/articles/10.3389/fbioe.2018.00043/pdf	English	null	Policy-Led Comparative Environmental Risk Assessment of Genetically Modified Crops: Testing for Increased Risk Rather Than Profiling Phenotypes Leads to Predictable and Transparent Decision-Making	Frontiers in bioengineering and biotechnology	2,018	cc-by	6,737	PERSPECTIVE published: 10 April 2018 doi: 10.3389/fbioe.2018.00043 *Correspondence: Alan Raybould alan.raybould@syngenta.com Specialty section: This article was submitted to Biosafety and Biosecurity, a section of the journal Frontiers in Bioengineering and Biotechnology Specialty section: This article was submitted to Biosafety and Biosecurity, a section of the journal Frontiers in Bioengineering and Biotechnology Keywords: risk assessment, genetically modiﬁed crops, regulatory policy, problem formulation, proﬁling, hypothesis testing Received: 30 January 2018 Accepted: 26 March 2018 Published: 10 April 2018 1 Syngenta Crop Protection AG, Basel, Switzerland, 2 Plant Health Science Services, Canadian Food Inspection Agency, Ottawa, ON, Canada Edited by: Randall Steven Murch, Virginia Tech, United States We describe two contrasting methods of comparative environmental risk assessment for genetically modiﬁed (GM) crops. Both are science-based, in the sense that they use science to help make decisions, but they differ in the relationship between science and policy. Policy-led comparative risk assessment begins by deﬁning what would be regarded as unacceptable changes when the use a particular GM crop replaces an accepted use of another crop. Hypotheses that these changes will not occur are tested using existing or new data, and corroboration or falsiﬁcation of the hypotheses is used to inform decision-making. Science-led comparative risk assessment, on the other hand, tends to test null hypotheses of no difference between a GM crop and a comparator. The variables that are compared may have little or no relevance to any previously stated policy objective and hence decision-making tends to be ad hoc in response to possibly spurious statistical signiﬁcance. We argue that policy-led comparative risk assessment is the far more effective method. With this in mind, we caution that phenotypic proﬁling of GM crops, particularly with omics methods, is potentially detrimental to risk assessment. Reviewed by: Jacqueline Fletcher, Oklahoma State University, United States Laura Adam, Ebiosec, Inc., United States Stephen Allen Morse, Centers for Disease Control and Prevention (CDC), United States Reviewed by: Jacqueline Fletcher, Oklahoma State University, United States Laura Adam, Ebiosec, Inc., United States Stephen Allen Morse, Centers for Disease Control and Prevention (CDC), United States Deﬁning Risk and Opportunity Risk may be expressed as a combination of the likelihood and severity of harm that may arise from hazardous properties of a proposed activity. Environmental risk assessors often think of risk in terms of the potential exposure to the hazard that can cause a harm, where potential exposure is the expression of likelihood. Seriousness of harm is related to the degree of hazard, but also contains subjective elements (see below). Risk is usually diﬃcult to quantify precisely, and most risk assessments rely on qualitative assessments and expert judgment. If severe harm is likely, risk is high; and if the most serious conceivable eﬀect is trivial and unlikely, then risk may be regarded as negligible. However, even a tiny probability of a harmful eﬀect may be regarded as high risk if the harmful eﬀect is serious. A severe decline in the population size of an endangered or iconic species might be one such eﬀect. Risk may also be regarded as non- negligible if low severity events are predicted to occur frequently (e.g., Slovic, 1999). Determining whether an activity poses acceptable risk requires several diﬃcult judgments. First, one must decide what would be regarded as harmful eﬀects of the activity and what would be regarded as beneﬁcial eﬀects. In addition, one must decide how to judge the severity of harm and the value of beneﬁts. While science may be used to limit the scope of discussions of harm and beneﬁt to plausible eﬀects of the proposed activity (Raybould, 2010a), the designation of an eﬀect as harmful, beneﬁcial or neither, and the severity and value ascribed respectively to harmful and beneﬁcial eﬀects of a particular size relies on non-scientiﬁc criteria. These criteria may be based on personal values, an organization’s objectives or public policy depending on who will make the decision. For brevity, hereafter we refer to these non-scientiﬁc criteria as “policy objectives.” Similar considerations apply to the opportunities that may arise from an activity. Opportunity is high if very valuable beneﬁts are likely to arise, such as shifts to more sustainable agricultural practices as have been seen in Canada with the widespread adoption of GM herbicide-tolerant (GMHT) canola varieties. Deﬁning Risk and Opportunity Use of tillage by growers prior to seeding for weed control for canola appears to have been eliminated and the signiﬁcant shift to minimum and zero tillage systems has reduced soil erosion, resulted in higher carbon sequestration in production areas, reduced the need for herbicide applications and created net economic beneﬁts for growers (Gusta et al., 2011; Smythe et al., 2011). Opportunity is negligible if the most valuable beneﬁt is unlikely and of low value, such as cultivation of a GM drought tolerant crop in an area where precipitation is almost never yield limiting. Opportunity may still be regarded as high if beneﬁcial eﬀects are unlikely, but would be hugely valuable if they arose. The reduction of a non-target eﬀect to a highly beneﬁcial or iconic insect species that may only rarely co-occur with crop production could be considered as highly beneﬁcial. This may occur if cultivation of the GM crop reduces the spraying of pesticides, either directly through endogenous insect protection or indirectly by carrying a disease tolerance that reduces the need to spray for an insect vector of the disease. Signiﬁcant opportunity may also accrue from frequent events of relatively low value. The second diﬃcult judgment is how one will weigh risk and opportunity. One must consider whether certain eﬀects should be unacceptable regardless of the size of the opportunity or whether the largest net opportunity will always be the preferred option. In addition, one will need a method for evaluating net opportunity when beneﬁts and harms may be very diﬀerent; how, for example, does one evaluate the net opportunity if growing a certain crop is expected to increase yield but reduce other ecosystem services (de Groot et al., 2010). The above considerations show the importance of setting clear policy objectives in order to ensure that the scientiﬁc parts of risk assessment answer questions that are useful for decision- makers rather than questions that scientists may ﬁnd interesting (Hill and Sendashonga, 2003; Evans et al., 2006). In practice, even with policy direction, such as a policy objective on the conservation of biodiversity, risk assessors rely on professional judgment when they weigh evidence in what is often a qualitative process and make a number of “micro policy judgments” while conducting the assessment. Citation: It follows that even severe risks may be acceptable provided the opportunities are high enough, and that an increase in risk many be acceptable provided it is outweighed by increased opportunity. In practice, determining the acceptability of risk for the cultivation of a GM crop is made diﬃcult by the need to balance complicated sector needs with a broader public good. The 1993 Canadian Regulatory Framework for Biotechnology (Industry Canada, 1998; Gabler, 2008), for example, attempts to articulate guiding principles for how decisions could be structured. The framework captures the idea that any regulatory decisions should enable innovation, but also protect the environment and the health and well-being of citizens. Governments often have competing internal interests where departments of environment may view the opportunities for cultivating GM crops diﬀerently from Departments of Agriculture who see the acceptable risks and beneﬁts of agriculture with a more commercial perspective. Citation: Raybould A and Macdonald P (2018) Policy-Led Comparative Environmental Risk Assessment of Genetically Modiﬁed Crops: Testing for Increased Risk Rather Than Proﬁling Phenotypes Leads to Predictable and Transparent Decision-Making. Front. Bioeng. Biotechnol. 6:43. doi: 10.3389/fbioe.2018.00043 Regulatory risk-management of GM crops often uses comparative risk assessment to inform decision-making. Decisions may include whether to allow cultivation or importation of a particular crop in the relevant jurisdiction, and whether any conditions need to be placed on those uses if they are permitted. Comparative risk assessment contextualizes the risk by comparing the risks posed by the cultivation of the GM crop with the risks posed by the cultivation of the non-GM counterpart. If the risk assessment indicates that cultivating a GM crop poses no greater environmental risk than cultivating the non-GM counterpart, then it might be thought that cultivating the GM crop April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org Policy-Led ERA for GM Crops Raybould and Macdonald decision-making, the course of action posing the highest net opportunity—the opportunity minus the risk—must be selected. It follows that even severe risks may be acceptable provided the opportunities are high enough, and that an increase in risk many be acceptable provided it is outweighed by increased opportunity. In practice, determining the acceptability of risk for the cultivation of a GM crop is made diﬃcult by the need to balance complicated sector needs with a broader public good. The 1993 Canadian Regulatory Framework for Biotechnology (Industry Canada, 1998; Gabler, 2008), for example, attempts to articulate guiding principles for how decisions could be structured. The framework captures the idea that any regulatory decisions should enable innovation, but also protect the environment and the health and well-being of citizens. Governments often have competing internal interests where departments of environment may view the opportunities for cultivating GM crops diﬀerently from Departments of Agriculture who see the acceptable risks and beneﬁts of agriculture with a more commercial perspective. poses no unacceptable risk. However, judging the acceptability of a risk goes beyond the scientiﬁc comparison of relative risks. In order to make this point, we discuss deﬁnitions of risk, opportunity and acceptability. We concentrate on environmental risk assessment and GM crops, but our discussion is pertinent to risk assessment and decision-making more generally. decision-making, the course of action posing the highest net opportunity—the opportunity minus the risk—must be selected. Deﬁning Risk and Opportunity Indeed, the promotion of “science-based risk assessment” (= science-led in our terms) (e.g., Andow and Hilbeck, 2004; Kuntz et al., 2013) could lead to the mistaken and pernicious idea that it is desirable to eliminate consideration of policy objectives and judgment from risk assessment. Such thinking is almost guaranteed to produce controversy and paralyze decision-making (e.g., Risk Hypotheses and Decision-Making Criteria At their most conservative, the risk hypotheses will be that no harmful eﬀect will result from the proposed activity. If these hypotheses are corroborated under rigorous testing using information from reputable sources, including data from laboratory or ﬁeld tests, the risk managers can be conﬁdent that the proposed activity poses negligible risk, and then use that conclusion in their decision-making. Less conservative risk hypotheses acknowledge the probability and contextualize the impact of any harmful eﬀect; that is, the hypotheses under test would be that the risk does not exceed a threshold of acceptability. The threshold may be set to be the same as the risk posed by similar activities, or higher risk could be tolerated if the activity provides greater opportunities; for example, greater risk might be acceptable for cultivation of a GM crop that provides higher yield or improved quality than the crops it will replace. Rigorous corroboration of the hypotheses would indicate that the risks could be placed in the context of those from comparable activities, such as the cultivation of a non–GM crop that has a similar trait, even though the risks may not be negligible. That conclusion would contribute to decision-making. In essence, regulatory risk assessments should test hypotheses that help risk managers to make good decisions about whether to permit particular activities. Problem formulation is the process FIGURE 1 \| Conceptual models of science-led and policy-led risk assessment. Judging the Acceptability of Risk g g p y Judging the acceptability of risk requires a method to weigh the opportunities against the risks of the activity under consideration (Sanvido et al., 2012). Under ethical decision-making, if a risk exceeds an acceptability threshold, then the risk is unacceptable regardless of the size of the opportunity. Under utilitarian April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 2 Policy-Led ERA for GM Crops Raybould and Macdonald Raybould, 2010b). Instead, “policy-led risk assessment” ought to be the aim (Figure 1). by which these risk hypotheses, and plans to test them, are devised. While we concentrate on environmental risk posed by the cultivation of GM crops, our comments are relevant to any regulatory decision-making that makes use of risk assessment. In this article, we explore the implications of a change of emphasis from science to policy on two aspects of comparative environmental risk assessment of GM crops that are of current interest: problem formulation and the use of proﬁling data from various omics techniques. While we focus on regulatory decision- making about GM crops, our remarks are relevant to all crops with novel phenotypes, however they are produced, and to other types of decision-making, such as choosing which products to develop (Macdonald, 2014). In regulatory environmental risk assessment, decision- making criteria should relate to the probability and severity of environmentally harmful eﬀects arising from the proposed activity covered by the regulations. In the case of GM crops, the proposed activity will be cultivation of a speciﬁed GM crop in a particular place, perhaps with other stipulations such as whether certain crop-protection chemicals will be applied to the crop. The deﬁnition of what is harmful is a matter for the risk managers based on their interpretation of the policy objectives of the legislation that the regulations are designed to implement. Placing Risks in Context of Current Practice Similarly, a hypothesis that growing a certain GM crop will poses no unacceptable risk, is really a hypothesis that any increase in risk caused by growing the GM crop will be acceptable, either because the increase falls below a threshold of acceptability or because the additional opportunities created by growing the crop are worth the risk. As “no additional harm” sets a higher standard than “no unacceptable increase in risk,” testing a hypothesis of no additional harm may be regarded as rigorous testing of a hypothesis of no unacceptable increase in risk provided other factors that determine acceptability of risk, such as the size of the opportunity, are unchanged. A hypothesis that growing a GM crop will cause no unacceptable increase in risk is useful in a least three respects. First, corroboration or falsiﬁcation of this hypothesis is valuable to risk managers. Second, it shows that GM regulation follows the Principles of Regulation by not treating GM crops diﬀerently from other agricultural practices. Finally, it is useful to risk assessors, because if “unacceptable risk” is suﬃciently operationalized, risk assessors have clarity about the data they need in order to conduct the risk assessment, namely data that test the hypothesis of no unacceptable risk. Policy-led risk assessment would approach the problem by deﬁning, at the very least, general trends that would be regarded as harmful changes in the weed ﬂora; harmful meaning detrimental to achieving policy objectives. One might deﬁne harm of cultivating the GMHT canola as an increase in the abundance of speciﬁc species of economically damaging weeds, or a decrease in abundance of speciﬁc species that may have aesthetic or nature-conservation value, compared with their abundance under conventional management (e.g., Pimentel et al., 2001). Another option would be the incorporation of some decision-making criteria into the deﬁnitions; thus, one might deﬁne the threshold of unacceptable harm as a 50% increase in the abundance of noxious weed X or as a 25% decrease in the abundance of endangered species Y. Consider a proposal to cultivate a new variety of GMHT canola that is likely to replace long-standing cultivation of a non-GM (“conventional”) canola. Also, suppose that the eﬀects of recommended herbicide applications to the GMHT canola fall under regulations covering GM crops and the eﬀects of recommended herbicide application to the conventional canola are covered by pesticide regulations. Placing Risks in Context of Current Practice In theory, regulations could specify that certain eﬀects are harmful if they are caused by the cultivation of GM crops but are not harmful if caused by other activities. However, such deﬁnitions of harm would violate accepted standards of good regulatory practice. The OECD (2014) describes eight Principles of Regulation, and deﬁning eﬀects as harmful only if they are caused by GM crops would violate at least three of them: Principle 2 that regulations must have a sound legal and empirical basis; Principle 4 that regulations must minimize market distortions; and Principle 7 that regulations should be consistent with other regulations and policies. Hence, deﬁnitions of acceptable risk for GM crops should consider what is regarded as acceptable for other agricultural practices. Many publications have concluded that conceivable harmful environmental eﬀects from cultivating GM crops are of the same type as those from growing non-GM crops (e.g., Tiedje et al., 1989; NRC, 2002; Perry et al., 2004; Lemaux, 2009). Hence, a hypothesis that growing a certain GM crop will cause no harm, is really a hypothesis that growing the GM crop will cause no FIGURE 1 \| Conceptual models of science-led and policy-led risk assessment. April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 3 Policy-Led ERA for GM Crops Raybould and Macdonald imaginations of the researchers in devising ways to categorize diﬀerence. However, few or even none of these diﬀerences may have any relevance to regulatory policy objectives. Consequently, cataloging diﬀerences is at best an ineﬃcient way to conduct risk assessment, because eﬀort is wasted on measurements of no value for decision-making. At worst it is ineﬀective and potentially counterproductive because decisions are made ad hoc in response to statistical signiﬁcance, which can easily be spurious when many variables are measured (Benjamini and Hochberg, 1995; Leek et al., 2017), rather than after serious consideration of what the objectives of agricultural and environmental policies ought to be. We could call this behavior PARKing—Policymaking After the Results are Known—based on Kerr’s (1998) term HARKing for Hypothesizing After the Results are Known. greater harm than the current practice that cultivation of the GM crop may replace. Placing Risks in Context of Current Practice A possible eﬀect of switching from conventional canola to the GMHT canola is a change in the abundance and species diversity of weeds owing to variation in their sensitivity to the diﬀerent herbicides used on these crops (e.g., Perry et al., 2004; Wilson et al., 2007). In assessing the risks posed by cultivating the GMHT canola, the Principles of Regulation suggest that it would be unreasonable to compare the weed ﬂora in the GMHT canola regime with the weed ﬂora if no herbicides were used; the comparison ought to be with the conventional herbicide management. Prior deﬁnition of decision-making criteria means that experiments can be designed to rigorously test risk hypotheses. One could envisage, for example, testing a hypothesis that the abundance of noxious weed X will not increase by more than 50% by testing a hypothesis that it is at least as sensitive to the herbicide that will be applied to the GMHT canola as it is to the herbicides applied to conventional canola. Such a targeted test of a policy-relevant hypothesis would be entail vastly more eﬃcient and eﬀective parameters for data collection than would untargeted comparisons of the weed ﬂoras of GMHT and conventional canola. With best practices, risk assessors will contextualize the risks for cultivating the GMHT canola and compare that with the harm from the cultivation of conventional canola. In the risk assessment, the risk assessor will consider that cultivation of a monoculture and the management of a crop in an agricultural production system reduces biodiversity and has an impact on the environment. The crop plant itself has a suite of traits that result in the production of compounds that create environmental eﬀects and inﬂuence ecosystem services. In the comparative risk assessment, the risk assessor will evaluate the relative impacts of the two phenotypes and evaluate whether the addition of the new trait creates harms that exceed those already imposed by the cultivation of the existing crop. In this scenario, the evaluation does not insist the results of growing the two crops be identical, only that the probability or severity of a harm is not increased. Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org April 2018 \| Volume 6 \| Article 43 Assessing Risks Rather Than Measuring Differences Identifying a fair comparator is only a partial solution to the problem of formulating a useful risk hypothesis. Countless changes in the weed ﬂora are theoretically possible when switching from conventional to GMHT weed management. Science-led risk assessment (Figure 1) might approach this problem by setting up multiple ﬁeld trials at many sites over many years to measure the change in the weed ﬂora when GMHT replaces conventional management; in eﬀect, the hypothesis under test would be one of no diﬀerence between the weed ﬂoras of conventional and GMHT canola. Comparing weed diversity and abundance between conventional and GMHT canola will almost inevitably reveal numerous statistically signiﬁcant diﬀerences (e.g., Heard et al., 2003a,b), with the number limited only by the size of the experiments, the sensitivity of the measuring techniques and the Policy-led risk assessment can target risk management to make interventions in order to realize beneﬁts and reduce harms. In testing the risk hypothesis that the endangered species Y will April 2018 \| Volume 6 \| Article 43 April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 4 Policy-Led ERA for GM Crops Raybould and Macdonald not decrease by more than 25%, testing may reveal that the species is more sensitive to the GMHT herbicide than to the conventional canola herbicide. This ﬁnding could trigger a search for changes to management techniques that ensure weeds are still adequately controlled while minimizing exposure of species Y to the herbicide, perhaps by altering the proposed timing, rate or method of its application (e.g., Thompson et al., 1991). In contrast, unfocussed risk assessment may reveal potential changes in the abundance of numerous species without any attempt to contextualize the risk. Faced with such a ﬁnding, risk managers may simply refuse to approve the GMHT canola (Sanvido et al., 2011), thereby foregoing opportunities and not necessarily reducing risk—although they may have reduced the probability of change. damage, and attributes of grain or ﬁber quality depending on the crop (Horak et al., 2007). The aim of these studies is to identify diﬀerences between the GM crop and its comparator that need further evaluation in order to characterize risk to human and animal health and to the environment from using the GM crop (Kuiper et al., 2001; Nap et al., 2003). Assessing Risks Rather Than Measuring Differences Although not routinely required for regulatory testing, proﬁling of GM crops can also be carried out at the molecular level, using transcriptomics, proteomics or metabolomics (Kuiper et al., 2003). The value of these methods, along with characterization of the epigenome, for crop improvement has recently been discussed by the National Academies of Sciences, Engineering and Medicine (NAS, 2016). Our purpose here is not to evaluate the technical feasibility of molecular proﬁling, but to discuss whether proﬁling approaches generally are valuable in risk assessment of GM crops. In summary, problem formulation for comparative risk assessment of GM crops should consider two important elements. First, the comparison should be consistent with the Principles of Regulation. The eﬀects of using the GM crop should be compared with agricultural practices that these uses will replace. Second, the selection of the hypotheses to be tested in the risk assessment should always be policy-led and informed by science. Policy-led risk assessment will guide risk assessors to develop hypotheses of known relevance to the ﬁnal regulatory decision and suggest experiments that are required to improve decision-making rather than satisfying scientiﬁc curiosity. The combination of hypotheses based on prior agreement of decision- making criteria and rigorous testing maximizes the chances that risk managers will make decisions that fulﬁll agricultural and environmental policy objectives. Risk communication will also be improved. Science-led risk assessment, on the other hand, leads to PARKing: ad hoc decision-making based on whatever diﬀerences happen to reach statistical signiﬁcance in comparisons of many variables. These decisions are unlikely to meet wider policy objectives. They are also likely to create controversy because decisions appear to be ﬁxed by selecting particular data rather than after a debate about what the objectives of policy ought to be (e.g., Sarewitz, 2004). A claimed advantage of proﬁling methods is that they are unbiased (Kuiper et al., 2003). They make no assumptions about how the GM crop might diﬀer from its non-GM counterpart. In addition, unbiased approaches make no judgment about what diﬀerences might be important in indicating that using the GM crop may pose greater risk than similar uses of the comparator. Hence, proﬁling approaches are science-led evaluations of potential diﬀerences with all the problems that entails (Figure 1). In the early days of GM crop development, there was signiﬁcant uncertainty about the extent to which transformation of plants could lead to unintended changes. PROFILING IN RISK ASSESSMENT In the example above, we proposed that rigorous testing of targeted hypotheses is a more eﬃcient and eﬀective approach to risk assessment than are untargeted tests of null hypotheses of no diﬀerence between a GM and a non-GM cropping system. The latter approach makes use of proﬁling—the characterization of a system by describing a combination of many of its attributes. It is important to recognize that comparing nutrients is policy- led risk assessment because protecting human and animal health is a policy objective. To keep the risk assessment policy-led, however, it is important that the substances tested really are determinates of health. If the most extreme conceivable change in the amount of a substance would have no material eﬀect on health, then that substance should be of no concern for policy- led risk assessment, and comparing its concentration in the GM and non-GM crop should not be necessary to determine risk. Assessing Risks Rather Than Measuring Differences Hence, compositional and phenotypic proﬁling of GM crops made sense as methods to explore the extent of these changes: testing the hypothesis that transformation introduces no unintended changes was a useful tool for basic research into the eﬀects of transgenesis and also for risk assessors struggling to characterize products of new technology. In retrospect, however, there was always a need to ensure that these studies were placed in context when used to inform the risk assessment. In practice, this has generally been the case when a GM crop and its non-modiﬁed counterpart are compared. For example, as changes in the nutritional value of a crop could be harmful to human and animal health, the risk assessor determines whether the amounts of key nutritional components are statistically diﬀerent between the GM and non-GM comparator. If statistically signiﬁcant diﬀerences are identiﬁed, the assessor will ask whether the amounts in the GM crop fall into the normal range for that crop. If they do, the diﬀerences will generally be disregarded. April 2018 \| Volume 6 \| Article 43 Proﬁling Using Omics Methods The introduction of molecular proﬁling methods into regulatory risk assessments would only increase the pervasiveness of unfocussed data generation rather than policy-led attitudes to risk assessment. Additional data generation will often pose questions for which there are no ready answers leading to a continuing need to produce yet more data. The ability to ﬁnd diﬀerences between a GM crop and its non-GM comparator is virtually limitless, creating endless opportunities for PARKing. Advocates of molecular proﬁling may argue that the methods could show that variation between GM and non-GM plants as a class is insigniﬁcant compared with variation among non- GM plants. However, this misses the point. The purpose of regulatory risk assessment is not to make general points about a technology or class of products, it is to evaluate whether the risks posed by a speciﬁc use of a speciﬁc product are acceptable. Acceptability of risk is ultimately a policy decision, and anything that promotes policymaking as an ad hoc response to possibly spurious statistically signiﬁcant diﬀerences, rather than careful deliberation about delivering agreed societal objectives, should be discouraged. However, while lack of bias in testing a hypothesis is a virtue in risk assessment, as in all basic and applied science, lack of bias in selecting the hypotheses to be tested is a grave weakness: we should be strongly biased toward hypotheses that help decision-making and realization of policy objectives. Without this bias, policy may be formulated in response to trivial diﬀerences, perhaps inﬂuenced by ill-informed indignation that a GM crop, unsurprisingly, diﬀers from a non-GM comparator in some respect. It is this very lack of bias that we believe makes science-led risk assessment vastly less eﬀective than the policy-led alternative. In advocating policy-led risk assessment, we do not underestimate the diﬃculties agreeing on policy objectives. Disagreement about what comprise beneﬁcial or harmful eﬀects of using certain GM crops is rife, even within organizations that develop and regulate them. However, sooner or later policy objectives have to be set in order to make decision-making feasible and hence risk assessment eﬃcient and eﬀective. While deﬁning these objectives may be controversial, such controversy is likely to be less than that produced by making policy ad hoc in response to possibly spurious statistically signiﬁcant diﬀerences identiﬁed by untargeted proﬁling methods. Ultimately, decision- makers have to decide based on their individual or organizational policy objectives. AUTHOR CONTRIBUTIONS Comparative risk assessment is a valuable method for making risk assessment tractable, provided that it is policy-led rather than science-led. Ideally, policy-led comparative risk assessment for a GM crop would deﬁne eﬀects that comprise unacceptable All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. Proﬁling Using Omics Methods This responsibility cannot be outsourced to statistical algorithms processing vast amounts of proﬁling data. Finally, our point is not that omics methods can never have value in regulatory risk assessment. If measurements of speciﬁc transcripts, proteins or metabolites are a good test of a hypothesis that a given use of a given GM crop does not pose an unacceptable increase in risk, then the measurements may have value for regulatory decision-making. However, using the methods simply to create proﬁles will be a serious impediment to moving from science-led to policy-led risk assessment and decision-making. Historic and Current Use of Proﬁling in Risk Assessment Proﬁling of GM crops is used widely in risk assessment. Compositional analysis typically tests for statistically signiﬁcant diﬀerences between the GM crop and a near-isogenic comparator variety in the amounts of 60–80 nutrients and anti-nutrients (Herman and Price, 2013). Phenotypic characterization compares 30 or more aspects of germination, plant growth and development, morphology, reproduction, disease and pest Without prior deﬁnitions of important changes, science-led proﬁling can encourage the idea that producing more data April 2018 \| Volume 6 \| Article 43 April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 5 Raybould and Macdonald Policy-Led ERA for GM Crops increases in risk from its use. The comparison would be with the acceptable eﬀects of a similar crop in a similar agricultural system that is likely to be replaced by use of the GM crop. inevitably leads to better risk assessment. Statistically non- signiﬁcant comparisons of thousands of substances may appear to be a more convincing demonstration of negligible risk than is the lack of diﬀerence in a few key nutrients. However, unless it is possible to specify values of particular variables that would show a policy-led risk hypothesis to be false, the data are of no relevance for drawing conclusions about risk. Finally, proﬁling may also understate the importance of policy in risk assessment and decision-making. It seems to promote the idea that if suﬃcient data are collected, uncertainty will be diminished and the “correct” policy toward the use of GMOs will become obvious. Deﬁning an unacceptable increase in risk enables the formulation of testable hypotheses for risk assessment. At their most conservative, the hypotheses will be that certain eﬀects are no more likely to occur, and if they do occur, are no more severe than those caused by use of the crop that will be replaced. Only data that test such hypotheses, that is, are able to show them to be false, are useful for such policy-led risk assessment. The alternative method of comparative risk assessment dispenses with policy objectives and makes numerous tests of the null hypothesis that the GM crop does not diﬀer from the crop that it will replace. Such “science-led” risk assessment makes no judgment about the importance of the variables being measured. Proponents of this method of risk assessment see this unbiased nature of the risk assessment as a strength (e.g., Kuiper et al., 2003). de Groot, R. S., Alkemade, R., Braat, L., Hein, L., and Willemen, L. (2010). Challenges in integrating the concept of ecosystem services and values in Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B 57, 289–300. Andow, D., and Hilbeck, A. (2004). Science-based risk assessment for nontarget eﬀects of transgenic crops. BioScience 54, 637–649. doi: 10.1641/0006- 3568(2004)054[0637:SRAFNE]2.0.CO;2 Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B 57, 289–300. Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. B 57, 289–300. de Groot, R. S., Alkemade, R., Braat, L., Hein, L., and Willemen, L. (2010). Challenges in integrating the concept of ecosystem services and values in REFERENCES GM Crops 1, 25–31. doi: 10.4161/gmcr.1.1.9776 Heard, M. S., Hawes, C., Champion, G. T., Clark, S. J., Firbank, L. G., Haughton, A. J., et al. (2003b). Weeds in ﬁelds with contrasting conventional and genetically modiﬁed herbicide-tolerant crops. II. Eﬀects on individual species. Philos. Trans. R. Soc. Lond. B 358, 1833–1846. doi: 10.1098/rstb.2003.1401 Sanvido, O., Romeis, J., and Bigler, F. (2011). Environmental change challenges decision-making during post-market environmental monitoring of transgenic crops. Transgenic Res. 20, 1191–1201. doi: 10.1007/s11248-011- 9524-8 Herman, R. A., and Price, W. D. (2013). Unintended compositional changes in genetically modiﬁed (GM) crops: 20 years of research. J. Agric. Food Chem. 61, 11695–11701. doi: 10.1021/jf400135r Hill, R. A., and Sendashonga, C. (2003). General principles for risk assessment of living modiﬁed organisms: lessons from chemical risk assessment. Environ. Biosafety Res. 2, 81–88. doi: 10.1051/ebr:2003004 Sanvido, O., Romeis, J., Gathmann, A., Gielkens, M., Raybould, A., and Bigler, F. (2012). Evaluating environmental risks of genetically modiﬁed crops – ecological harm criteria for regulatory decision-making. Environ. Sci. Policy 9, 82–91. doi: 10.1016/j.envsci.2011.08.006 Horak, M. J., Rosenabum, E. W., Woodrum, C. L., Martens, A. B., Mery, R. F., Cothren, J. T., et al. (2007). Characterization of Roundup Ready ﬂex cotton, “MON 88913”, for use in ecological risk assessment: evaluation of seed germination, vegetative and reproductive growth, and ecological interactions. Crop Sci. 47, 268–277. doi: 10.2135/cropsci2006.02.0063 Sarewitz, D. (2004). How science makes environmental controversies worse. Environ. Sci. Policy 7, 385–403. doi: 10.1016/j.envsci.2004.06.001 Slovic, P. (1999). Are trivial risks the greatest risks of all? J. Risk Res. 2, 281–288. doi: 10.1080/136698799376727 Crop Sci. 47, 268–277. doi: 10.2135/cropsci2006.02.0063 Industry Canada (1998). The 1998 Canadian Biotechnology Strategy: An Ongoing Renewal Process. Ottawa, ON: Industry Canada. Smythe, S. J., Gusta, M., Belcher, K., Phillips, P., and Castle, D. (2011). Environmental impacts from herbicide tolerant canola production in Western Canada. Agric. Syst. 104, 403–410 doi: 10.1016/j.agsy.2011.01.004 Kerr, N. L. (1998). HARKing: hypothesizing after the results are known. Pers. Soc. Psychol. Rev. 2, 196–217. doi: 10.1207/s15327957pspr0203_4 Thompson, J. F., Staﬀord, J. V., and Miller, P. C. H. (1991). Potential for automatic weed detection and selective herbicide application. Crop Protection 10, 254–259. doi: 10.1016/0261-2194(91)90002-9 Kuiper, H. A., Kleter, G. A., Noteborn, H. P., and Kok, E. J. (2001). Assessment of the food safety issues related to genetically modiﬁed foods. Plant J. 27, 503–528. doi: 10.1046/j.1365-313X.2001.01119.x Tiedje, J. M., Colwell, R. K., Grossman, Y. L., Hodson, R. E., Lenski, R. E., Mack, R. REFERENCES Andow, D., and Hilbeck, A. (2004). Science-based risk assessment for nontarget eﬀects of transgenic crops. BioScience 54, 637–649. doi: 10.1641/0006- 3568(2004)054[0637:SRAFNE]2.0.CO;2 Andow, D., and Hilbeck, A. (2004). Science-based risk assessment for nontarget eﬀects of transgenic crops. BioScience 54, 637–649. doi: 10.1641/0006- 3568(2004)054[0637:SRAFNE]2.0.CO;2 April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 6 Raybould and Macdonald Policy-Led ERA for GM Crops landscape planning, management and decision making. Ecol. Complex. 7, 260–272. doi: 10.1016/j.ecocom.2009.10.006 NAS (National Academies of Sciences, Engineering, and Medicine) (2016). Genetically Engineered Crops: Experiences and Prospects. Washington, DC: The National Academies Press. Evans, J., Wood, G., and Miller, A. (2006). The risk assessment–policy gap: an example from the UK contaminated land regime. Environ. Int. 32, 1066–1071. doi: 10.1016/j.envint.2006.06.002 NRC (National Research Council) (2002). Environmental Eﬀects of Transgenic Plants. Washington, DC: National Academy Press. Gabler, M. (2008). Intergovernmental Relations in Food Biotechnology Governance: Complementary Disentanglement in Regulation with Collaboration in Food Safety and Inspection. Queens’ University Institute of Governmental Relations working paper. Available online at http://www.queensu. ca/iigr/sites/webpublish.queensu.ca.iigrwww/ﬁles/ﬁles/WorkingPapers/ PublicHealthSeries/gablerpaperﬁnal.pdf (Accessed 2 April 2018). OECD (Organisation for Economic Cooperation and Development) (2014). The Governance of Regulators, OECD Best Practice Principles for Regulatory Policy. Paris: OECD Publishing. Perry, J. N., Firbank, L. G., Champion, G. T., Clark, S. J., Heard, M. S., May, M. J., et al. (2004). Ban on triazine herbicides likely to reduce but not negate the relative beneﬁts of GMHT maize cropping. Nature 428, 313–316. doi: 10.1038/nature02374 Gusta, M., Smyth, S., Belcher, K., Phillips, P., and Castle, D. (2011). Economic beneﬁts of genetically-modiﬁed herbicide-tolerant canola for producers. AgBioForum 14, 1–13. Available online at: http://agbioforum.org/v14n1/ v14n1a01-smyth.pdf Pimentel, D., McNair, S., Janecka, J., Wightman, J., Simmonds, C., O’Connell, C., et al. (2001). Economic and environmental threats of alien plant, animal, and microbe invasions. Agric. Ecosyst. Environ. 84, 1–20. doi: 10.1016/S0167-8809(00)00178-X Heard, M. S., Hawes, C., Champion, G. T., Clark, S. J., Firbank, L. G., Haughton, A. J., et al. (2003a). Weeds in ﬁelds with contrasting conventional and genetically modiﬁed herbicide-tolerant crops. I. Eﬀects on abundance and diversity. Philos. Trans. R. Soc. Lond. B 358, 1819–1832. doi: 10.1098/rstb.2003.1402 Raybould, A. (2010a). The bucket and the searchlight: formulating and testing risk hypotheses about the weediness and invasiveness potential of transgenic crops. Environ. Biosafety Res. 9, 123–133. doi: 10.1051/ebr/2011101 Raybould, A. (2010b). Reducing uncertainty in regulatory decision-making for transgenic crops: more ecological research or clearer environmental risk assessment? REFERENCES N., et al. (1989). The planned introduction of genetically engineered organisms: ecological considerations and recommendations. Ecology 70, 298–315. doi: 10.2307/1937535 Kuiper, H. A., Kok, E. J., and Engel, K.-H. (2003). Exploitation of molecular proﬁling techniques for GM food safety assessment. Curr. Opin. Biotechnol. 14, 238–243. doi: 10.1016/S0958-1669(03)00021-1 Wilson, R. G., Miller, S. D., Westra, P., Kniss, A. R., Stahlman, P. W., Wicks, G. W., et al. (2007). Glyphosate-induced weed shifts in glyphosate-resistant corn or a rotation of glyphosate-resistant corn, sugarbeet, and spring wheat. Weed Technol. 21, 900–909. doi: 10.1614/ WT-06-199.1 Kuntz, M., Davison, J., and Ricroch, A. E. (2013). What the French ban of Bt MON810 maize means for science-based risk assessment. Nat. Biotechnol. 31, 498–500. doi: 10.1038/nbt.2613 Leek, J., McShane, B. B., Gelman, A., Colquhoun, D., Nuijten, M. B., and Goodman, S. N. (2017). Five ways to ﬁx statistics. Nature 551, 557–559. doi: 10.1038/d41586-017-07522-z Conﬂict of Interest Statement: During the writing of this paper AR was employed by Syngenta and PM was employed by the Canadian Food Inspection Agency. Lemaux, P. G. (2009). Genetically engineered plants and foods: a scientist’s analysis of the issues (part II). Annu. Rev. Plant Biol. 60, 511–559. doi: 10.1146/annurev.arplant.043008.092013 Macdonald, P. (2014). Genetically modiﬁed organisms regulatory challenges and science: a Canadian perspective. J. Verbraucher Lebensmittelsicherheit 9(Suppl. 1), S59–S64. doi: 10.1007/s00003-014-0893-9 Copyright © 2018 Raybould and Macdonald. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Nap, J. P., Metz, P. L., Escaler, M., and Conner, A. J. (2003). The release of genetically modiﬁed crops into the environment. Part I. Overview of current status and regulations. Plant J. 33, 1–18. doi: 10.1046/j.0960-7412.2003.01602.x April 2018 \| Volume 6 \| Article 43 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org
https://openalex.org/W2844672055	https://openresearch.surrey.ac.uk/view/delivery/44SUR_INST/12139840760002346/13140569430002346	English	null	Towards connected autonomous driving: review of use-cases	Vehicle system dynamics/Vehicle System Dynamics	2,018	cc-by	23,638	Vehicle System Dynamics International Journal of Vehicle Mechanics and Mobility Vehicle System Dynamics International Journal of Vehicle Mechanics and Mobility ISSN: 0042-3114 (Print) 1744-5159 (Online) Journal homepage: http://www.tandfonline.com/loi/nvsd20 KEYWORDS Connected autonomous vehicles; cooperative driving; use-case analysis Towards connected autonomous driving: review of use-cases Umberto Montanaro a, Shilp Dixit a, Saber Fallaha, Mehrdad Dianatib, Alan Stevensc, David Oxtobyd and Alexandros Mouzakitisd aCentre for Automotive Engineering, University of Surrey, Guildford, UK; bWarwick Manufacturing Group, University of Warwick, Coventry, UK; cTransport Research Laboratory, Wokingham, UK; dJaguar Land Rover Limited, Coventry, UK Umberto Montanaro, Shilp Dixit, Saber Fallah, Mehrdad Dianati, Alan Stevens, David Oxtoby & Alexandros Mouzakitis Umberto Montanaro, Shilp Dixit, Saber Fallah, Mehrdad Dianati, Alan Stevens, David Oxtoby & Alexandros Mouzakitis To cite this article: Umberto Montanaro, Shilp Dixit, Saber Fallah, Mehrdad Dianati, Alan Stevens, David Oxtoby & Alexandros Mouzakitis (2018): Towards connected autonomous driving: review of use-cases, Vehicle System Dynamics, DOI: 10.1080/00423114.2018.1492142 Published online: 09 Jul 2018. Submit your article to this journal View Crossmark data VEHICLE SYSTEM DYNAMICS https://doi.org/10.1080/00423114.2018.1492142 ABSTRACT ARTICLE HISTORY Received 7 November 2017 Revised 22 May 2018 Accepted 1 June 2018 Connected autonomous vehicles are considered as mitigators of issues such as traffic congestion, road safety, inefficient fuel consumption and pollutant emissions that current road trans- portation system suffers from. Connected autonomous vehicles utilise communication systems to enhance the performance of autonomous vehicles and consequently improve transportation by enabling cooperative functionalities, namely, cooperative sensing and cooperative manoeuvring. The former refers to the ability to share and fuse information gathered from vehicle sensors and road infrastructures to create a better understanding of the surrounding environment while the latter enables groups of vehicles to drive in a co-ordinated way which ultimately results in a safer and more effi- cient driving environment. However, there is a gap in understanding howandtowhat extent connectivity can contribute to improving the efficiency, safety and performance of autonomous vehicles. There- fore, the aim of this paper is to investigate the potential benefits that can be achieved from connected autonomous vehicles through analysing five use-cases: (i) vehicle platooning, (ii) lane changing, (iii) intersection management, (iv) energy management and (v) road friction estimation. The current paper highlights that although con- nectivity can enhance the performance of autonomous vehicles and contribute to the improvement of current transportation system per- formance, the level of achievable benefits depends on factors such as the penetration rate of connected vehicles, traffic scenarios and the way of augmenting off-board information into vehicle control systems. 1. Introduction Road safety, ineffective use of the roadway infrastructure, inefficient fuel consumption and pollutant emissions are the main challenges associated with the current transportation sys- tem [1]. In terms of road safety, it is noted that about 92% of the road crashes are mainly caused by human recognition errors (e.g. drivers inattention, drivers distractions and inad- equate surveillance) and human decision errors (e.g. driving too fast, false assumption of others actions and misjudgement of gap or others speed) [2,3]. Also, it is remarked CONTACT Saber Fallah s.fallah@surrey.ac.uk © 2018 Informa UK Limited, trading as Taylor & Francis Group CONTACT Saber Fallah s.fallah@surrey.ac.uk © 2018 Informa UK Limited, trading as Taylor & Francis Group 2 U. MONTANARO ET AL. 2 2 that erroneous drivers decisions are not only responsible for road fatalities but also for the underuse of road infrastructure and excessive fuel consumption. For instance, it was shown in [4] that erroneous driving styles which include excessive acceleration/braking and engine idling have a notable negative impact on fuel consumption (e.g. an increase of about 3–5 L per 100 km) and emission pollutions. Also, it has been noted that current road infrastructures are not operated at their maximum capacity by human-driven vehi- cles and only 11% of the road lane length of highways are occupied by vehicles while the remaining 89% represents the gaps that the drivers need to maintain behind other vehi- cles in order to feel safe while driving at high speed [1]. Connected Autonomous Vehicles (CAVs) which benefit from the autonomous vehicle and connected vehicle technologies are considered as a potential mitigator of current transportation system challenges. It is noted that an autonomous vehicle is able to perceive its environment, and use control sys- tems to autonomously plan vehicle motion and decide vehicle manipulation commands. Autonomous vehicles can improve road safety through precise control of the position and velocity (e.g. tight control of inter-vehicular distances) and smaller reaction time compared with human-driven vehicles [1]. Furthermore, they can operate the engine and vehicle powertrain in regions with high efficiency resulting in a reduction of fuel consumption and pollutant emissions. However, the performance of non-connected autonomous vehicles in different manoeuvring scenarios and traffic conditions are restricted due to the limitation of on-board sensing systems (e.g. direct line of sight, sensor accuracy in the wide range of environment conditions, sensor operation range, etc.). 1. Introduction On the other hand, connected vehi- cles refer to those vehicles which are capable of exchanging information with other vehicles or roadside infrastructures by using wireless communication technologies [5]. Vehicle-to- Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication technologies enable vehicles to access external information that cannot be directly obtained through on-board vehicle sensing systems (e.g. the position of vehicles beyond the line of sight, traffic data ahead and roadway conditions). This external information can be fused with on-board observation to create a better perception of the surrounding environment. It is reported in [6,7] that equipping autonomous vehicles with communication systems significantly enhance their performance and consequently the efficiency of transportation systems. Also, it enables cooperative manoeuvring which is the ability to jointly plan trajectories and decisions so that a common goal for the fleet, e.g. minimisation of fuel consumption while preserving safety, can be achieved [8]. It is noted that cooperative manoeuvring is not possible neither with the sole use of connected vehicles nor with autonomous vehi- cles but it is a feature that emerges when connectivity is added to autonomous vehicles. The potential of a Vehicle-to-Everything (V2X) platform in expanding the capabilities of autonomous vehicles was also initially investigated through European research projects such as Cybercars-2 [9], SARTRE [10] and AutoNet2030 [11], just to name a few. The project Cybercars-2 (2006–2009) focused on the design, development, prototyping and on-road testing of cooperative driving manoeuvers, such as overtaking and crossing inter- sections not operated by traffic lights [9]. The results showed that in a cooperative driving i f b d hil f i i ll h d that erroneous drivers decisions are not only responsible for road fatalities but also for the underuse of road infrastructure and excessive fuel consumption. For instance, it was shown in [4] that erroneous driving styles which include excessive acceleration/braking and engine idling have a notable negative impact on fuel consumption (e.g. an increase of about 3–5 L per 100 km) and emission pollutions. Also, it has been noted that current road infrastructures are not operated at their maximum capacity by human-driven vehi- cles and only 11% of the road lane length of highways are occupied by vehicles while the remaining 89% represents the gaps that the drivers need to maintain behind other vehi- cles in order to feel safe while driving at high speed [1]. 1. Introduction Connected Autonomous Vehicles (CAVs) which benefit from the autonomous vehicle and connected vehicle technologies are considered as a potential mitigator of current transportation system challenges. It is noted that an autonomous vehicle is able to perceive its environment, and use control sys- tems to autonomously plan vehicle motion and decide vehicle manipulation commands. Autonomous vehicles can improve road safety through precise control of the position and velocity (e.g. tight control of inter-vehicular distances) and smaller reaction time compared with human-driven vehicles [1]. Furthermore, they can operate the engine and vehicle powertrain in regions with high efficiency resulting in a reduction of fuel consumption and pollutant emissions. However, the performance of non-connected autonomous vehicles in different manoeuvring scenarios and traffic conditions are restricted due to the limitation of on-board sensing systems (e.g. direct line of sight, sensor accuracy in the wide range of environment conditions, sensor operation range, etc.). On the other hand, connected vehi- cles refer to those vehicles which are capable of exchanging information with other vehicles or roadside infrastructures by using wireless communication technologies [5]. Vehicle-to- Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication technologies enable vehicles to access external information that cannot be directly obtained through on-board vehicle sensing systems (e.g. the position of vehicles beyond the line of sight, traffic data ahead and roadway conditions). This external information can be fused with on-board observation to create a better perception of the surrounding environment. It is reported in [6,7] that equipping autonomous vehicles with communication systems significantly enhance their performance and consequently the efficiency of transportation systems. Also, it enables cooperative manoeuvring which is the ability to jointly plan trajectories and decisions so that a common goal for the fleet, e.g. minimisation of fuel consumption while preserving safety, can be achieved [8]. It is noted that cooperative manoeuvring is not possible neither with the sole use of connected vehicles nor with autonomous vehi- cles but it is a feature that emerges when connectivity is added to autonomous vehicles. The potential of a Vehicle-to-Everything (V2X) platform in expanding the capabilities of autonomous vehicles was also initially investigated through European research projects such as Cybercars-2 [9], SARTRE [10] and AutoNet2030 [11], just to name a few. The project Cybercars-2 (2006–2009) focused on the design, development, prototyping and on-road testing of cooperative driving manoeuvers, such as overtaking and crossing inter- sections not operated by traffic lights [9]. 1. Introduction The results showed that in a cooperative driving environment, safety can be guaranteed while performing potentially hazardous manoeu- vres such as overtaking a stationary obstacle in a single carriageway. The SARTRE project (2009–2012) focused on developing strategies to operate vehicle platoons on normal pub- lic highways with the aim to improve fuel efficiency, safety and reduce congestion. Results from the project showed that fuel savings (potentially up to 20%) can be achieved as a VEHICLE SYSTEM DYNAMICS 3 result of cooperative platooning. The AutoNet2030 (2013–2016) project studied how con- nected autonomous vehicles can negotiate manoeuvers and interact with manually driven vehicles in a safe and reliable way by two demonstrations [11], i.e. (i) vehicle platooning in mixed traffic and (ii) vehicle cooperation in low speed driving scenarios (i.e. car following, merging, lane changing and intersections). Currently, there are research projects in the UK which aim to further investigate the use of V2X for autonomous vehicles. For instance, the target of the project UKCITE ( 2016–2018) is to create an environment for real-time test- ing of connected autonomous vehicles [12]. It involves equipping over 40 miles of urban roads, dual-carriageways and motorways within Coventry and Warwickshire with V2X technologies. The i-MOTORS project (2016–2018) is devoted to developing a vehicular cloud computing platform that fuses data from vehicles with information from the road environment to create dynamic maps and real-time alerts of possible roadway hazards [13]. Expected benefits from the i-MOTORS project are twofold, (i) reduction in fuel consump- tion and travel time by considering real-time traffic data for active route planning, and (ii) improvement in road safety via car platooning. The G-ACTIVE project (2016–2019) targets a reduction of fuel consumption for passenger and light duty road vehicles for a range of drivetrain architectures (conventional, electric and hybrid electric) by leveraging off-board data including traffic condition and timing of traffic lights [14]. This off-board information will be used to simultaneously optimise drivetrain energy and vehicle driving speed. The aim of the CARMA project (2016–2021) is to develop and test a cooperative automated driving technology based on a distributed control system enabled by an ultra- low latency and highly reliable cloud-based infrastructure [15]. Although there are several past and on-going research activities in the domain of CAV technology, the potentials and limitations of this technology in addressing the issues of current transportation system are not well investigated [16]. 2. Vehicle platooning A vehicle platoon is a group of two or more consecutive automated vehicles, also denoted as a string of vehicles, travelling along a highway in the same lane with a short inter-vehicle distance and at the same velocity. Typically, spacing strategy (also denoted as spacing pol- icy) has been used to define the required inter-vehicular distance, while the target speed is decided by either a lead vehicle (the first vehicle in the string which can be either an autonomous or human-driven vehicle) or, if available, by the road infrastructure which acts as a virtual leader [18]. There are several benefits of organising the road vehicles in platoons, for instance (i) increasing traffic flow while reducing traffic shock waves, (ii) reducing fuel consumption and pollutant emissions and (iii) improving road safety. Vehicle platooning also improves drivers comfort through (i) providing drivers the opportunity to focus on other activities during the trip, and (ii) generating smooth speed variations result- ing in having less jerk compared with vehicles under human control. It is noted that the key parameters for increasing traffic flow through platooning are the inter-vehicular dis- tance and the number of vehicle participating in the platoon (i.e. platoon lenght). The traffic flow increases when the inter-vehicular gap reduces or the number of the vehicles in a string increases. In the case that all vehicles of a highway are organised in sets of platoons, it is pos- sible to increase traffic flow about four times and reach a value up to 8000 vehicle/hour/lane for automated highways [19]. However, in a mixed traffic scenario, the achievable traffic flow and reduction of shock waves increase as the fraction of vehicles grouped in platoons increases [20,21]. Shortening the inter-vehicular distances also results in a reduction of fuel consumption of the platoon due to the reduction of aerodynamic drag force acting on each vehicle in the fleet. As documented by the earliest outcome of the PATH program [22,23], the aerodynamic drag force experienced by platoons reduces when the inter-vehicular gap reduces or the number of vehicles in the platoon increases. Furthermore, the average reduc- tion in the drag force approaches a limit as the number of platoon members increase and this limit is a function of inter-vehicular spacing. 1. Introduction Therefore, this paper is devoted to analysing achievable benefits of exploiting off-board data gathered via V2X communications, and inter-vehicular coop- eration on autonomous vehicles. To investigate the potentials and limitations of CAVs, a set of five use-cases is chosen and analysed through the results presented in the current technical literature. The first four use cases (i.e. vehicle platooning, lane change, intersec- tion management and energy management) have been selected to show examples of how connectivity can support cooperative manoeuvring thereby improving road transporta- tion whereas the last use-case (i.e. cooperative road friction estimation) is dedicated to demonstrating how perception of the surrounding environment can be improved when vehicles cooperatively share their local perception knowledge. The analysis of cooperative localisation systems has been performed in a separate work by the authors and is reported in [17]. The remainder of the paper is organised as follow. Section 2 and Section 3 investigate the benefits of cooperative driving in highway scenarios. Section 2 is devoted to the analysis of vehicle platooning while Section 3 examines lane change manoeuvres supported by com- munication systems. Intersection management is analysed in Section 4, to study achievable benefits through V2X communication systems in urban environments. In Section 5, it is discussed how off-board information gathered via communication systems can be used to reduce energy consumption. Cooperative estimation of the road friction for supporting on-board safety systems of CAVs is analysed in Section 6. Finally, conclusions are drawn in Section 7. 4 U. MONTANARO ET AL. 2. Vehicle platooning The reduction of aerodynamic drag force results in the reduction of required drive torque and therefore less fuel consumption, which is expected to be on average 20% for the inter-vehicular gap of 0.2 vehicle length (about 1 m). Similar conclusion regarding fuel efficiency of vehicles operating in platoons have also been drawn more recently in [24,25]. It is noted that, as human driver reaction time (of about 0.25 s) is too large to guarantee collision avoidance while driving with such reduced inter-vehicular distances, high vehicle automation is required to implement platoon sce- narios [26] which also improve road safety through tight control of the inter-vehicular gap and speed [27]. However, it is worth mentioning that excessive shortening of inter- vehicular distance can have a negative impact on the passengers comfort [28]. Platoons of autonomous vehicles must guarantee two stability criteria: (i) individual vehicle stability and (ii) string stability. Individual stability requires that the difference between the refer- ence inter-vehicular gap and the actual one (spacing error) converges to zero when the lead vehicle has a constant speed [29,30]. The convergence speed of the spacing error to zero, known as ‘stability margin’ is used as an index to evaluate the performance of the platoon in establishing the required cooperative motion [31,32]. String stability refers to the stability of all vehicles travelling together in the platoon and needs to be robust against perturbations of motion of the lead vehicle [33]. Unstable strings of vehicles can induce traffic waves which force the following vehicles into sudden accelerations/decelerations VEHICLE SYSTEM DYNAMICS 5 slowdown or standstill, thus reducing traffic flow and drivers comfort, while increasing at the same time fuel consumption [34] and the possibility of rear-end collisions [35]. It is noted that autonomous vehicles that adjust the inter-vehicular distance based only on the position of the lead vehicle (measured through local sensors) have limited capability to be operated in platoons [35,36]. Furthermore, the limitation of autonomous vehicles to be organised in platoons depends also on the platoon spacing strategies. In the con- text of vehicle platooning, spacing strategies are mainly classified as: (i) constant policy, and (ii) headway policy (also known as velocity-dependent spacing policy). When a con- stant policy is adopted, the inter-vehicular gap is independent from the velocities of the vehicles. 2. Vehicle platooning This policy can increase throughput more than the headway policy and has been used initially to show the benefits of automated highways to increase road capacity [25]. On the other hand, in the headway policy, each vehicle adapts the inter-vehicular distance linearly based on its own speed and the time headway where reducing the time headway will increase the achievable benefits from platooning. However, it is noted that, for both constant and headway policies, maintaining the string stability is a major concern. For the constant policy, it was proven that the string stability cannot be guaranteed when pla- toon control systems make decision based only on the inter-vehicular gap [37,38] while for the headway policy the string stability suffers when reducing the time headway [33,35]. Through V2X communication, it is possible to mitigate the limitations of both constant and headway policies. For instance, a multitude of information obtained from all or a frac- tion of the vehicles in the platoon (e.g. acceleration of the lead vehicle, spacing error of other vehicles in the string, vehicles position, velocity and accelerations, etc.) can guar- antee the string stability of constant policy [37]. On the other hand, platoons of CAVs can be operated also with a time-headway equal to 0.6 s [39], which is also accepted by passengers as experimentally shown in [28]. Moreover, by increasing the range of infor- mation shared by vehicles in a platoon, better performance in terms of string stability can be achieved. For instance, in [39], it was shown that by including information from the vehicle ahead the lead vehicle (defined as second lead vehicle) into the control actions of subject vehicle1 (see Figure 1(a)), it is possible to reduce the minimum time headways with respect to the case where only information from the lead vehicle was used without jeopar- dising string stability. Another example is documented in [40] where it was proven that by including information from the following vehicle into the computation of the acceleration of the subject vehicle (see Figure 1(b)), it is possible to preserve the cooperative motion also when one of the vehicles in the string has limited speed performance compared to the other vehicles in the fleet. Thus, there is a proof from the literature that varying the num- ber of communication links has an effect on the string stability of a platoon. 2. Vehicle platooning The idea of improving the performance and robustness of vehicle platoons by increasing the number (a) (b) Figure 1. (a) Platoon scenario presented in [39] and (b) platoon scenario presented in [40]. Grey arrows indicate the communication links carrying data among vehicles. Figure 1. (a) Platoon scenario presented in [39] and (b) platoon scenario presented in [40]. Grey arrows indicate the communication links carrying data among vehicles. U. MONTANARO ET AL. 6 6 of communication links between vehicles has recently motivated researchers to investigate the possibility of coordinating vehicle motion by using different communication network topologies2 between vehicles of a platoon; see, for instance, [42,43] and references therein. The focus of the platoon control system is then to guarantee platoon formation and stabil- ity independently from the underlying communication network topology while assuring robustness to uncertain and time varying communication delays. However, the current literature does not provide a systematic approach either for the selection of the underlying network topology or for the information that vehicles in the platoon should share to achive the synchronised motion, and usually only platoons with homogenous vehicles (i.e. with equal acceleration/deceleration capability, equal actuator delays, etc.) with the same preferences (such as spacing policy, time headway, etc.) are considered. The information shared among the vehicles for maintaining the synchronised motion of the platoon depend on the vehicle dynamic model used for the design of platoon control algorithms. Usually, when the control objective is to impose individual and string stability, only the longitudinal vehicle dynamics are considered. According to the automotive lit- erature, the longitudinal control system architecture for vehicle platooning is hierarchical where it is composed by an upper-level controller, known also as Cooperative Adaptive Cruise Control (CACC) system, and a low-level controller [44,45]. The former determines either the desired acceleration or driving/braking torque for each vehicle based on infor- mation collected from its neighbours, while the latter is instead locally used to generate for each vehicle the throttle and/or brake commands required to exactly track the desired references as planned by the upper-level CACC. 2. Vehicle platooning Figure 2 depicts the two-level platoon con- trol architecture where the Estimation & Ambient Condition block is used to reconstruct information not directly measured (such as the air mass incoming into the engine) by on- board sensors and provide ambient data (such as road slope) to the Inner-loop low-level controller, while the Localisation block is used to localise the vehicle on a Local Dynamic Map (LDM). The reference speed is provided to the leader by planning controllers, e.g. speed optimisation controllers for the minimisation of the fuel consumption of the fleet (see Section 5 for further details on speed optimisation supported by V2X communica- tion), and the V2X devices are used to gather information from the neighbouring vehicle in the platoon. Figure 2. General control architecture for vehicle platooning. Figure 2. General control architecture for vehicle platooning. 7 7 VEHICLE SYSTEM DYNAMICS It is noted that both the upper and the low-level controllers are designed based on the longitudinal vehicle dynamics. However, models with different fidelity have been used to capture the longitudinal motion, for instance, detailed models are adopted for design- ing inner-loop controllers while simple models are exploited to design CACC algorithms [42]. Furthermore, it is noted that several simplified longitudinal vehicle models have been adopted in the technical literature for devising of CACC strategies. These models can be classified based on the order of the resulting dynamical system (i.e. the number of differ- ential equations describing the vehicle state or motion) and they might contain nonlinear terms such as drag force and rolling resistance. Typically, simplified longitudinal vehicle dynamics are described using (i) first-order models (single integrators), (ii) second-order models (double integrators and linear/nonlinear damper-mass systems) and third-order models (linear/nonlinear third-order systems). When single integrators (first-order mod- els) are used the vehicle state is the vehicle position and the vehicle velocity is used as control action [46]. On the other hand, second-order models describe the vehicle as a point mass where the vehicle state variables are the longitudinal vehicle position and velocity while the longitudinal acceleration is used as control action [31,47–49]. In case, the drag force and rolling resistance are neglected, second-order models reduce to double integra- tors [31]. On the contrary, if the drag force is relevant (e.g. 2. Vehicle platooning in the case of platoons of trucks or cars driving at high speed), the longitudinal vehicle dynamics is described through a damper-mass system [47], which is linear when a linear approximation of the drag action is used [48]. In the case of third-order models, the vehicle is still described as a point mass but an additional state is introduced to consider the time-lag in the longitudinal acceler- ation due to, for instance, the vehicle powertrain and engine dynamics, thus the control action is usually the reference driving/braking torque or the desired acceleration [32,50]. First- and second-order models have been considered in the literature to simplify the control design and the closed-loop analysis of platoon control system especially with respect of the platoon length, i.e. the number of vehicles in the platoon. For example, in [46] the single-integrator model was used to analytically show that the magnitude of the control action can scale with respect to the root of the platoon length. In [31], the double-integrator model was adopted to prove that the stability margin goes to zero as the inverse of the square of the platoon length when the bidirectional topology3 is used. In [48] it was shown that string stability can be preserved in presence of communication latency if the information shared through the V2X links are augmented with the leader velocity information. This result was achieved by using a second-order vehicle model which included also the drag force. However, to facilitate the closed-loop analysis, a linear drag force was considered together with the perfect knowledge of the vehicle parame- ters. Consequently, unavoidable parameter mismatches and unmodelled nonlinearities can jeopardise string stability and safety in real working conditions. The need to design robust platoon algorithms with respect to vehicle nonlinearities and parameter uncertainties has been discussed for instance in [47] where a nonlinear damper-mass system with unknown parameters was used to model the longitudinal motion. In [47], authors suggested using a nonlinear controller with adaptive parameters to tackle vehicle uncertainties and dis- turbances while achieving string stability for bidirectional topologies. It is noted that when first- and second-order models are exploited for the design of CACC strategies, it is implic- itly assumed that the control action can be instantaneously imposed to the vehicle. On the other hand, this assumption might not be fulfilled due to inevitable parasitic delays and U. MONTANARO ET AL. 2. Vehicle platooning 8 8 lags of the powertrain, sensors and actuators, which can jeopardise stability and perfor- mance of platoon control systems. In accordance to [35], a lumped parasitic delay and lag are the combined result of pure time delays and lags in (i) the engine response, (ii) the throttle actuator, (iii) the brake actuator and (iv) low-pass filters used for sensors such as engine manifold pressure sensor, wheel speed sensor, etc. In [35], authors also proved that when vehicles in the platoon receive only information from the predecessor vehicle, the time-headway has to be at least double of the sum of the lumped parasitic delay and lag to guarantee string stability, thus increasing inter-vehicular distance which results in less fuel efficiency and road usage [51]. For a constant spacing policy and vehicles modelled as a second-order point-mass system, authors have proven in [52] that string stability is pre- served in presence of unknown but upper-bounded time-lags by reducing the magnitude of the control gains which might reduce platoon stability margins. g g p y g To systematically consider delays and time-lags, third-order vehicle models are used where an additional state is added to the aforementioned point-mass systems to capture either the dynamics of the acceleration of the vehicle or its torque. Usually, the additional equation is a low-pass linear filter with unitary gain and time constant equal to a lumped value of the lags of the powertrain, sensors and actuators. The input to the filter is either the desired vehicle acceleration [32,39,53], desired driving/braking torque [50,54], or a delayed version of these quantities to model also a lumped delay of the powertrain, sensors and actuators [35]. The analysis of the technical literature have shown that the lumped time-lag usually ranges from 200 to 800 ms, while the time-delay is typically within 20 and 250 ms [55]. Furthermore, both linear and nonlinear third-order vehicle models have been used for platoon control design. A nonlinear third-order model has been recently considered in [50] for the design of an optimal network-based platoon strategy which also guaran- tees that the desired driving/braking torque is confined to a preassigned set. 2. Vehicle platooning An additional example of the use of nonlinear third-order systems is provided in [54] where a nonlinear controller was designed to guarantee the convergence to zero of the inter-vehicular error in a finite time and it was shown that the converging time depends on the network topology. It is noted that, although nonlinear systems have been used, linear third-order models are more often adopted for CACC design as they simplify the stability analysis of the closed- loop system. For instance, in [39] it was proven that by including the desired acceleration of the second lead vehicle (see also Figure 1(a)) into the computation of the CACC action it is possible to ensure string stability also for time-headway smaller than the double of the sum of the vehicle time-lags and delays, thus improving the result in [35]. More recently, authors in [32] have proposed a systematic approach for design of the platoon control parame- ters based on the study of the roots of a third-order algebraic equation whose coefficients depend on the vehicle time-lag, network structure and control gains, thus showing the coupling among control parameters, network topology and vehicle model to achieve the cooperative platoon motion. In [53], authors extended the previous analysis to prove that by enlarging information topology (i.e. adding communication links to the bidirectional topology) stability margin can be made independent on the number of the vehicles in the platoon, thus improving the result in [31] for double-integrators vehicle model. It is remarked that the longitudinal vehicle models discussed above are used to design CACC strategies to impose the synchronised motion to a fleet of consecutive vehicles. The output of these control algorithms is then imposed to the vehicle by low-level controllers [44,45] (see also Figure 2). As the control variables of low-level controllers are the inputs to 9 VEHICLE SYSTEM DYNAMICS the vehicle actuators (e.g. throttle opening and the braking system), their design requires more detailed longitudinal vehicle models which also include the engine dynamics, engine maps, speed density functions, throttle body dynamics, dynamics of the braking system, nonlinear drag forces, rolling resistance and road slope, etc. [44,45]. 2. Vehicle platooning The ability of the low-level controller to impose to the vehicle the commands provided by the upper-level controller is fundamental to achieve the cooperative motion, and controllers with poor tracking capability might induce larger transients and steady-state spacing errors, such as those numerically documented in [56], which might threaten road safety. It is noted that parameter uncertainties, unmodeled dynamics (e.g. engine/vehicle unmodeled dynam- ics), disturbances (e.g. wind) and changes in operating speed are unavoidable and prevent the perfect tracking of the commands generated by platoon controllers. Consequently, the dynamics of the vehicle with the low-level controller can differ from those predicted by using simplified longitudinal vehicle models. The mismatch between the simplified models and those provided by vehicle/low-level controllers motivates the use of uncertain systems subjected to bounded unknown disturbances for the CACC design. A systematic tech- nique to model the low-level controller and the vehicle as a third-order linear system with unknown but bounded parameters and disturbances for platoon applications has been recently proposed in [45]. Specifically, based on the range of the possible variations of the vehicle parameters, in [45], the authors used a detailed longitudinal vehicle with a nonlin- ear low-level controller to obtain a set of vehicle acceleration profiles. These acceleration profiles were then used to tune multiple third-order linear models with a multiplicative uncertainty to be exploited for the design of robust and adaptive platoon controllers, such as those presented in [57–59], which guarantee stability also in the presence of such model uncertainties. It is noted that, for the design of control algorithms to impose the cooperative platoon motion, usually, only the longitudinal vehicle dynamics is considered. However, for additional platoon operations, such as side merging into a platoon and leaving a platoon, also lateral vehicle dynamics must be modelled for the generation and tracking of lateral feasible trajectories [60,61]. Lateral vehicle dynamic models are discussed in Section 3 for the planning and tracking of lateral manoeuvres for lane changes. These models can also be used for platoon functionalities that require lateral movements, thus the reader is referred to Section 3 for further details on lateral vehicle models for CAVs. 2. Vehicle platooning Finally, it is remarked that although V2X communication systems can help in ensuring string stability with reduced inter-vehicular gaps, the control of platoons in the presence of network imperfections (such as time varying and uncertain communication delays, and packet loss) is still a challenging open problem. Furthermore, even though several numeri- cal studies have confirmed that organising vehicles in platoons increases fuel economy and road capacity; systematic platoon strategies for achieving a suitable compromise between safety and performance based on system parameters (e.g. inter-vehicular distance, platoon speed, network latency and packet drop rate) to the best of the authors knowledge are not available in the current literature. 3. Lane change Lane changes are common manoeuvres on highways and have a significant impact on transportation traffic flow and safety as they can induce speed and traffic flow oscilla- tions especially in dense traffic situations [62]. Furthermore, vehicles performing lane 10 U. MONTANARO ET AL. 10 changes act as moving obstructions which increase traffic shock waves and reduce safety [62]. Vehicles that can autonomously perform high-speed manoeuvres have the poten- tial to reduce the risk of lane change crashes, thus improving traffic safety [63]. However, to autonomously perform a lane change, it is essential that a vehicle can determine if, when, and how to perform a lane change. The decisions that govern ‘if’ a lane change is required are dependent on factors such as (i) legislation, (ii) route plan, (iii) traffic con- ditions, etc. Once the decision to perform a lane change has been made, it is the need to identify the right instant i.e. ‘when’ to initiate and terminate this manoeuvre to maintain both safety, comfort of the occupants and other road users. Finally, once the time and spa- tial boundaries are computed, the important task of actually performing i.e. ‘how’ the lane change needs to be tackled by the CAV. Various approaches proposed in the literature for performing these tasks are discussed in further detail below. p g According to [64], two types of lane change are possible: (a) discretionary lane change (DLC) and (b) mandatory lane change (MLC). MLCs are either imposed by the traffic rules or by vehicles destinations to follow their desired route, e.g. for the correct selection of the lane for merging onto the highway from an on-ramp or taking an exit off-ramp. Contrary, for DLCs the change of lane is performed if the traffic situation is perceived as better in the destination lane (also denoted as the target lane); for example, allowing the subject vehi- cle to maintain or increase speed or to pass a sequence of moving vehicles with a possible return to the origin lane (overtaking). A fundamental step in any DLC is the Lane Selec- tion [64], which is defined as the process of choosing the target lane through foreseeing an improvement in the driving condition. Autonomous vehicles can select the target lane based only on the data collected from nearest vehicles, i.e. those within the line of sight of on-board sensors [65,66]. 3. Lane change Numerical results show that in the case where the subject vehicle does not have a view of traffic conditions on the road segment beyond its sight (case (i)), it must reduce its veloc- ity from about 58 to 35 km/h (a velocity reduction of about 40%) before initiating a lane change manoeuvre at t =30 s. But, when traffic information is included (case (ii)), the lane change manoeuvre initiates at t =16 s and the velocity for the subject vehicle is reduced at most by 17.25%. Intuitively, the latter vehicle response improves both the average vehicle speed and fuel economy as well. When the optimal lane change is used, the average veloc- ity increases about 7.72% while fuel consumption is reduced about 40%. For accurately evaluating fuel reduction in more realistic conditions, a mixed scenario was also simulated where a road segment of 2 km was considered with three lanes including a merging lane. Energy consumption on the road segment was evaluated for different penetration rates. In addition, results were compared with respect to the cases: (a) no vehicle is equipped with the optimal selection of the target lane and (b) some vehicles adopt Cooperative Adaptive Cruise Control (CACC). Numerical results showed that the optimal lane selection based on traffic information always provides better performance in terms of fuel economy with respect to the CACC by improving full economy up to 14%. Also, with low penetration rate, the optimal lane selection provides consistent improvement of fuel economy compared with no vehicle so equipped. For instance, with a penetration rate of 10%, a fuel econ- omy of about 8.8% is obtained. Controlled vehicles in a mixed scenario also improve fuel economy of conventional vehicles. For instance, when the penetration rate is 25%, the fuel reduction of all vehicles is 9.2%. Such additional benefit is due to the intrinsic cooperation among vehicles that share the same road section. Indeed, some of the conventional vehi- cles must adapt their velocity and target lane with respect to the controlled ones to avoid collisions. For the lane selection method described above, each vehicle selects a sequence of target lanes using off-board data collected via V2X communication by optimising a cost func- tion that weights only individual performance indices (fuel consumption, travel time, etc.). 3. Lane change Nevertheless, limited knowledge of the traffic ahead is a key issue to lane selection as it might generate incoherent lane change decision making (e.g. non-beneficial lane change). By assuming a V2X environment, where traffic information for both the current lane and adjacent lanes are known in advance, each connected vehicle can select a sequence of lane changes over a given horizon with the aim to optimise an individual cost function (travel time, driver comfort, etc.) This approach for lane selection has been used in [67,68] where an optimal strategy has been exploited to compute both the optimal lane and acceleration of the subject vehicle, over a finite horizon to minimise a cost function related to the driving efficiency and comfort. The effectiveness of including traffic information of an entire road segment for the optimal selection of the target lane has been proven numerically in [68] in which two simulation case-studies were considered. In the first scenario, shown in Figure 3, there is only one subject vehicle indicated as SV. The SV is initially in lane 1 and is preceded by the vehicles c1, c2, . . . ,c7. Only vehicle c1, also Figure 3. Thetraﬃcscenario,wherealeadingvehicle(c7)intheprecedingtraﬃcrunsslowlythataﬀects the traﬃc ﬂows on lane 1 (based on the scheme in [68]). Figure 3. Thetraﬃcscenario,wherealeadingvehicle(c7)intheprecedingtraﬃcrunsslowlythataﬀects the traﬃc ﬂows on lane 1 (based on the scheme in [68]). 11 VEHICLE SYSTEM DYNAMICS 11 indicated as PV (Preceding Vehicle) in Figure 3, is in view of the subject vehicle (see also Figure 3). The 7th preceding vehicle (c7), which is beyond the sight of the SV, suddenly decelerates between t =2 s to t =5 s and its velocity drops from 60 km/h to 35 km/h and then continues moving slowly. This slow vehicle gradually affects the traffic downstream, e.g. from the 6th vehicle (c6) towards the 1st vehicle c1. The scenario described in Figure 3 is then analysed considering the following cases, (i) on road vehicles are not connected and the subject vehicle senses only the position of vehicle c1 and lane selection is performed by using a set of rules; (ii) a connected environment is assumed and the subject vehicle can sense in advance the velocity reduction of vehicle c7 via communication networks and use this information to change its longitudinal and lateral motions before vehicle c1 starts to slow down. 3. Lane change Hence, this approach does not exploit the potential of V2X communication to create coop- erative driving scenarios aiming to improve the traffic flow, pollutant emissions and fuel consumption of road networks. To improve the use of road infrastructures, in [69–71], cooperative lane selection methods have been proposed. The common idea underlying these methods is to measure the performance of a group of vehicles and select the set of lanes that guarantees the minimisation of a common cost function. When these methods are adopted, the road is divided in segments and at the beginning of each segment the following steps are performed: 12 U. MONTANARO ET AL. U. MONTANARO ET AL. • Data collection: A group of vehicles is formed and information from each vehicle is col- lected through communication links. These data can include location, speed, lane and desired driving speed of each individual vehicle. • Optimal target lane determination: Based on the data collected in the previous step the target lane for each vehicle in the group is computed by minimising a collective cost function (e.g. a cost function that includes the energy consumption or the travel time of the group). g p • Lane change execution: Each vehicle in the group executes the required lane change within the current road segment. • Lane change execution: Each vehicle in the group executes the required lane change within the current road segment. For cooperative lane selection, both Centralised and Distributed solutions have been pro- posed. Centralised algorithms use one control unit for selecting the target lanes for the vehicles in the group and data are collected via V2I communication links, while in the case of Distributed solutions, the set of the vehicles in the group cooperate to select the target lanes to minimise the common cost function by using V2V communication links. In the following, some examples from the literature are reported. In [69,70], a distributed optimal lane selection for minimising the travel time of a vehicle which travels on a highway has been proposed. The effectiveness of reducing the travel time has been shown numerically in simulations for four-lane highway with three entries and three exit points. Numerical results indicated that the average travel time reduction of the proposed solution depends on traffic density and the percentage of cooperative vehicles. 3. Lane change For instance, it was shown that if the traffic density is 2000 vehicle/hour with a penetration rate of 100%, the average travel time reduction would be 14.3% compared with the case that no vehicle is connected, while it would be 6.25% if the penetration rate is 50%. On the other hand, with considering a traffic density of 6000 vehicle/hour with 100% penetration, the maximum fuel reduction would be 8.8% compared with the case where no vehicle is connected. In [71], a centralised cooperative lane selection has been proposed to minimise the travel time of the group. The method was tested numerically for a highway segment 2000 m long. Furthermore, differ- ent volume-to-capacity (V/C) ratios, ranging from 0.5 to 0.95, were considered. Notice that the V/C ratio is a measurement of the traffic status based on the current traffic volume and capacity of the road segment, which is defined as the maximum number of vehicles per unit time which can be accommodated on the road segment under given conditions, e.g. a given time headway between consecutive vehicles [72]. Simulation results confirmed a travel time reduction of up to 3.79% when the V/C ratio is 0.7. The reduced travelling time has an effect also on the fuel consumption and pollutant emission with reductions about 2.2% and 19%, respectively. Analysis of the literature has shown that lane changes resulting from limited informa- tion and lack of cooperation among of road vehicles have an impact of the traffic flow, travel time and fuel consumption. Through the use of CAVs, it is possible to; (i) improve the lane selection process of individual vehicles by using off-board traffic data and (ii) enable coop- erative scenario where a group of vehicles collaboratively choose their target lane. However, lane selection solutions which use V2X communication only for gathering additional off- board information do not completely exploit the potential of communication channels to create cooperative driving scenarios where sets of vehicles decide together future actions to improve the overall traffic. The results presented in this section have shown that coopera- tive lane selection is possible and it is beneficial for fleets of vehicle to increase fuel efficiency VEHICLE SYSTEM DYNAMICS 1 13 and to reduce travel time thereby increasing road throughput. 3. Lane change On the other hand, the con- cept of deciding a set of target lanes by optimising a single cost function for a group of vehicles implies that some of the vehicles in the group might worsen their performance (e.g. fuel consumption, travelling time, etc.) with respect to the case of no cooperation. Conse- quently, this approach might be used by road authorities to improve traffic but it might not be suitable for controlling individual vehicles which might aim to get individual benefit (unless individual benefits can be demonstrated as well). Finally, even though individually based and cooperative lane selections supported by V2X communication have shown to provide similar benefits, it is not possible to compare them quantitatively. The results pre- sented depend on a range of factors such as the simulator, traffic scenario, control strategy, etc. which makes a systematic comparison difficult. y pfi Once the decision (either MLC or DLC) to perform a lane change has been made, the autonomous vehicle needs to (i) generate a feasible collision-free trajectory in real time, and (ii) track this trajectory as closely as possible by applying appropriate amount steer- ing, acceleration and braking action. Performing an autonomous lane change manoeuvre is a challenging task as it combines both lateral and longitudinal motion of a subject vehi- cle while avoiding collisions with other road users [73]. The two control tasks mentioned above for achieving this goal are divided into (i) trajectory planning, and (ii) trajectory tracking. A general control architecture to depict the closed-loop structure is shown in Figure 4 [74–79]. The trajectory planning controller perceives the environment, moni- tors vehicle states (longitudinal and lateral positions, longitudinal and lateral velocities, longitudinal and lateral accelerations, and heading) and computes safe and feasible trajec- tories (e.g. reference velocity, vref, longitudinal and lateral position, denoted in Figure 4 as Xref and Yref, respectively) for the vehicle to track [80]. The trajectory tracking controller then computes, via feedback algorithms based on the tracking error, the necessary torque (τref) and steering inputs (δref) required to track the reference, despite possible measure- ment noise, un-modelled dynamics, parametric uncertainties which may or may not be accounted for by the trajectory planning controller. An autonomous vehicle performing lane changes at high speeds requires that vehicle dynamics and constraints are considered while planning trajectories so that in addi- tion to being safe are also feasible. 3. Lane change Thus for autonomous high-speed driving, the vehicle model provides two pivotal functions, (i) vehicle system simulation, and (ii) design and Figure 4. General control architecture for an autonomous lane change (V2X block with dot-dash boundary: optional functionality). Figure 4. General control architecture for an autonomous lane change (V2X block with dot-dash boundary: optional functionality). 14 U. MONTANARO ET AL. 14 investigation of controller properties [81]. A variety of vehicle models ranging from point- mass model all the way to multi-body models have been developed and the ones relevant for controller design have been documented in [30]. Since, the lateral dynamics of a vehicle has a nonlinear dependence on the longitudinal velocity, vehicle models that provide a suitable trade-off between model accuracy and fidelity need to be used for trajectory planning. A comprehensive review of different trajectory planning techniques in [82] shows that there are three vehicle models namely (i)the point-mass model, (ii) the linear kinematic bicycle model and (iii) the non-linear kinematic bicycle model which are most commonly used by researchers. Among these, the point-mass vehicle model does not incorporate any lat- eral and yaw dynamics of the vehicle and hence trajectories involving lateral motion that were generated using this model were often found to be unfeasible under certain condi- tions (especially at high speeds and/or low friction coefficient). The two bicycle models mentioned above capture the lateral and yaw dynamics by modelling them using geomet- rical relationships. This additional system dynamics means that kinematic bicycle models are suitable for trajectory planning even in conditions of high-speed and low friction coefficient [78]. fi As explained above, the planned trajectory is provided as the reference to the track- ing controller to be accurately tracked while maintaining also state and input constraints. A comprehensive review of trajectory tracking control on the aspects of choice of vehicle model, control strategies, and controller performance criteria has been performed in [81]. The review demonstrated that geometric models based on Ackermann steering are not suitable for high-speed trajectory tracking due to their inability to include vehicle dynam- ics (e.g. acceleration and velocity). Additionally, it is highlighted that kinematic models (bicycle or four-wheel) are also unsuitable for high-speed trajectory tracking as they are inaccurate due to assumptions such as (i) no side-slip in tyres and/or vehicle, and (ii) lin- ear tyre behaviour throughout the entire range of operation. 4. Intersection management Intersections are shared areas (or conflict areas) and represent bottlenecks in the traffic flow. They can be classified as (i) Signalised Intersections which are equipped with traffic lights, and (ii) Non-Signalised Intersections which do not have traffic lights and drivers must interact with each other. Although intersections represent a small fraction of the road system, a noteworthy amount of fatalities occur within the area jointly used by the intersecting streets and are mainly caused by human errors [90]. For example, during the 10-year time period from 2005 to 2014, over 20% of the fatalities on EU roads took place at intersections [91]. Furthermore, ineffective intersection management (e.g. traffic lights with fixed timing, stop signs etc.) can increase the time that vehicles are stationary at junctions thereby resulting in an increase of travel time, fuel consumption and pollutant emissions. Hence, the correct management of intersections is of utmost importance for keeping traffic safe, improving traffic flow, and at the same time reducing energy consump- tion, pollutant emissions, and travel time. Autonomous vehicles and their cooperation with intelligent road intersections can reduce human mistakes and improve the efficiency of intersection management systems, thereby improving safety, energy and traffic efficiency [92]. In addition to traffic timing, V2X communication systems can be used to provide detailed vehicle information and driver’s intentions of vehicles approaching the intersec- tion. Such information can be collected and exploited to anticipate the behaviour of the vehicles, particularly those that are not within the line of sight of the on-board systems of the subject vehicle. Therefore, this section is devoted to analysing the working principle of Cooperative Intersection Management (CIM) systems. Hereafter, a CIM system refers to those intersection control systems which exploit V2X communications to provide solutions to the problem of coordinating the motion of a set of vehicles through intersections safely and efficiently. The management algorithms for CIM solutions are mainly based on heuris- tic rule-based methods [92] or optimisation based methods [93]. In the case of heuristic rule-based methods, the vehicles cooperate to pass the intersection by using a set of fixed rules implemented as an interaction protocol among the vehicles and the coordination unit. On the other hand, optimisation-based methods mathematically formulate a CIM prob- lem and solve it by using tools from control theory. It is noted that the decision variables (also known as control variables) depend on the adopted method. 3. Lane change Dynamic vehicle models (full vehicle model, half vehicle model and bicycle model) attempt to address these issues by incorporating additional states such as (i) vehicle side-slip angle (β), (ii) tyre side-slip (α) and (iii) linear or non-linear tyre models and they were found to provide a more accu- rate representation of a vehicle during high-speed driving [81]. Moreover, for manoeuvres (lane-change manoeuvre, overtaking manoeuvre, highway driving) that require small lat- eral accelerations (≤0.5 g) and low vehicle side-slip angles (≤5◦) the tyres remain within the linear region of operation and hence, a dynamic bicycle model (linear) is sufficient to capture the relevant dynamics of a vehicle [81,83]. As a result, a majority of papers in lit- erature have used a single-track vehicle model (bicycle model) for developing a tracking controller for performing lane change manoeuvres since this manoeuvre is performed well within the dynamic limits of the vehicle (i.e. lateral acceleration, vehicle side-slip and yaw- rate) where both the vehicle as well as tyre dynamics can be approximated by linear models. However, while performing lane changes at high-speeds and/or in low road friction con- ditions, the tyres might operate in the non-linear region which might lead to the lateral and yaw dynamics of the system to exhibit substantial non-linear behaviour and therefore for appropriate scenarios either nonlinear models, linear parameter varying (LPV) models or multiple models can be used to capture the relevant dynamic behaviour of the system [83,84]. Out of the wide variety of vehicle models available in literature a kinematic bicy- cle model and dynamic bicycle model have been found to provide a good compromise between model complexity and accuracy for controller design related to highway driving 15 VEHICLE SYSTEM DYNAMICS 15 applications [85,86]. For a more detailed discussion on different vehicle models, the reader is directed towards the work by [81,87–89]. applications [85,86]. For a more detailed discussion on different vehicle models, the reader is directed towards the work by [81,87–89]. 4. Intersection management They can be the tim- ing of traffic lights, time slots to each vehicle to pass the intersection, intersection passing sequence, acceleration and deceleration, etc. In what follows, CIM solutions for signalised intersection and non-sigalised intersection are analysed to identify potential benefits and limitations of each method for autonomous driving systems. l d h f ll d fi d d In signalised intersections, the safety requirement is usually assumed satisfied under the implicit hypothesis that all the vehicles respect the traffic light signals. Therefore, the objective of CIM systems can be to reduce idling at red lights to maximise traffic flow or to improve fuel efficiency. The idea behind these controlling systems is to use informa- tion collected by means of V2X communication links to adjust vehicles motion with the aim to reduce idling and fuel consumption at red lights. Typically, it is assumed that vehi- cles approaching the intersection can communicate with the infrastructure and receive U. MONTANARO ET AL. 16 information from the upcoming traffic light. This information can be either the current state of the traffic light or the Signal Phase and Timing (SPaT) signal, i.e. the current state and timing of the upcoming traffic light. Based on this information, the velocity profile of each vehicle is accordingly modified to pass the intersection during green light windows. Example of rule based methods for computing the vehicles speed profile can be found in [94,95]. Usually, rule based methods compute the vehicle speed profile based on the current state of the upcoming traffic light and do not consider vehicle dynamics within the design criteria. Conversely, when optimisation methods are used, the speed profiles are the solu- tions of dynamic constrained optimisation problems where the objective is to reduce the fuel consumption over the whole trip [96,97] or to minimise the travel time [98,99]. In this case, common constraints of an optimisation problem are longitudinal vehicle dynamics, maximum allowed vehicle speed, inter-vehicular distance and acceleration/deceleration capability of a vehicle. Furthermore, the accuracy of the optimisation algorithms can be enhanced if the precise position of traffic lights along the path, the SPaT information of traffic lights, and upcoming road parameters (e.g. road surface friction, maximum speed velocity) are known. Therefore, V2X communication systems have the potential to enhance the accuracy and efficiency of current optimisation algorithms by providing to a subject vehicle upcoming environmental and SPaT information. 4. Intersection management However, the SPaT information cannot be accurately obtained from traffic lights that adapt their timing based on the vehicle queue in each road segment approaching the intersection [100,101]. The current literature conforms that if all on-road vehicles can adjust their speed based on SPaT information, considerable fuel reduction can be achieved compared to the case where no vehicle is connected. For instance, in [97], it was shown that when all vehicles are connected and follow precisely the speed profile based on the SPaT data, in the case that the traffic density is 600 vehicles/hour/lane, the average fuel reduction would be 30% compared with the case where all vehicles are unconnected. Also, it was shown that in the case of mixed traffic scenario (i.e. a mixture of connected and human-driven vehicles), the achievable increase of fuel efficiency depends on the fraction of connected vehicles (i.e. penetration rate). Moreover, by increasing the fraction of connected vehicles in a traffic segment, unconnected vehicles are more likely to adapt their motion to match the speed of surrounding connected vehicles where higher synchronisation of vehicle motions results in reduced stoppages at the intersection, thus bringing down overall fuel consumption for connected as well as unconnected vehicles. However, assessing the real benefits of CIMs for signalised intersections in a mixed scenario is not trivial as results are sensitive to the simulated scenario (e.g. number of lanes, traffic flow, etc.), adopted method for computing the speed profile, and the numerical tool used to simulate the human-driven vehicles (e.g. Car-following model; Lane-changing model). Therefore, discordant results on fuel reduc- tion as a function of the fraction of the equipped vehicles are reported in the literature. For instance, in [95] the trend of the fuel consumption for connected, unconnected, and all on-road vehicles were decreasing as a function of the penetration rate. Furthermore, fuel consumption of connected vehicles was larger than that of unconnected vehicles for a pen- etration rate less than 60%. In [102], it was noted that if the penetration rate of connected autonomous vehicles was less than 25%, an increase in fuel consumption for all road users, up to 5%, was measured. However, the fuel efficiency of the all road users became a pos- itive increasing function for penetration rates higher than 30%. 4. Intersection management In [97], it was observed that the average fuel consumption of connected vehicles was always much smaller than 17 VEHICLE SYSTEM DYNAMICS 17 those unconnected. However, for high density traffic of 900 vehicles/hour/lane, the fuel consumption of connected vehicles increased with the penetration rate. This unwanted effect was caused by the impossibility of connected vehicles to implement their optimal speed profiles in high-traffic density. It is noted that in the above-mentioned research papers, only the traffic light information was considered to optimise the speed of con- nected autonomous vehicles. However, V2V communication systems can be also used to gather additional information such as neighbouring vehicle states to further improve fuel efficiency. For instance, in [103,104], an optimisation control algorithm which utilises both V2I (SPaT) and V2V communication was proposed. The idea behind the approach was to adjust the motion of each connected autonomous vehicle with respect to the velocity of its successor to possibly allow both vehicles to cross the intersection in the same green light window. In the scenario of having a string of 15 vehicles on a road and traffic lights located every 500 m, the simulation results showed that the fuel economy of traffic utilising both V2I and V2V information can be improved by 22% compared to the case when just V2I information is available and 50% compared to the case that the cooperation is not used. In [102], V2X communication systems were also used to broadcast the length of the queue stopped at the upcoming traffic light, thus to adjust the vehicle motion to avoid stoppages behind traffic. Numerical results showed that by including traffic data in the computation of the speed profile, it is possible to further increase fuel efficiency with respect to the case where only SPaT information was used in mixed traffic. However for the this case study, the increase was limited to 2%. In non-signalised intersection, as there are no signals or sign-posts available, vehicles must coordinate the use of the common area within the junction. This coordination prob- lem is usually formulated as calculating the trajectories for individual vehicles that allow them to safely reach their destination in a finite time. According to [105], any solution should meet the basic requirements of safety and liveness (defined below), while optimis- ing performance metrics (e.g. fuel consumption, travel time, etc.). 4. Intersection management Here, safety means that there is no collision between pairs of vehicle crossing the intersection whereas liveness guarantees that all vehicles enter and exit the coordination area in finite time so that per- manent stops and traffic deadlocks within the intersection are avoided. It is noted that despite the management of signalised intersections in which vehicles can be either just connected or connected autonomous, for the management of a non-signalised intersec- tion, a common assumption is that all vehicles are connected and autonomous. For the management of a non-signalised intersection, Ref. [106] proposed a rule-based approach known as resource-reservation scheme. Achievable mobility and environmental benefits were reported in [107] through testing the method for a four-way intersection. The average delay to pass the intersection, fuel consumption, and pollutant emissions were compared with traffic lights and roundabouts. The simulation results in [107] indicated that the resource-reservation method always outperform current traffic lights and roundabouts regardless of traffic density. The average reduction in travel delay and fuel consumption with respect to traffic lights and roundabouts were 50% and 20%, respectively. Also, pollu- tant emissions such as HC, NOx, CO were reduced by replacing traffic lights with CIMs. On average, the HC reduction was 35%, while NOx and CO reductions were 33% and 43%, respectively. The application of resource-reservation method was the subject of other research as well, e.g. [108,109]. However, their findings and results are in agreement with the outcome of [107]. Although a rule-based method can be more effective with respect U. MONTANARO ET AL. 18 18 to the roundabouts and traffic lights, its performance and the safety and liveness require- ments are not guaranteed for events which have not been considered in rule definitions. As a result, the general lack of formal guarantees in terms of the object and constraints form the main weakness of rule-based solutions [105]. On the other hand, the main idea behind optimisation-based methods is to reformulate the problem of coordinating a set of vehicles through an intersection as an optimisation control problem [105]. The require- ments of safety and liveness are recast as constraints of the problem while minimising the performance metrics. Collision-free and deadlock-free solutions are achieved by impos- ing constraints on the motion trajectories of vehicles. In this approach, for each pair of vehicles, it is required that the travel trajectories do not cross each other (safety require- ment). 4. Intersection management In addition, each travel trajectory must enter and leave the common area of the intersection in a finite time (liveness requirement). Examples of optimisation based strate- gies can be found in [110–113]. In [110,111], the objectives of the minimisation was the travel time and it was numerically proven that optimal-based CIMs can outperform inter- sections controlled thought traffic lights, roundabouts and all-way stop control in terms of fuel efficiency and travel time and for different road condition (dry, rainy and snowy). In [112] the target was to minimise the overlap of the vehicles position inside the inter- section zone. Namely, the acceleration profiles of the vehicles were computed such that only a limited number of vehicles are present inside the intersection at each time instant. Numerical results showed that optimal-based CIMs have the potential to reduce fuel con- sumption with respect to intelligent traffic light, i.e. those where the timing depends on the traffic queues behind the stop bar, especially for the case when traffic volumes exceed the capacities of the roads. In [113], a multi-objective optimisation approach was used in which the cost function included a term to penalise speed variation with respect to a speed reference, and a term to penalise excessive acceleration/deceleration. It was numerically proven that the optimal CIM was able to double the traffic flow compared to the case where traffic lights were used for coordinating a four-way intersection. Although opti- misation based approaches consider vehicle dynamics and physical constraints to assure collision-free and deadlock-free solutions, the time to reach the optimal solution expo- nentially increases with the number of conflict relationships among the vehicles [105,114]. For these reasons, the optimisation-based approach becomes numerically intractable for real-time scenarios. To address the computation issues associated with optimisation-based methods, in [114] authors have proposed a sub-optimal solution in which a controller cal- culates the occupancy time interval (i.e. the optimal time instants that must be used by each vehicle to safely enter and leave the intersection) based on shared information of traf- fic vehicles. Then, based on the allocated time instances, each vehicle computes its own optimal trajectory so that a local objective is minimised. The effectiveness of the method to solve the intersection problem was demonstrated through simulations. Furthermore, a set of 1000 simulations was analysed to assess the average time to solve the optimisation problem with and without the proposed solution. 4. Intersection management Analysis of the simulation results showed that the average time to solve the optimal control problem was about 10.14 s with a standard deviation of 24 s On the other hand when the proposed method was applied the average to the roundabouts and traffic lights, its performance and the safety and liveness require- ments are not guaranteed for events which have not been considered in rule definitions. As a result, the general lack of formal guarantees in terms of the object and constraints form the main weakness of rule-based solutions [105]. On the other hand, the main idea behind optimisation-based methods is to reformulate the problem of coordinating a set of vehicles through an intersection as an optimisation control problem [105]. The require- ments of safety and liveness are recast as constraints of the problem while minimising the performance metrics. Collision-free and deadlock-free solutions are achieved by impos- ing constraints on the motion trajectories of vehicles. In this approach, for each pair of vehicles, it is required that the travel trajectories do not cross each other (safety require- ment). In addition, each travel trajectory must enter and leave the common area of the intersection in a finite time (liveness requirement). Examples of optimisation based strate- gies can be found in [110–113]. In [110,111], the objectives of the minimisation was the travel time and it was numerically proven that optimal-based CIMs can outperform inter- sections controlled thought traffic lights, roundabouts and all-way stop control in terms of fuel efficiency and travel time and for different road condition (dry, rainy and snowy). In [112] the target was to minimise the overlap of the vehicles position inside the inter- section zone. Namely, the acceleration profiles of the vehicles were computed such that only a limited number of vehicles are present inside the intersection at each time instant. Numerical results showed that optimal-based CIMs have the potential to reduce fuel con- sumption with respect to intelligent traffic light, i.e. those where the timing depends on the traffic queues behind the stop bar, especially for the case when traffic volumes exceed the capacities of the roads. In [113], a multi-objective optimisation approach was used in which the cost function included a term to penalise speed variation with respect to a speed reference, and a term to penalise excessive acceleration/deceleration. 4. Intersection management It was numerically proven that the optimal CIM was able to double the traffic flow compared to the case where traffic lights were used for coordinating a four-way intersection. Although opti- misation based approaches consider vehicle dynamics and physical constraints to assure collision-free and deadlock-free solutions, the time to reach the optimal solution expo- nentially increases with the number of conflict relationships among the vehicles [105,114]. For these reasons, the optimisation-based approach becomes numerically intractable for real-time scenarios. To address the computation issues associated with optimisation-based methods, in [114] authors have proposed a sub-optimal solution in which a controller cal- culates the occupancy time interval (i.e. the optimal time instants that must be used by each vehicle to safely enter and leave the intersection) based on shared information of traf- fic vehicles. Then, based on the allocated time instances, each vehicle computes its own optimal trajectory so that a local objective is minimised. The effectiveness of the method to solve the intersection problem was demonstrated through simulations. Furthermore, a set of 1000 simulations was analysed to assess the average time to solve the optimisation problem with and without the proposed solution. Analysis of the simulation results showed that the average time to solve the optimal control problem was about 10.14 s with a standard deviation of 24 s. On the other hand, when the proposed method was applied, the average computational time reduced to 0.043 s with a standard deviation of 0.022 s. Regarding sub- optimality of the solution, numerical results showed that the approximated algorithm gave results 20% less accurate than an optimal solution for about 85% of the random realisations. The aforementioned survey for intersection management showed that integration of V2X VEHICLE SYSTEM DYNAMICS 19 communication systems improved the performance of autonomous vehicles; however, the level of achievable benefits depends on the case study, simulation scenario, type of inter- section, etc., in both signalised and non-signalised intersections. Also, it depends on the quality of communication signals in terms of packets loss, packet errors and latency. Contrary to the use-cases discussed in previous sections, the main focus of intersection management is more towards accurate modelling of intersection area and traffic flow rather than modelling the detailed motion of each vehicle at the intersection. Furthermore, the primary concern for non-signalised intersections is to guarantee a collision-free path for each vehicle while for signalised intersections, it is to reduce energy consumption. 4. Intersection management As a result, vehicle models at different levels of abstraction are utilised for each problem and are discussed in further detail below. For non-signalised intersections, where collision avoidance is the primary objective, it is imperative to ascertain a vehicle’s position on the road with respect to time. Further- more, as the subject vehicle can either maintain its lane or turn at the intersection, it is important to employ vehicle models that capture both longitudinal and lateral motion of the vehicle. For studies based on non-signalised intersections where the vehicles cross the intersection without turning, point-mass vehicle models, which are similar to third-order models discussed in Section 2, are most commonly used for this purpose as they provide a reasonable approximation of the planar motion (no vertical motion) [92,107,111,113]. The position, velocity and acceleration of the subject vehicle are the vehicle states used to con- trol the vehicle on given portions of the road segment [105,112]. The road-load equation of a vehicle to generate a state-space model of the vehicle is another technique proposed for capturing the longitudinal dynamics of a vehicle. The application of this vehicle model is demonstrated in [111]. Moreover, some studies also consider the possibility that the vehi- cles can turn at the intersection and the researchers propose the use of kinematic bicycle model as discussed in Section 3 to capture the non-holonomic motion of a vehicle with sufficient accuracy [109]. fi Signalised intersections pose a different challenge for vehicle modelling because of the twin requirements of (i) optimising a vehicle’s velocity and acceleration to reduce time spent waiting at the signal, and (ii) reducing the amount of energy consumed in successfully navigating through a signalised intersection. Thus, vehicle models with simple dynamics to capture longitudinal motion and its effect on fuel consumption are commonly employed by researchers. The longitudinal motion of the vehicle is often modelled using the road load equation with velocity and acceleration being the two states of the system, i.e. the second- order model discussed in Section 2 [98,99,103,104]. Moreover, an optional third state of the system is the energy consumption that is modelled as a function of the first two states which is then used within the optimisation routines discussed above in this section to optimise the velocity of the vehicle while minimising fuel consumption as part of a multi-objective optimisation problem [96,97]. 5. Vehicle energy management The fuel consumption reduction was computed by considering the fuel consumption obtained via the advisor following with respect to the baseline driving, and the fuel consumption of the intelligent adaptive cruise control with respect to the advi- sor following case. With respect to the baseline, the SAS improved fuel efficiency for the highway driving from 10.6% to 14.4% (with an average gain of 12.6%), while in the case of urban driving scenario, the fuel improvement was in the range of 8.1–20.9% with an aver- age improvement of 12.5%. The analysis provided in [119] also shows that by combining SAS with ACC, additional fuel reduction can be achieved. By removing the human variabil- ity in following the advice speed, on average an additional 2% fuel reduction is obtained for highway driving while an improvement of 6.3% can be achieved for urban driving. The SAS proposed in [119] has been recently enhanced by IBM and Clemson University by includ- ing traffic light information [120]. Furthermore, to tackle the computational complexity, a parallel computing system was used as an off-board system to compute the global optimal speed profile. The optimisation algorithm implemented by exploiting the parallel comput- ing framework was denoted as FastVO (Fast Velocity Optimisation). The effectiveness of the enhanced FastVO was also experimentally tested by the authors. The fuel consumption of FastVO was compared with a global optimal solution proposed in [119] and a Predictive Cruise Control (PCC) presented in [121]. It is important to point out that the global opti- mal solution was also implemented in the off-board system but without taking advantage of parallel computing and it does not consider the state of the traffic lights (in accordance to [119]). In addition, when PCC is used, the vehicle receives the optimal speed profile as a function of the traffic light ahead. This solution might be different from the global optimal solution as the optimal velocity profile was computed only when the vehicle is sufficiently systems. These predictions, also known as preview information, can be then used for opti- mising the vehicle speed or vehicle powertrain to minimise fuel consumption over the entire trip [116]. Systems optimising the global vehicle speed trajectory are known in the literature as Speed Advisory Systems (SAS). Global optimal speed advisory systems provide the optimal speed profile over the entire trip, ideally for the entire source-destination route. 5. Vehicle energy management However, global optimal solutions are not easily tractable since the computational burden increases exponentially with the number of states and control variables. This drawback is known in optimisation literature as the curse of dimensionality [117,118]. In addition, as the traffic on a road is highly dynamic and unpredictable, each vehicle needs to periodically update its optimal velocity profile based on the current traffic condition. For these rea- sons, the use of off-board computing for transportation systems has been proposed to assist vehicles with computation of the optimal speed profile [119,120]. When this approach for speed optimisation is adopted, each vehicle uploads its information, e.g. destination, cur- rent velocity and position, to an off-board computing system which computes the optimal velocity profile and sends it back to the vehicle. For example, in [119], a cloud-based system was used as an off-board computing system. The system was tested experimentally both for urban and highway scenarios. For both scenarios, three runs were considered : (i) the driver drove using his normal driving style and the speed advisory system was turned-off (these runs are referred to as baseline driving); (ii) the vehicle was connected to the cloud and received the optimal speed profile but the driver was responsible for controlling the vehi- cle to impose the optimal speed (these runs are referred to as advisor following); and (iii) the vehicle was assumed to be connected and autonomous and the optimal speed profile was imposed through an ACC system (in this paper, these runs are referred to as intelligent adaptive cruise control). The fuel consumption reduction was computed by considering the fuel consumption obtained via the advisor following with respect to the baseline driving, and the fuel consumption of the intelligent adaptive cruise control with respect to the advi- sor following case. With respect to the baseline, the SAS improved fuel efficiency for the highway driving from 10.6% to 14.4% (with an average gain of 12.6%), while in the case of urban driving scenario, the fuel improvement was in the range of 8.1–20.9% with an aver- age improvement of 12.5%. The analysis provided in [119] also shows that by combining SAS with ACC, additional fuel reduction can be achieved. By removing the human variabil- ity in following the advice speed, on average an additional 2% fuel reduction is obtained for highway driving while an improvement of 6.3% can be achieved for urban driving. 5. Vehicle energy management CAVs have the potential to increase fuel economy by operating vehicles engines in regions with high efficiency which can result in a reduction of fuel consumption [115]. Communi- cation channels can be exploited to provide partial or complete trip information augmented with real time information (traffic condition and weather, etc.) to the on-board control U. MONTANARO ET AL. 20 systems. These predictions, also known as preview information, can be then used for opti- mising the vehicle speed or vehicle powertrain to minimise fuel consumption over the entire trip [116]. Systems optimising the global vehicle speed trajectory are known in the literature as Speed Advisory Systems (SAS). Global optimal speed advisory systems provide the optimal speed profile over the entire trip, ideally for the entire source-destination route. However, global optimal solutions are not easily tractable since the computational burden increases exponentially with the number of states and control variables. This drawback is known in optimisation literature as the curse of dimensionality [117,118]. In addition, as the traffic on a road is highly dynamic and unpredictable, each vehicle needs to periodically update its optimal velocity profile based on the current traffic condition. For these rea- sons, the use of off-board computing for transportation systems has been proposed to assist vehicles with computation of the optimal speed profile [119,120]. When this approach for speed optimisation is adopted, each vehicle uploads its information, e.g. destination, cur- rent velocity and position, to an off-board computing system which computes the optimal velocity profile and sends it back to the vehicle. For example, in [119], a cloud-based system was used as an off-board computing system. The system was tested experimentally both for urban and highway scenarios. For both scenarios, three runs were considered : (i) the driver drove using his normal driving style and the speed advisory system was turned-off (these runs are referred to as baseline driving); (ii) the vehicle was connected to the cloud and received the optimal speed profile but the driver was responsible for controlling the vehi- cle to impose the optimal speed (these runs are referred to as advisor following); and (iii) the vehicle was assumed to be connected and autonomous and the optimal speed profile was imposed through an ACC system (in this paper, these runs are referred to as intelligent adaptive cruise control). 5. Vehicle energy management The SAS proposed in [119] has been recently enhanced by IBM and Clemson University by includ- ing traffic light information [120]. Furthermore, to tackle the computational complexity, a parallel computing system was used as an off-board system to compute the global optimal speed profile. The optimisation algorithm implemented by exploiting the parallel comput- ing framework was denoted as FastVO (Fast Velocity Optimisation). The effectiveness of the enhanced FastVO was also experimentally tested by the authors. The fuel consumption of FastVO was compared with a global optimal solution proposed in [119] and a Predictive Cruise Control (PCC) presented in [121]. It is important to point out that the global opti- mal solution was also implemented in the off-board system but without taking advantage of parallel computing and it does not consider the state of the traffic lights (in accordance t [119]) I dditi h PCC i d th hi l i th ti l d fil 21 VEHICLE SYSTEM DYNAMICS 21 close to the traffic lights. In terms of fuel consumption, the outcome of the experimental analysis showed that FastVO is always better than that of the global optimal solution and the PCC strategy. In addition, the final fuel reduction of FastVO with respect to the global optimal solution was about 32% while it was about 11% with respect to the PCC strategy. These experimental results can be explained as follows. As the solution in [119] does not consider traffic lights, vehicles can stop at intersections, thereby reducing fuel economy. Consequently, the additional idling phases increases the fuel consumption compared with the solution in [120]. When the PCC is used, the subject vehicle lacks the global informa- tion, and it can only determine its own velocity profile according to the traffic light ahead and hence cannot achieve a globally optimal solution. FastVO considers traffic lights and achieves global optimal velocity profiles, thus producing the least fuel consumption. The speed advisory system based on the use of off-board computing has been extended to a platoon of vehicles to further increase fuel reduction in [122–124]. In [122], the optimal platoon speed was selected as a trade-off among minimisation of the travel time, fuel con- sumption and pollutant emissions. In addition to the speed, the authors adapted also the time headway along a planned route in hilly terrain. 5. Vehicle energy management Numerical results showed that for a homogenous platoon (all vehicle being equal) tight platoon formation is the optimal choice, but when controlling a fleet of heterogeneous vehicles, the optimal spacing depends on the road slope and the position of the vehicle in the string. Furthermore, it was shown that, by optimising the platoon speed, the overall fuel consumption was reduced by about 10% with respect to the case of vehicles driving alone with no platooning. In [124] the authors assumed that information about the traffic state ahead can be broadcasted through V2X communications systems. Platoons can then be informed in advance of the presence of congestion along the road where the velocity must be reduced in accordance to the traffic level. By including velocity information of the traffic ahead in the computation of the opti- mal profile, it was shown by numerical simulation that a fuel reduction up to 80% could be possible. p It is expected that future autonomous vehicles will be either fully electric or hybrid- electric which makes analysis of cooperative energy management developed for cur- rent electric and hybrid electric vehicle important [125]. It is noted that in the case of Hybrid Electric Vehicles (HEVs), efficiency highly depends on the strategy for determin- ing the split of power request between the combustion engine and the electric machine [126]. According to the technical literature [127], energy consumption of electric and hybrid/electric vehicles depends on the state of charge (SoC) of the battery. Using this variable, it is possible to optimise the power split ratio between power sources. When pre- view knowledge of driving profile is known in advance, the SoC profile can be scheduled in advanced through an off-line optimisation. However, these methods perform poorly when the driving route conditions change due to variations in traffic conditions. This draw- back can be overcome by connectivity as it provides access to real-time traffic data which can be used to improve the prediction of future driving cycles. The opportunity to use real-time traffic data gathered via V2X communications to reduce energy consumption has recently been investigated for example in [128]. In this study, the average velocity of traffic was provided to the subject vehicle as preview of the traffic situation ahead. 5. Vehicle energy management Based on these data, the subject vehicle computed the optimal SoC profile and, then, the SoC profile was used to optimise the power split with the aim of minimising energy consump- tion. In this study, the preview information was provided to the optimisation algorithm as 22 U. MONTANARO ET AL. 22 (1) static traffic information where the vehicle obtained the traffic velocity information only once at the beginning of the trip (i.e. the first generated SoC reference trajectory was assumed to be relevant until the end of the trip) and (2) dynamic traffic informa- tion where vehicle obtained the traffic velocity information periodically (every 300 s). The energy consumption provided by the strategy was evaluated by considering a real-world highway driving scenario based on collected traffic flow data from the Mobile Century project [129]. To better point out the benefit in terms of achievable energy reduction, the proposed controller was compared with a heuristic-based control algorithm (i.e. when no traffic information is available). Simulation results showed that the static solution (i.e. algo- rithms with static information) reduced the fuel consumption by 1.84% compared with a heuristic-based method. The fuel reduction was 5% with the dynamic strategy (i.e. algo- rithms with dynamic information). Numerical results confirmed that better fuel reduction can be achieved by including traffic information during computation of the SoC pro- file. Furthermore, the highly dynamic nature of the traffic must be considered to further increase fuel economy. When traffic information is used to re-compute the optimal state of charge trajectory, the fuel economy can more than double with respect to the static solution. It is remarked that, the use of traffic data to improve fuel efficiently of connected hybrid vehicles has been documented also in some recent review papers on the energy manage- ment of hybrid powertrains such as [130–132] and can be directly applied to CAVs with hybrid or electric powertrains. Independently from the optimisation technique, the optimisation of vehicle’s fuel con- sumption starts by defining vehicle models, consisting of its longitudinal vehicle dynamics, engine and fuel consumption models. Usually, the longitudinal dynamics is modelled by second-order systems as those discussed in Section 2 for the design of low-level controllers. 5. Vehicle energy management These models are obtained by applying Newton’s Second Law of motion to a vehicle, thus establishing a relation among vehicle velocity, vehicle acceleration, vehicle’s tractive force and all opposing forces acting on the vehicle, with the most significant being the aero- dynamic drag, rolling resistance and the gravitational force. However, additional forces can be considered to further improve model fidelity, e.g. the braking force provided by the mechanical friction brakes, retarder or exhaust brakes [119,123] or estimated auxiliary power losses coming from the clutch or drivetrain or other auxiliary devices [123]. Once the longitudinal dynamics has been defined, it is possible to relate it to the engine speed and torque; as the tractive force is a function of the engine torque, the gear ratio and the gearbox efficiency. A gear shifting model or logic can also be added such as the work done by [119] in order to schedule the gear shifting as well as to evaluate its feasibility in rela- tion to the engine speed and torque and additional power loss sources such as those due to engine, transmission and driveshaft rotational inertia [133]. Due to their highly nonlin- ear nature and parameter variation of each vehicle, fuel consumption models are obtained by interpolating experimental data. Usually they are expressed as functions of the engine speed and torque maps. However, several approaches have been used for modelling the fuel consumption functions. Linear functions with coefficients depending on the engine speed, i.e. Willans line approximation models, have been designed in [134], while a piecewise con- stant function has been proposed in [133] where the engine map was divided into sub-areas with a fixed fuel consumption value assigned to each zone. Polynomials are also employed, e.g. in [119], a third-order polynomial function with coefficients experimentally tuned was VEHICLE SYSTEM DYNAMICS 2 23 designed. Empirical models can also be exploited, such as the Virginia Tech Comprehen- sive Power-Based Fuel Model (VT-CPFM) which claims to estimate fuel consumption rates with actual field measurements with a 2% error [135]. It is noted that, if Dynamic Pro- gramming (DP) or Model Predictive Control (MPC) are used to optimise fuel efficiency, no specific advantages or disadvantages were noted among the fuel consumption models. 5. Vehicle energy management However, if the Pontryagin Maximum Principle (PMP) is adopted to compute the optimal solution, the VT-CPFM has the advantage that the derivative of the fuel consumption with respect to the engine torque still depends on that torque, thus preventing bangbang solu- tions where the optimal control action switches between the maximum and the minimum admissible control values [135]. Finally, the optimisation problem is completed by defining the cost function based on the fuel consumption model and constraints such as maximum accelerations and decelerations, speed limits and minimum inter-vehicular distances with respect to the preceding vehicle, which are imposed to the longitudinal vehicle dynamics. A common method for minimising the cost function over the entire trip is the DP algorithm. However, this method suffers from the aforementioned ‘curse of dimension- ality’ which makes it difficult to consider the motion of the surrounding vehicles in real time, e.g. in terms of time headway with respect to the vehicle in front. Alternatively, if pre- dictions of the accelerations of the vehicle ahead are available over a given time horizon, e.g. through V2X links, it is possible to use models of the longitudinal vehicle dynam- ics in Section 2 to predict its motion. The prediction of the preceding vehicle motion is then included in the optimisation of the fuel consumption of the subject vehicle as colli- sion avoidance constraints, and MPC techniques are used to find the optimal solution over the given time horizon. However, optimisation methods that consider also the dynam- ics of the surrounding vehicles cannot be used for searching global optimal solutions that extend over the entire trip. They can be exploited only over a limited time horizon where prediction of the traffic motion is valid. fi To overcome the aforementioned drawbacks, recently the authors in [134,136] have focused on the design of two stages optimisation strategies which can use vehicle models with different level of accuracy for each optimisation layer. The upper level optimises the global vehicle’s speed profile. To reduce the computational complexity in this layer simpli- fied longitudinal vehicle dynamics and fuel consumption models are used and other road participants are not considered. Then, the low-level optimiser locally adjusts the velocity profile received by the upper level over time horizons where predictions of the motion of the surrounding vehicles are available with the aim to further optimise fuel consumption while preserving road safety. 6. Road friction estimation Autonomous vehicles are equipped with various safety systems where their precise acti- vation is highly depended on an accurate knowledge of road friction conditions (which depends on the road surface type and prevailing weather conditions). Autonomous vehi- cles can use several techniques involving on-board sensors, ranging from optical, acoustic, camera to tyre sensors and/or data fusion methods, to estimate the road friction. However, critical reviews, such as those reported in [137,138], have pointed out that each technique has a limited estimation accuracy due to external noises, input frequencies, system models etc.. Furthermore, the current techniques available to autonomous vehicles estimate the instantaneous road friction coefficient and are not capable of estimating upcoming road conditions. This drawback can be mitigated if vehicles on a road segment can share their knowledge of the road condition. CAVs have the potential of improving the accuracy of the estimation as well as providing preview information of the upcoming road segment condition via cooperative road friction estimations enabled by V2X comunications. The general idea behind cooperative estimation is that every vehicle on a road section can act as a potential but imprecise sensor of road condition. Using communication systems, the road friction condition sensed by a group of vehicles on a road section can be shared and combined through data fusion algorithms to reduce the uncertainty of the estimate and increase its accuracy. Furthermore, a cooperative estimate might be stored and then dis- tributed to multiple vehicles. Consequently, cooperation among vehicles based on V2X communication allows (in principle) a solution to the main drawbacks of current esti- mation methods and their application in safety systems for autonomous vehicles [139]. Moreover, conventional vehicles, which are not equipped with any road friction estimation system, can benefit from an anticipated knowledge of the road friction condition as drivers can change in advance their driving style to prevent emergency situations. For instance, V2X communications have been used by Volvo to design a system for warning vehicles about the presence of slippery spots on roads [140]. A fleet of vehicles is used to monitor the state of the road and the presence of hazardous conditions, e.g. presence of icy patches, is detected and distributed through a communication network to other road users as warn- ings. An example of cooperative estimation of the road friction condition was proposed in [141]. 5. Vehicle energy management It is noted that the lower layer can exploit more detailed mod- els of the vehicle, engine and fuel consumption by considering, for instance, power losses due to the clutch, drivetrain and engine pumping losses. Moreover, the reduction of the drag force acting on the longitudinal vehicle dynamics while driving with reduced inter- vehicular distances can also be modelled in the low-level optimisation stage to increase the fidelity of the vehicle dynamics and the accuracy of the optimisation. It is noted that the two-stage optimisation methods are particularly suitable in a V2X environment sup- ported by cloud computing where the cloud can be used for the upper optimisation stage while road side units, which collect real-time information of the positon and velocity of CAVs, can host the lower layer of the optimisation technique. Cloud assisted solutions for cooperative driving applications are also currently under investigation within the CARMA project [15]. 24 U. MONTANARO ET AL. U. MONTANARO ET AL. 6. Road friction estimation The idea behind the method is that vehicles travelling through the same road section will experience similar road conditions. Hence, when any vehicle traverses the road, its esti- mation can be collected and fused to the estimations calculated by vehicles passing over the same road segment to create a common and better hypothesis of the road condition. To achieve a common estimation of the road condition, the following assumptions were made: (i) the road friction coefficients are modelled as a random variable normally dis- tributed with an unknown mean and known bounded variance, (ii) each vehicle travelling through the road section is equipped with a system for estimating the road friction coef- ficient and (iii) the instantaneous estimation of the road friction for each vehicle is also normally distributed and the upper bound for its variance is known. The road friction was modelled as a random variable to consider possible variations of this coefficient from one vehicle to another due to different types of vehicles. The authors proposed to estimate a lower bound of the friction condition such that the road friction coefficient experienced by each vehicle is above this bound with a given probability. The authors pointed out that VEHICLE SYSTEM DYNAMICS 2 25 number of participating vehicles and therefore the proportion of connected vehicles. For instance, in the case of traffic flow of 2000 vehicle/h with 10% of the vehicle participating in the common estimation process, about 13 minutes are required to re-establish a satis- factory estimation of the road condition after a sudden drop in road friction. However, a detailed analysis of converge time of the proposed algorithm in different mixing traffic scenarios is missing. Furthermore, authors did not specify the on-board road friction esti- mator for the implementation of the cooperative fusion technique. However, to meet the requirement of normally distributed on-board estimates, Kalman-based data fusion meth- ods can be exploited as they guarantee that the estimation error is a Gaussian, i.e. normal, random variable. Consequently, several algorithms available in the literature, such as those presented in [142–146] can be adopted for implementing the cooperative strategy in [141]. Usually, when Kalman strategy are used, the road friction coefficient is estimated together with other measures for vehicle dynamics such as vehicle side-slip angle, wheel sideslip angles and wheel slip ratios, and forces acting on tyres. 6. Road friction estimation Consequently, more detailed vehi- cle models compared to those discussed in Sections 2 and 3 are designed for reproducing the vehicle motion while capturing tyre dynamics. For instance, in [143], the second-order nonlinear longitudinal vehicle model discussed in Section 2 is augmented with the dynam- ics of the wheels, i.e. the wheel’s angular velocity and the wheel’s longitudinal slip, which in turn provide the longitudinal tyre force in accordance to the Pacejka model. Moreover, detailed lateral vehicle dynamics for the road friction estimation via Kaman filtering strate- gies has been considered in [142,144–146]. In these works four wheel vehicle models have been designed to capture the dynamics of the yaw rate, lateral forces acting on the wheels, sideslip angles and slip ratios of the wheels, thus providing a more detailed description of the vehicle behaviour compared to that given by bicycle vehicle models discussed for lane change in Section 3. As another example, an approach to cooperatively estimating road friction was pre- sented in [147,148]. The idea is to increase the precision of the estimation of the road friction coefficient available to a given fleet of vehicles that can exchange information by exploiting V2V communication channels. In the proposed framework, each vehicle in the fleet runs a dual-rate estimation scheme composed of (i) a low-level individual vehicle dynamics based estimation scheme, and (ii) an upper-level cooperative estima- tion scheme. The low-level individual parameter identification algorithm generates the individual high-rate estimate by using a dynamic model of the tyres and velocity-based signals. The upper-level cooperative estimation scheme is fed periodically by the individ- ual high-rate crude estimates, and it is used to converge to a common estimate among the vehicles. It is noted that, a longitudinal slip-based road friction estimation method was used as the low-level estimation system. As required by this estimation technique, a regression model of the longitudinal vehicle slips and longitudinal tyre forces, which is linear in the friction coefficient, was found and exploited to design a Least-Square (LS) identification strategy (the reader is referred to [138] for an overview of longitudi- nal and lateral slip-based road friction estimation methods). It is noted that, compared to Kalman filter methods, the design of slip-based strategies require less detailed vehicle dynamic models as they only focus on the estimation of the road friction coefficient rather than the entire vehicle state. 6. Road friction estimation So far, it is expected that the spread is not very significant, therefore, the cooperative methods can provide at least some general warning of possible danger on the road ahead [151,152]. However, besides tyre-road friction also the hydroplaning phenomenon should be consid- ered. Hydroplaning might be observed as very low friction, but it is a different physical phenomenon and it strongly depends on velocity, tread depth and vehicle weight. Conse- quently larger differences might occur between vehicles in hydroplaning cases. Therefore ideally, besides a friction number also the road condition should be estimated and commu- nicated between vehicles. Hence, the question of understanding if cooperative estimation of road friction is beneficial for autonomous driving scenarios is still open. As a possi- ble approach for creating a cooperative estimate of road condition for a fleet of vehicles with large difference in the tyre-road friction coefficients is to augment the information available to on-board vehicle dynamics based estimation systems with environmental sen- sors data gathered by the fleet. Environmental sensors, such as cameras, optical and radio frequency-based sensors utilise changes in the signal reflectance, polarisation and absorp- tion properties caused by the road surface. For instance, optical sensors can sense if a road is slippery by analysing how beams of light are scattered and absorbed by the road surface, while cameras can discern the road type based on pixel luminance levels. The advantage of using environmental sensors is that in principle they might provide an estimate of the road condition independently from the tyre condition and vehicle motion (e.g. they work also for stationary vehicles). Consequently, data gathered from a fleet of vehicles can be The effectiveness of the cooperative method in [147,148] to improve the estimation of the road friction coefficient provided by slip-based identification tecniques was proven numerically where a fleet of five vehicles was considered. A numerical investigation for different road conditions revealed that the proposed cooperative estimation can enhance the estimation of the road friction coefficient by up to 34% with respect to the case the slip-based identificaiton method is performed by each vehicle individually and without any shared information. Furthermore, as a case study, a cooperative algorithm was employed by a collision avoidance controller to generate feasible longitudinal and lateral vehicle acceler- ations. The controller was then numerically tested for a scenario where the subject vehicle had to avoid two consecutive fixed obstacles. 6. Road friction estimation For this scenario, numerical results showed that an accurate estimation of the road friction coefficient was fundamental to avoid col- lisions for slippery roads. It was shown that when the low-level estimation method was used without any correction, the estimation error during the evasive manoeuvre was about 10%, but this accuracy was not sufficient to avoid a collision. On the other hand, when a cooperative approach was exploited, the residual estimation error was smaller than 3% and the vehicle could successfully perform the evasive manoeuvre. Thus, the literature exists with some attempts to cooperatively estimate the road friction coefficient for autonomous vehicles which are supported by promising numerical results. However, no experimental evidence of the effectiveness of these cooperative schemes has been provided to the best of the authors’ knowledges. Furthermore, it is noted that the tyre-road friction not only depends on road conditions but also on tyre conditions. It might be interesting to see how much the tyre-road friction coefficient varies for a variety of tyres and vehicles [150]. So far, it is expected that the spread is not very significant, therefore, the cooperative methods can provide at least some general warning of possible danger on the road ahead [151,152]. However, besides tyre-road friction also the hydroplaning phenomenon should be consid- ered. Hydroplaning might be observed as very low friction, but it is a different physical phenomenon and it strongly depends on velocity, tread depth and vehicle weight. Conse- quently larger differences might occur between vehicles in hydroplaning cases. Therefore ideally, besides a friction number also the road condition should be estimated and commu- nicated between vehicles. Hence, the question of understanding if cooperative estimation of road friction is beneficial for autonomous driving scenarios is still open. As a possi- ble approach for creating a cooperative estimate of road condition for a fleet of vehicles with large difference in the tyre-road friction coefficients is to augment the information available to on-board vehicle dynamics based estimation systems with environmental sen- sors data gathered by the fleet. Environmental sensors, such as cameras, optical and radio frequency-based sensors utilise changes in the signal reflectance, polarisation and absorp- tion properties caused by the road surface. For instance, optical sensors can sense if a road is slippery by analysing how beams of light are scattered and absorbed by the road surface, while cameras can discern the road type based on pixel luminance levels. 6. Road friction estimation However, the estimation accuracy depends on the amplitude and the frequency spectrum of the vehicle slips. For instance, the accuracy reduces for small values of the slips when the amplitude of measurement noise of wheel-speed sensors U. MONTANARO ET AL. 26 become comparable to the slip values [149]. Furthermore, LS methods might fail in case the variability of the vehicle slip profiles is not large enough (i.e. when vehicle slip pro- files do not verify persistent excitation conditions) thus preventing the convergence of the LS-algorithms [147]. LS algorithms [147]. The effectiveness of the cooperative method in [147,148] to improve the estimation of the road friction coefficient provided by slip-based identification tecniques was proven numerically where a fleet of five vehicles was considered. A numerical investigation for different road conditions revealed that the proposed cooperative estimation can enhance the estimation of the road friction coefficient by up to 34% with respect to the case the slip-based identificaiton method is performed by each vehicle individually and without any shared information. Furthermore, as a case study, a cooperative algorithm was employed by a collision avoidance controller to generate feasible longitudinal and lateral vehicle acceler- ations. The controller was then numerically tested for a scenario where the subject vehicle had to avoid two consecutive fixed obstacles. For this scenario, numerical results showed that an accurate estimation of the road friction coefficient was fundamental to avoid col- lisions for slippery roads. It was shown that when the low-level estimation method was used without any correction, the estimation error during the evasive manoeuvre was about 10%, but this accuracy was not sufficient to avoid a collision. On the other hand, when a cooperative approach was exploited, the residual estimation error was smaller than 3% and the vehicle could successfully perform the evasive manoeuvre. Thus, the literature exists with some attempts to cooperatively estimate the road friction coefficient for autonomous vehicles which are supported by promising numerical results. However, no experimental evidence of the effectiveness of these cooperative schemes has been provided to the best of the authors’ knowledges. Furthermore, it is noted that the tyre-road friction not only depends on road conditions but also on tyre conditions. It might be interesting to see how much the tyre-road friction coefficient varies for a variety of tyres and vehicles [150]. 6. Road friction estimation The advantage of using environmental sensors is that in principle they might provide an estimate of the road condition independently from the tyre condition and vehicle motion (e.g. they work also for stationary vehicles). Consequently, data gathered from a fleet of vehicles can be 27 VEHICLE SYSTEM DYNAMICS 27 cooperatively fused and become then the input of machine learning algorithms to provide a preliminary common hypothesis on the road friction level and road condition. How- ever, as environmental based sensors capture only road features without considering tyre dynamics, the common estimates must be further adjusted for each vehicle in the fleet with on-board sensor data to create an individual and customised estimation of the friction level. For instance, if slip-based methods are locally used, the cooperative estimate can become the initial guess of LS algorithms to improve their convergence, while if Kalman-based methods are exploited, the cooperative hypothesis on the road condition can be considered as the output of a noisy sensor of the road friction coefficient to be fused together with other on-board measurements to reduce its uncertainty. Furthermore, each vehicle might store a database containing time histories of the local friction estimate, on-board sensor readings (e.g. tyre slips and estimate tyre forces), cooperative friction estimates and environmental readings. This friction database can then be used to train machine learning techniques to predict the road friction of the upcoming road section based on the environmental sensor data of CAVs travelling on the next road segment. It is noted that the possibility of merg- ing on-board and environmental sensor data together has been investigated for instance in [149] within the European project FRICTION. In this project, information from optical sensors, camera, tyre-sensors is fused together with the road friction estimates provided by vehicle dynamics based methods with the aim of improving the estimation accuracy. The practical benefit of an improved friction estimation was then demonstrated for the case of collision mitigation systems. On the other hand, the authors in [153] have recently pro- posed a machine learning-based method which uses environmental sensor data for road friction predictions from a fleet of connected vehicles. The problem was formulated as a classification task to predict the friction class (slippery or non-slippery) for a sequence of road sections. However, the output of the classifier was used only for warning and its use for CAVs applications in adverse conditions was not investigated. 6. Road friction estimation Hence, additional research towards the merging of approaches such as those presented in [149,153] are expected in the near future. 7. Conclusion In this paper, five use-cases have been analysed with the aim investigating the potential ben- efits and associated limitations of connected autonomous vehicles which leverage off-board vehicle data obtained through communication channels. The use-cases analysed cover dif- ferent areas of connected autonomous vehicles and common features of these use-cases are drawn below: • V2X channels can broaden the sensory horizon of autonomous vehicles by provid- ing additional off-board information (vehicles and features beyond the line of sight, timing of traffic lights, preview of road friction coefficient, etc.). However, the achiev- able benefits in terms of traffic safety, fuel efficiency and traffic flow in a connected environment depend on how this additional information is used by a vehicles control system. Therefore, control algorithms play a pivotal role in intelligently utilising V2X channels. • In addition to a vehicle’s internal states, off-board information (e.g. traffic light timing and velocity of traffic ahead) can help improve energy efficiency. Furthermore, it has • In addition to a vehicle’s internal states, off-board information (e.g. traffic light timing and velocity of traffic ahead) can help improve energy efficiency. Furthermore, it has U. MONTANARO ET AL. 28 been shown that fuel reduction can also be achieved by performing manoeuvres cooper- atively. For instance, in the case of non-signalised intersections, vehicles can collaborate to ensure collision avoidance. In addition, cooperation also reduces the stationary wait times at signalised and non-signalised intersections which helps in reduction of fuel consumption. Similarly, cooperative techniques to equalise the velocities of vehicles over a road segment remove excessive accelerations/decelerations which result in an increase in the global fuel efficiency. fi • In the case of a mixed scenario (of cooperative and non-cooperative vehicles), achiev- able benefits through cooperation not only depend on the control/estimation algo- rithms, but also on the penetration rate and the traffic scenario (e.g. traffic density, numbers of lane, etc.). In addition, human-driven vehicles can also benefit from the presence of connected autonomous vehicles (for instance, in terms of fuel reduction). Such additional benefit is due to the intrinsic cooperation among vehicles that share the same road section.i • Although results discussed throughout the paper confirm that connected autonomous vehicles have the potential to improve traffic safety, fuel efficiency and traffic flow, most of the results have been obtained through simulation under assumptions that might be not completely fulfilled in a real environment, e.g. Notes 1. Subject vehicle refers to the autonomous vehicle, performing a given task or manoeuver, e.g. a lane change, crossing an intersection, cooperative driving in a platoon, etc.i 1. Subject vehicle refers to the autonomous vehicle, performing a given task or manoeuver, e.g. a lane change, crossing an intersection, cooperative driving in a platoon, etc. g g g 2. A communication network topology is defined as the connectivity structure of the vehicu- lar network [41], i.e. the pattern in which vehicles are connected in the platoon via wireless communications. 2. A communication network topology is defined as the connectivity structure of the vehicu- lar network [41], i.e. the pattern in which vehicles are connected in the platoon via wireless communications. 3. The topology is said to be bidirectional when each platoon vehicle gets information only from the predecessor vehicle and the follower vehicle (see also Figure 1(b)). 3. The topology is said to be bidirectional when each platoon vehicle gets information only from the predecessor vehicle and the follower vehicle (see also Figure 1(b)). Disclosure statement No potential conflict of interest was reported by the authors. VEHICLE SYSTEM DYNAMICS 29 29 Funding This work was supported by Jaguar Land Rover and the UK-EPSRC [grant number EP/N01300X/1] as part of the jointly funded Towards Autonomy: Smart and Connected Control (TASCC) Pro- gramme. ORCID Umberto Montanaro http://orcid.org/0000-0003-0620-1906 Shilp Dixit http://orcid.org/0000-0001-9378-442X 7. Conclusion (i) ideal working conditions of the communication channel (e.g. no packet loss, communication failure, noise, etc.), (ii) perfect knowledge of vehicle dynamics (vehicle parameters, road friction condition, etc.), (iii) perfect knowledge of the positions of the vehicles. Hence, additional investiga- tion is required to understand how the aforementioned uncertainties affect cooperating driving scenarios. Although beyond the scope of the paper, it is noted that, from a computer sci- ence perspective, cybersecurity is an active research area for connected autonomous vehicles. Information shared among CAVs in cooperative driving scenarios must be protected from cyber-attacks to guarantee road safety and privacy of CAVs and other road users. The ongoing research on CAVs as cooperative mobile computing systems focuses to identify the cyber threats and to design strategies for preventing dam- ages caused by such cyber-attacks. Cyber threats and attacks studied in the literature include: (a) impersonation attack (the attacker pretends to be a legitimate vehicle with the aim to send false messages); (b) message spoofing (the attacker sends false messages to spread wrong information in the network); (c) spamming attack (use- less messages are spread to increase the transmission latency and bandwidth usage); (d) sybil attack (the attacker pretends to have multiple identities and act as if it were a large fleet of CAVs) and (e) message tampering (the attacker aims to drop, mod- ify or corrupt the messages sent by legitimate vehicles to prevent other vehicles to know the original data). Requirements for cybersecurity solutions for improving trust- worthiness of information source ranges from (i) authentication (vehicles must use messages transmitted only by legitimate network), (ii) non-repudiation (if required a sender must not deny a transmission of a message), to (iii) integrity (received messages are the same as the original messages and they have not been altered dur- ing the transmission). The reader is referred to recent surveys on cybersecurity for CAVs [154–157] available in the computer science literature on mobile computing for a detailed and comprehensive analysis of the cyber threats and the corresponding solutions. VEHICLE SYSTEM DYNAMICS 29 References g y y p g [20] Van Arem B, Van Driel C, Visser R. The impact of cooperative adaptive cruise control on traffic-flow characteristics. IEEE Trans Intell Transp Syst. 2006;7(4):429–436. fil [21] Arnaout GM, Arnaout JC. Exploring the effects of cooperative adaptive cruise con- trol on highway traffic flow using microscopic traffic simulation. Transp Plann Technol. 2014;37(2):186–199. [22] Zabat M, Stabile N, Frascaroli S, et al. The aerodynamic performance of platoons a final report; 1995. Available from http://www.its.berkeley.edu. y ] Shladover SE. PATH at 20-history and major milestones. IEEE Trans Intell Transp Sys 2007;8(4):584–592. ] Davila A, Aramburu E, Freixas A. Making the best out of aerodynamics: platoons. SAE 201 World Congress and Exhibition; Detroit; 2013.fi [25] Turri V, Besselink B, Johansson KH. Cooperative look-ahead control for fuel-efficient and safe heavy-duty vehicle platooning. IEEE Trans Control Syst Technol. 2017;25(1):12–28. [26] Varaiya P. Smart cards on smart roads: problems of control. IEEE Trans Automat Contr. 1993;38(2):195–207.i [27] Kianfar R, Falcone P, Fredriksson J. Safety verification of automated driving system. IEEE Intell Transp Syst Mag. 2013;5(4):73–86. [28] Nowakowski C, O’Connell J, Shladover SE, et al. Cooperative adaptive cruise control: driver acceptance of following gap settings less than one second. Proceedings of the Human Factors and Ergonomics Society 54th Annual Meeting; 2010. p. 2033–2037. g y g [29] Swaroop D, Hedrick JK, Chien CC, et al. A comparision of spacing and headway control laws for automatically controlled vehicles. Vehicle Syst Dyn. 1994;23(1):597–625. y y y [30] Rajamani R. Vehicle dynamics and control. New York: Springer; 2012. [ ] j y p g [31] Hao H, Barooah P. Stability and robustness of large platoons of vehicles with double integrator models and nearest neighbor interaction. Int J Robust Nonlin. 2013;23(18):2097–2122. [32] Zheng Y, Li SE, Wang J, et al. Stability and scalability of homogeneous vehicular pla- toon: study on the influence of information flow topologies. IEEE Trans Intell Transp Syst. 2016;17(1):14–26. ( ) [33] Naus GJL, Vugts RPA, Ploeg J, et al. String-stable CACC design and experimental validation: a frequency-domain approach. IEEE Trans Veh Technol. 2010;59(9):4268–4279.fil q y pp [34] Treiber M, Kesting A. Traffic flow dynamics: data, models and simulation. Heidelberg: Springer; 2013. p g [35] Xiao L, Gao F. Practical string stability of platoon of adaptive cruise control vehicles. IEEE Trans Intell Transp Syst. 2011;12(4):1184–1194. [36] Milanés V, Shladover SE. Handling cut-in vehicles in strings of cooperative adaptive cruise control vehicles. References [1] Shladover SE. Cooperative vehicle-highway automation systems. IEEE Intell Transp Syst Mag. 2009;1(1):10–19. [2] Thomas P, Morris A, Talbot R, et al. Identifying the causes of road crashes in Europe. Ann Adv Automot Med Assoc Adv Automot Med Sci Conf. 2013;57:13–22. [3] NHTSA. Critical reasons for crashes investigated in the national motor vehicle crash causation survey. National Highway Traffic Safety Administration; 2015. p. 1–2. (Report No. DOT HS 812 115). [4] Ferreira JC, Almeida J, Rodrigues A. The impact of driving styles on fuel consumption: a data warehouse and data mining based discovery process. IEEE Trans Intell Transp Syst. 2015;16(5):2653–2662. [5] 5G-PPP. 5G Automotive Vision; 2015. Available from: https://5g-ppp.eu/white-papers/. ] Kim S-W, Liu W, Ang MH, et al. The impact of cooperative perception on decision makin and planning of autonomous vehicles. IEEE Intell Transp Syst Mag. 2015;7(3):39–50. ] Casavola A, Mosca E, Papini M. Connected and autonomous vehicles introducing the futur mobility; 2002. Available from: http://www.atkinsglobal.com. [8] During M, Lemmer K. Cooperative maneuver planning for cooperative driving. IEEE Intell Transp Syst Mag. 2016;8(3):8–22.i p y g ] INRIA. Cybercars 2 final report; 2009. Available from: http://www.transport-research.info. i [10] SARTRE. Safe road trains for the environment; developing strategies and technologies to allow vehicle platoons to operate on normal public highways; 2012. Available from: http://www.sartre-project.eu. [11] AutoNet2030. Cooperation among vehicles with or without a driver; 2016. Available from: http://www.autonet2030.eu/. [12] UKCITE. UK connected intelligent transport environment; 2016. Available from: https:// www.ukcite.co.uk. [13] I-MOTORS. i-Motors driving innovation; 2016. Available from: http://www.i-motors.cloud/. [13] I-MOTORS. i-Motors driving innovation; 2016. Available from: http://www.i-motors.cloud/. U. MONTANARO ET AL. 30 [14] G-ACTIVE. Green adaptive control for future interconnected vehicles. Available from: http://g-active.uk/. [15] Stevens A, Dianati M, Katsaros K, et al. Cooperative automation through the cloud: the CARMA project. ITS European Congress; 2017. [16] House of Lords Science and Technology Select Committee. Connected and autonomous vehicles: the future?; 2017. Available from: https://publications.parliament.uk. [17] Kuutti S, Fallah S, Katsaros K, et al. A survey of the state-of-the-art localisation tech- niques and their potentials for autonomous vehicle applications. IEEE Internet Things J. 2018;5(2):829–846. [18] Xu L, Wang L, Yin G. Coordinated control and communication for enhanced safety of highway vehicle platoons. Proceedings of the 2013 International Conference on Connected Vehicles and Expo (ICCVE); 2013 Dec 2–6; Las Vegas, NV, USA; 2013. p g [19] Ioannou PA. Automated highway systems. New York: Springer; 1997. References J Intell Transp Syst. 2016;20(2):178–191. [37] Seiler P, Pant A, Hedrick K. IEEE transactions on automatic control. Veh Syst Dyn. 2004;49(10):1835–1841. [38] Shaw E, Hedrick JK. String stability analysis for heterogeneous vehicle strings. American Control Conference; 2007 Jul 11–13; New York; 2007. p. 3118–3125. 31 VEHICLE SYSTEM DYNAMICS 31 [39] Ploeg J, Van De Wouw N, Nijmeijer H. Fault tolerance of cooperative vehicle platoons subject to communication delay. Proceedings of the 12th IFAC Workshop on Time Delay Systems (TDS 2015); 2015 Jun 28–30; Ann Arbor, MI, United States; 2015. p. 352–357. [40] Zegers JC, Semsar-Kazerooni E, Ploeg J, et al. Consensus control for vehicular platoon- ing with velocity constraints. IEEE Transactions on Control System Technology; 2017. doi:10.1109/TCST.2017.2720141 [41] Jia Dongyao D, Ngoduy D. Enhanced cooperative car-following traffic model with the combination of V2V and V2I communication. Transp Res B Methodol. 2016;90:172–191. [42] Li SE, Zheng Y, Li K, et al. An overview of vehicular platoon control under the four-component framework. Proceedings of the 2015 IEEE Intelligent Vehicles Symposium (IV); 2015 Jun 28–Jul 1; Seoul, South Korea; 2015. p. 286–291. p [43] Montanaro U, Tufo M, Fiengo G, et al. On convergence and robustness of the extended coop- erative cruise control. Proceedings of the 53rd IEEE Conference on Decision and Control; 2014 Dec 15–17; Los Angeles, CA, USA; 2014. p. 4083–4088. g p [44] Gerdes JC, Hedrick JK. Vehicle speed and spacing control via coordinated throttle and brake actuation. Control Eng Pract. 1997;5(11):1607–1614. g [45] Li SE, Gao F, Cao D, et al. Multiple-model switching control of vehicle longitudinal dynamics for platoon-level automation. IEEE Trans Veh Technol. 2016;65(6):4480–4492. p [46] Lin F, Fardad M, Jovanovic MR. Optimal control of vehicular formations with nearest neighbor interactions. IEEE Trans Automat Contr. 2012;57(9):2203–2218. [47] Kwon J-W, Chwa D. Adaptive bidirectional platoon control using a coupled sliding mode [46] Lin F, Fardad M, Jovanovic MR. Optimal control of vehicular formations with nearest neighbor interactions. IEEE Trans Automat Contr. 2012;57(9):2203–2218. ] Kwon J-W, Chwa D. Adaptive bidirectional platoon control using a coupled sliding mod control method. IEEE Trans Intell Transp Syst. 2014;15(5):2040–2048. ] Peters AA, Middleton RH, Mason O. Leader tracking in homogeneous vehicle platoons wit broadcast delays. Automatica. 2014;50(1):64–74. [49] Montanaro U, Tufo M, Fiengo G, et al. A novel cooperative adaptive cruise control approach: theory and hardware in the loop experimental validation. References Proceedings of the 22nd Mediter- ranean Conference on Control and Automation; 2014 Jun 16–19; Palermo, Italy; 2014. p. 37–42. [50] Zheng Y, Li SE, Li K, et al. Distributed model predictive control for heterogeneous vehicle pla- toons under unidirectional topologies. IEEE Trans Control Syst Technol. 2017;25(3):899–910. [51] Shladover SE, Su D, Lu X-Y. Impacts of cooperative adaptive cruise control on freeway trafficl [50] Zheng Y, Li SE, Li K, et al. Distributed model predictive control for heterogeneous vehicle pla- toons under unidirectional topologies. IEEE Trans Control Syst Technol. 2017;25(3):899–910.fi [51] Shladover SE, Su D, Lu X-Y. Impacts of cooperative adaptive cruise control on freeway traffic flow. J Transp Res. 2012;2324:63–70. l [52] Konduri S, Pagilla PR, Darbha S. Vehicle platooning with multiple vehicle look-ahead information. Proceedings of the IFAC World Congress; Toulouse; 2017. p. 5768–5773. [53] Zheng Y, Li SE, Li K, et al. Stability margin improvement of vehicular platoon con- sidering undirected topology and asymmetric control. IEEE Trans Control Syst Technol. 2016;24(4):1253–1265. [54] Wu Y, Li SE, Zheng Y, et al. Distributed sliding mode control for multi-vehicle systems with positive definite topologies. Proceedings of the 2016 IEEE 55th Conference on Decision and Control (CDC); Las Vegas, NV, USA; 2016. p. 5213–5219. g p [55] Al-Jhayyish AMH, Schmidt KW. Feedforward strategies for cooperative adaptive cruise control in heterogeneous vehicle strings. IEEE Trans Intell Transp Syst. 2018;19(1):113–122.f [56] Montanaro U, Fiengo G, Tufano A, et al. On the effectiveness of the extended coopera- tive adaptive control for vehicles platooning. Proceedings of the 2016 European Control Conference (ECC); 2016 Jun 29–Jul 1; Aalborg, Denmark; 2016. p. 2453–2458. g p [57] Gao F, Li SE, Zheng Y, et al. Robust control of heterogeneous vehicular platoon with uncertain dynamics and communication delay. IET Intell Transp Syst. 2016;10(7):503–513. [58] Gao F, Dang DF, Huang SS, et al. Decoupled robust control of vehicular platoon with identical controller and rigid information flow. Int J Automot Technol. 2017;18(1):157–164. [59] Harfouch YA, Yuan S, Baldi S. Adaptive control of interconnected networked systems with application to heterogeneous platooning. Proceedings of the 2017 13th IEEE International Conference on Control & Automation (ICCA); 2017 Jul 3–6; Ohrid, Macedonia; 2017. p. 212–217. 32 U. MONTANARO ET AL. 32 [60] Rajamani R, Tan HS, Law B, et al. Demonstration of integrated longitudinal and lateral control for the operation of automated vehicles in platoons. IET Intell Transp Syst. 2000;8: 695–708. [61] Lim EHM, Hedrick JK. References Lateral and longitudinal vehicle control coupling for automated vehi- cle operation. Proceedings of the 1999 American Control Conference (Cat. No. 99CH36251); 1999 Jun 2–4; San Diego, CA, USA; 1999. p. 3676–3680.fi g [62] Sasoh A, Ohara T. Shock wave relation containing lane change source term for two-lane traffic flow. J Phys Soc Japan. 2002;71(9):2339–2347. l [63] Center for Urban Transportation Research University of South Florida. Impact of automated vehicles on highway safety and operations; 2013.ffi g y y [64] Moridpour S, Rose G, Sarvi M. Effect of surrounding traffic characteristics on lane changing behavior. J Transp Eng. 2010;136(11):973–985. [65] Ardelt M, Coester C, Kaempchen N. Highly automated driving on freeways in real traffic using a probabilistic framework. IEEE Trans Intell Transp Syst. 2012;13(4):1576–1585. [66] Ulbrich S, Maurer M. Situation assessment in tactical lane change behavior planning for auto- mated vehicles. Proceedings of the 2015 IEEE 18th International Conference on Intelligent Transportation Systems; 2015 Sept 15–18; Las Palmas, Spain; 2015. p. 975–981.fi [67] Kamal MAS, Taguchi S, Yoshimura T. Efficient vehicle driving on multi-lane roads using model predictive control under a connected vehicle environment. Proceedings of the 2015 IEEE Intelligent Vehicles Symposium (IV); 2015 Jun 8–Jul 1; Seoul, South Korea; 2015. p. 736–741. [68] Kamal MAS, Taguchi S, Yoshimura T. Efficient driving on multilane roads under a connected vehicle environment. IEEE Trans Intell Transp Syst. 2016;17(9):2541–2551.fi [69] Dao T, Clark CM, Huissoon JP. Distributed platoon assignment and lane selection for traffic flow optimization. Proceedings of the 2008 IEEE Intelligent Vehicles Symposium; 2008 Jun 4–8; Eindhoven, Netherlands; 2008. p. 739–744. p [70] Dao T. A decentralized approach to dynamic collaborative driving coordination [PhD thesis]. University of Waterloo; 2008.fi [71] Jin Q, Wu G, Boriboonsomsin K, et al. Improving traffic operations using real-time optimal lane selection with connected vehicle technology. Proceedings of the 2014 IEEE Intelligent Vehicles Symposium Proceedings; 2014 Jun 8–11; Dearborn, MI, USA; 2014. p. 70–75. [72] Roess RP, Prassas ES. The highway capacity manual: a conceptual and research history. Basel: Springer; 2014. [73] Milanés V, Llorca DF, Villagrá J, et al. Intelligent automatic overtaking system using vision for vehicle detection. Expert Syst Appl. 2012;39(3):3362–3373. p y pp ] Dixit S, Fallah S, Montanaro U, et al. Trajectory planning and tracking for autonomou overtaking: state-of-the-art and future prospects. Annu Rev Control. 2018;45:76–86. [75] Kim B, Kim D, Park S, et al. References Automated complex urban driving based on enhanced envi- ronment representation with GPS/map, radar, lidar and vision. Proceedings of the IFAC Symposium on Advances in Automotive Control; Norrköping; 2016. p. 190–195. [76] Kitazawa S. Control target algorithm for direction control of autonomous vehicles in consider- ation of mutual accordance in mixed traffic conditions. Proceedings of the 2016 International Symposium on Advanced Vehicle Control (AVEC ’16); 2016 Sept 13–16; Munich; 2016. y p p [77] Naranjo JE, González C, García R, et al. Lane-change fuzzy control in autonomous vehicles for the overtaking maneuver. IEEE Trans Intell Transp Syst. 2008;9(3):438–450. [78] Gao Y, Gray A, Frasch JV, et al. Spatial predictive control for agile semi-autonomous ground vehicles. Proceedings of the 2012 International Symposium on Advanced Vehicle Control (AVEC ’12); 2012 Sept 9–12; Seoul, Korea; 2012. [79] Nilsson J, Gao Y, Carvalho A, et al. Manoeuvre generation and control for automated highway driving. Proceedings of the 19th World Congress The International Federation of Automatic Control; 2014 Aug 24–29; Cape Town, South Africa; 2014. p. 6301–6306. g p p [80] Glaser S, Vanholme B, Mammar S, et al. Maneuver-based trajectory planning for highly autonomous vehicles on real road with traffic and driver interaction. IEEE Trans Intell Transp Syst. 2010;11(3):589–606. 3 VEHICLE SYSTEM DYNAMICS 33 [81] Amer NH, Zamzuri H, Hudha K, et al. Modelling and control strategies in path tracking control for autonomous ground vehicles: a review of state of the art and challenges. J Intell Robot Syst. 2016;86(2):225–254. [82] Katrakazas C, Quddus M, Chen W, et al. Real-time motion planning methods for autonomous on-road driving: state-of-the-art and future research directions. Transp Res C Emerg Technol. 2015;60:416–442. [83] Kim C, Kim M, Lee K. Development of a full speed range path-following system for the autonomous vehicle. Proceedings of the International Conference on Control, Automation and Systems; 2015 Oct 13–16; Busan, Korea; 2015. p. 710–715. y p [84] Filev D, Lu J, Hrovat D. Future mobility: integrating vehicle control with cloud computing. Mech Eng. 2013;135(3):18–24.f [85] Kong J, Pfeiffer M, Schildbach G, et al. Kinematic and dynamic vehicle models for autonomous driving control design. Proceedings of the 2015 IEEE Intelligent Vehicles Symposium (IV); 2015 Jun 28–Jul 1; Seoul, South Korea; 2015. p. 1094–1099. p [86] Kang CM, Lee SH, Chung CC. Comparative evaluation of dynamic and kinematic vehicle models. References Proceedings of the 53rd IEEE Conference on Decision and Control; 2014 Dec 15–17; Los Angeles, CA, USA; 2014. p. 648–653. g p [87] Snider JM. Automatic steering methods for autonomous automobile path tracking. Pitts- burgh, PA: Robotics Institute; 2009. (Tech. Rep. CMU-RITR-09-08). [88] Sorniotti A, Barber P, De Pinto S. Path tracking for automated driving: a tutorial on con- trol system formulations and ongoing research. Automated Driving: Safer and More Efficient Future Driving; 2017. p. 71–140. g p [89] Rupp A, Stolz M. Survey on control schemes for automated driving on highways. Automated Driving; 2017. p. 43–69. [90] NHTSA. National motor vehicle crash causation survey. U.S. Department of Transportation, National Highway Traffic Safety Administration; 2008.fi fi [91] European Commission. Traffic safety basic facts: junction; 2016. Available from: https://ec. europa.eu. [92] Rios-Torres JM, Malikopoulos AA. A survey on coordination of connected and automated vehicles at intersections and merging at highway on-ramps. IEEE Trans Intell Transp Syst. 2017;18(5):1066–1077. [93] Chen L, Englund C. Cooperative intersection management: a survey. IEEE Trans Intell Transp Syst. 2016;17(2):570–586.fi y [94] Lebre M, Mouël F, Ménard E, et al. Real scenario and simulations on GLOSA traffic light system for reduced CO2 emissions, waiting time and travel time. Proceedings of the 22nd ITS World Congress; 2015 Oct; Bordeaux, France; 2015. g [95] Katsaros K, Kernchen R, Dianati M, et al. Performance study of a green light optimal speed advisory (GLOSA) application using an integrated cooperative ITS simulation plat- form. Proceedings of the 7th International Wireless Communications and Mobile Computing Conference; 2011 Jul 4–11; Istanbul, Turkey; 2011. p. 918–923. [96] Seredynski M, Dorronsoro B, Khadraoui D. Comparison of green light optimal speed advisory approaches. Proceedings of the 16th International IEEE Conference on Intelligent Trans- portation Systems (ITSC 2013); 2013 Oct 6–9; The Hague, Netherlands; 2013. p. 2187–2192. p y g p [97] Wan N, Vahidi A, Luckow A. Optimal speed advisory for connected vehicles in arterial roads and the impact on mixed traffic. Transp Res C: Emerg Technol. 2016;69:548–563.fi fi [98] Mahler G, Vahidi A. An optimal velocity-planning scheme for vehicle energy efficiency through probabilistic prediction of traffic-signal timing. IEEE Trans Intell Transp Syst. 2014;15(6):2516–2523. [99] Mahler G, Vahidi A. Reducing idling at red lights based on probabilistic prediction of traffic signal timings. Proceedings of the 2012 American Control Conference (ACC); 2012 Oct 2; Montréal, QC, Canada; 2012. p. 6557–6562. References p [100] Kamal MAS, Imura J, Ohata A, et al. Control of traffic signals in a model predictive control framework. Proceedings of the 13th IFAC Symposium on Control in Transportation Systems U. MONTANARO ET AL. 34 The International Federation of Automatic Control; 2012 Sept 12–14; Sofia, Bulgaria; 2012. p. 38–43. The International Federation of Automatic Control; 2012 Sept 12–14; Sofia, Bulgaria; 2012. p. 38–43. [101] Kamal MAS, Imura J, Hayakawa T, et al. Traffic signal control of a road network using MILP in the MPC framework. Int J Intell Transp Syst Res. 2015;13(2):107–118. [102] Yang H, Rakha H, Ala MV. Eco-cooperative adaptive cruise control at signalized intersections considering queue effects. IEEE Trans Intell Transp Syst. 2016;18(6):1–11. [103] HomChaudhuri B, Vahidi A, Pisu P. A fuel economic model predictive control strategy for a group of connected vehicles in urban roads. Proceedings of the 2015 American Control Conference (ACC); 2015 Jul 1–3; Chicago, IL, USA; 2015. p. 2741–2746.fi [104] HomChaudhuri B, Vahidi A, Pisu P. Fast model predictive control-based fuel efficient control strategy for a group of connected vehicles in urban road conditions. IEEE Trans Control Syst Technol. 2017;25(2):760–767. [105] Hult R, Campos GR, Steinmetz E, et al. Coordination of cooperative autonomous vehicles. IEEE Signal Process Mag. 2016;33(6):74–84. g g [106] Dresner K, Stone P. Multiagent traffic management: a reservation-based intersection control mechanism. Proceedings of the 3rd International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS); 2004 Jul 23; New York, USA; 2004. p. 530–537.i g y p [107] Huang S, Sadek AW, Zhao Y. Assessing the mobility and environmental benefits of reservation-based intelligent intersections using an integrated simulator. IEEE Trans Intell Transp Syst. 2012;13(3):1201–1214.fi [108] Dresnert K, Stone P. Multiagent traffic management: an improved intersection control mech- anism. Proceedings of the 4th International Joint Conference on Autonomous Agents and Multiagent Systems (AAMAS); 2005 Jul; Utrecht, The Netherlands; 2005. p. 471–477. [109] Dresner K, Stone P. A multiagent approach to autonomous intersection management. J Artif Intell Res. 2008;31:591–656. [110] Zohdy IA, Rakha HA. Intersection management via vehicle connectivity: the intersection cooperative adaptive cruise control system concept. J Intell Transp Syst. 2016;20(1):17–32. [111] Zohdy IH, Kamalanathsharma RK, Rakha H. Intersection management for autonomous vehi- cles using iCACC. Proceedings of the 2012 15th International IEEE Conference on Intelligent Transportation Systems; 2012 Sept 16–19; Anchorage, AK, USA; 2012. p. 1109–1114. [112] Lee J, Park B. References Development and evaluation of a cooperative vehicle intersection con- trol algorithm under the connected vehicles environment. IEEE Trans Intell Transp Syst. 2012;13(1):81–90. [113] Kamal MAS, Imura J, Ohata A, et al. Coordination of automated vehicles at a traffic-lightless intersection. Proceedings of the 16th International IEEE Conference on Intelligent Trans- portation Systems (ITSC 2013); 2013 Oct 6–9; The Hague, Netherlands; 2013. p. 922–927. p y g p [114] Hult R, Campos GR, Falcone P, et al. An approximate solution to the optimal coordination problem for autonomous vehicles at intersections. Proceedings of the 2015 American Control Conference (ACC); 2015 Jul 1–3; Chicago, IL, USA; 2015. p. 763–768. g p [115] U.S. Energy Information Administration. Study of the potential energy consumption impacts of connected and automated vehicles; 2017. Available from: www.eia.gov. [116] Zhang C. Predictive energy management in connected vehicles: utilizing route information preview for energy saving [PhD thesis]. Clemson University; 2010. [117] Ansarey M, Shariat Panahi M, Ziarati H, et al. Optimal energy management in a dual-storage fuel-cell hybrid vehicle using multi-dimensional dynamic programming. J Power Sources. 2014;250:359–371. [118] Bellman R. Dynamic programming. Princeton: Princeton University Press; 1957.i [119] Ozatay E, Onori S, Wollaeger J, et al. Cloud-based velocity profile optimization for everyday driving: a dynamic-programming-based solution. IEEE Trans Intell Transp Syst. 2014;15(6):2491–2505. [120] Qiu C, Shen H, Sarker A, et al. Towards green transportation: fast vehicle velocity optimization for fuel efficiency. Proceeding of the 2016 IEEE International Conference on Cloud Com- puting Technology and Science (CloudCom); 2016 Dec 12–15; Luxembourg City; 2017. p. 59–66. 3 VEHICLE SYSTEM DYNAMICS 35 [121] Asadi B, Vahidi A. Predictive cruise control: utilizing upcoming traffic signal information for improving fuel economy and reducing trip time. IEEE Trans Control Syst Technol. 2011;19(3):707–714. [122] Németh B, Csikós A, Varga I, et al. Design of platoon velocity based on multi-criteria optimization. Proceedings of the 7th IFAC Symposium on Robust Control Design The Inter- national Federation of Automatic Control; 2012 Jun 20–22; Aalborg, Denmark; 2012. p. 523–528. [123] Murgovski N, Egardt B, Nilsson M. Cooperative energy management of automated vehicles. Control Eng Pract. 2016;57:84–98. [124] Johansson I, Jin J, Ma X, et al. Look-ahead speed planning for heavy-duty vehicle platoons using traffic information. Transp Res Procedia. 2017;22:561–569. fi [125] Wu G, Dong Z. Design, analysis and modeling of a novel hybrid powertrain system based on hybridized automated manual transmission. Mech Syst Signal Process. 2017;93: 688–705. [126] Sciarretta A, Guzzella L. References Control of hybrid electric vehicles. IEEE Cont Syst Mag. 2007;27(2):60–70. [127] Ehsani M, Gao Y, Emadi A. Modern electric, hybrid electric, and fuel cell vehicles: fundamen- tals, theory, and design. Boca Raton: CRC Press; 2009.fi y g [128] Sun C, Moura SJ, Hu X, et al. Dynamic traffic feedback data enabled energy manage- ment in plug-in hybrid electric vehicles. IEEE Trans Control Syst Technol. 2015;23(3):1075– 1086. [129] Herrera JC, Work DB, Herring R, et al. Evaluation of traffic data obtained via GPS- enabled mobile phones: the mobile century field experiment. Transp Res C Emerg Technol. 2010;18(4):568–583. [130] Martnez CM, Cao D, Hu X, et al. Energy management in plug-in hybrid electric vehi- cles: recent progress and a connected vehicles perspective. IEEE Trans Veh Technol. 2017;66:4534–4549. [131] Malikopoulos AA. Supervisory power management control algorithms for hybrid electric vehicles: a survey. IEEE Trans Intell Transp Syst. 2014;15(5):1869–1885. [132] Huang Y, Wang H, Khajepour A, et al. Model predictive control power management strategies for HEVs: a review. J Power Sources. 2017;341:91–106.i [133] Cheng Q, Nouvelire L, Orfila O. A new eco-driving assistance system for a light vehicle: energy management and speed optimization. Proceedings of the 2013 IEEE Intelligent Vehicles Symposium (IV); 2013 Jun 23–26; Gold Coast, QLD, Australia; 2013. p. 1434–1439. y p p [134] Lim H, Mi CC, Su W. A distance-based two-stage ecological driving system using an estima- tion of distribution algorithm and model predictive control. IEEE Trans Veh Technol. 2017 Aug;66(8):6663–6675. g [135] Rakha HA, Ahn K, Moran K, et al. Virginia tech comprehensive power-based fuel consumption model: model development and testing. Transp Res D Transp Environ. 2011;16(7):492–503. [136] Lim H, Su W, Mi CC. Distance-based ecological driving scheme using a two-stage hierar- chy for long-term optimization and short-term adaptation. IEEE Trans Veh Technol. 2017 Mar;66(3):1940–1949. [137] Muller S, Uchanski M, Hedrick K. Estimation of the maximum tire-road friction coefficient. J Dyn Syst Meas Control. 2003;125(4):607.fi y y [138] Rajamani R, Piyabongkarn N, Lew J, et al. Tire-road friction-coefficient estimation. IEEE Control Syst Mag. 2010;30(4):54–69. y g [139] Gustafsson F. Automotive safety systems. IEEE Signal Process Syst. 2009;26(4):32–47 [140] Volvo. Volvo cars puts 1000 test cars to use: Scandinavian cloud-based project for shar- ing road-condition information becomes a reality; 2014. Available from: https://www.media. volvocars.com. [141] Müller S, Uchanski M, Hedrick JK. Cooperative estimation using vehicle communications. Veh Syst Dyn. 2003;39(2):121–133. U. MONTANARO ET AL. U. MONTANARO ET AL. References 36 [142] Antonov S, Fehn A, Kugi A. Unscented Kalman filter for vehicle state estimation. Veh Syst Dyn. 2011;49(9):1497–1520. [143] Enisz K, Fodor D, Szalay I, et al. Improvement of active safety systems by the extended Kalman filter based estimation of tire-road friction coefficient. IEEE International Electric Vehicle Conference; Florence; 2014. ] Xu Y, Deng B, Xu G. Estimation of vehicle states and road friction based on DEKF. Interna tional Conference on Power Electronics Systems and Applications. Hong Kong; 2015. [145] Cheng Q, Correa-Victorino A, Charar A. A new nonlinear observer using unscented Kalman filter to estimate sideslip angle, lateral tire road forces and tire road friction coefficient. Pro- ceedings of the 2011 IEEE Intelligent Vehicles Symposium (IV); 2011 Jun 5–9; Baden-Baden, Germany; 2011. p. 709–714. y p [146] Chen L, Bian M, Luo Y, et al. Estimation of road-tire friction with unscented Kalman filter and MSE-weighted fusion based on a modified Dugoff tire model. SAE World Congress and Exhibition; Detroit; 2015. [147] Jalalmaab M, Pirani M, Fidan B, et al. Cooperative least square parameter identification by consensus within the network of autonomous vehicles. SAE Int J Passeng Cars — Electron Electr Syst. 2016;9(2):255–261. y [148] Jalalmaab M, Pirani M, Fidan B, et al. Cooperative road condition estimation for an adaptive model predictive collision avoidance control strategy. Proceedings of the 2016 IEEE Intelligent Vehicles Symposium (IV); 2016 Jun 19–22; Gothenburg, Sweden; 2016. p. 1072–1077. ] Koskinen S. Sensor data fusion based estimation of tyre road friction to enhance collisio avoidance [PhD thesis]. Technical Research Centre of Finland; 2010. [ ] ; [150] Pacejka H. Tire and vehicle dynamics. Oxford: Butterworth-Heinemann; 2005. [150] Pacejka H. Tire and vehicle dynamics. Oxford: Butterworth-Heinemann; 2005. [151] Prokes J. Realtime estimation of tyre-road friction for vehicle state estimator [MSc Thesis]. Chalmers University; 2015. y [152] Michelin. The tyre, grip. Société de Technologie Michelin; 2001. [153] Panahandeh G, Ek E, Mohammadiha N. Road friction estimation for connected vehicles using supervised machine learning. IEEE Intelligent Vehicles Symposium; 2017. [154] Mokhtar B, Azab M. Survey on security issues in vehicular ad hoc networks. Alexandria Eng J. 2015;54(4):1115–1126. [155] Petit J, Shladover SE. Potential cyberattacks on automated vehicles. IEEE Trans Intell Transp Syst. 2015;16(2):546–556. [156] He Q, Meng X, Qu R. Survey on cyber security of cav. Forum on Cooperative Positioning and Service Survey; 2017. y [157] De La Torre G, Rad P, Choob K-KR. References Driverless vehicle security: challenges and future research opportunities. Future Generation Computer System; 2018. doi:10.1016/j.future.2017.12.041.
W3165440811.txt	https://www.mdpi.com/1422-0067/22/11/5808/pdf?version=1622454620	en		Gender Differences in Diabetic Kidney Disease: Focus on Hormonal, Genetic and Clinical Factors	International journal of molecular sciences	2,021	cc-by	16,804	International Journal of Molecular Sciences Review Gender Differences in Diabetic Kidney Disease: Focus on Hormonal, Genetic and Clinical Factors Annalisa Giandalia 1 , Alfio Edoardo Giuffrida 2 , Guido Gembillo 2,3 , Domenico Cucinotta 1 , Giovanni Squadrito 1 , Domenico Santoro 2 and Giuseppina T. Russo 1, * 1 2 3 * Citation: Giandalia, A.; Giuffrida, A.E.; Gembillo, G.; Cucinotta, D.; Squadrito, G.; Santoro, D.; Russo, G.T. Gender Differences in Diabetic Kidney Disease: Focus on Hormonal, Genetic and Clinical Factors. Int. J. Mol. Sci. 2021, 22, 5808. https:// doi.org/10.3390/ijms22115808 Department of Clinical and Experimental Medicine, University of Messina, 98125 Messina, Italy; agiandalia@yahoo.it (A.G.); domenico.cucinotta@unime.it (D.C.); giovanni.squadrito@unime.it (G.S.) Unit of Nephrology and Dialysis, Department of Clinical and Experimental Medicine, University of Messina, 98125 Messina, Italy; alfiogiuffrida91@libero.it (A.E.G.); guidogembillo@live.it (G.G.); domenico.santoro@unime.it (D.S.) Department of Biomedical, Dental, Morphological and Functional Imaging Sciences, University of Messina, 98125 Messina, Italy Correspondence: giuseppina.russo@unime.it Abstract: Diabetic kidney disease (DKD) is one of the most serious complications of both type 1 (T1DM) and type 2 diabetes mellitus (T2DM). Current guidelines recommend a personalized approach in order to reduce the burden of DM and its complications. Recognizing sex and genderdifferences in medicine is considered one of the first steps toward personalized medicine, but the gender issue in DM has been scarcely explored so far. Gender differences have been reported in the incidence and the prevalence of DKD, in its phenotypes and clinical manifestations, as well as in several risk factors, with a different impact in the two genders. Hormonal factors, especially estrogen loss, play a significant role in explaining these differences. Additionally, the impact of sex chromosomes as well as the influence of gene–sex interactions with several susceptibility genes for DKD have been investigated. In spite of the increasing evidence that sex and gender should be included in the evaluation of DKD, several open issues remain uncovered, including the potentially different effects of newly recommended drugs, such as SGLT2i and GLP1Ras. This narrative review explored current evidence on sex/gender differences in DKD, taking into account hormonal, genetic and clinical factors. Academic Editors: Kyoichiro Tsuchiya and Lisa M. Keywords: gender; sex; diabetic kidney disease; estrogens; gene polymorphisms Harrison-Bernard Received: 12 April 2021 Accepted: 25 May 2021 1. Introduction Published: 28 May 2021 Diabetic kidney disease (DKD) is one of the most common microvascular complication of diabetes mellitus (DM), affecting ~30% of subjects with type 1 (T1DM) and ~40% of those with type 2 (T2DM) [1]. DKD is diagnosed according to the presence of albuminuria, reduced estimated glomerular filtration rate (eGFR), or both [2]. In the last few decades, research on DKD has witnessed enormous activity, a revolution encompassing epidemiology, diagnosis, clinical manifestations, risk factors and treatment options [3]. It has become clear that while in T1DM individuals, the natural history of DKD progresses from microalbuminuria, the first sign of renal impairment, to macroalbuminuria, and eventually to the decline of GFR toward end stage renal disease (ESRD), the path in T2DM is more heterogeneous. Thus, T2DM patients may present with impaired eGFR even only a few years after diagnosis, and they may progress to ESRD without ever developing albuminuria [4]. Accordingly, the recent advances in DKD’s pathophysiological and clinical aspects have prompted the use of the term “DKD” to include all types of renal injury occurring in diabetic individuals: the classical albuminuric phenotype, the “nonalbuminuric renal impairment” and the “progressive renal decline” [5]. Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Int. J. Mol. Sci. 2021, 22, 5808. https://doi.org/10.3390/ijms22115808 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2021, 22, 5808 2 of 24 These advances have been highlighted in a joint document of the Italian Diabetes Society (SID) and the Italian Society of Nephrology (SIN), providing an extensive review of the available evidence as well as updated treatment recommendations [5]. However, whether these phenotypes and their evolution and/or management have the same course in DM men and women is still a matter of debate. Considering sex- and gender-specific aspects is one of the first and simpler step toward personalized and patient-centered care also in the management of DM and its complications. Gender medicine analyzes the differences between men and women in human physiology, pathophysiology, and the clinical features of diseases, specifically evaluating the impact of sex as a biological and functional marker, and that of gender, which refers to a complex interrelation and integration of sex with psychological, social, ethnical and cultural behavior. Despite over 20 years of gender medicine and the insistent recommendations of scientific societies and research institutions [6], sex-based differences in biology, genetics, biomedical and clinical aspects of major diseases, including DM and its complications, are still poorly explored so far, a gap that may impair the efficacy of our diagnostic and therapeutic efforts, ultimately exposing patients to unwanted outcomes [7]. In this review, we explored potential sex- and gender-based differences in DKD prevalence and evolution, risk factors, clinical manifestations and treatment options, including the role of hormonal and genetic factors. 2. Gender Differences in the Prevalence of DKD and its Phenotypes Gender-related differences have been reported in non-diabetic chronic kidney disease (CKD) [8]. Overall, CKD seems to have a higher prevalence in women than in men [9–11]. However, a review including a large number of studies found 38 studies reporting a higher CKD prevalence in women and 13 among men [12]. On the contrary, the risk of progression to ESRD appears to be higher among males [13,14]. A recent analysis from the nationwide Swedish Renal Registry-CKD (SRR-CKD), showed that among adult patients with incident CKD stage G3b-5, women had a lower risk of CKD progression (sub hazard ratio [SHR] 0.88 (0.85–0.92)), and a lower all-cause (SHR 0.90, 95% CI 0.85–0.94) and cardiovascular mortality (SHR 0.83, 95% CI 0.76–0.90), compared to men [15]. Accordingly, a large metaanalysis confirmed a more rapid decline in men than in women with non-diabetic CKD [16]. Women live longer than men and age and post-menopausal status appear to modify the association between sex and nondiabetic kidney disease [17]; in this regard, Jafar et al. reported that old post-menopausal women had a faster renal progression compared to age-matched men [18]. How CKD is defined may also play a relevant role in determining the effect of gender on CKD incidence and progression risk. Thus, when using eGFR-based definitions of CKD, the incidence of CKD was significantly higher in women than men [19]. In cohorts of DM subjects, several large epidemiological studies have explored sex differences in the prevalence of DKD and its phenotypes, specifically evaluating low eGFR, micro- or macroalbuminuria or both (Table 1). Studies performing a separate gender analysis varied in design, sample size, type of diabetes, length of follow-up, and how DKD was reported, as shown in Table 1 [13,20–58] and in two recently published reviews [59,60]. Int. J. Mol. Sci. 2021, 22, 5808 3 of 24 Table 1. Sex/gender differences in DKD phenotypes. Sex Specific Association Reference Ethnic Group Study Design MAU Low eGFR Gall, 1997 [20] Denmark Prospective Lewis J, 2001 [21] Multi-ethnic Intervention study Keane WF, 2003 [22] Multi-ethnic Intervention study Female Rossing K, 2004 [23] Denmark Prospective Both sexes ESRD Male Higher risk NR NR Female Higher risk NR NR Higher risk Higher risk Higher risk NR Higher risk Higher risk NR Studies on T2DM subjects Retkaran, 2006 [24] UK Prospective Male/Female Higher risk male Higher risk female Penno, 2011 [25] Italy Cross-sectional Male/Female Higher risk male Higher risk female NR Yu M, 2012 [26] USA Cross-sectional Male/Female Higher risk male Higher risk male Higher risk female Jardine, 2012 [27] UK Intervention study Male NR NR Higher risk Zoppini, 2012 [28] Italy Prospective Both sexes NR Higher risk NR Altemtan,2012 [29] UK Retrospective Both sexes NR Higher risk NR Elley, 2013 [30] New Zealand Nationwide cohort Female NR NR Higher risk de Hautecloque 2014 [31] France Prospective Male NR Higher risk Higher risk Kaiwara 2016 [32] Japan Prospective Female NR Higher risk NR Male Higher risk NR NR Studies on T1DM subjects Orchard, 1990 [33] US Prospective Lovshin, 1990 [34] US Cross-sectional Male Higher risk NR NR Holl, 1999 [35] Germany Retrospective Female Higher risk NR NR Jacobsen, 1999 [36] Denmark Prospective Male NR Higher risk NR Rossing, 2002 [37] Denmark Prospective Both sexes Higher risk NR NR Zhang, 2003 [38] US Prospective Male Higher risk NR NR Hovind, 2004 [39] Denmark Prospective Male Higher risk NR NR Finne, 2005 [40] Finland Register Male NR NR Higher risk Sibley, 2006 [41] US Prospective Male Higher risk NR NR RAile, 2007 [42] Germany Prospective Male Higher risk NR NR Monti, 2007 [43] US Cross-sectional Both sexes Higher risk Higher risk NR Mollsten, 2010 [44] Sweden Population Male NR NR Higher risk Costacou, 2011 [45] US Prospective Male 1950-1964 Higher risk NR Higher risk Costacou, 2011 [45] US Prospective Female1965-1980 Higher risk NR Higher risk Harjutsalo, 2011 [46] Finland Prospective Male NR NR Higher risk Kautzy-Willer 2013 [47] Austria Cross-sectional both sexes Higher risk Higher risk NR Skupien, 2019 [48] Multi-ethnic Prospective Male NR NR Higher risk Studies including T1DM/T2DM subjects Dick, 1994 [49] Canada Prospective Female NR NR Higher risk Xue, 2007 [51] USA Prospective Female NR NR Higher risk Yamagotha, 2007 [52] Japan Prospective Male Higher risk Both sexes NR Hippsley-Cox,2010 [53] UK Registrative data Female NR Both sexes Higher risk Hoffman F, 2011 [54] Germany Claims data Female NR NR Higher risk Johnson, 2011 [55] USA Retrospective Male NR NR Higher risk Tohidi, 2012 [56] Iran Prospective Female NR Higher risk NR Kei, 2013 [57] Japan Prospective Both sexes Higher risk NR NR van Blijderveen, 2014 [58] Netherlands Retrospective Male/Female Higher risk Male Both sexes Higher risk Female Haroun, 2003 [50] USA Prospective Female NR NR Higher risk F Ricardo,2018 [13] USA Prospective Male Higher risk Male Both sexes Higher risk Male Abbreviations: DKD, diabetic kidney disease; eGFR, estimated glomerular filtration rate; MAU, micro/macroalbuminuria; ESRD, End stage renal disease; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; NR, sex/gender differences not reported. A recent meta-analysis of 10 studies including data from >5 million subjects evaluated the relative effect of diabetes on CKD and ESRD in women compared with men. The pooled adjusted risk ratio of DKD was 3.34 (95 % CI 2.27, 4.93) in women and 2.84 (95 % Int. J. Mol. Sci. 2021, 22, 5808 4 of 24 CI 1.73, 4.68) in men, without any difference in diabetes-related risk of DKD, with the exception of ESRD [61]. Furthermore, in the SURDIAGENE study, survival without ESRD was higher in T2DM women than in men, who showed an higher risk of steeper eGFR decline [31]. Among the studies that evaluated DKD phenotypes specifically in T2DM cohorts, the UK Prospective Diabetes Study 74 (UKPDS) found that T2DM men had a higher risk of microalbuminuria, while T2DM women were at higher risk of developing low eGFR (<60 mL/min/1.73 m2 ) [24]. Similarly, a population study from UK showed a higher prevalence of impaired eGFR in DM women compared to men [62], and a higher prevalence of the non-albuminuric phenotype among T2DM women was also reported in the large cohort from the RIACE study, in Italy [63] Age may impact the associations among T2DM, gender and DKD risk. In T2DM patients, the prevalence of both eGFR and albuminuria increases with age, with male sex also positively associated with albuminuria rise and negatively with eGFR impairment in older subjects [64]. The impact of age (<60 and ≥60 years) and gender on the prevalence of DKD, advanced DKD (eGFR <30 mL/min/1.73 m2 ), and common risk factors, was specifically analyzed in the 4839 participants of the Pathways Study [26]. This study showed that women had, overall, a 28% decreased OR of DKD (OR 0.72, 95% CI 0.62–0.83), but a higher prevalence of advanced DKD (OR 1.67, 95% CI 1.05–2.64). Although the prevalence of microalbuminuria was higher among men with T2DM, women presented a greater risk of advanced renal dysfunction and higher prevalence of common DKD risk factors, with these differences being most evident amongst older subjects. Male gender also appears to be an independent risk factor for DKD incidence, especially when the albuminuric phenotype is taken into account. A prospective observational study of 191 T2DM patients followed for a median period of 5.8 years [20] found that male sex was the second risk factor after albuminuria for the development of incipient or overt DKD. Moreover, an association between sex and incident DKD was also found in 1464 patients with diabetes and normal renal function at baseline, followed-up for almost 10 years [65]. Risk factors for prevalence, incidence and progression of DKD and its phenotypes have been recently evaluated in the large cohort of the AMD Annals Initiative, an observational study including >400,000 T2DM and >25,000 adult T1DM subjects, from 251 diabetologists’ centers collected across Italy (Table 2). Figure 1 [66–73] shows gender-specific DKD phenotypes distribution in T1DM and T2DM from the AMD Annals Initiative. DKD, defined as eGFR <60 mL/min and/or micro- macroalbuminuria, affected 47.3% T2DM men (29.8% isolated micro- macroalbuminuria, 6.6% isolated low eGFR, 11% both) and 43.8% T2DM women (18.3% micro-macroalbuminuria, 14.5% isolated low eGFR, 11.2% both) [66]. Furthermore, when predictors of DKD were analyzed in >120,000 T2DM patients over a 4 year follow-up period, male gender was a positive risk factor for the incidence of albuminuria and a negative risk factor for the development of eGFR decline (eGFR <60 mL/min) [72]. Table 2. Age- and gender differences in DKD prevalence, incidence and progression in T1DM and T2DM participants in AMD Annals Initiative. T1DM Prevalence Incidence 5 years 4 years Progression eGFR<60 mL/min or >30% reduction Male Sex DKD 1.01 (0.91–1.11) p = 0.901 Low eGFR 0.64 (0.55–0.74) p < 0.001 MAU 1.26 (1.14–1.40) p < 0.001 DKD 1.01 (0.81–1.27) p = 0.913 Low eGFR 0.96 (0.59–1.56) p = 0.873 MAU 1.04 (0.82–1.32) p = 0.760 0.59 (0.46–0.76) p < 0.001 Age By 10 year DKD 1.15 (1.10–1.19) p < 0.001 Low eGFR 1.85 (1.74–1.96) p < 0.001 MAU 0.93 (0.89–0.87) p < 0.001 By 10 year DKD 1.07 (0.96–1.18) p = 0.203 Low eGFR 1.95 (1.57–2.43) p < 0.001 MAU 0.93 (0.84–1.04) p = 0.227 By 10 year 1.46 (1.30–1.63) p < 0.001 Int. J. Mol. Sci. 2021, 22, 5808 5 of 24 Table 2. Cont. Int. J. Mol. Sci. 2021, 22, 5808 5 of 23 T2DM Prevalence Incidence 5 years 4 years Progression eGFR <60 mL/min or >30% reduction Low1.89 eGFR(1.81–1.98) 0.69 (0.64–0.73) Low eGFR1.075 1.373(1.033–1.118) (1.326–1.422) pp<<0.001 In subjects with DM and hypertension MAU MAU 0.001 MAU 1.89 (1.81–1.98) MAU 1.075 (1.033–1.118) p < 0.001 hypertension BothBoth 1.521.52 (1.42–1.63) Both 1.381 (1.306–1.460) p < 0.001 0.78 (0.72–0.86) p 0.001 < 0.001 (1.42–1.63) Both 1.381 (1.306–1.460) p < 0.001 0.78 (0.72–0.86) p < By 1 year By 10 year By 1 year By 10 year By10 10year year Low Low eGFR 1.121.12 (1.11–1.12) Low 0.767(0.681–0.864) (0.681–0.864) < 0.001 eGFR (1.11–1.12) LoweGFR eGFR 0.767 p < p0.001 By AgeAge MAU (1.01–1.01) MAU 1.355 p <p0.001 1.49 (1.41–1.58) p <p 0.001 MAU 1.011.01 (1.01–1.01) MAU 1.355(1.220–1.504) (1.220–1.504) < 0.001 1.49 (1.41–1.58) < 0.001 Both 1.12 (1.12–1.13) Both 1.090 (0.926–1.283) p = 0.30 Both 1.12 (1.12–1.13) Both 1.090 (0.926–1.283) p = 0.30 Abbreviations: DKD, diabetic kidney disease;eGFR, eGFR, estimated filtration rate;rate; MAU, microalbuminuria. Abbreviations: DKD, diabetic kidney disease; estimatedglomerular glomerular filtration MAU, microalbuminuria. Male Sex Although the overall epidemiological Although the theavailable availableinformation informationisisstill stillnot notconclusive, conclusive, the overall epidemiologidata indicate that that the risk developing DKD DKD is higher in DMin men, have a higher cal data indicate the of risk of developing is higher DMwho men,also who also have a risk of DKD progression. However, when DKD specific phenotypes are taken into account, higher risk of DKD progression. However, when DKD specific phenotypes are taken into DM men are higher of developing albuminuric while women are at account, DMatmen arerisk at higher risk of the developing the phenotype, albuminuric phenotype, while higher risk of eGFR impairment, and of developing ESRD, especially at older ages. These women are at higher risk of eGFR impairment, and of developing ESRD, especially at differences apply to both of diabetes, important for the older ages. seem Theseto differences seemtype to apply to both with type of diabetes,implications with important imdiagnosis and management of DKD in the clinical practice. plications for the diagnosis and management of DKD in the clinical practice. The reported sex sex and and gender-disparities in DKD stillare largely The reasons reasonsbehind behindthese these reported gender-disparities in are DKD still unknown, but hormonal or genetic or differences, as well as differences the prevalence or largely unknown, but hormonal genetic differences, as well asindifferences in the impact of major risk factors seem to play a relevant role. prevalence or impact of major risk factors seem to play a relevant role. Figure 1.1. Renoprotective Renoprotective effects effects of offemale femalehormones hormoneson ondiabetic diabetickidney kidneydisease. disease. Abbreviations: Abbreviations: DKD, DKD, diabetic diabetic kidney kidney Figure disease; ER, estrogen receptor; HRT, hormone replacement therapy; GPER-1, G protein-coupled estrogen receptor disease; ER, estrogen receptor; HRT, hormone replacement therapy; GPER-1, G protein-coupled estrogen receptor 1.1. 3. Impact of Female Sex Hormones on DKD Sex hormones play an important role in the pathophysiology of diabetes and its complications, especially in DM women, who seem to lose the protective effects of estrogens on the cardiovascular bed, even before menopause. In the last few decades, the pleiotropic effects of estrogens beyond those on the reproductive system have been ob- Int. J. Mol. Sci. 2021, 22, 5808 6 of 24 3. Impact of Female Sex Hormones on DKD Sex hormones play an important role in the pathophysiology of diabetes and its complications, especially in DM women, who seem to lose the protective effects of estrogens on the cardiovascular bed, even before menopause. In the last few decades, the pleiotropic effects of estrogens beyond those on the reproductive system have been objects of intense research [74–76], including their potential role in DKD (Figure 1). The activity of estrogens is closely related to the presence of specific receptors that are ubiquitously localized, with particular reference to the vascular district and endothelial cells [77]. Estrogens and their metabolites operate through a classical pathway with nuclear Estrogen Receptors (ERs) such as ER-α and ER-β [78,79] exerting their genomic actions. These effects are guaranteed by different signaling pathways such as MAPK/ERK, PI3K and the important NF-KB via [80]. These pathways lead to decreased apoptosis processes, cellular growth, differentiation, and inflammation [81,82]. The ER-α and ER-β are localized in several areas of crucial importance. One of these is the hypothalamus, especially the nuclei arcuate, paraventricular, lateral, and ventral regions, where estrogens produce relevant effects on food intake and thirsty [83] Another site is the skeletal muscle, where both ER-α and ER-β are expressed and contribute to glucose homeostasis, reducing the expression of GLUT4 [84]. This role has also been confirmed in studies on mice with ER-α knockout; these mice showed insulin resistance and alteration in glucose levels [85]. Several receptors have been identified in adipose tissue, whereas only ER-α is expressed in brown adipose tissue [86], suggesting its key-role in obesity onset [87]. ER-α is also the most predominant estrogenic receptor in the liver [88] while during the fetal phase ER-β is more represented [89]. The former has especially predominant anti-inflammatory effects on the liver [90]. The cardiovascular system is not exempt from ERs’ presence, and their interaction with estrogens can ameliorate heart failure and inhibit apoptosis and fibrosis [91]. Furthermore, activation of ER-β pathway leads to the reduction of cardiac fibrosis in women [92]. Experimental studies demonstrated that ER-α expression is abnormally represented in the diabetic kidney [93]. Estrogen’s most important metabolite is the 17 β Estradiol (E2) [94]; in healthy conditions, this metabolite acts as a vasodilator, increasing the endothelial expression of nitric oxide synthase and resulting in phosphorylation and nitric oxide production via the ER-α receptor [95]. Moreover, E2 seems to attenuate glomerulosclerosis and tubulointerstitial fibrosis [96]. In animal models estrogens seem to counter fibrosis and apoptosis in the kidney [97], while testosterone promotes pro-inflammatory, proapoptotic and pro-fibrotic processes [98–100]. These findings are partially in contrast with evidence from human studies showing an association of oral contraceptives and estrogen replacement therapy with an increased risk of microalbuminuria and kidney function decline [101,102]. Estrogens may exert their effects even through another receptor, associated with protein G, the G protein-coupled estrogen receptor 1 (GPER-1), which explains their rapid non-genomic effects [103]. This is a membrane receptor encoded by the GPER gene located at chromosome 7p22.3, and its expression has been demonstrated in the hypothalamus, hypophysis, adrenal gland, ovary and particularly in the renal pelvis [104,105]. Some studies have also evidenced a predominant expression of GPER-1 in renal tubular cells [106,107], and therapy with GPER-1 agonist in female mice with salt-sensitive disease has been reported to improve glomerular function and hypertrophy and to reduce proteinuria [108], thus suggesting a key role of this receptor in kidney disorders and, probably, in DKD. This hypothesis is further supported by data coming from the sustained use of icarine, a GPER-1 activator. This metabolite seems to improve the nephropathy of T1DM mice, through GPER mediated p62-dependent Keap1 degradation and Nrf2 activation, also attenuating mesangial expansion [109]. GPER is implicated in different pathways by several receptors such as serotonin 1A receptor [110–113]. It also works with GPER/TGF-β1 via inducing cellular proliferation or increasing the expression of type IV collagen [114] and NO production [115]. Int. J. Mol. Sci. 2021, 22, 5808 7 of 24 Other membrane receptors were identified: ER-α 36, a variant of ER-α and ER-β. ER-α 36 is another splicing variant; it plays an important role, inhibiting wild-type ER-α (ER-α66) and ER-β, being involved in the resistance of breast cancer to hormonal treatment [116] and in testosterone-initiated carcinogenesis [117]. The experimental data on estrogen supplementations in murine models of diabetes also seem to confirm the role of female hormones in renal protection. Thus, Wells et al. demonstrated a reduction in circulating estrogen levels and an increase in the renal ER-α—ER-β expression ratio in diabetic rats. Supplementation with E2 restored this rate, supporting the hypothesis of nephroprotection exerted by estrogens in DM [118]. Furthermore, in diabetic mice, estrogen pellet implantation was able to inhibit glomerulosclerosis, collagen IV deposition and albuminuria, even in animals with an advanced stage of renal injury, suggesting the efficacy of E2 treatment in the reduction of DKD progression [119]. On the other hand, the inhibition of estradiol synthesis using the anastrozole, an aromatase inhibitor, partially attenuated renal injury in male streptozotocininduced diabetic rats [120]. Both the levels of E2 and the androgen-to-E2 ratio seem to be crucial factors for progression of renal injury in diabetic subjects; estrogenic activity is modulated by androgens, probably through the accessibility of ER to E2. In humans, DKD is associated with an increase in estrogen concentration and a decrease in testosterone levels in male patients [121], but not in women [122,123]. The potential beneficial effect of hormonal replacement therapy (HRT) on DKD has also been explored, with some studies supporting the efficacy of estrogenic therapy in improving insulin sensitivity, plasma lipid levels and creatinine clearance in postmenopausal diabetic women [124–126]. Maric et al. reviewed the current literature in this field, underlining the protective role of estrogens in DKD and the role of oral supplementation of 17 β estradiol in attenuating its evolution [127]. Further investigations have shown that treatment with 17 β estradiol decreased albuminuria, tubulointerstitial fibrosis and glomerulosclerosis in the diabetic population [128]. Oral therapy with selective estrogen receptor modulators showed similar benefits. Additionally, a provisional regimen with raloxifene, a selective estrogen receptor modulator, reduced albuminuria levels in post-menopausal women with T2DM [129]. Furthermore, improvements of important kidney outcomes, including a reduction in the progression of the albumin–creatinine ratio in post-menopausal T2DM women [130] has been reported in a randomized trial with raloxifen. Accordingly, the oral combination therapy with estradiol and norgestrel improved eGFR and proteinuria in T2DM postmenopausal women with hypertension, and this nephroprotective action was not related to the modification of conventional risk factors such as blood pressure and lipids [126]. Conversely, a recent meta-analysis demonstrated the short-term benefits of HRT on lipid profile in young women with CKD, although a reduction in CV morbidity and/or mortality was not observed [131]. Vitamin D could also exert a synergetic action with estrogens in renal protection and diabetes control [132]. Thus, vitamin D acts as a real steroid hormone and its level is influenced by estrogen status [133]. It exerts a modulating action at both tubular [134] and glomerular level [135], with peculiar protective action in DKD [136]. Moreover, preand post-menopausal DM women with an adequate vitamin D status seem to have a better glycemic control [137]. Estrogens interfere with vitamin D immunomodulatory activities [138] and, in turn, vitamin D down-regulates aromatase action, with a reduction of the adverse events linked to peripheral estrogen overexpression [139]. Combined therapy with vitamin D supplementations and sex steroids seems to protect endothelium integrity, contrasting the cardiovascular damage that contributes to CKD and DKD progression [140,141]. Multiple mechanisms behind the reno-protective effects of oral estrogen supplementation in DM women have been reported. Chronic hyperglycemia leads to an increase of reactive oxygen species (ROS), and an impairment of nitric oxide (NO) secretion [142]. Int. J. Mol. Sci. 2021, 22, 5808 8 of 24 Additionally, the polyol pathways are involved in the DKD genesis [143] Another pathway of glycemic damage is represented by the accumulation of Advanced Glycation End Products (AGEs), which are derived from nonenzymatic glycosylation of product proteins or lipids [144,145]. AGEs reduce the efficiency of anti-oxidant systems, downregulating several protective molecules, such as AGER1 and SIRT1 [146,147]. E2 therapy can interfere with these pathways at different levels. It reduces the expression of transforming growth factor-beta (TGF β), AT1 receptors, and endothelines with a decreased production of collagen (especially I and IV) [148–151] and a reduction of apoptotic phenomena [152]. The increased activity of nitric oxide synthase at the glomerular level may be another effect of estrogens in the kidney, improving vascular permeability and glomerular function [153]. A study on ovariectomized rats has shown an increased expression in glomerular tissue of SIRT1 mRNA in those treated with E2. In addition, E2 increased ER α -mRNA expression in the glomerular mesangium and reduced the fibrotic process and TGF-β levels [154]. TGF-β plays a key role in the genesis of DKD, increasing the production of extracellular matrix, and that of collagen IV in the podocytes and expanding the mesangial area [155,156]. Its expression is upregulated by a state of chronic hyperglycemia, leading to glomerulosclerosis [157–159], and TGF-β levels are usually increased in men and reduced in women [160], thus indirectly suggesting a crucial role of estrogens in regulating TGF-β. Accordingly, ER-α, the major estrogenic receptor expressed at the renal level, has also been shown to bind to several target molecules [161]. Other metabolic pathways are involved in the regulation of TGF β levels, such as the protein kinase system (CK2) [162] and the renin–angiontensin–aldosterone system (RAAS) that increase TGF-β production [163], and estrogens have been also reported to reduce the activity of RAAS and, therefore, stimulate TGF-β [164]. Beyond the role of estrogens, progesterone also seems to play an important role in kidney protection. Thus, progesterone receptors are mainly localized in the epithelial cells of distal tubule [164], both in the medulla and cortex kidney of male and female subjects [165]. Loss of renal function related to the ageing process and damage to of the proximal tubule could be prevented by the administration of estrogen alone or even by the addition of progesterone replacement therapy [166], which also demonstrated beneficial effects on the ischemic tubular damage [167], further supporting their role in nephroprotection. Accordingly, Baha et al. highlighted that the administration of progesterone for 10 weeks in ovariectomized diabetic mice improves the different outcomes of diabetic nephropathy, reducing glomerulosclerosis and profibrotic/angiogenetic factors (TGF-β, vascular endothelial growth factor -A, type 1 receptor of angiotensin II), downregulating podocyte markers such as nephrin and podocin [168]. In spite of this increasing experimental evidence of the renoprotective role of estrogens alone or in combination with progesterone, large prospective trials with HT in postmenopausal T1DM and T2DM women with variable stages of CKD are still needed in order to evaluate their possible role in the treatment of this high-risk population in future. 4. Impact of Sex Genes Interactions on DKD Sex may impact the effect of genes at different levels. Recent evidence supports the role of sex chromosomes on renal impairment in non-diabetic individuals, as reported for Alport syndrome, arising from a mutation in COL4A5 on chromosome X [169], and showing a different renal prognosis in affected males and females [170,171]. However, the influence of sex is far more complex, and sex is an important modifier of the influence of genetic background on several chronic diseases, including CKD. Overall, the genetic heritability of CKD has been estimated to range from 30 to 75% [172,173] and several lines of evidence confirmed the relevant role of genetic predisposition in the initiation and progression of renal complications both in T1DM and T2DM subjects. Epidemiological studies have revealed familial clustering of DKD in both types Int. J. Mol. Sci. 2021, 22, 5808 9 of 24 of diabetes [174,175], and a relevant influence of ethnical background [176–178]. More than 150 genes have been associated with DKD in T1DM and T2DM, although with a different biological relevance, and most of them have been identified through genomewide association studies (GWAS) [179–183]. A recent GWAS analysis on DKD, involving large T2DM and T1DM cohorts and taking into account eight complementary dichotomous and quantitative DKD phenotypes, identified a novel loci (near GABRR1, rs9942471) specifically associated with microalbuminuria in European T2DM case subjects only, with no signal in Asian diabetic subjects or in those with T1DM irrespective of ethnical origin [184]. These findings indicate that a phenotype-, ethnicity- and type of diabetes-driven analysis may be more specific in the identification of genetic susceptibility to DKD. Accordingly, in T1DM, a GWAS analysis of 19,406 subjects using various definitions of DKD, based on renal function and albuminuria, recognized 16 loci, including protective variants (the rs55703767 minor allele, Asp326Tyr) and a variant (rs55703767), on the collagen type IV alpha 3 chain (COL4A3) gene, with the most significant association with DKD in T1DM patients [185]. Furthermore, the minor C allele of rs17389016 of the 11β-Hydroxysteroid dehydrogenase 1 (HSD11B1) was recently associated with the “fast decliner” phenotype and overt DKD in a cohort of 466 T1DM subjects (OR = 2.10; CI 95% = 1.14–3.89; p = 0.018) [186]. Despite this overwhelming evidence, the identified genes and single nucleotide polymorphisms (SNPs) only explain a minor part of the genetic susceptibility to DKD. Thus, epigenetic mechanisms, i.e., DNA methylation, chromosome histone modification and noncoding RNA (ncRNA) regulation [187,188] and their potential interactions with personal or environmental factors, including sex, may also play an important role [189]. Although several genetic loci in genes implied in the RAAS, inflammation, oxidation, glucose and lipid metabolism have been associated with DKD, the potential role of sex–gene interaction has been evaluated for only a few of them [8]. In one of the first studies of gene–gender interaction on DKD, a case–control study from the Joslin Diabetes center, the M235T variant in the angiotensinogen gene, which is associated with a greater expression of this gene, increased DKD risk only in T1DM men [190]. In T2DM study subjects from the Health Professionals Follow-Up Study (HPFS) and the Nurses’ Health Study (NHS), sex-specific associations were found between the angiotensin II type 1 receptor gene AGT1R 1166 C-allele and AGT 235T and coronary heart disease (CHD), whereas the AGT1R T573 C-allele variant was not associated with CHD or DKD [191]. Furthermore, a case–control study in a large cohort of T1DM patients from Denmark, Finland, France and Sweden, found that the AA genotype of the rs5186 AGTR1 polymorphism significantly increased the DKD risk in male patients (OR = 1.27; 95% CI = 1.02–1.58, p = 0.03), after adjustment for multiple confounders, whereas no significant associations were noted in women [192]. Sex differences were also noted for common variants in carnosinase genes on chromosome 18q, CNDP1 and CNDP2 [193], with the 5-leucine repeat (5L-5L) variant of the CNDP1 gene being associated with a reduced prevalence of DKD in T2DM women [194], whereas the rs12604675-A variant in CNDP1 was shown to confer higher susceptibility to overt proteinuria in T2DM women from Japan [195]. The common angiotensin-converting enzyme (ACE) polymorphism (I = insertion, D = deletion) has also been extensively studied as a DKD susceptibility gene in both T1DM and T2DM. While this variant was not able to explain the observed gender differences in ESDR occurrence in black DM individuals [196], it seemed to have an independent impact on survival in DM patients on dialysis [197]. Furthermore, T2DM women carriers of the ACE D allele were found to be at increased risk of DKD progression, whereas no difference was found in T2DM men, even after adjustment for multiple confounders [198]. Sex–genes interactions were also reported for genes implied in inflammation and oxidation, as well as in other DKD-related risk factors. Int. J. Mol. Sci. 2021, 22, 5808 10 of 24 Significant interaction with sex was also noted for the IL-6 (rs1800795) genetic variant in a study on predictive genetic models of microalbuminuria in relatives of subjects with DKD [199]. In 1120 T1DM subjects (529 men and 591 women), the SNP rs11915160 SOX2 gene, located in chromosome 3q26.33, was significantly associated with DKD and ESRD in women but not in men, with a combined effect with the adiponectin promoter polymorphism rs266729 [200]. Sex-specific associations were also reported for two SNPs in the regulatory regions of CYBB (NOX2, coding, respectively, for superoxide-generating nicotinamide adenine dinucleotide phosphate-oxidase 2) and GPX4 (glutathione peroxidase 4), involved in the redox status. The minor A-allele of CYBB rs6610650 was associated with DKD in women, whereas the minor T-allele of GPX4 rs713041 showed an inverse association with DKD in T1DM men of South America and European origins [201]. T1DM male carriers of the 59029G allele and those with the 32-bp deletion on the secreted (RANTES) receptor gene (CCR5) variant, which is associated with the diminished expression of CCR5 on immunocompetent cells, had a greater risk of DKD, compared with non-carriers. Furthermore, the distribution of the combined haplotypes with these two variants differed significantly in men with and without DKD, but not in women [202]. GWAS studies in T1M cohorts also identified sex specific associations with DKD susceptibility variants. Thus, a GWAS study in the large cohort of the Finnish Diabetic Nephropathy (FinnDiane) Study found that a common variant, the rs4972593 on chromosome 2q31.1, that was associated with ESRD in women but not in men, and this was confirmed in a meta-analysis of three independent T1DM cohorts [203]. Notably, this variant (rs4972593) is able to interact with ERα, modulating the expression of genes implicated in glomerular function and cell proliferation, thus potentially contributing to the sex-specific protection against ESRD [203]. Another study conducted on three cohorts including T1DM of European descent found that the IGF2BP2 polymorphism, a variant protein that binds to 5’-UTR of the imprinting IGF2 gene, was associated with DKD only in male T1DM subjects. In the same cohorts, a genetic interaction between IGF2BP2 and IGF2 gene was also identified, suggesting a protective role against DKD in male T1DM subjects [204]. In T2DM women, the PNPLA3 rs738409 polymorphism, a genetic risk factor for non-alcoholic fatty liver disease, was associated with impaired eGFR values and DKD prevalence, irrespective of the presence of NAFLD and common cardio–renal risk factors [205]. Polymorphisms in cholesteryl ester transfer protein (CETP) gene, a key enzyme in triglyceride metabolism, were also investigated as genetic risk factors for DKD. In a group of T2DM women, followed up for ~9 years, the Taq1B variant was not associated with the risk of developing DKD, whereas it predicted the onset of diabetic retinopathy [206]. Conversely, another study, conducted on a total of 3023 Taiwanese individuals (1383 without and 1640 with T2DM) found that the A-allele of rs1800775 in the CETP gene was significantly related to a lower DKD risk (OR, 0.78; 95% CI, 0.64-0.96) [207]. Unfortunately, no gene–gender specific analysis was conducted in this study and ethnic differences may also have played a role in differentiating these results. No sex differences were found according to five SNPs in other genes involved in lipid metabolism such as the SLC2A9 and ABCG2 genes in DKD [208]. Genetic background has a relevant role in increasing or decreasing the risk of developing DKD; however, the relative importance of each genetic variant seems to vary according to type of diabetes, study design, DKD phenotypes (albuminuric/low eGFR), and ethnic background. Overall, the available evidence suggests that sex should be considered among the variables potentially influencing the impact of genes on DKD risk. Thus, when a sex-specific analysis was conducted, some of these genetic risk factors were exclusively or more strongly associated with DKD in either DM men or women (Table 3) [180–186,189–195,198,200–205,207]. Future research should persevere to confirm Int. J. Mol. Sci. 2021, 22, 5808 11 of 24 sex-specific associations and basic research should investigate the mechanisms behind these observed results. Table 3. Studies reporting sex specific associations of selected gene variants with DKD phenotypes. Reference Ethnic Group Locus Study Design Lin, 2009 [191] US AGT1R (1166) cohort study Tien, 2009 [198] Taiwan ACE D/I Mooyaart, 2011 [181] Multi-ethnic 24 gene variants: ACE, AKR1B1 (two variants), APOC1, APOE, EPO,NOS3 (two variants), HSPG2, VEGFA, FRMD3 (two variants), CARS (two variants), UNC13B, CPVL and CHN2, and GREM1, plus 3 variants not near genes. Ahluwalia, 2011 [193] Sweden Ahluwalia, 2011 [193] Sex Specific Association * MAU Low eGFR DKD Male/Female Not significant Higher risk Higher risk Prospective Female Higher risk Higher risk Higher risk GWAS NR Higher risk Higher risk Higher risk CNDP1 (rs2346061) Case– control Male/Female Higher risk Not significant Higher risk Sweden CNDP2 (rs7577) Case– control Female Higher risk Not significant Higher risk Kurashige, 2013 [195] Japan CNDP1 (rs12604675) Case– control Female Higher risk Not significant Higher risk Alkhalaf, 2015 [194] Netherlands CNDP1 (5L-5L) prospective Male/Female Not significant Not significant Not significant Teumer, 2015 [183] European RAB38/CTSC (rs649529), HS6ST1 (rs13427836), CUBN (rs10795433) GWAS NR NR NR NR Prudente, 2017 [180] Italy UMOD (rs12917707) Crosssectional NR NR NR NR Russo, 2017 [206] Italy CETP Taq1B cohort Female Not significant Not significant Not significant van Zuydam, 2018 [184] European GABRR1 (rs9942471) GWAS NR Higher risk Not significant Higher risk Huang, 2019 [207] Taiwan CETP rs1800775 Crosssectional NR Higher risk Higher risk Higher risk Mantovani, 2019 [205] Italy PNPLA3 rs738409 Crosssectional Female NR Higher risk Higher risk Vujkovic, 2020 [189] Multi-ethnic UMOD GWAS Not significant Not significant Higher risk Higher risk US AGT (M235T) case–control Male Higher risk Not significant Higher risk Male Higher risk Higher risk Higher risk Studies on T2DM subjects Studies on T1DM subjects Freire, 1998 [190] Miynarski, 2005 [202] US CCR5 (A59029G) Case– control Miynarski, 2005 [202] US CCR5 (32bp deletion) Case– control Male Higher risk Higher risk Higher risk Mollsten, 2011 [192] Denmark, Finland, France and Sweden AGTR1 (rs5186) case–control Male Higher risk Not significant Higher risk Gu/Horova, 2012 [204] European IGF2 (rs10770125) Case– control Male Higher risk Higher risk Higher risk Gu/Horova, 2012 [204] European IGF2BP2(rs4402960) Case– control Male Higher risk Higher risk Higher risk Montero, 2013 [201] Brazil/France/ Belgium CYBB (rs6610650) Crosssectional Female Higher risk Higher risk Higher risk Montero, 2013 [201] Brazil/France/ Belgium GPX4 (rs713041) Cross -sectional Male Higher risk Higher risk Higher risk Saldholm, 2013 [203] Finland Chr2q31.1 (rs4972593) GWAS Female NR Higher risk Higher risk Sanholm, 2017 [182] European AFF3, CNTNAP2, NRG3, and PTPN13, ELMO1, 13q, and SIK1 GWAS NR Higher risk Higher risk Higher risk Int. J. Mol. Sci. 2021, 22, 5808 12 of 24 Table 3. Cont. Reference Ethnic Group Locus Study Design Sex Specific Association * MAU Low eGFR DKD Gu, 2019 [200] Sweden SOX2 (rs11915160) Case– control Female Higher risk Higher risk Higher risk Salem, 2019 [185] European COL4A3 (rs55703767) GWAS Male Higher risk Higher risk Higher risk Mori, 2020 [186] Brazil HSD11B1 (rs17389016) cohort NR Not significant Higher risk Higher risk Abbreviations: DKD, diabetic kidney disease; * Sex specific association: sex with the strongest association is reported; eGFR, estimated glomerular filtration rate; MAU, micro/macroalbuminuria; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; NR, sex/gender differences not reported. Not significant, no significant association. 5. Gender Differences in DKD Risk Factors and Renoprotective Drugs Cardiovascular disease (CVD) and DKD share several common risk factors, including hyperglycemia, high blood pressure (BP) values, atherogenic dyslipidemia, obesity, inflammatory markers, high uric acid levels and others. Furthermore, both albuminuria and low eGFR values independently increase CVD morbidity and mortality risk in diabetic individuals [209]. Sex and gender differences have been largely reported in CVD and its major risk factors, both in T2DM and T1DM [70,210–218]. Thus, while the CVD risk is overall higher in DM men than in women, the relative risk of developing CVD complications is 2-4 times higher in DM females than in males [210–212]. Ageing seems to reduce the impact of diabetes on CVD risk, being particularly high in subjects 35-59 years [213,214]. Recent data from the UK estimated the excess CVD risk related to T2DM being approximately 50% higher in women (HR 1.96 (95% CI 1.60, 2.41)) than in men (HR 1.33 (95% CI 1.18, 1.51)) [215], although some more contemporary data attenuated these findings [216]. Notably, an Italian cohort of >11,000 T2DM subjects, followed up for 4 years, reported that, the impact of common risk factors on coronary heart disease (CHD) was different in T2DM men and women, with microvascular complications, including DKD, being a stronger risk factor in females [217]. Thus, sex and gender differences were also reported for several DKD risk factors, with T2DM women generally showing a greater prevalence of out-of-target values [70,218]. In the AMD Annali Initiative, women were 14% more likely than men to have HbA1c >9.0, 42% more likely to have LDL-cholesterol ≥130 mg/dl, and 50% more likely to have BMI ≥30 kg/m2 despite appropriate treatment [1]. Furthermore, older T2DM women were those at the highest risk of uncontrolled lipid values [219]. Beside LDL-C levels, also atherogenic dyslipidemia, i.e., high triglycerides (TG) and low HDL-C values, the typical lipid profile in subjects with insulin-resistance and T2DM, has been recognized as an independent risk factor for DKD [220]. In 15,362 Italian T2DM patients, with eGFR ≥60 mL/min/1.73 m2 , normoalbuminuria, and LDL-C ≤130 mg/dL at baseline, followed-up for 4 years, low HDL-C and high TG levels were independent risk factors for the development of DKD, defined as either low eGFR (<60 mL/min/1.73 m2 ), eGFR reduction >30% and/or albuminuria. After stratification for multiple risk factors, the association of low HDL-C levels with low eGFR was more pronounced for male gender, whereas no other sex interactions were noted for high triglycerides and low eGFR risk and for both lipid fractions and microalbuminuria (MAU) [221]. Notably, another Japanese observational cohort study found an association of atherogenic dyslipidemia with a higher risk of developing DKD in men only [222]. The heterogeneity of HDL particles in their lipid and protein composition, as well as in their function may have contributed to these results. In order to clarify this issue, we evaluated the effect of atherogenic dyslipidemia, the HDL subclasses distribution and the common cholesteryl ester transfer protein (CETP)TaqIB variant on the incidence or the progression of DKD and diabetic retinopathy (DR) in a group of T2DM women followed up for ~9 years. In this study, atheroprotective HDL subclasses together with BMI and LDL/HDL ratio were associated with an increased risk of developing DKD; although these associations were attenuated at multivariate analysis [206]. Int. J. Mol. Sci. 2021, 22, 5808 13 of 24 BMI, LDL/HDL ratio and low levels of α-1 HDL particles were associated to the occurrence of DKD at univariate analysis, although BMI was the only significant predictor at stepwise multivariate regression analysis [206]. Gender differences were also noted in the impact of serum uric acid on DKD risk [223–225]. In a large Chinese cohort [226], hyperuricemia was independently associated with an increased risk of DKD in both genders, but, after adjustment for traditional DKD risk factors, the association remained significant only in men. Data on potential gender differences in the response to drugs recommended for DKD are even more sparse. Thus, recent guidelines recommend the use of inhibitors of sodiumglucose cotransporter 1 (SGLT2i) and glucagonlike protein 1 receptor agonists (GLP1Ras) in T2DM subjects with DKD, on the wave of the encouraging results of many cardiovascular outcome trials (CVOTs) and dedicated investigations in patients with CKD with or without T2DM, but very few of them have specifically evaluated sex or gender differences [227,228]. When renal outcomes were investigated as a safety issue in a retrospective cohort study [229], no differences between T2DM male and females in acute kidney injury (AKI) were reported for SGLT2i (female: 20.9 cases/1000 patients; male: 20 cases/1000 patients) and for GLP1Ras (female: 17.8 cases on 1000 GLP-1RA users; male: 36 cases/1000 patients). Furthermore, no sex interaction was reported in another study exploring the risk of serious renal events among SGLT2i- and GLP1Ras- users [230]. Overall, trials with empagliflozin, canagliflozin, dapagliflozin in patients with and without T2DM, also showed no significant sex gender interactions on renal outcomes, indicating a similarly beneficial effect of SGLT2i in the two genders, although the number of women included in the trial population was different among these studies [231–233]. Collectively these data indicate that the emerging evidence of a gender-difference in the impact of DKD risk factors is yet to be confirmed, but the literature data consistently indicate that DM women, especially those with T2DM, do not reach targets for major CVD/DKD risk factors as easily as men. The reasons behind these gender differences are still only partly explored, and they may include differences in drug prescriptions, adherence and/or drug response, research areas that should be further explored. Whatever the reason, out-of-target risk factors may contribute to the worst outcomes observed in DM women even in terms of DKD-related mortality [234]. 6. Conclusions DKD is one of the most burdensome complications of both T1DM and T2DM. While the prevalence of CVD and other chronic DM complications have shown a progressive decrease in industrialized countries, the prevalence of DKD has not, in spite of the recent diagnostic and therapeutic advances, mostly because of the impact of population ageing and the continuous increase in the prevalence of T1DM and T2DM worldwide. In order to cope with this challenge, the guidelines recommend a personalized approach, where sex and gender differences need to be taken into account. Sex and gender differences have been reported in the prevalence of specific DKD phenotypes, as well as in the prevalence, impact and control of common DKD risk factors. It has become well recognized that both albuminuria and eGFR are important factors for DKD diagnosis, with the non-albuminuric phenotype more prevalent in DM women and the albuminuric one in DM men. As for progression toward ESDR, the data indicate a faster progression and overall worse outcomes in DM women, especially those with T2DM in the late decades of life. Hormonal and genetic factors have also been shown to play a relevant role in explaining these differences in DKD, as demonstrated by several experimental studies, showing an overall protective role for estrogens and progesterone (Figure 1). Conversely, although many genetic loci have been associated with an higher DKD risk both in T1DM and T2DM, their impact seems to vary according to selected gene variants, type of DM, DKD phenotype, study design and according to sex (Table 3). In spite of this increasing evidence, many areas of knowledge still need to be covered. No gender-specific guidance on im- Int. J. Mol. Sci. 2021, 22, 5808 14 of 24 portant diagnostic and therapeutic aspects is available to date. Particularly, it is still not known whether the reported gender differences in DKD phenotypes will result in different outcomes in terms of renal progression and/or CVD mortality. Moreover, the potential protective effect of estrogen replacement therapy on renal outcomes is still under debate. Additionally, the relative impact of socio-economical differences, i.e., the gender disparities, on DKD incidence and progression has not been evaluated so far. Potential sex/gender differences in the therapeutic approach to DKD is another area that needs to be covered. Thus, it is well recognized that women usually experience more drug side effects, and that female gender is poorly represented in randomized controlled trials testing new drugs, an issue that also applies to drugs recommended for patients with DKD, which witness the predominance of the male gender among trials’ participants. Although the available evidence suggests similar renal benefits and/or side effects from the new hypoglycemic drugs with renal benefit in T2DM men and women, the lower representation of women in these trials and the overall lack of a gender-specific analysis prevent us from drawing firm conclusions on these important efficacy and safety issues. In conclusion, sex and gender differences embrace several aspects of DKD pathogenesis, diagnosis, and management, leaving a number of unanswered questions that should be addressed by future research in order to reduce the burden of this serious complication. Funding: This research received no external funding. Conflicts of Interest: The authors declare no conflict of interest. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Rossi, M.C.; Cristofaro, M.R.; Gentile, S.; Lucisano, G.; Manicardi, V.; Mulas, M.F.; Napoli, A.; Nicolucci, A.; Pellegrini, F.; Suraci, C.; et al. Sex disparities in the quality of diabetes care: Biological and cultural factors may play a different role for different outcomes: A cross-sectional observational study from the AMD Annals initiative. Diabetes Care 2013, 36, 162–168. [CrossRef] De Boer, I.H.; Caramori, M.L.; Chan, J.C.N.; Heerspink, H.J.L.; Hurst, C.; Khunti, K.; Liew, A.; Michos, E.D.; Navaneethan, S.D.; Olowu, W.A.; et al. KDIGO 2020 clinical practice guideline for diabetes management in chronic kidney disease. Kidney Int. 2020, 98, S1–S115. [CrossRef] Bolignano, D.; Cernaro, V.; Gembillo, G.; Baggetta, R.; Buemi, M.; D’Arrigo, G. Antioxidant agents for delaying diabetic kidney disease progression: A systematic review and meta-analysis. PLoS ONE 2017, 12, e0178699. [CrossRef] Viazzi, F.; Russo, G.T.; Ceriello, A.; Fioretto, P.; Giorda, C.; De Cosmo, S.; Pontremoli, R. Natural history and risk factors for diabetic kidney disease in patients with T2D: Lessons from the AMD-annals. J. Nephrol. 2019, 32, 517–525. [CrossRef] Pugliese, G.; Penno, G.; Natali, A.; Barutta, F.; Di Paolo, S.; Reboldi, G.; Gesualdo, L.; De Nicola, L. Italian Diabetes Society and the Italian Society of Nephrology. Diabetic kidney disease: New clinical and therapeutic issues. Joint position statement of the Italian Diabetes Society and the Italian Society of Nephrology on “The natural history of diabetic kidney disease and treatment of hyperglycemia in patients with type 2 diabetes and impaired renal function”. J. Nephrol. 2020, 9–35. [CrossRef] Lee, S.K. Sex as an important biological variable in biomedical research. BMB Rep. 2018, 51, 167–173. [CrossRef] Gemmati, D.; Varani, K.; Bramanti, B.; Piva, R.; Bonaccorsi, G.; Trentini, A.; Manfrinato, M.C.; Tisato, V.; Carè, A.; Bellini, T. “Bridging the Gap” Everything that Could Have Been Avoided If We Had Applied Gender Medicine, Pharmacogenetics and Personalized Medicine in the Gender-Omics and Sex-Omics Era. Int. J. Mol. Sci. 2019, 21, 296. [CrossRef] [PubMed] Cañadas-Garre, M.; Anderson, K.; Cappa, R.; Skelly, R.; Smyth, L.J.; McKnight, A.J.; Maxwell, A.P. Genetic Susceptibility to Chronic Kidney Disease–Some More Pieces for the Heritability Puzzle. Front. Genet. 2019, 10, 453. [CrossRef] Siligato, R.; Gembillo, G.; Cernaro, V.; Torre, F.; Salvo, A.; Granese, R.; Santoro, D. Maternal and Fetal Outcomes of Pregnancy in Nephrotic Syndrome Due to Primary Glomerulonephritis. Front. Med. 2020, 7, 563094. [CrossRef] [PubMed] Bang, H.; Vupputuri, S.; Shoham, D.A.; Klemmer, P.J.; Falk, R.J.; Mazumdar, M.; Gipson, D.; Colindres, R.E.; Kshirsagar, A.V. SCreening for Occult REnal Disease (SCORED): A simple prediction model for chronic kidney disease. Arch. Intern. Med. 2007, 167, 374–381. [CrossRef] [PubMed] Süleymanlar, G.; Utaş, C.; Arinsoy, T.; Ateş, K.; Altun, B.; Altiparmak, M.R.; Ecder, T.; Yilmaz, M.E.; Çamsari, T.; Başçi, A.; et al. A population-based survey of Chronic REnal Disease In Turkey–the CREDIT study. Nephrol. Dial. Transplant. 2010, 26, 1862–1871. [CrossRef] [PubMed] Hill, N.R.; Fatoba, S.T.; Oke, J.L.; Hirst, J.A.; O’Callaghan, C.A.; Lasserson, D.S.; Hobbs, F.D. Global Prevalence of Chronic Kidney Disease-A Systematic Review and Meta-Analysis. PLoS ONE 2016, 11, e0158765. [CrossRef] [PubMed] Ricardo, A.C.; Yang, W.; Sha, D.; Appel, L.J.; Chen, J.; Krousel-Wood, M.; Manoharan, A.; Steigerwalt, S.; Wright, J.; Rahman, M.; et al. Sex-Related Disparities in CKD Progression. J. Am. Soc. Nephrol. 2019, 30, 137–146. [CrossRef] [PubMed] Int. J. Mol. Sci. 2021, 22, 5808 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 15 of 24 Yang, W.; Xie, D.; Anderson, A.H.; Joffe, M.M.; Greene, T.; Teal, V.; Hsu, C.Y.; Fink, J.C.; He, J.; Lash, J.P.; et al. Association of kidney disease outcomes with risk factors for CKD: Findings from the Chronic Renal Insufficiency Cohort (CRIC) study. Am. J. Kidney Dis. 2014, 63, 236–243. [CrossRef] Swartling, O.; Rydell, H.; Stendahl, M.; Segelmark, M.; Lagerros, Y.T.; Evans, M. CKD Progression and Mortality Among Men and Women: A Nationwide Study in Sweden. Am. J. Kidney Dis. 2021, 9. [CrossRef] Neugarten, J.; Acharya, A.; Silbiger, S.R. Effect of gender on the progression of nondiabetic renal disease: A meta-analysis. J. Am. Soc. Nephrol. 2000, 11, 319–329. [CrossRef] Iseki, K. Gender differences in chronic kidney disease. Kidney Int. 2008, 74, 415–417. [CrossRef] Jafar, T.H.; Schmid, C.; Stark, P.C.; Toto, R.D.; Remuzzi, G.; Ruggenenti, P.; Marcantoni, C.; Becker, G.J.; Shahinfar, S.; De Jong, P.E.; et al. The rate of progression of renal disease may not be slower in women compared with men: A patient-level meta-analysis. Nephrol. Dial. Transplant. 2003, 18, 2047–2053. [CrossRef] [PubMed] Bash, L.D.; Coresh, J.; Köttgen, A.; Parekh, R.S.; Fulop, T.; Wang, Y.; Astor, B.C. Defining incident chronic kidney disease in the research setting: The ARIC Study. Am. J. Epidemiol. 2009, 170, 414–424. [CrossRef] Gall, M.A.; Hougaard, P.; Borch-Johnsen, K.; Parving, H.H. Risk factors for development of incipient and overt diabetic nephropathy in patients with non-insulin dependent diabetes mellitus: Prospective, observational study. BMJ 1997, 314, 783–788. [CrossRef] Lewis, E.J.; Hunsicker, L.G.; Rodby, R.A. A clinical trial in type 2 diabetic nephropathy. Am. J. Kidney Dis. 2001, 38 (Suppl. S1), S191–S194. [CrossRef] [PubMed] Keane, W.F.; Brenner, B.M.; de Zeeuw, D.; Grunfeld, J.P.; McGill, J.; Mitch, W.E.; Ribeiro, A.B.; Shahinfar, S.; Simpson, R.L.; Snapinn, S.M.; et al. The risk of developing end-stage renal disease in patients with type 2 diabetes and nephropathy: The RENAAL study. Kidney Int. 2003, 63, 1499–1507. [CrossRef] [PubMed] Rossing, K.; Christensen, P.K.; Hovind, P.; Tarnow, L.; Rossing, P.; Parving, H.H. Progression of nephropathy in type 2 diabetic patients. Kidney Int. 2004, 66, 1596–1605. [CrossRef] [PubMed] Retnakaran, R.; Cull, C.A.; Thorne, K.I.; Adler, A.I.; Holman, R.R.; UKPDS Study Group. Risk factors for renal dysfunction in type 2 diabetes: U.K. Prospective Diabetes Study 74. Diabetes 2006, 55, 1832–1839. [CrossRef] Penno, G.; Solini, A.; Bonora, E.; Fondelli, C.; Orsi, E.; Zerbini, G.; Trevisan, R.; Vedovato, M.; Gruden, G.; Laviola, L.; et al. Gender differences in cardiovascular disease risk factors, treatments and complications in patients with type 2 diabetes: The RIACE Italian multicentre study. J. Intern. Med. 2013, 274, 176–191. [CrossRef] Yu, M.K.; Lyles, C.R.; Bent-Shaw, L.A.; Young, B.A. Pathways Authors. Risk factor, age and sex differences in chronic kidney disease prevalence in a diabetic cohort: The pathways study. Am. J. Nephrol. 2012, 36, 245–251. [CrossRef] Jardine, M.J.; Hata, J.; Woodward, M.; Perkovic, V.; Ninomiya, T.; Arima, H.; Zoungas, S.; Cass, A.; Patel, A.; Marre, M.; et al. Prediction of kidney-related outcomes in patients with type 2 diabetes. Am. J. Kidney Dis. 2012, 60, 770–778. [CrossRef] Zoppini, G.; Targher, G.; Chonchol, M.; Ortalda, V.; Negri, G.; Stoico, V.; Bonora, E. Predictors of Estimated GFR Decline in Patients withType 2 Diabetes and Preserved Kidney Function. Clin. J. Am. Soc. Nephrol. 2012, 7, 401–408. [CrossRef] Altemtam, N.; Russell, J.; El Nahas, M. A study of the natural history of diabetic kidney disease (DKD). Nephrol. Dial. Transplant. 2012, 27, 1847–1854. [CrossRef] Elley, C.R.; Robinson, T.; Moyes, S.A.; Kenealy, T.; Collins, J.; Robinson, E.; Orr-Walker, B.; Drury, P.L. Derivation and validation of a renal risk score for people with type 2 diabetes. Diabetes Care 2013, 36, 3113–3120. [CrossRef] De Hauteclocque, A.; Ragot, S.; Slaoui, Y.; Gand, E.; Miot, A.; Sosner, P.; Halimi, J.-M.; Zaoui, P.; Rigalleau, V.; Roussel, R.; et al. The influence of sex on renal function decline in people with type 2 diabetes. Diabet Med. 2014, 31, 1121–1128. [CrossRef] [PubMed] Kajiwara, A.; Kita, A.; Saruwatari, J.; Miyazaki, H.; Kawata, Y.; Morita, K.; Oniki, K.; Yoshida, A.; Jinnouchi, H.; Nakagawa, K. Sex differences in the renal function decline of patients with type 2 diabetes. J. Diabetes Res. 2016, 2016, 4626382. [CrossRef] Orchard, T.J.; Dorman, J.S.; Maser, R.E.; Becker, D.J.; Drash, A.L.; Ellis, D.; LaPorte, R.E.; Kuller, L.H. Prevalence of complications in IDDM by sex and duration. Pittsburgh Epidemiology of Diabetes Complications Study II. Diabetes 1990, 39, 1116–1124. [CrossRef] Lovshin, J.A.; Škrtić, M.; Bjornstad, P.; Moineddin, R.; Daneman, D.; Dunger, P.D.; Reich, H.N.; Mahmud, F.H.; Scholey, J.W.; Cherney, D.Z.I.; et al. Hyperfiltration, urinaryalbumin excretion, and ambulatory blood pressure in adolescents with Type 1 diabetes mellitus. Am. J. Physiol. Renal. Physiol. 2018, 314, F667–F674. [CrossRef] Holl, R.W.; Grabert, M.; Thon, A.; Heinze, E. Urinary excretion of albumin in adolescents with type 1 diabetes: Persistent versus intermittent microalbuminuria and relationship to duration of diabetes, sex, and metabolic control. Diabetes Care 1999, 22, 1555–1560. [CrossRef] Jacobsen, P.; Rossing, K.; Tarnow, L.; Rossing, P.; Mallet, C.; Poirier, O.; Cambien, F.; Parving, H.-H. Progression of diabetic nephropathy in normotensive type 1 diabetic patients. Kidney Int Suppl. 1999, 71, S101–S105. [CrossRef] [PubMed] Rossing, P.; Hougaard, P.; Parving, H.H. Risk factors for development of incipient and overt diabetic nephropathy in type 1 diabetic patients: A 10-year prospective observational study. Diabetes Care 2002, 25, 859–864. [CrossRef] Zhang, L.; Krzentowski, G.; Albert, A.; Lefebvre, P.J. Factors predictive of nephropathy in DCCT Type 1 diabetic patients with good or poor metabolic control. Diabet Med. 2003, 20, 580–585. [CrossRef] [PubMed] Int. J. Mol. Sci. 2021, 22, 5808 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 16 of 24 Hovind, P.; Tarnow, L.; Rossing, P.; Graae, M.; Torp, I.; Binder, C.; Parving, H.-H. Predictors for the development of microalbuminuria and macroalbuminuria in patients with type 1 diabetes: Inception cohort study. BMJ 2004, 328, 1105. [CrossRef] Finne, P.; Reunanen, A.; Stenman, S.; Groop, P.H.; GronhagenRiska, C. Incidence of end-stage renal disease in patients with type 1 diabetes. JAMA 2005, 294, 1782–1787. [CrossRef] [PubMed] Sibley, S.D.; Thomas, W.; de Boer, I.; Brunzell, J.D.; Steffes, M.W. Gender and elevated albumin excretion in the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) cohort: Role of central obesity. Am. J. Kidney Dis. 2006, 47, 223–232. [CrossRef] Raile, K.; Galler, A.; Hofer, S.; Herbst, A.; Dunstheimer, D.; Busch, P.; Holl, R.W. Diabetic nephropathy in 27,805 children, adolescents, and adults with type 1 diabetes: Effect of diabetes duration, A1C, hypertension, dyslipidemia, diabetes onset, and sex. Diabetes Care 2007, 30, 2523–2528. [CrossRef] Monti, M.C.; Lonsdale, J.T.; Montomoli, C.; Montross, R.; Schlag, E.; Greenberg, D.A. Familial risk factors for microvascular complications and differential male-female risk in a large cohort of American families with type 1 diabetes. J. Clin. Endocrinol. Metab. 2007, 92, 4650–4655. [CrossRef] Mollsten, A.; Svensson, M.; Waernbaum, I.; Berhan, Y.; Schon, S.; Nystrom, L.; Arnqvist, H.J.; Dahlquist, G. Cumulative risk, age at onset and sex-specific differences for developing endstage renal disease in young patients with type 1 diabetes. A nationwide population based cohort study. Diabetes 2010, 59, 1803–1808. Costacou, T.; Fried, L.; Ellis, D.; Orchard, T.J. Sex differences in the development of kidney disease in individuals with type 1 diabetes mellitus: A contemporary analysis. Am. J. Kidney Dis. 2011, 58, 565–573. [CrossRef] Harjutsalo, V.; on behalf of the FinnDiane Study Group; Maric, C.; Forsblom, C.; Thorn, L.; Wadén, J.; Groop, P.H. Sex-related differences in the long-term risk of microvascular complications by age at onset of type 1 diabetes. Diabetologia 2011, 54, 1992–1999. [CrossRef] [PubMed] Kautzky-Willer, A.; Stich, K.; Hintersteiner, J.; Kautzky, A.; Kamyar, M.R.; Saukel, J.; Johnson, J.; Lemmens-Gruber, R. Sexspecificdifferences in cardiometabolic risk in type 1 diabetes: A cross-sectional study. Cardiovasc. Diabetol. 2013, 12, 78. [CrossRef] [PubMed] Skupien, J.; Smiles, A.M.; Valo, E.; Ahluwalia, T.S.; Gyorgy, B.; Sandholm, N.; Croall, S.; Lajer, M.; McDonnell, K.; Forsblom, C.; et al. Variations in Risk of EndStage Renal Disease and Risk of Mortality in an International Study of Patients With Type 1 Diabetes and Advanced Nephropathy. Diabetes Care 2019, 42, 93–101. [CrossRef] [PubMed] Dyck, R.F.; Tan, L. Rates and outcomes of diabetic end-stage renal disease among registered native people in Saskatchewan. CMAJ 1994, 150, 203–208. Haroun, M.K.; Jaar, B.G.; Hoffman, S.C.; Comstock, G.W.; Klag, M.J.; Coresh, J. Risk factors for chronic kidney disease: A prospective study of 23,534 men and women in Washington County, Maryland. J. Am. Soc. Nephrol. 2003, 14, 2934–2941. [CrossRef] [PubMed] Xue, J.L.; Eggers, P.W.; Agodoa, L.Y.; Foley, R.N.; Collins, A.J. Longitudinal study of racial and ethnic differences in developing end-stage renal disease among aged medicare beneficiaries. J. Am. Soc. Nephrol. 2007, 18, 1299–1306. [CrossRef] Yamagata, K.; Ishida, K.; Sairenchi, T.; Takahashi, H.; Ohba, S.; Shiigai, T.; Narita, M.; Koyama, A. Risk factors for chronic kidney disease in a community-based population: A 10-year follow-up study. Kidney Int. 2007, 71, 159–166. [CrossRef] Hippisley-Cox, J.; Coupland, C. Predicting the risk of chronic Kidney Disease in men and women in England and Wales: Prospective derivation and external validation of the QKidney Scores. BMC Fam. Pract. 2010, 11, 49. [CrossRef] Hoffmann, F.; Haastert, B.; Koch, M.; Giani, G.; Glaeske, G.; Icks, A. The effect of diabetes on incidence and mortality in end-stage renal disease in Germany. Nephrol. Dial. Transplant. 2011, 26, 1634–1640. [CrossRef] Johnson, E.S.; Smith, D.H.; Thorp, M.L.; Yang, X.; Juhaeri, J. Predicting the risk of end-stage renal disease in the population-based setting: A retrospective case-control study. BMC Nephrol. 2011, 12, 17. [CrossRef] Tohidi, M.; Hasheminia, M.; Mohebi, R.; Khalili, D.; Hosseinpanah, F.; Yazdani, B.; Nasiri, A.A.; Azizi, F.; Hadaegh, F. Incidence of chronic kidney disease and its risk factors, results of over 10 year follow up in an Iranian cohort. PLoS ONE 2012, 7, e45304. [CrossRef] Nagai, K.; Saito, C.; Watanabe, F.; Ohkubo, R.; Sato, C.; Kawamura, T.; Uchida, K.; Hiwatashi, A.; Kai, H.; Ishida, K.; et al. Annual incidence of persistent proteinuria in the general population from Ibaraki annual urinalysis study. Clin. Exp. Nephrol. 2013, 17, 255–260. [CrossRef] [PubMed] van Blijderveen, J.C.; Straus, S.M.; Zietse, R.; Stricker, B.H.; Sturkenboom, M.C.; Verhamme, K.M. A population-based study on the prevalence and incidence of chronic kidney disease in the Netherlands. Int. Urol. Nephrol. 2014, 46, 583–592. [CrossRef] [PubMed] Maric-Bilkan, C. Sex differences in diabetic kidney disease. Mayo Clin. Proc. 2020, 95, 587–599. [CrossRef] [PubMed] Shepard, B.D. Sex differences in diabetes and kidney disease: Mechanisms and consequences. Am. J. Physiol. Renal. Physiol. 2019, 317, F456–F462. [CrossRef] [PubMed] Shen, Y.; Cai, R.; Sun, J.; Dong, X.; Huang, R.; Tian, S.; Wang, S. Diabetes mellitus as a risk factor for incident chronic kidney disease and end-stage renal disease in women compared with men: A systematic review and meta-analysis. Endocrine 2017, 55, 66–76. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 17 of 24 Nag, S.; Bilous, R.; Kelly, W.; Jones, S.; Roper, N.; Connolly, V. All-cause and cardiovascular mortality in diabetic subjects increases significantly with reduced estimated glomerular filtration rate (eGFR): 10 years’ data from the South Tees Diabetes Mortality study. Diabet Med. 2007, 24, 10–17. [CrossRef] [PubMed] Pugliese, G.; Solini, A.; Bonora, E.; Fondelli, C.; Orsi, E.; Nicolucci, A.; Penno, G.; RIACE Study Group. Chronic kidney disease in type 2 diabetes: Lessons from the Renal Insufficiency and Cardiovascular Events (RIACE) Italian Multicentre Study. Nutr. Metab. Cardiovasc. Dis. 2014, 24, 815–822. [CrossRef] [PubMed] Russo, G.T.; De Cosmo, S.; Viazzi, F.; Mirijello, A.; Ceriello, A.; Guida, P.; Giorda, C.; Cucinotta, D.; Pontremoli, R.; Fioretto, P.; et al. Diabetic kidney disease in the elderly: Prevalence and clinical correlates. BMC Geriatr. 2018, 18, 38. [CrossRef] [PubMed] Yu, M.K.; Katon, W.; Young, B.A. Associations between sex and incident chronic kidney disease in a prospective diabetic cohort. Nephrology 2015, 20, 451–458. [CrossRef] [PubMed] A-AMD Annals: A model of continuous monitoring and improvement of the quality of diabetes care. Epidemiol. Prev. 2011, 35, 18–26. Mirijello, A.; Viazzi, F.; Fioretto, P.; Giorda, C.; Ceriello, A.; Russo, G.T.; Guida, P.; Pontremoli, R.; De Cosmo, S. Association of kidney disease measures with risk of renal function worsening in patients with type 1 diabetes. BMC Nephrol. 2018, 19, 347. [CrossRef] [PubMed] Piscitelli, P.; Viazzi, F.; Fioretto, P.; Giorda, C.; Ceriello, A.; Genovese, S.; Russo, G.; Guida, P.; Pontremoli, R.; De Cosmo, S. Predictors of chronic kidney disease in type 1 diabetes: A longitudinal study from the AMD Annals initiative. Sci. Rep. 2017, 7, 3313. [CrossRef] Pacilli, A.; Viazzi, F.; Fioretto, P.; Giorda, C.; Ceriello, A.; Genovese, S.; Russo, G.; Guida, P.; Pontremoli, R.; De Cosmo, S.; et al. Epidemiology of diabetic kidney disease in adult patients with type 1 diabetes in Italy: The AMD-Annals initiative. Diabetes Metab. Res. Rev. 2017, 33. [CrossRef] Manicardi, V.; Russo, G.; Napoli, A.; Torlone, E.; Li Volsi, P.; Giorda, C.B.; Musacchio, N.; Nicolucci, A.; Suraci, C.; Lucisano, G.; et al. Gender- Disparities in Adults with Type 1 Diabetes: More Than a Quality of Care Issue. A Cross-Sectional Observational Study from the AMD Annals Initiative. PLoS ONE 2016, 11, e0162960. [CrossRef] Viazzi, F.; Piscitelli, P.; Giorda, C.; Ceriello, A.; Genovese, S.; Russo, G.T.; Fioretto, P.; Guida, P.; De Cosmo, S.; Pontremoli, R. Association of kidney disease measures with risk of renal function worsening in patients with hypertension and type 2 diabetes. J. Diabetes Complicat. 2017, 31, 419–426. [CrossRef] De Cosmo, S.; Viazzi, F.; Pacilli, A.; Giorda, C.; Ceriello, A.; Gentile, S.; Russo, G.; Rossi, M.C.; Nicolucci, A.; Guida, P.; et al. Predictors of chronic kidney disease in type 2 diabetes: A longitudinal study from the AMD Annals initiative. Medicine 2016, 95, e4007. [CrossRef] De Cosmo, S.; Rossi, M.C.; Pellegrini, F.; Lucisano, G.; Bacci, S.; Gentile, S.; Ceriello, A.; Russo, G.; Nicolucci, A.; Giorda, C.; et al. Kidney dysfunction and related cardiovascular risk factors among patients with type 2 diabetes. Nephrol. Dial. Transplant. 2014, 29, 657–662. [CrossRef] Acconcia, F.; Kumar, R. Signaling regulation of genomic and nongenomic functions of estrogen receptors. Cancer Lett. 2006, 238, 1–14. [CrossRef] [PubMed] Prossnitz, E.R.; Arterburn, J.B. International Union of Basic and Clinical Pharmacology. XCVII. G protein-coupled estrogen receptor and its pharmacologic modulators. Pharmacol. Rev. 2015, 67, 505–540. [CrossRef] Kiyama, R.; Wada-Kiyama, Y. Estrogenic endocrine disruptors: Molecular mechanisms of action. Environ. Int. 2015, 83, 11–40. [CrossRef] Venkov, C.D.; Rankin, A.B.; Vaughan, D.E. Identification of authentic estrogen receptor in cultured endothelial cells—A potential mechanism for steroid hormone regulation of endothelial function. Circulation 1996, 94, 727–733. [CrossRef] Meyer, M.R.; Barton, M. Estrogens and coronary artery disease: New clinical perspectives. Adv. Pharmacol. 2016, 77, 307–360. Burns Katherine, A.; Korach Kenneth, S. Estrogen receptors and human disease: An update. Arch. Toxicol. 2012, 10, 1491–1504. [CrossRef] Ribeiro, J.R.; Freiman, R.N. Estrogen signaling crosstalk: Implications for endocrine resistance in ovarian cancer. J. Steroid. Biochem. Mol. Biol. 2014, 143, 160–173. [CrossRef] Xing, D.; Nozell, S.; Chen, Y.F.; Hage, F.; Oparil, S. Estrogen and mechanisms of vascular protection. Arterioscler. Thromb. Vasc. Biol. 2009, 29, 289–295. [CrossRef] Lewis-Wambi, J.S.; Jordan, V.C. Estrogen regulation of apoptosis: How can one hormone stimulate and inhibit? Breast Cancer Res. 2009, 11, 206. [CrossRef] Roepke, T.A. Oestrogen modulates hypothalamic control of energy homeostasis through multiple mechanisms. J. Neuroendocrinol. 2009, 21, 141–150. [CrossRef] Barros, R.P.; Machado, U.F.; Warner, M.; Gustafsson, J.A. Muscle GLUT4 regulation by estrogen receptors ERbeta and ERalpha. Proc. Natl. Acad. Sci. USA 2006, 103, 1605–1608. [CrossRef] Heine, P.A.; Taylor, J.A.; Iwamoto, G.A.; Lubahn, D.B.; Cooke, P.S. Increased adipose tissue in male and female estrogen receptor-alpha knockout mice. Proc. Natl. Acad. Sci. USA 2000, 97, 12729–12734. [CrossRef] Rodriguez-Cuenca, S.; Monjo, M.; Frontera, M.; Gianotti, M.; Proenza, A.M.; Roca, P. Sex steroid receptor expression profile in brown adipose tissue. Effects of hormonal status. Cell Physiol. Biochem. 2007, 20, 877–886. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 18 of 24 Casazza, K.; Page, G.P.; Fernandez, J.R. The Association Between the rs2234693 and rs9340799 Estrogen Receptor α Gene Polymorphisms and Risk Factors for Cardiovascular Disease: A Review. Biol. Res. For. Nurs. 2010, 12, 84–97. [CrossRef] Gao, H.; Fält, S.; Sandelin, A.; Gustafsson, J.A.; Dahlman-Wright, K. Genome-wide identification of estrogen receptor alphabinding sites in mouse liver. Mol. Endocrinol. 2008, 22, 10–22. [CrossRef] Brandenberger, A.W.; Tee, M.K.; Lee, J.Y.; Chao, V.; Jaffe, R.B. Tissue distribution of estrogen receptors alpha (ER-alpha) and beta (ER-beta) mRNA in the midgestational human fetus. J. Clin. Endocrinol. Metab. 1997, 82, 3509–3512. Evans, M.J.; Lai, K.; Shaw, L.J.; Harnish, D.C.; Chadwick, C.C. Estrogen receptor α inhibits IL-1β induction of gene expression in the mouse liver. Endocrinology 2002, 7, 2559–2570. [CrossRef] Fliegner, D.; Schubert, C.; Penkalla, A.; Witt, H.; Kararigas, G.; Dworatzek, E.; Staub, E.; Martus, P.; Ruiz Noppinger, P.; Kintscher, U.; et al. Female sex and estrogen receptor-beta attenuate cardiac remodeling and apoptosis in pressure overload. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2010, 298, R1597–R1606. [CrossRef] Gürgen, D.; Hegner, B.; Kusch, A.; Catar, R.; Chaykovska, L.; Hoff, U.; Dragun, D. Estrogen receptor-β signals left ventricular hypertrophy sex differences in normotensive deoxycorticosterone acetate-salt mice. Hypertension 2011, 57, 648–654. [CrossRef] Lovegrove, A.S.; Sun, J.; Gould, K.A.; Lubahn, D.B.; Korach, K.S.; Lane, P.H. Estrogen receptor alpha-mediated events promote sex-specific diabetic glomerular hypertrophy. Am. J. Physiol. Renal. Physiol. 2004, 287, F586–F591. [CrossRef] Shimizu, Y. [Estrogen: Estrone (E1), estradiol (E2), estriol (E3) and estetrol (E4)]. Nihon Rinsho. 2005, 63 (Suppl. S8), 425–438. (In Japanese) [PubMed] Dantas, A.P.V.; Fortes, Z.B.; de Carvalho, M.H.C. Vascular disease in diabetic women: Why do they miss the female protection? Exp. Diabetes Res. 2012, 570598. [CrossRef] [PubMed] Maric, C.; Sandberg, K.; Hinojosa-Laborde, C. Glomerulosclerosis and tubulointerstitial fibrosis are attenuated with 17betaestradiol in the aging Dahl salt sensitive rat. J. Am. Soc. Nephrol. 2004, 15, 1546–1556. [CrossRef] [PubMed] Stringer, K.D.; Komers, R.; Osman, S.A.; Oyama, T.T.; Lindsley, J.N.; Anderson, S. Gender hormones and the progression of experimental polycystic kidney disease. Kidney Int. 2005, 68, 1729–1739. [CrossRef] Elliot, S.J.; Berho, M.; Korach, K.; Doublier, S.; Lupia, E.; Striker, G.E.; Karl, M. Gender-specific effects of endogenous testosterone: Female alpha-estrogen receptor-deficient C57Bl/6J mice develop glomerulosclerosis. Kidney Int. 2007, 72, 464–472. [CrossRef] [PubMed] Metcalfe Peter, D.; Leslie, J.A.; Campbell, M.T.; Meldrum, D.R.; Hile, K.L.; Meldrum, K.K. Testosterone exacerbates obstructive renal injury by stimulating TNF-α production and increasing proapoptotic and profibrotic signaling. Am. J. Physiol. Endocrinol. Metab. 2008, 294, E435–E443. [CrossRef] Reckelhoff, J.F.; Zhang, H.; Srivastava, K. Gender differences in development of hypertension in spontaneously hypertensive rats: Role of the renin-angiotensin system. Hypertension 2000, 35 Pt 2, 480–483. [CrossRef] Monster, T.B.; Janssen, W.M.; de Jong, P.E.; de Jong-van den Berg, L.T. Prevention of Renal and Vascular End Stage Disease Study Group. Oral contraceptive use and hormone replacement therapy are associated with microalbuminuria. Arch. Intern. Med. 2001, 161, 2000–2005. [CrossRef] Ahmed, S.B.; Culleton, B.F.; Tonelli, M.; Klarenbach, S.W.; Macrae, J.M.; Zhang, J.; Hemmelgarn, B.R.; Alberta Kidney Disease Network. Oral estrogen therapy in postmenopausal women is associated with loss of kidney function. Kidney Int. 2008, 74, 370–376. [CrossRef] Alexander, S.P.; Mathie, A.; Peters, J.A. Guide to Receptors and Channels (GRAC), 5th edition. Br. J. Pharmacol. 2011, 164 (Suppl. S1), S1-324. [CrossRef] [PubMed] Hazell, G.G.; Yao, S.T.; Roper, J.A.; Prossnitz, E.R.; O’Carroll, A.M.; Lolait, S.J. Localisation of GPR30, a novel G protein-coupled oestrogen receptor, suggests multiple functions in rodent brain and peripheral tissues. J. Endocrinol. 2009, 202, 223–236. [CrossRef] [PubMed] Soltysik, K.; Czekaj, P. Membrane estrogen receptors—Is it an alternative way of estrogen action? J. Physiol. Pharmacol. 2013, 64, 129–142. Cheng, S.B.; Graeber, C.T.; Quinn, J.A.; Filardo, E.J. Retrograde transport of the transmembrane estrogen receptor, G-proteincoupled-receptor-30 (GPR30/GPER) from the plasma membrane towards the nucleus. Steroids 2011, 76, 892–896. [CrossRef] Cheng, S.B.; Dong, J.; Pang, Y.; LaRocca, J.; Hixon, M.; Thomas, P.; Filardo, E.J. Anatomical location and redistribution of G protein-coupled estrogen receptor-1 during the estrus cycle in mouse kidney and specific binding to estrogens but not aldosterone. Mol. Cell Endocrinol. 2014, 382, 950–959. [CrossRef] Lindsey, S.H.; Yamaleyeva, L.M.; Brosnihan, K.B.; Gallagher, P.E.; Chappell, M.C. Estrogen receptor GPR30 reduces oxidative stress and proteinuria in the salt-sensitive female mRen2.Lewis rat. Hypertension 2011, 58, 665–671. [CrossRef] Wang, K.; Zheng, X.; Pan, Z.; Yao, W.; Gao, X.; Wang, X.; Ding, X. Icariin Prevents Extracellular Matrix Accumulation and Ameliorates Experimental Diabetic Kidney Disease by Inhibiting Oxidative Stress via GPER Mediated p62-Dependent Keap1 Degradation and Nrf2 Activation. Front Cell Dev Biol. 2020, 8, 559. [CrossRef] Li, Q.; Sullivan, N.R.; McAllister, C.E.; Van de Kar, L.D.; Muma, N.A. Estradiol accelerates the effects of fluoxetine on serotonin 1A receptor signaling. Psychoneuroendocrinology 2013, 38, 1145–1157. [CrossRef] Xu, H.; Qin, S.; Carrasco, G.A.; Dai, Y.; Filardo, E.J.; Prossnitz, E.R.; Battaglia, G.; Doncarlos, L.L.; Muma, N.A. Extra-nuclear estrogen receptor GPR30 regulates serotonin function in rat hypothalamus. Neuroscience 2009, 158, 1599–1607. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 19 of 24 112. McAllister, C.E.; Creech, R.D.; Kimball, P.A.; Muma, N.A.; Li, Q. GPR30 is necessary for estradiol-induced desensitization of 5-HT1A receptor signaling in the paraventricular nucleus of the rat hypothalamus. Psychoneuroendocrinology 2012, 37, 1248–1260. [CrossRef] 113. Akama, K.T.; Thompson, L.I.; Milner, T.A.; McEwen, B.S. Post-synaptic density95 (PSD-95) binding capacity of G-protein-coupled receptor 30 (GPR30), an estrogen receptor that can be identified in hippocampal dendritic spines. J. Biol. Chem. 2013, 288, 6438–6450. [CrossRef] [PubMed] 114. Li, Y.C.; Ding, X.S.; Li, H.M.; Zhang, C. Icariin attenuates high glucose-induced type IV collagen and fibronectin accumulation in glomerular mesangial cells by inhibiting transforming growth factor-β production and signalling through G protein-coupled oestrogen receptor 1. Clin. Exp. Pharmacol. Physiol. 2013, 40, 635–643. [CrossRef] 115. Rowlands, D.J.; Chapple, S.; Siow, R.C.; Mann, G.E. Equol-stimulated mitochondrial reactive oxygen species activate endothelial nitric oxide synthase and redox signaling in endothelial cells: Roles for F-actin and GPR30. Hypertension 2011, 57, 833–840. [CrossRef] 116. Rao, J.; Jiang, X.; Wang, Y.; Chen, B. 2011. Advances in the understanding of the structure and function of ER-α36, a novel variant of human estrogen receptor-alpha. J. Steroid Biochem. Mol. Biol. 2011, 127, 231–237. [CrossRef] 117. Lin, S.L.; Yan, L.Y.; Liang, X.W.; Wang, Z.B.; Wang, Z.Y.; Qiao, J.; Schatten, H.; Sun, Q.Y. A novel variant of ER-alpha, ER-alpha36 mediates testosterone-stimulated ERK and Akt activation in endometrial cancer Hec1A cells. Reprod. Biol. Endocrinol. 2009, 7, 102. [CrossRef] 118. Wells, C.C.; Riazi, S.; Mankhey, R.W.; Bhatti, F.; Ecelbarger, C.; Maric, C. Diabetic nephropathy is associated with decreased circulating estradiol levels and imbalance in the expression of renal estrogen receptors. Gend. Med. 2005, 2, 227–237. [CrossRef] 119. Inada, A.; Inada, O.; Fujii, N.L.; Nagafuchi, S.; Katsuta, H.; Yasunami, Y.; Matsubara, T.; Arai, H.; Fukatsu, A.; Nabeshima, Y.I. Adjusting the 17β-Estradiol-to-Androgen Ratio Ameliorates Diabetic Nephropathy. J. Am. Soc. Nephrol. 2016, 27, 3035–3050. [CrossRef] [PubMed] 120. Manigrasso, M.B.; Sawyer, R.T.; Marbury, D.C.; Flynn, E.R.; Maric, C. Inhibition of estradiol synthesis attenuates renal injury in male streptozotocin-induced diabetic rats. Am. J. Physiol. Renal Physiol. 2011, 301, F634–F640. [CrossRef] 121. Maric, C.; Forsblom, C.; Thorn, L.; Waden, J.; Groop, P.H.; FinnDiane Study Group. Association between testosterone, estradiol and sex hormone binding globulin levels in men with type 1 diabetes with nephropathy. Steroids 2010, 75, 772–778. [CrossRef] 122. Matsushita, M.; Tamura, K.; Osada, S.; Kogo, H. Effect of troglitazone on the excess testosterone and LH secretion in thyroidectomized, insulin-resistant, type 2 diabetic Goto-Kakizaki rats. Endocrine 2005, 27, 301–305. [CrossRef] 123. Salonia, A.; Lanzi, R.; Scavini, M.; Pontillo, M.; Gatti, E.; Petrella, G.; Licata, G.; Nappi, R.E.; Bosi, E.; Briganti, A.; et al. Sexual function and endocrine profile in fertile women with type 1 diabetes. Diabetes Care 2006, 29, 312–316. [CrossRef] [PubMed] 124. Andersson, B.; Mattsson, L.A.; Hahn, L.; Mårin, P.; Lapidus, L.; Holm, G.; Bengtsson, B.A.; Björntorp, P. Estrogen replacement therapy decreases hyperandrogenicity and improves glucose homeostasis and plasma lipids in postmenopausal women with noninsulin-dependent diabetes mellitus. J. Clin. Endocrinol. Metab. 1997, 82, 638–643. [CrossRef] 125. Brussaard, H.E.; Gevers Leuven, J.A.; Frölich, M.; Kluft, C.; Krans, H.M. Short-term oestrogen replacement therapy improves insulin resistance, lipids and fibrinolysis in postmenopausal women with NIDDM. Diabetologia 1997, 40, 843–849. [CrossRef] [PubMed] 126. Szekacs, B.; Vajo, Z.; Varbiro, S.; Kakucs, R.; Vaslaki, L.; Acs, N.; Mucsi, I.; Brinton, E.A. Postmenopausal hormone replacement improves proteinuria and impaired creatinine clearance in type 2 diabetes mellitus and hypertension. Br. J. Obstet. Gynaecol. 2000, 107, 1017–1021. [CrossRef] 127. Maric, C.; Sullivan, S. Estrogens and the diabetic kidney. Gender Med. 2008, 5, S103–S113. [CrossRef] [PubMed] 128. Mankhey, R.W.; Bhatti, F.; Maric, C. 17β-Estradiol replacement improves renal function and pathology associated with diabetic nephropathy. Am. J. Physiol. Renal Physiol. 2005, 288, F399–F405. [CrossRef] 129. Chin, M.; Isono, M.; Isshiki, K.; Araki, S.; Sugimoto, T.; Guo, B.; Sato, H.; Haneda, M.; Kashiwagi, A.; Koya, D. Estrogen and raloxifene, a selective estrogen receptor modulator, ameliorate renal damage in db/db mice. Am. J. Pathol. 2005, 166, 1629–1636. [CrossRef] 130. Hadjadj, S.; Gourdy, P.; Zaoui, P.; Guerci, B.; Roudaut, N.; Gautier, J.F.; Chabin, M.; Mauco, G.; Ragot, S.; RADIAN Study Group. Effect of raloxifene—A selective oestrogen receptor modulator—On kidney function in post-menopausal women with Type 2 diabetes: Results from a randomized, placebo-controlled pilot trial. Diabet Med. 2007, 24, 906–910. [CrossRef] [PubMed] 131. Ramesh, S.; Mann, M.C.; Holroyd-Leduc, J.M.; Wilton, S.B.; James, M.T.; Seely, E.W.; Ahmed, S.B. Hormone therapy and clinical and surrogate cardiovascular endpoints in women with chronic kidney disease: A systematic review and meta-analysis. Menopause 2016, 23, 1028–1037. [CrossRef] 132. Hu, X.; Liu, W.; Yan, Y.; Liu, H.; Huang, Q.; Xiao, Y.; Gong, Z.; Du, J. Vitamin D protects against diabetic nephropathy: Evidence-based effectiveness and mechanism. Eur. J. Pharmacol. 2019, 845, 91–98. [CrossRef] 133. Cutolo, M.; Plebani, M.; Shoenfeld, Y.; Adorini, L.; Tincani, A. Vitamin D endocrine system and the immune response in rheumatic diseases. Vitam. Horm. 2011, 86, 327–351. [CrossRef] 134. Gembillo, G.; Cernaro, V.; Siligato, R.; Curreri, F.; Catalano, A.; Santoro, D. Protective Role of Vitamin D in Renal Tubulopathies. Metabolites 2020, 10, 115. [CrossRef] 135. Gembillo, G.; Siligato, R.; Amatruda, M.; Conti, G.; Santoro, D. Vitamin D and Glomerulonephritis. Medicina 2021, 57, 186. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 20 of 24 136. Gembillo, G.; Cernaro, V.; Salvo, A.; Siligato, R.; Laudani, A.; Buemi, M.; Santoro, D. Role of Vitamin D Status in Diabetic Patients with Renal Disease. Medicina 2019, 55, 273. [CrossRef] 137. Fondjo, L.A.; Sakyi, S.A.; Owiredu, W.K.B.A.; Laing, E.F.; Owiredu, E.W.; Awusi, E.K.; Ephraim, R.K.D.; Kantanka, O.S. Evaluating Vitamin D Status in Pre- and Postmenopausal Type 2 Diabetics and Its Association with Glucose Homeostasis. Biomed. Res. Int. 2018, 2018, 9369282. [CrossRef] 138. Agmon-Levin, N.; Theodor, E.; Segal, R.M. Vitamin D in systemic and organ-specific autoimmune diseases. Clin. Rev. Allergy Immunol. 2013, 45, 256–266. [CrossRef] 139. Cutolo, M.; Paolino, S.; Sulli, A.; Smith, V.; Pizzorni, C.; Seriolo, B. Vitamin D, steroid hormones, and autoimmunity. Ann. N. Y. Acad. Sci. 2014, 1317, 39–46. [CrossRef] 140. Santoro, D.; Gembillo, G.; Andò, G. Glomerular Filtration Rate as a Predictor of Outcome in Acute Coronary Syndrome Complicated by Atrial Fibrillation. J. Clin. Med. 2020, 9, 1466. [CrossRef] 141. Gangula, P.R.; Dong, Y.L.; Al-Hendy, A.; Richard-Davis, G.; Montgomery-Rice, V.; Haddad, G.; Millis, R.; Nicholas, S.B.; Moseberry, D. Protective cardiovascular and renal actions of vitamin D and estrogen. Front. Biosci. 2013, 5, 134–148. [CrossRef] [PubMed] 142. Meyer, M.R.; Clegg, D.J.; Prossnitz, E.R.; Barton, M. Obesity, insulin resistance and diabetes: Sex differences and role of oestrogen receptors. Acta Physiol. 2011, 203, 259–269. [CrossRef] [PubMed] 143. Dunlop, M. Aldose reductase and the role of the polyol pathway in diabetic nephropathy. Kidney Int. Suppl. 2000, 77, S3–S12. [CrossRef] 144. Conti, G.; Caccamo, D.; Siligato, R.; Gembillo, G.; Satta, E.; Pazzano, D.; Carucci, N.; Carella, A.; Campo, G.D.; Salvo, A.; et al. Association of Higher Advanced Oxidation Protein Products (AOPPs) Levels in Patients with Diabetic and Hypertensive Nephropathy. Medicina 2019, 55, 675. [CrossRef] 145. Semba, R.D.; Nicklett, E.; Ferrucci, L. Does accumulation of advanced glycation end products contribute to the aging phenotype? J. Gerontol. Ser. A Biomed. Sci. Med Sci. 2010, 65, 963–975. [CrossRef] 146. Cai, W.; Ramdas, M.; Zhu, L.; Chen, X.; Striker, G.E.; Vlassara, H. Oral advanced glycation endproducts (AGEs) promote insulin resistance and diabetes by depleting the antioxidant defenses AGE receptor-1 and sirtuin 1. Proc. Natl. Acad. Sci. USA 2012, 109, 15888–15893. [CrossRef] 147. Uribarri, J.; Cai, W.; Ramdas, M.; Goodman, S.; Pyzik, R.; Chen, X.; Zhu, L.; Striker, G.E.; Vlassara, H. Restriction of advanced glycation end products improves insulin resistance in human type 2 diabetes: Potential role of AGER1 and SIRT1. Diabetes Care 2011, 34, 1610–1616. [CrossRef] 148. Elliot, S.J.; Karl, M.; Berho, M.; Xia, X.; Pereria-Simon, S.; Espinosa-Heidmann, D.; Striker, G.E. Smoking induces glomerulosclerosis in aging estrogen-deficient mice through cross-talk between TGF-beta1 and IGF-I signaling pathways. J. Am. Soc. Nephrol. 2006, 17, 3315–3324. [CrossRef] 149. Blush, J.; Lei, J.; Ju, W.; Silbiger, S.; Pullman, J.; Neugarten, J. Estradiol reverses renal injury in Alb/TGF-beta1 transgenic mice. Kidney Int. 2004, 66, 2148–2154. [CrossRef] 150. Gross, M.-L.; Adamczak, M.; Rabe, T.; Harbi, N.A.; Krtil, J.; Koch, A.; Ritz, E. Beneficial effects of estrogens on indices of renal damage in uninephrectomized SHRsp rats. J. Am. Soc. Nephrol. 2004, 15, 348–358. [CrossRef] 151. Dean, S.A.; Tan, J.; O’Brien, E.R.; Leenen, F.H. 17beta-estradiol downregulates tissue angiotensin-converting enzyme and ANG II type 1 receptor in female rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2005, 288, R759–R766. [CrossRef] 152. Negulescu, O.; Bognar, I.; Lei, J.; Devarajan, P.; Silbiger, S.; Neugarten, J. Estradiol reverses TGF-β1–induced mesangial cell apoptosis by a casein kinase 2-dependent mechanism. Kidney Int. 2002, 62, 1989–1998. [CrossRef] 153. Xiao, S.; Gillespie, D.G.; Baylis, C.; Jackson, E.K.; Dubey, R.K. Effects of estradiol and its metabolites on glomerular endothelial nitric oxide synthesis and mesangial cell growth. Hypertension 2001, 37, 645–650. [CrossRef] [PubMed] 154. Pereira-Simon, S.; Rubio, G.A.; Xia, X.; Cai, W.; Choi, R.; Striker, G.E.; Elliot, S.J. Inhibition of Advanced Glycation End Products (AGEs) Accumulation by Pyridoxamine Modulates Glomerular and Mesangial Cell Estrogen Receptor α Expression in Aged Female Mice. PLoS ONE 2016, 11, e0159666. [CrossRef] 155. Nakamura, T.; Miller, D.; Ruoslahti, E.; Border, W.A. Production of extracellular matrix by glomerular epithelial cells is regulated by transforming growth factor-beta 1. Kidney Int. 1992, 41, 1213–1221. [CrossRef] 156. Miner, J.H. Renal basement membrane components. Kidney Int. 1999, 56, 2016–2024. [CrossRef] 157. Patek, C.E.; Fleming, S.; Miles, C.G.; Bellamy, C.O.; Ladomery, M.; Spraggon, L.; Mullins, J.; Hastie, N.D.; Hooper, M.L. Murine Denys-Drash syndrome: Evidence of podocyte de-differentiation and systemic mediation of glomerulosclerosis. Hum. Mol. Genet. 2003, 12, 2379–2394. [CrossRef] 158. Lee, H.S.; Song, C.Y. Differential role of mesangial cells and podocytes in TGF-ß-induced mesangial matrix synthesis in chronic glomerular disease. Histol. Histopathol. 2009, 24, 901–908. [CrossRef] 159. Lee, H.S. Pathogenic role of TGF-ß in the progression of podocyte diseases. Histol. Histopathol. 2011, 26, 107–116. [CrossRef] 160. Lin, Y.; Nakachi, K.; Ito, Y.; Kikuchi, S.; Tamakoshi, A.; Yagyu, K.; Watanabe, Y.; Inaba, Y.; Tajima, K.; Jacc Study Group. Variations in serum transforming growth factor-beta1 levels with gender, age and lifestyle factors of healthy Japanese adults. Dis. Markers 2009, 27, 23–28. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 21 of 24 161. Ito, I.; Hanyu, A.; Wayama, M.; Goto, N.; Katsuno, Y.; Kawasaki, S.; Nakajima, Y.; Kajiro, M.; Komatsu, Y.; Fujimura, A.; et al. Estrogen inhibits transforming growth factor beta signaling by promoting Smad2/3 degradation. J. Biol. Chem. 2010, 285, 14747–14755. [CrossRef] 162. Zdunek, M.; Silbiger, S.; Lei, J.; Neugarten, J. Protein kinase CK2 mediates TGF-β; 1-stimulated type IV collagen gene transcription and its reversal by estradiol. Kidney Int. 2001, 60, 2097–2108. [CrossRef] 163. Ruggenenti, P.; Cravedi, P.; Remuzzi, G. The RAAS in the pathogenesis and treatment of diabetic nephropathy. Nat. Rev. Nephrol. 2010, 6, 319. [CrossRef] 164. Yanes, L.L.; Sartori-Valinotti, J.C.; Reckelhoff, J.F. Sex steroids and renal disease: Lessons from animal studies. Hypertension 2008, 51, 976–981. [CrossRef] 165. Bumke-Vogt, C.; Bähr, V.; Diederich, S.; Herrmann, S.M.; Anagnostopoulos, I.; Oelkers, W.; Quinkler, M. Expression of the progesterone receptor and progesterone- metabolising enzymes in the female and male human kidney. J. Endocrinol. 2002, 175, 349–364. [CrossRef] [PubMed] 166. Nielsen, C.B.; Flyvbjerg, A.; Bruun, J.M.; Forman, A.; Wogensen, L.; Thomsen, K. Decreases in renal functional reserve and proximal tubular fluid output in conscious oophorectomized rats: Normalization with sex hormone substitution. J. Am. Soc. Nephrol. 2003, 14, 3102–3110. [CrossRef] 167. Sandhi, J.; Singh, J.P.; Kaur, T.; Ghuman, S.S.; Singh, A.P. Involvement of progesterone receptors in ascorbic acid-mediated protection against ischemia-reperfusion-induced acute kidney injury. J. Surg. Res. 2014, 187, 278–288. [CrossRef] [PubMed] 168. Al-Trad, B.; Ashankyty, I.M.; Alaraj, M. Progesterone ameliorates diabetic nephropathy in streptozotocin-induced diabetic Rats. Diabetol. Metab. Syndr. 2015, 7, 1–13. [CrossRef] [PubMed] 169. Kashtan, C. (2017). Alport syndrome: Facts and opinions. F1000Research 2017, 6, 50. [CrossRef] 170. Temme, J.; Kramer, A.; Jager, K.J.; Lange, K.; Peters, F.; Müller, G.-A.; Kramar, R.; Heaf, J.G.; Finne, P.; Palsson, R.; et al. Outcomes of male patients with alport syndrome undergoing renal replacement therapy. Clin. J. Am. Soc. Nephrol. 2012, 7, 1969–1976. [CrossRef] [PubMed] 171. Savige, J.; Colville, D.; Rheault, M.; Gear, S.; Lennon, R.; Lagas, S. Special feature alport syndrome in women and girls. Clin. J. Am. Soc. Nephrol. 2016, 11, 1713–1720. [CrossRef] [PubMed] 172. O’Seaghdha, C.M.; Fox, C.S. Genome-wide association studies of chronic kidney disease: What have we learned? Nat. Rev. Nephrol. 2011, 8, 89–99. [CrossRef] 173. Regele, F.; Jelencsics, K.; Shiffman, D.; Paré, G.; McQueen, M.J.; Mann, J.F. Genome-wide studies to identify risk factors for kidney disease with a focus on patients with diabetes. Nephrol. Dial. Transplant. 2015, 30 (Suppl. S4), iv26–iv34. [CrossRef] 174. Seaquist, E.R.; Goetz, F.C.; Rich, S.; Barbosa, J. Familial clustering of diabetic kidney disease. Evidence for genetic susceptibility to diabetic nephropathy. N. Engl. J. Med. 1989, 320, 1161–1165. [CrossRef] 175. Quinn, M.; Angelico, M.C.; Warram, J.H.; Krolewski, A.S. Familial factors determine the development of diabetic nephropathy in patients with IDDM. Diabetologia 1996, 39, 940–945. [CrossRef] 176. Vijay, V.; Snehalatha, C.; Shina, K.; Lalitha, S.; Ramachandran, A. Familial aggregation of diabetic kidney disease in Type 2 diabetes in south India. Diabetes Res. Clin. Pract. 1999, 43, 167–171. [CrossRef] 177. Pettitt, D.J.; Saad, M.F.; Bennett, P.H.; Nelson, R.G.; Knowler, W.C. Familial predisposition to renal disease in two generations of Pima Indians with type 2 (noninsulin- dependent) diabetes mellitus. Diabetologia 1990, 33, 438–443. [CrossRef] [PubMed] 178. Freedman, B.I.; Spray, B.J.; Tuttle, A.B.; Buckalew, V.M., Jr. The familial risk of endstage renal disease in African Americans. Am. J. Kidney Dis. 1993, 21, 387–393. [CrossRef] 179. Gu, H.F. Genetic and Epigenetic Studies in Diabetic Kidney Disease. Front. Genet. 2019, 10, 507. [CrossRef] 180. Prudente, S.; Di Paola, R.; Copetti, M.; Lucchesi, D.; Lamacchia, O.; Pezzilli, S. The rs12917707 polymorphism at the UMOD locus and glomerular filtration rate in individuals with type 2 diabetes: Evidence of heterogeneity across two different European populations. Nephrol. Dial. Transplant. 2017, 32, 1718–1722. [CrossRef] [PubMed] 181. Mooyaart, A.L.; Valk, E.J.J.; Van Es, L.A.; Bruijn, J.A.; De Heer, E.; Freedman, B.I.; Dekkers, O.M.; Baelde, H.J. Genetic associations in diabetic nephropathy: A meta-analysis. Diabetologia 2011, 54, 544–553. [CrossRef] [PubMed] 182. Sandholm, N.; Van Zuydam, N.; Ahlqvist, E.; Juliusdottir, T.; Deshmukh, H.A.; Rayner, N.W.; Di Camillo, B.; Forsblom, C.; Fadista, J.; Ziemek, D.; et al. The genetic landscape of renal complications in type 1 diabetes. J. Am. Soc. Nephrol. 2017, 28, 557–574. [CrossRef] 183. Teumer, A.; Tin, A.; Sorice, R.; Gorski, M.; Yeo, N.C.; Chu, A.Y.; Li, M.; Li, Y.; Mijatovic, V.; Ko, Y.A.; et al. Genome-wide Association Studies Identify Genetic Loci Associated With Albuminuria in Diabetes. Diabetes 2016, 65, 803–817. [CrossRef] 184. van Zuydam, N.R.; Ahlqvist, E.; Sandholm, N.; Deshmukh, H.; Rayner, N.W.; Abdalla, M.; Ladenvall, C.; Ziemek, D.; Fauman, E.; Robertson, N.R.; et al. A Genome-Wide Association Study of Diabetic Kidney Disease in Subjects With Type 2 Diabetes. Diabetes 2018, 67, 1414–1427. [CrossRef] [PubMed] 185. Salem, R.M.; Todd, J.N.; Sandholm, N.; Cole, J.B.; Chen, W.-M.; Andrews, D.; Pezzolesi, M.G.; McKeigue, P.M.; Hiraki, L.T.; Qiu, C.; et al. Genome-Wide Association Study of Diabetic Kidney Disease Highlights Biology Involved in Glomerular Basement Membrane Collagen. J. Am. Soc. Nephrol. 2019, 30, 2000–2016. [CrossRef] [PubMed] 186. Mori, R.C.; Santos-Bezerra, D.P.; Pelaes, T.S.; Admoni, S.N.; Perez, R.V.; Monteiro, M.B.; Machado, C.G.; Queiroz, M.S.; Machado, U.F.; Correa-Giannella, M.L. Variants in HSD11B1 gene modulate susceptibility to diabetes kidney disease and to insulin resistance in type 1 diabetes. Diabetes Metab. Res. Rev. 2021, 37, e3352. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 22 of 24 187. Kato, M.; Natarajan, R. Diabetic nephropathy–emerging epigenetic mechanisms. Nat. Rev. Nephrol. 2014, 10, 517–530. [CrossRef] 188. Allis, C.D.; Jenuwein, T. The molecular hallmarks of epigenetic control. Nat. Rev. Genet. 2016, 17, 487–500. [CrossRef] 189. Vujkovic, M.; Keaton, J.M.; Lynch, J.A.; Miller, D.R.; Zhou, J.; Tcheandjieu, C.; Huffman, J.E.; Assimes, T.L.; Lorenz, K.; Zhu, X.; et al. Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1.4 million participants in a multi-ancestry meta-analysis. Nat. Genet. 2020, 52, 680–691. [CrossRef] [PubMed] 190. Freire, M.B.; Ji, L.; Onuma, T.; Orban, T.; Warram, J.H.; Krolewski, A.S. Gender-specific association of M235T polymorphism in angiotensinogen gene and diabetic nephropathy in NIDDM. Hypertension 1998, 31, 896–899. [CrossRef] 191. Lin, J.; Hu, F.B.; Qi, L.; Curhan, G.C. Genetic polymorphisms of angiotensin-2 type 1 receptor and angiotensinogen and risk of renal dysfunction and coronary heart disease in type 2 diabetes mellitus. BMC Nephrol. 2009, 10, 9. [CrossRef] 192. Möllsten, A.; Vionnet, N.; Forsblom, C.; Parkkonen, M.; Tarnow, L.; Hadjadj, S.; Marre, M.; Parving, H.H.; Groop, P.H. A polymorphism in the angiotensin II type 1 receptor gene has different effects on the risk of diabetic nephropathy in men and women. Mol. Genet. Metab. 2011, 103, 66–70. [CrossRef] [PubMed] 193. Ahluwalia, T.S.; Lindholm, E.; Groop, L.C. Common variants in CNDP1 and CNDP2, and risk of nephropathy in type 2 diabetes. Diabetologia 2011, 54, 2295–2302. [CrossRef] 194. Alkhalaf, A.; Landman, G.W.; van Hateren, K.J.; Groenier, K.H.; Mooyaart, A.L.; De Heer, E.; Gans, R.O.; Navis, G.J.; Bakker, S.J.; Kleefstra, N.; et al. Sex specific association between carnosinase gene CNDP1 and cardiovascular mortality in patients with type 2 diabetes (ZODIAC-22). J Nephrol. 2015, 28, 201–207. [CrossRef] [PubMed] 195. Kurashige, M.; Imamura, M.; Araki, S.; Suzuki, D.; Babazono, T.; Uzu, T.; Umezono, T.; Toyoda, M.; Kawai, K.; Imanishi, M.; et al. The influence of a single nucleotide polymorphism within CNDP1 on susceptibility to diabetic nephropathy in Japanese women with type 2 diabetes. PLoS ONE 2013, 8, e54064. [CrossRef] [PubMed] 196. Crook, E.D.; Genous, L.; Oliver, B. Angiotensin-converting enzyme genotype in blacks with diabetic nephropathy: Effects on risk of diabetes and its complications. J. Investig Med. 2003, 51, 360–365. [CrossRef] 197. Sakka, Y.; Babazono, T.; Sato, A.; Ujihara, N.; Iwamoto, Y. ACE gene polymorphism, left ventricular geometry, and mortality in diabetic patients with end-stage renal disease. Diabetes Res. Clin. Pract. 2004, 64, 41–49. [CrossRef] 198. Tien, K.J.; Hsiao, J.Y.; Hsu, S.C.; Liang, H.T.; Lin, S.R.; Chen, H.C.; Hsieh, M.C. Gender-dependent effect of ACE I/D and AGT M235T polymorphisms on the progression of urinary albumin excretion in Taiwanese with type 2 diabetes. Am. J. Nephrol. 2009, 29, 299–308. [CrossRef] 199. Zambrano-Galván, G.; Reyes-Romero, M.A.; Lazalde, B.; Rodríguez-Morán, M.; Guerrero-Romero, F. Risk of microalbuminuria in relatives of subjects with diabetic nephropathy: A predictive model based on multivariable dimensionality reduction approach. Clin. Nephrol. 2015, 83, 86–92. [CrossRef] 200. Gu, H.F.; Alvarsson, A.; Efendic, S.; Brismar, K. SOX2 has gender-specific genetic effects on diabetic nephropathy in samples from patients with type 1 diabetes mellitus in the GoKinD study. Gend Med. 2009, 6, 555–564. [CrossRef] 201. Monteiro, M.B.; Patente, T.A.; Mohammedi, K.; Queiroz, M.S.; Azevedo, M.J.; Canani, L.H.; Parisi, M.C.; Marre, M.; Velho, G.; Corrêa-Giannella, M.L. Sex-specific associations of variants in regulatory regions of NADPH oxidase-2 (CYBB) and glutathione peroxidase 4 (GPX4) genes with kidney disease in type 1 diabetes. Free Radic. Res. 2013, 47, 804–810. [CrossRef] 202. Mlynarski, W.M.; Placha, G.P.; Wolkow, P.P.; Bochenski, J.P.; Warram, J.H.; Krolewski, A.S. Risk of diabetic nephropathy in type 1 diabetes is associated with functional polymorphisms in RANTES receptor gene (CCR5): A sex-specific effect. Diabetes. 2005, 54, 3331–3335. [CrossRef] [PubMed] 203. Sandholm, N.; McKnight, A.J.; Salem, R.M.; Brennan, E.P.; Forsblom, C.; Harjutsalo, V.; Mäkinen, V.P.; McKay, G.J.; Sadlier, D.M.; Williams, W.W.; et al. Chromosome 2q31.1 associates with ESRD in women with type 1 diabetes. J. Am. Soc. Nephrol. 2013, 24, 1537–1543. [CrossRef] 204. Gu, T.; Horová, E.; Möllsten, A.; Seman, N.A.; Falhammar, H.; Prázný, M.; Brismar, K.; Gu, H.F. IGF2BP2 and IGF2 genetic effects in diabetes and diabetic nephropathy. J. Diabetes Complicat. 2012, 26, 393–398. [CrossRef] [PubMed] 205. Mantovani, A.; Zusi, C.; Sani, E.; Colecchia, A.; Lippi, G.; Zaza, G.L.; Valenti, L.; Byrne, C.D.; Maffeis, C.; Bonora, E.; et al. Association between PNPLA3rs738409 polymorphism decreased kidney function in postmenopausal type 2 diabetic women with or without non-alcoholic fatty liver disease. Diabetes Metab. 2019, 45, 480–487. [CrossRef] [PubMed] 206. Russo, G.T.; Giandalia, A.; Romeo, E.L.; Muscianisi, M.; Ruffo, M.C.; Alibrandi, A.; Bitto, A.; Forte, F.; Grillone, A.; Asztalos, B.; et al. HDL subclasses and the common CETP TaqIB variant predict the incidence of microangiopatic complications in type 2 diabetic women: A 9years follow-up study. Diabetes Res. Clin. Pract. 2017, 132, 108–117. [CrossRef] [PubMed] 207. Huang, Y.-C.; Chen, S.-Y.; Liu, S.-P.; Lin, J.-M.; Lin, H.-J.; Lei, Y.-J.; Chung, Y.-C.; Chen, Y.-C.; Wang, Y.-H.; Liao, W.-L.; et al. Cholesteryl Ester Transfer Protein Genetic Variants Associated with Risk for Type 2 Diabetes and Diabetic Kidney Disease in Taiwanese Population Genes. Genes (Basel) 2019, 10, 782. [CrossRef] 208. Bartáková, V.; Kuricová, K.; Pácal, L.; Nová, Z.; Dvořáková, V.; Švrčková, M.; Malúšková, D.; Svobodová, I.; Řehořová, J.; Svojanovský, J.; et al. Hyperuricemia contributes to the faster progression of diabetic kidney disease in type 2 diabetes mellitus. J. Diabetes Complicat. 2016, 30, 1300–1307. [CrossRef] 209. Fox, C.S.; Matsushita, K.; Woodward, M.; Bilo, H.J.; Chalmers, J.; Heerspink, H.J.; Lee, B.J.; Perkins, R.M.; Rossing, P.; Sairenchi, T.; et al. Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without diabetes: A meta-analysis. Lancet 2012, 380, 1662–1673. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 23 of 24 210. Huxley, R.; Barzi, F. Woodward. Excess risk of fatal coronary heart disease associated with diabetes in men and women: Metaanalysis of 37 prospective cohort studies. BMJ 2006, Russo GT, Baggio G, Rossi MC, Kautzky-Willer, A. Type 2diabetes and cardiovascular risk in women. Int. J. Endocrinol. 2015, 2015, 832484. [CrossRef] 211. Huebschmann, A.G.; Huxley, R.R.; Kohrt, W.M.; Zeitler, P.; Regensteiner, J.G.; Reusch, J.E.B. Sex differences in the burden of type 2 diabetes and cardiovascular risk across the life course. Diabetologia 2019, 62, 1761–1772. [CrossRef] 212. Ballotari, P.; Ranieri, S.C.; Luberto, F.; Caroli, S.; Greci, M.; Rossi, P.G.; Manicardi, V. Sex differences in cardiovascular mortality in diabetics and nondiabetic subjects: A population-based study (Italy). Int. J. Endocrinol. 2015, 2015, 914057. [CrossRef] [PubMed] 213. Rospective Studies Collaboration and Asia Pacific Cohort Studies Collaboration. Sex-specific relevance of diabetes to occlusive vascular and other mortality: A collaborative meta-analysis of individual data from 980 793 adults from 68 prospective studies. Lancet Diabetes Endocrinol. 2018, 6, 538–546. [CrossRef] 214. Kalyani, R.R.; Lazo, M.; Ouyang, P.; Turkbey, E.; Chevalier, K.; Brancati, F.; Becker, D.; Vaidya, D. Sex differences in diabetes and risk of incident coronary artery disease in healthy young and middle-aged adults. Diabetes Care 2014, 37, 830–838. [CrossRef] 215. Millett, E.R.C.; Peters, S.A.E.; Woodward, M. Sex differences in risk factors for myocardial infarction: Cohort study of UK Biobank participants. BMJ 2018, 363, k4247. [CrossRef] [PubMed] 216. Wright, A.K.; Kontopantelis, E.; Emsley, R.; Buchan, I.; Mamas, M.A.; Sattar, N.; Ashcroft, D.; Rutter, M.K. Cardiovascular risk and risk factormanagement in type 2 diabetes:a population-based cohort study assessing sex disparities. Circulation 2019, 139, 2742–2753. [CrossRef] [PubMed] 217. Avogaro, A.; Giorda, C.; Maggini, M.; Mannucci, E.; Raschetti, R.; Lombardo, F.; Spila-Alegiani, S.; Turco, S.; Velussi, M.; Ferrannini, E. Incidence of coronary heart disease in type 2 diabetic men and women: Impact of microvascular complications, treatment, and geographic location. Diabetes Care 2007, 30, 1241–1247. [CrossRef] 218. Ferrara, A.; Mangione, C.M.; Kim, C.; Marrero, D.G.; Curb, D.; Stevens, M.; Selby, J.V.; Translating Research Into Action for Diabetes Study Group. Sex disparities in control and treatment of modifiable cardiovascular disease risk factors among patients with diabetes: Translating Research Into Action for Diabetes (TRIAD) Study. Diabetes Care 2008, 31, 69–74. [CrossRef] 219. Russo, G.; Pintaudi, B.; Giorda, C.; Lucisano, G.; Nicolucci, A.; Cristofaro, M.R.; Suraci, C.; Mulas, M.F.; Napoli, A.; Rossi, M.C.; et al. Age- and Gender-Related Differences in LDL-Cholesterol Management in Outpatients with Type 2 Diabetes Mellitus. Int. J. Endocrinol. 2015, 2015, 957105. [CrossRef] 220. Russo, G.; Piscitelli, P.; Giandalia, A.; Viazzi, F.; Pontremoli, R.; Fioretto, P.; De Cosmo, S. Atherogenic dyslipidemia and diabetic nephropathy. J. Nephrol. 2020, 33, 1001–1008. [CrossRef] 221. Russo, G.T.; De Cosmo, S.; Viazzi, F.; Pacilli, A.; Ceriello, A.; Genovese, S.; Guida, P.; Giorda, C.; Cucinotta, D.; Pontremoli, R.; et al. Plasma Triglycerides and HDL-C Levels Predict the Development of Diabetic Kidney Disease in Subjects With Type 2 Diabetes: The AMD Annals Initiative. Diabetes Care 2016, 39, 2278–2287. [CrossRef] 222. Hanai, K.; Babazono, T.; Yoshida, N.; Nyumura, I.; Toya, K.; Hayashi, T.; Bouchi, R.; Tanaka, N.; Ishii, A.; Iwamoto, Y. Gender differences in the association between HDL cholesterol and the progression of diabetic kidney disease in type 2 diabetic patients. Nephrol. Dial. Transplant. 2012, 27, 1070–1075. [CrossRef] 223. Wan, H.; Wang, Y.; Chen, Y.; Fang, S.; Zhang, W.; Xia, F.; Wang, N.; Lu, Y. Different associations between serum urate and diabetic complications in men and postmenopausal women. Diabetes Res. Clin. Pract. 2020, 160, 108005. [CrossRef] [PubMed] 224. Pisano, A.; Cernaro, V.; Gembillo, G.; D’Arrigo, G.; Buemi, M.; Bolignano, D. Xanthine Oxidase Inhibitors for Improving Renal Function in Chronic Kidney Disease Patients: An Updated Systematic Review and Meta-Analysis. Int. J. Mol. Sci. 2017, 18, 2283. [CrossRef] [PubMed] 225. De Cosmo, S.; Viazzi, F.; Pacilli, A.; Giorda, C.; Ceriello, A.; Gentile, S.; Russo, G.; Rossi, M.C.; Nicolucci, A.; Guida, P.; et al. (2015). Serum Uric Acid and Risk of CKD in Type 2 Diabetes. Clin. J. Am. Soc. Nephrol. CJASN 2015, 10, 1921–1929. [CrossRef] [PubMed] 226. Guo, M.; Niu, J.Y.; Li, S.R.; Ye, X.W.; Fang, H.; Zhao, Y.P.; Gu, Y. Gender differences in the association between hyperuricemia and diabetic kidney disease in community elderly patients. J. Diabetes Complicat. 2015, 29, 1042–1049. [CrossRef] 227. Chung, W.K.; Erion, K.; Florez, J.C.; Hattersley, A.T.; Hivert, M.F.; Lee, C.G.; McCarthy, M.I.; Nolan, J.J.; Norris, J.M.; Pearson, E.R.; et al. Precision Medicine in Diabetes: A Consensus Report From the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care 2020, 43, 1617–1635. [CrossRef] 228. Kautzky-Willer, A.; Harreiter, J.; Pacini, G. Sex and Gender Differences in Risk, Pathophysiology and Complications of Type 2 Diabetes Mellitus. Endocr. Rev. 2016, 37, 278–316. [CrossRef] [PubMed] 229. Ueda, P.; Svanström, H.; Melbye, M.; Eliasson, B.; Svensson, A.M.; Franzén, S.; Gudbjörnsdottir, S.; Hveem, K.; Jonasson, C.; Pasternak, B. Sodium glucose cotransporter 2 inhibitors and risk of serious adverse events: Nationwide register based cohort study. BMJ 2018, 363, k4365. [CrossRef] 230. Pasternak, B.; Wintzell, V.; Melbye, M.; Eliasson, B.; Svensson, A.M.; Franzén, S.; Gudbjörnsdottir, S.; Hveem, K.; Jonasson, C.; Svanström, H.; et al. Use of sodium-glucose co-transporter 2 inhibitors and risk of serious renal events: Scandinavian cohort study. BMJ 2020, 369, m1186. [CrossRef] [PubMed] 231. Zinman, B.; Wanner, C.; Lachin, J.M.; Fitchett, D.; Bluhmki, E.; Hantel, S.; Mattheus, M.; Devins, T.; Johansen, O.E.; Woerle, H.J.; et al. Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes. N. Engl. J. Med. 2015, 373, 2117–2128. [CrossRef] Int. J. Mol. Sci. 2021, 22, 5808 24 of 24 232. Perkovic, V.; Jardine, M.J.; Neal, B.; Bompoint, S.; Heerspink, H.J.L.; Charytan, D.M.; Edwards, R.; Agarwal, R.; Bakris, G.; Bull, S.; et al. Canagliflozin and Renal Outcomes in Type 2 Diabetes and Nephropathy. N. Engl. J. Med. 2019, 380, 2295–2306. [CrossRef] [PubMed] 233. Heerspink, H.J.L.; Stefánsson, B.V.; Correa-Rotter, R.; Chertow, G.M.; Greene, T.; Hou, F.F.; Mann, J.F.E.; McMurray, J.J.V.; Lindberg, M.; Rossing, P.; et al. Dapagliflozin in Patients with Chronic Kidney Disease. N. Engl. J. Med. 2020, 383, 1436–1446. [CrossRef] [PubMed] 234. Carrero, J.J.; de Mutsert, R.; Axelsson, J.; Dekkers, O.M.; Jager, K.J.; Boeschoten, E.W.; Krediet, R.T.; Dekker, F.W.; NECOSAD Study Group. Sex differences in the impact of diabetes on mortality in chronic dialysis patients. Nephrol. Dial. Transplant. 2011, 26, 270–276. [CrossRef] [PubMed]
https://openalex.org/W4312705658	https://ejournal.stiepancasetia.ac.id/kindai/article/download/201/170	Indonesian	null	ANALISIS BEBAN KERJA PARA MERCHANDISER DISPLAY (MD) UNTUK MENURUNKAN TINGKAT PERPUTARAN KARYAWAN PADA PT. FRISIAN FLAG INDONESIA (FFI) CABANG BANJARMASIN	Kindai	2,018	cc-by	3,691	ANALISIS BEBAN KERJA PARA MERCHANDISER DISPLAY (MD) UNTUK MENURUNKAN TINGKAT PERPUTARAN KARYAWAN PADA PT. FRISIAN FLAG INDONESIA (FFI) CABANG BANJARMASIN Muhammad Ardi* & Arifia Nurriqli Mahasiswa Sekolah Tinggi Ilmu Ekonomi Pancasetia Banjarmasin Jl. A Yani Km. 5,5 Banjarmasin, Kalimantan Selatan e-mail: ardy0697@gmail.com Abstrak: Tujuan dari penelitian ini adalah untuk mengetahui Analisis Beban Kerja Para Merchandiser Display (MD) Untuk Menurunkan Tingkat Perputaran Karyawan pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin. Metode penelitian yang digunakan adalah deskriptif kualitatif. Teknik pengumpulan data menggunakan dokumentasi, wawancara dan observasi. Berdasarkan hasil penelitian diketahui bahwa kerap terjadi penolakan dari pemilik toko untuk melakukan pendisplayan barang atau pemasangan spanduk ditoko. Padahal hal tersebut merupakan tugas dari para karyawan Merchandiser Display (MD) sebagai bukti pelaporan ke kantor pusat. Hal tersebut mengakibatkan terhambatnya pekerjaan seorang MD. Sebaiknya PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin mempertahankan kualitasnya dan konsisten dengan garansi serta teknik penjualan produk yang sesuai. Keywords: Workload, Merchandiser Display (MD), Employee Turnover Keywords: Workload, Merchandiser Display (MD), Employee Turnover memaksimalkan sumber daya manusia yang dimiliki. Kata Kunci: Beban Kerja, Merchandiser Display (MD), Perputaran Karyawan Abstract: The purpose of this study is to know the Workload Analysis of Merchandiser Display (MD) To Reduce Employee Turnover Rate at PT. Frisian Flag Indonesia (FFI) Banjarmasin Branch. The research method used is descriptive qualitative. Data collection techniques use documentation, interviews and observation. Based on the results of the research note that there is often a refusal from the shop owner to perform pendisplayan goods or installation of banners ditoko. Though it is the duty of employees Merchandiser Display (MD) as evidence of reporting to the headquarters. This results in the inhibition of an MD's job. We recommend PT. Frisian Flag Indonesia (FFI) Banjarmasin Branch maintains its quality and is consistent with the warranty and appropriate product sales techniques. Latar Belakang Keberhasilan g Keberhasilan manajemen dalam mengelola perusahaan akan sangat ditentukan oleh kemampuan dalam mengolah serta mendayagunakan sumber daya manusia yang dimiliki perusahaan. Walaupun sebuah perusahaan telah menggunakan teknologi yang sangat canggih pada peralatan kerjanya, serta penerapan metode-metode yang baik dalam pekerjaan semuanya tidak akan akan banyak membantu dalam mencapai tujuan yang sudah ditetapkan dalam 116 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 116 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 instruksi. Di samping itu, memberi para karyawan dengan beragam pekerjaan. Pelatihan silang itu, membantu perusahaan ketika lowongan, ketidakhadiran, penyusutan usaha, atau pengunduran diri terjadi. Partisipasi peserta pelatihan dan kemampu- pindahan pekerjaan yang tinggi merupakan keunggulan belajar dari perputaran pekerjaan. Perputaran adalah paling terkait dengan karyawan yang sudah bekerja dengan lamanya mereka bekerja. Mereka diharapkan dapat melakukan pekerjaan-pekerjaan pada ragam tingkat atau posisi tertentu. Beban kerja sangat mempengaruhi kinerja karyawan dalam bekerja. Beban kerja yang berat dapat membuat karyawan merasa tidak nyaman dan menjadi tidak betah dalam bekerja. Oleh karena itu, diperlukan perhitungan beban kerja untuk karyawan. Perhitungan beban kerja dapat dilihat dari 3 aspek, yakni fisik, mental dan penggunaan waktu. Aspek fisik meliputi beban kerja berdasarkan kriteria-kriteria fisik manusia. Aspek mental merupakan perhitungan beban kerja dengan mempertimbangkan aspek mental (psikologis). Sedangkan aspek pemanfaatan waktu lebih mempertimbangkan pada aspek pengunaan waktu untuk bekerja (Adipradana, 2014:78). instruksi. Di samping itu, memberi para karyawan dengan beragam pekerjaan. Pelatihan silang itu, membantu perusahaan ketika lowongan, ketidakhadiran, penyusutan usaha, atau pengunduran diri terjadi. Partisipasi peserta pelatihan dan kemampu- pindahan pekerjaan yang tinggi merupakan keunggulan belajar dari perputaran pekerjaan. Perputaran adalah paling terkait dengan karyawan yang sudah bekerja dengan lamanya mereka bekerja. Mereka diharapkan dapat melakukan pekerjaan-pekerjaan pada ragam tingkat atau posisi tertentu. PT Frisian Flag Indonesia merupakan perusahaan multinasional, sebagai bagian dari keluarga besar Cooperatve Condensfabriek Friesland, yang kini berubah nama menjadi Royal FrieslandCampina, yang bergerak dalam bidang industri produk nutrisi berbasis susu. Friesland Campina merupakan koperasi peternak sapi perah terbesar di dunia yang berpusat di Belanda dan beranggotakan 16 ribu peternak yang tersebar di tiga negara, serta memiliki karyawan tidak kurang dari 22 ribu orang yang tersebar di 100 perusahaan di seluruh dunia. Frisian Flag pertama kali muncul di Indonesia pada tahun 1922, ketika produk susu Friesche Vlag di impor ke Indonesia melalui Cooperative Condens-Fabriek. Sekitar tahun 1930-an, produk Frisian Flag dan Friesche Flag mulai dipromosikan dengan nama Soesoe Tjap Bendera. PT Frisian Flag Indonesia merupakan perusahaan multinasional, sebagai bagian dari keluarga besar Cooperatve Condensfabriek Friesland, yang kini berubah nama menjadi Royal FrieslandCampina, yang bergerak dalam bidang industri produk nutrisi berbasis susu. Friesland Campina merupakan koperasi peternak sapi perah terbesar di dunia yang berpusat di Belanda dan beranggotakan 16 ribu peternak yang tersebar di tiga negara, serta memiliki karyawan tidak kurang dari 22 ribu orang yang tersebar di 100 perusahaan di seluruh dunia. Keberhasilan Sebagian besar lingkungan kerja yang baik dapat membantu meningkatkan kepuasan kerja serta berhubungan positif antara lingkungan kerja dan kepuasan kerja ada untuk semua jenis kelompok pekerjaan, namun lingkungan fisik dan non fisik merupakan faktor yang mempengaruhi kepuasan kerja selain kompensasi, promosi jabatan serta karakteristik dari pekerjaan yang bersangkutan (Sardzoska, 2012:14). 115 116 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 116 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 Biasanya tiap kepindahan didahului dengan pelatihan Tingginya jam kerja serta peraturan yang diberikan perusahaan berdampak besar terhadap kerja para karyawan sebagai contoh, karyawan diharuskan bekerja dengan cepat untuk emmenuhi target namun hal tersebut tidak berimbang dengan kemampuan yang Ardi & Nurriqli, Analisis Beban Kerja Pada Merchandiser Display (MD) Pada …. 117 117 dimiliki para karyawan Merchandiser Display (MD). MD atau Merchandiser Display adalah salah satu bagian dari team promosi yang bertugas mendisplay atau memajang produk di etalase toko dengan baik. Umumnya karyawan bagian Merchandiser Display bertugas untuk memastikan produk yang tertata rapi di etalase selain sedap dipandang mata juga akan meningkatkan penjualan jika produk ditata sedemikian rupa sehingga mudah dijangkau dan ditemukan oleh pembeli. Tugas Merchandiser Display tidak hanya bertugas di modern market tetapi juga di retail, di pasar-pasar tradisional akan kita lihat pajangan shampo, deterjen, pelembut pakaian yang tertata rapi. Itu juga hasil kerja Mercandiser Display. kepuasaan kerja meningkat, maka perputaran karyawan dan absensi menurun atau sebaliknya. Turnover yang tinggi pada suatu bidang dalam suatu organisasi, menunjukkan bahwa bidang yang bersangkutan perlu diperbaiki kondisi kerjanya atau cara pembinaannya. Hal tersebut tidak berbanding dengan kemampuan yang dimiliki oleh karyawan Merchandiser Display, selain waktu yang disediakan untuk satu buah toko hanya selama 25 menit, kendala lain ditemukan dari pemilik toko yang tidak mau toko merekan dipasang atau dirubah-rubah tempat penjualan sehingga para karyawan harus meyakinkan para pemilik toko serta melepas kembali atribut dan mengembalikan display ke awal setelah melakukan pendokumentasian ke kantor pusat. Selain itu tugas Merchandiser Display (MD) lainnya adalah memasang alat promosi produk seperti striker, spanduk, banner dan lain sebagainya sebagai bentuk promosi “di darat” untuk menunjang promosi produk yang telah dilakukan “di udara” lewat iklan di TV, radio atau internet. Namun, pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin, Merchandiser Display juga memiliki target serta memantau penjualan atau mengambil sebagian pekerjaan sales yakni menghitung jumlah penjualan serta memantau produk pesaing. Alasan yang telah dipaparkan tersebut merupakan alasan yang membuat sebagian karyawan Merchandiser Display memilih untuk berhenti bekerja akibat ketidakmampuan dalam tuntutan pekerjaan yang diberikan. Berdasarkan lingkup fenomena tersebut, maka peneliti tertarik untuk meneliti “Analisis Beban Kerja Para Merchandiser Display (MD) Untuk Menurunkan Tingkat Perputaran Karyawan pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin”. PT. 116 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 Frisian Flag pertama kali muncul di Indonesia pada tahun 1922, ketika produk susu Friesche Vlag di impor ke Indonesia melalui Cooperative Condens-Fabriek. Sekitar tahun 1930-an, produk Frisian Flag dan Friesche Flag mulai dipromosikan dengan nama Soesoe Tjap Bendera. Tidak jarang kinerja kayawan mengalami fluktuasi. Kondisi seperti itu diduga ada hubungannya dengan terlalu lamanya seseorang dalam periode kerja di satu unit atau di satu pekerjaan saja. Akibatnya timbul kebosanan dan bahkan kejenuhan di kalangan mereka. Dalam konteks pengembangan sumberdaya manusia kondisi seperti itu tidak sehat. Maka manajemen seharusnya menerapkan kebijakan rotasi pekerjaan/karyawan secara berkala. Rotasi atau perputaran pekerjaan tidak selalu berjalan mulus. Bisa saja tindakan seperti itu menuai protes dari karyawan yang merasa dirinya sudah mapan pada posisi yang sekarang. Karena itu kebijakan seperti itu harus didasarkan pada data dan informasi akurat mengenai kinerja individu, pengalaman kerja di unit, keterlibatan pelatihan, dan perilaku karyawan. Kemudian perlu dilakukan sosialisasi agar para karyawan tidak merasa diperlakukan secara tidak adil. Hal lain yang penting juga dipertimbangkan bahwa rotasi pekerjaan harus berbasis kompetensi dari karyawan bersangkutan. Mereka harus disiapkan lebih dahulu paling tidak dalam bentuk orientasi di tempat pekerjaan yang baru. PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin merupakan salah satu anak cabang dari PT. Frisian Flag Indonesia (FFI) Indonesia yang bertugas untuk mendistrbutor produk-produk dari PT. Frisian Flag Indonesia (FFI) seperti susu kental manis (SKM), susu bubuk anak dan susu UHT. Dalam perananannya, PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin memberlakukan jam kerja yang tinggi serta tanggung jawab yang tinggi bagi seluruh karyawan khususnya para karyawan Merchandiser Display (MD). PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin merupakan salah satu anak cabang dari PT. Frisian Flag Indonesia (FFI) Indonesia yang bertugas untuk mendistrbutor produk-produk dari PT. Frisian Flag Indonesia (FFI) seperti susu kental manis (SKM), susu bubuk anak dan susu UHT. Dalam perananannya, PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin memberlakukan jam kerja yang tinggi serta tanggung jawab yang tinggi bagi seluruh karyawan khususnya para karyawan Merchandiser Display (MD). Karena itu rotasi pekerjaan dapat dikategorikan sebagai bentuk pelatihan “on- the-job”. Karyawan mengikuti pelatihan silang antarjenis pekerjaan. Di sini, pelatih memindahkan karyawan dari satu pekerjaan ke pekerjaan lainnya. 116 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 Frisian Flag Indonesia (FFI) Cabang Banjarmasin memberikan target bagi seorang Merchandiser Display waktu selama 8 jam untuk mendisplay, mendokumentasikan (berupa foto), membawa sebagian produk yang tidak ada, memasang stiker/wallpaper atau banner produk serta memantau penjualan produk pesaing dengan jumlah took atau mini market sebanyak 14 buah dengan cakupan wilayah Banjarmasin, Banjarbaru, Martapura (Kab. Banjar) dan Kab. Barito Kuala (Marabahan hingga Kapuas). PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin memberikan target bagi seorang Merchandiser Display waktu selama 8 jam untuk mendisplay, mendokumentasikan (berupa foto), membawa sebagian produk yang tidak ada, memasang stiker/wallpaper atau banner produk serta memantau penjualan produk pesaing dengan jumlah took atau mini market sebanyak 14 buah dengan cakupan wilayah Banjarmasin, Banjarbaru, Martapura (Kab. Banjar) dan Kab. Barito Kuala (Marabahan hingga Kapuas). Studi Literatur Manajemen pada dasarnya berasal dari kata to manage yang artinya mengatur. Manajemen adalah ilmu dan seni yang mengatur proses pemanfaatan sumber daya manusia dan sumber-sumber daya lainnya secara efektif dan efesien untuk mencapai suatu tujuan tertentu (Hasibuan, 2015:141). Manajemen ini terdiri dari enam unsur yaitu, men, money, method, materialis, machines, market. MSDM adalah suatu bidang manajemen yang khusus mempelajari hubungan dan peranan manusia dalam organisasi perusahaan. Unsur MSDM adalah manusia yang merupakan tenaga kerja, dengan demikian fokus yang dipekajarai MSDM ini hanyalah masalah yang berhubungan tenaga kerja manusia saja. Manajemen sumber daya manusia adalah seni dan ilmu pengadaan, pemngambangan Dampak dari perputaran karyawan meliputi beberapa hal diantaranya terhadap produktifitas, keabsenan dan pengunduran diri. Meskipun hanya merupakan salah satu faktor dari banyak faktor pengaruh lainnyaa, kepuasaan kerja mempengaruhi tingkat perputaran karyawan dan keabsensi. perusahaan bisa mengharapkan bahwa bila 118 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 118 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 Judul penelitian ini adalah analisis beban kerja para Merchandiser Display (MD) untuk menurunkan tingkat perputaran karyawan pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin. Dalam melakukan penelitian ini penulis menggunakan jenis penelitian deskriptif kualitatif yaitu untuk mengetahui gambaran dari objek penelitian dengan menggunakan wawancara, observasi dan dokumentasi. Populasi dalam penelitian ini adalah seluruh karyawan PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin yang dijadikan sebagai sampel penelitian yaitu sebanyak 5 orang yang terdiri dari 1 orang Team Leader (TL), 2 karyawan Merchandiser Display (MD) dan 2 orang pemilik toko. dan manfaat sumber daya manusia sehingga tujuan perusahaan dapat dihasilkan secara daya guna dan kegairahan kerja (Manulang, 2014:32). Penjualan (selling) adalah suatu kegiatan yang ditujukan untuk mencari pembeli, mempengaruhi, dan memberi petunjuk agar pembeli dapat menyesuaikan kebutuhannya dengan produksi yang ditawarkan serta mengadakan perjanjian mengenai harga yang menguntungkan bagi kedua pihak (Moekijat, 2014:488). Robbins (2015:24) menjelaskan bahwa perputaran karyawan (employee turnover) adalah pengunduran diri permanen secara sukarela maupun tidak sukarela dari suatu organisasi atau proses dimana karyawan- karyawan meninggalkan organisasi dan harus digantikan. Hasil Penelitian dan Pembahasan Perputaraan karyawan yang tinggi mengakibatkan bengkaknya biaya perekrutmen, seleksi, dan pelatihan. Sementara itu keinginan berpindah (Turnover) yang berujung pada keputusan karyawan untuk meninggalkan pekerjaannya. Meningkatnya tinggi turnover pada perusahaan karyawan akan semakin banyak menimbulkan berbagai potensi biaya, baik itu biaya pelatihan yang sudah di investasikan pada karyawan, tingkat kinerja yang mesti di korbankan, maupun biaya rekruitmen dan pelatihan kembali (Agustina, 2015:48). 1. Beban Kerja Yang Dilaksanakan oleh Para Karyawan Merchandiser Display (MD) pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin yang Selama ini Krisis global yang melanda seluruh Negara pada rentan tahun 2013-2017 ini memicu pergerakan-pergerakan ekonomi yang cukup cepat disetiap sektornya dan mengharuskan seluruh perusahaan dan organisasi perekonomian di seluruh dunia melakukan tindakan strategi ekonomi yang tepat, cepat dan akurat untuk menanggulangi krisis global tersebut, dan resiko dari hal itu semua adalah penambahannya stress kerja dan tekanan kerja untuk para pekerja yang terlibat, disadari atau tidaknya oleh para pekerja hal tersebut dapat memicu terjadinya kenaikan beban kerja. Menurut Danang Sunyoto (2012:64), beban kerja adalah yang terlalu banyak dapat menyebabkan ketegangan dalam diri seseorang sehingga menimbulkan stress. Hal ini bisa disebabkan oleh tingkat keahlian yang dituntut terlalu tinggi, kecepatan kerja mungkin terlalu tinggi, volume kerja mungkin terlalu banyak dan sebagainya. Hasil penelitian oleh Alvin (2016) berdasarkan survey yang di lakukan oleh Canada life lebih dari setengah karyawan di Inggris merasakan tingkat stress yang menigkat dibandingkan dengan tahun lalu. Survey ini di lakukan kepada 1.100 karyawan di seluruh Inggris 48% karyawan mengatakan terlalu banyak bekerja dan tidak memiliki work life balance dan 22% karyawan menuturkan mereka terlalu takut untuk menjelaskan hal tersebut kepada managemen Beban kerja adalah istilah yang mulai dikenal sejak tahun 1970-an. Banyak ahli yang telah mengemukakan definisi beban kerja sehingga terdapat beberapa definisi yang berbeda mengenai beban kerja. Ia merupakan suatu konsep yang multi-dimensi, sehingga sulit diperoleh satu kesimpulan saja mengenai definisi yang tepat (Cain, 2016:68). Metode Penelitian Metode Penelitian Ardi & Nurriqli, Analisis Beban Kerja Pada Merchandiser Display (MD) Pada …. 1 119 persusahaan dan meminta bantuan kepada atasan dan rekan kerjanya. waktu, maka para Team Leader harus membagi waktu untuk memberikan training dan arahan kepada para karyawan baru. Penelitian tersebut mengungkap bahwa 17% mereka tidak mendapatkan dukungan dari atasan dan managemen walaupun hal tersebut telah mereka jelaskan dan hanya 15% dari atasan yang memahami kecemasan terhadap stress yang karyawan alami. Dari hal itu semua terdapat 40% karyawan yang memilih melakukan cuti, ijin sakit ataupun alasan lainnya hanya untuk mendapatkan waktu senggang diluar jam kerja yang selama ini mereka dapatkan. Masalah yang sering dihadapi para Merchandiser Display (MD) yang mengakibatkan tingginya perputaran karyawan di lapangan yakni kerap terjadi penolakan dari pemilik toko untuk melakukan pendisplayan barang atau pemasangan spanduk ditoko. Padahal hal tersebut merupakan tugas dari para karyawan Merchandiser Display (MD) sebagai bukti pelaporan ke kantor pusat. Hal tersebut mengakibatkan terhambatnya pekerjaan seorang MD. Hal tersebut hampir serupa dengan apa yang dialami oleh para karyawan Merchandiser Display (MD) pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin yang selama ini. Dari hasil observasi sistem kerja yang terjadi pada para karyawan Merchandiser Display (MD) pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin yakni setiap karyawan terlihat bekerja dengan sangat cekatan, cepat dan sering di temukan karyawan yang terlihat memegang jidat, tampak kesal, kelelahan dan sering mengeluh setiap kali harus menyerahkan laporan atau absen di kantor. Masalah lain juga sering dilakukan oleh para karyawan Merchandiser Display (MD) seperti mendisplay barang toko yang tidak ada dengan membawa barang dari perusahaan dan memnjamkannya hanya untuk sementara waktu selama proses pelaporan kerja ke kantor pusat berlangsung, namun kecurangan tersebut sifatnya kesengajaan dan demi cepatnya kerja para karyawan Merchandiser Display (MD) serta tidak memberatkan pemilik toko untuk membeli produk yang tidak ada karena menurut pemilik toko, pembelian produk barang dagangan dilakukan pada waktu yang telah ditentukan serta menunggu stok barang yang lain habis agar tidak terjadi penumpukan barang. Pergantian tersebut tidak serta merta terjadi dalam waktu bersamaan namun bertahap. Berdasarkan haisl dokumentasi yang ditemukan peneliti di bagian HRD PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin, dalam setiap bulan sebanyak 2-5 orang karyawan Merchandiser Display (MD) mengundurkan dengan berbagai alasan, salah satunya yaitu jadwal kerja yang terlalu padat dan tidak berimbang dengan kemampuan karyawan. Metode Penelitian Posisi karyawan yang kosong harus segera digantikan oleh karyawan baru, namun hal ini dapat berdampak pada kerugian perusahaan baik secara operasional dan secara teknis karena dalam merekrut karyawan baru perusahaan harus mengalokasikan dana untuk memajang poster rekrutmen tenaga kerja, memasang iklan dan biaya operasional lain yang dibutuhkan dalam rekrutmen tenaga kerja baru khususnya bagian Merchandiser Display (MD), namun ditinjau dari segi 2 Penerapan Beban Kerja yang Dapat Menurunkan Tingkat Perputaran Karyawan pada PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin Dalam mengurangi dampak beban kerja bagi karyawan yang mengakibatkan tingginya perputaran karyawan maka sebaiknya PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin melalui pimpinan yang ada menambah jumlah karyawan untuk mengurangi jumlah pekerjaan yang dilakukan. Untuk mengurangi absennya karyawan seperti cuti, sakit, izin urusan keluarga, izin urusan perkuliahan akibat karyawan sering pulang larut malam dengan jam kerja yang berlebih sebaiknya pihak 120 KINDAI, Vol 14, Nomor 2, April 2018, halaman 115-121 gratis bagi perusahaan sehingga karyawan tidak perlu memalsukan data dan tidak pula menurunkan prestasi karyawan sebagai Merchandiser Display (MD). perusahaan memasukan hal tersebut ke dalam jam lebur maka karyawan sebagai ganti dari jam yang berlebih diberikan uang insentif sebagai bentuk imbalan atau bonus dengan tujuanmampu memberikan motivasi bagi karyawan dalam bekerja. Sebaiknya PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin menambah jumlah MD sehingga satu orang MD hanya memiliki target 7 toko dalam seharinya sehingga target perusahaan terpenuhi dengan baik. Kesimpulan Berdasarkan analisis beban kerja saat ini, sering dilakukan oleh para karyawan Merchandiser Display (MD) seperti mendisplay barang toko yang tidak ada dengan membawa barang dari perusahaan dan memnjamkannya hanya untuk sementara waktu selama proses pelaporan kerja ke kantor pusat berlangsung, namun kecurangan tersebut sifatnya kesengajaan dan demi cepatnya kerja para karyawan Merchandiser Display (MD) serta tidak memberatkan pemilik toko. Selain itu, untuk mengurangi terjadinya peningkatan perputaran kerja perusahaan sebaiknya memberikan reward abagi karyawan yang berprestasi dan memberikan kesempatan peningkatan jenjang karir bagi karyawan guna meningkatkan ilmu pengetahuan karyawan dalam bekerja sehinggga kinerja karyawan lebih baik dalam bekerja serta perusahaan juga dapat memberikan training atau pelatihan tambahan bagi seluruh karyawan guna meningkatkan kualitas kerja dan meningkatkan penjualan perusahaan. Penting bagi Frisian Flag untuk menambah jumlah MD agar tidak ada lagi MD yang kelabakan dalam bekerja dan lebih menghargai para MD melalui pemberian reward yang sesuai dengan kinerja mereka. Setiap ada pekerjaan yang membutuhkan lembur maka harus disiapkan uang lemburnya. Pihak PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin sebaiknya melakukan analisis lagi tentang produk yang lebih diminati oleh masyarakat. PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin juga dapat memberikan display atau sampel produk yang tidak untuk dijual kepada para pemilik toko yang nantinya akan digunakan untuk pelaporan kegiatan para Merchandiser Display (MD) Untuk mengurangi terjadinya kasus kecurangan pendisplayan barang dagangan, PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin melalui bagian sales giat melakukan pengecekkan kepada toko-toko dan meminta agar para pengecer aatau pemilik toko membeli produk best seller kepada pemilik toko dengan jumlah yang lebih banyak. Serta pihak PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin sebaiknya melakukan analisis lagi tentang produk yang lebih diminati oleh masyarakat. PT. Frisian Flag Indonesia (FFI) Cabang Banjarmasin juga dapat memberikan display atau sampel produk yang tidak untuk dijual kepada para pemilik toko yang nantinya akan digunakan untuk pelaporan kegiatan para Merchandiser Display (MD) atau memberikan hadiah secara Cuma-Cuma beberapa produk susu yang telah habis sesuai dengan keperluan pendisplayan bagi toko penjual produk PT. Frisian Flag Indonesia (FFI) guna meningkatkan loyalitas terhadap para para agen dan pengecer serta guna promosi DAFTAR PUSTAKA Abdurahman. 2012. Abdurahman. 2012. Paduan Praktis Memahami Penelitian Manajemen. Pustaka Setia: Bandung. Adipradana. 2014. Amnalisis Beban Kerja pada Karyawan. Jurnal Psikologi. :Edisi Kedua Vol. 3. Universitas Muhammadiyah Surakarta. Adipradana. 2014. Amnalisis Beban Kerja pada Karyawan. Jurnal Psikologi. :Edisi Kedua Vol. 3. Universitas Muhammadiyah Surakarta. Anindya. 2012. Analisis Beban Kerja Untuk Menurunkan Tingkat Perputaran Karyawan (Studi Di PT. Bank Negara Indonesia JPK Bandung). Jurnal Manajemen. Bandung: Universitas Pajajaran. Arifin. 2013. Analisis SWOT dalam Menentukan Strategi Pemasaran Arifin. 2013. Analisis SWOT dalam Menentukan Strategi Pemasaran Ardi & Nurriqli, Analisis Beban Kerja Pada Merchandiser Display (MD) Pada …. 121 Sepeda Motor pada PT. Samekarindo Indah di Samarinda. Administrasi Bisnis, 56-70. Pearson. 2015. Sistem Informasi Manajemen. Jakarta: Salemba. Pridjominto. 2013. Disiplin Kiat Sukses. Jakarta: Abadi Arika. 2011. Analisis Bauran Pemasaran Terhadap Keputusan Pembelian Motor Merek Suzuki Pada PT. Sinar Galesong Pratama Makassar. Skripsi Teknik Industri Universitas Hasanuddin. Prihatini. 2016. Analisis Hubungan baban Kerja dengan Stres Kerja Perawat di Tiap Ruang Rawat Inap RSUD Sidikalang. Jurnal Manajemen: Vol. 12. Medan. Universitas Sumatera Arikunto, S. 2014. Prosedur Penelitian Suatu Pendekatan Praktik. Rieneka Cipta: Jakarta Rangkuti. F. 2014. Analisis SWOT Teknik Membedah Kasus Bisnis. PT. Gramedia Pustaka Utama: Jakarta. Cain. 2016. Sciening, An Involvement Approach to Elementary Science Methods. Columbus: Memi Publishing Co. Sardzoska. 2012. Organizational behavior in diverses work settings. Psychological thought, (1) 2, 121-136 Danang Sunyoto. 2012. Metodologi Penelitian Akutansi. PT. Refika. Bandung. Sehwarat dan Narang. 2016. Production Management. Jakarta: Salemba Sugiyono. 2014. Metode Penelitian Kualitatif dan Kuantitatif R & D. Alfabeta: Jakarta. Hasibuan. 2015. Manajemen Sumber Daya Manusia. Erlangga. Jakarta. Hendra T. 2012. Manajemen Pemasaran. Jakarta: Rieneka Cipta. Sinta Monica. 2015. Analisis Beban Kerja Untuk Menurunkan Tingkat Perputaran Karyawan Pada PT. Galamedia Bandung Perkasa. Jurnal Manajemen. Bandung: Universitas Pajajaran. Kotler. 2014. Definisi Pelayanan. Salemba Medika: Jakarta Kotler. 2015. Manajemen Sumber Daya Manusia. Salemba Medika: Jakarta Suma’mur. 2009. Hiegiene Perusahaan dan Keselamatan Kerja. Jakarta : CV. Sagung Seto. Manulang. 2014. Manajemen Sumber Daya Manusia. BPFE. Jogjakarta. Swastha. 2014. Manajemen Pemasaran Modern. Yogyakarta: Liberty. Mariyani. 2015. Upaya Meningkatkan Penjualan Melalui Disiplin Kerja Karyawan PT. Marga Nusantara Jaya Cb. Samarinda. Skripsi Manajamen. Universitas Shahid. Swasta. 2016. Manajemen Penjualan: Pelaksanaan Penjualan, BPFE-. Yogyakarta. McCarthy dan Perreault. 2015. Envoiremental Manajemen. Jakarta: Erlangga Tarwaka. 2015. Keselamatan dan Kesehatan Kerja Manajemen dan Implementasi. K3 di Tempat Kerja, Harapan Press, Surakarta; Moekijat. 2014. Manajemen Sumber Daya Manusia. CV. Mandar Maju. Bandung. Wijaya, Tony. 2015. Manajemen Kualitas Jasa, Jakarta :PT. Indeks
https://openalex.org/W2923527943	https://www.jmir.org/2019/3/e11732/PDF	English	null	A Human(e) Factor in Clinical Decision Support Systems	JMIR. Journal of medical internet research/Journal of medical internet research	2,019	cc-by	8,482	A Human(e) Factor in Clinical Decision Support Systems Tim Bezemer1, MSc; Mark CH de Groot1, PhD; Enja Blasse1, BSc; Maarten J ten Berg1, PhD; Teus H Kappen2, MD, PhD; Annelien L Bredenoord3, PhD; Wouter W van Solinge1, PhD; Imo E Hoefer1, MD, PhD; Saskia Haitjema1, MD, PhD Corresponding Author: Saskia Haitjema, MD, PhD Laboratory of Clinical Chemistry and Haematology University Medical Center Utrecht Room G03.550 Heidelberglaan 100 Utrecht, 3584 CX Netherlands Phone: +31 88 755 5555 Email: S.Haitjema@umcutrecht.nl KEYWORDS clinical decision support; big data; artificial intelligence; machine learning; deep learning; precision medicine; expert systems; data science; health care providers clinical decision support; big data; artificial intelligence; machine learning; deep learning; precision medicine; expert systems; data science; health care providers medical training is dedicated to learning how to distinguish relevant from irrelevant information to ultimately make the best decision possible. Yet, the overwhelming amount, production speed, multidimensionality, and potential value of data currently available (often simplified and referred to as big data) exceed the limits of understanding of the human brain. Abstract The overwhelming amount, production speed, multidimensionality, and potential value of data currently available—often simplified and referred to as big data —exceed the limits of understanding of the human brain. At the same time, developments in data analytics and computational power provide the opportunity to obtain new insights and transfer data-provided added value to clinical practice in real time. What is the role of the health care professional in collaboration with the data scientist in the changing landscape of modern care? We discuss how health care professionals should provide expert knowledge in each of the stages of clinical decision support design: data level, algorithm level, and decision support level. Including various ethical considerations, we advocate for health care professionals to responsibly initiate and guide interprofessional teams, including patients, and embrace novel analytic technologies to translate big data into patient benefit driven by human(e) values. JOURNAL OF MEDICAL INTERNET RESEARCH JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 1 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ 1Laboratory of Clinical Chemistry and Haematology, University Medical Center Utrecht, Utrecht University, Utrecht, Netherlands 2Department of Anesthesiology, University Medical Center Utrecht, Utrecht University, Utrecht, Netherlands 3Department of Medical Humanities, University Medical Center Utrecht, Utrecht University, Utrecht, Netherlands Introduction clinically relevant outcomes based on input data. A decision can be supported by a rule or a model as simple as an if-then rule (eg, built-in reference values for laboratory measurements) or a complex prediction model (eg, artificial intelligence [AI] pointing radiologists to possible incidental findings). The corresponding output of a CDS system varies from showing the generated prediction as input for a clinical decision (eg, automatically generated early warning scores) to acting upon the decision without human interference (eg, an implantable cardioverter defibrillator). clinically relevant outcomes based on input data. A decision can be supported by a rule or a model as simple as an if-then rule (eg, built-in reference values for laboratory measurements) or a complex prediction model (eg, artificial intelligence [AI] pointing radiologists to possible incidental findings). The corresponding output of a CDS system varies from showing the generated prediction as input for a clinical decision (eg, automatically generated early warning scores) to acting upon the decision without human interference (eg, an implantable cardioverter defibrillator). Recent reports on CDS systems in radiology and pathology are promising. Computers can, for example, support radiologists in interpreting mammograms or help pathologists in the classification of brain tumors [2,3]. Google recently also received the Food and Drug Administration’s approval for the introduction of a diabetic retinopathy algorithm based on retinal imaging [4]. Moreover, development of complex algorithms now starts to transcend beyond imaging specialties [5]. If the computer seems to know better anyhow, should we fully abolish medical curricula and focus on data scientists who Table 1. This table shows the 3 levels in the building process of a clinical decision support system and some examples of where clinical expert knowledge of health care professionals plays a role in each of these levels. Table 1. This table shows the 3 levels in the building process of a clinical decision support system and some examples of where clinical expert knowledge of health care professionals plays a role in each of these levels. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 2 (page number not for citation purposes) Introduction Although medical data collection and interpretation used to be the domain of health care professionals, the broad availability of health data in unprecedented amounts has significantly and irrevocably changed the landscape of modern care. Even patients now start to collect their own health data using, for instance, smart watches or apps, which may become an important source of health data in the future. Conversely, developments in data analytics and computational power provide the opportunity to obtain new insights and transfer data-provided added value to clinical practice in real time. Such systems are called clinical decision support (CDS) and can broadly be defined as “information systems designed to aid in the clinical decision-making process, by integrating different sources of health information such as Electronic Health Records, laboratory test results, etc” [1]. CDS systems come in many forms and functions, but all share the aim of generating The craft of translating information into the right diagnosis and corresponding treatment is daily routine for health care professionals. It entails collecting the relevant data for each individual patient, integrating this information with pre-existing knowledge, drawing a conclusion, and initiating appropriate treatment in dialogue with the patient. A significant portion of http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al develop CDS systems, with lay people gathering the information required for them, entitling the computer to do the interpretation instead? Probably not. There are at least two human beings present when a medical decision is made: a patient and a health care professional. Recently, the role of the patient as a disease experience expert [6] and his or her role in shared decision making have come into awareness. Here, we focus on the role of health care professionals and their expert knowledge. Throughout this paper, we will briefly touch upon various ethical issues. However, we strongly feel ethical considerations pertaining to algorithmic decision making deserve a discussion of their own, and kindly refer the reader to a recent overview on this topic [7]. In this paper, we show that a well-designed CDS system needs expert knowledge of health care professionals in all 3 phases of development: data, algorithm, and decision support (Table 1). Moreover, in the era of CDS, we advocate for health care professionals to responsibly initiate and guide interprofessional teams, including patients, and embrace novel analytic technologies to translate big data into patient benefit driven by human(e) values. Data Developing a CDS system starts with data. Data come in many forms and sets (Table 2). Structured data such as numeric data (eg, laboratory measurements and blood pressure) or categorical data (eg, hypertension yes/no or educational level) are easiest to work with in a model. This is the first point at which expert knowledge of health care professionals may enter CDS development process. However, a substantial part of day-to-day clinical decisions is based on unstructured free-text entries, encompassing, for example, patient history and physical examination observations by doctors or regular notes from nurses. Although discouraged in modern electronic health record (EHR) systems, unstructured free-text clinical notes still provide irreplaceable information and context to health care professionals. Using free text introduces a number of challenges. Aside from the obvious ones, such as writing style and typos, medical text is incredibly site specific and can be highly biased. This phenomenon ranges from language- and country-specific abbreviations to jargon differences between 2 wards within the same hospital (eg, AF for atrial flutter and amniotic fluid or MS for mitral stenosis and multiple sclerosis). This is an understandable effect of rapid communication between health care professionals or of health care professionals taking personal notes to capture their train of thought. However, this leads to a given phrase, term, acronym, or abbreviation being context specific and having different meanings in different situations. Introduction Poor: free text contains jargon-specif- ic, ambiguous abbreviations (eg, PCI: percutaneous coronary interven- tion/prophylactic cranial irradiation). Excellent: data are recorded in a standardized way, designed to prevent noise. Excellent: data are recorded in a standardized way, designed to prevent noise. Very poor: These type of data are very sensitive to interobserver noise (eg, personal abbreviations, spelling mistakes, and personal focus in recording certain types of informa- tion). Good: data are often machine-derived or recorded in a standardized way. However, bias because of differences in information-recording habits among health care professionals may arise. Limited: trials are designed and con- ducted for one specific research question. Limited: trials are designed and con- ducted for one specific research question. Excellent: these type of data are readily available, contain a lot of context (see Context completeness), and can thus be repurposed for a vari- ety of applications. Excellent: these type of data are readily available and can thus be used for a plethora of purposes. Poor: study design is hit-or-miss. Bias cannot be corrected after the data recording process. Excellent: study design can be revisit- ed if unanticipated bias effects arise. In this sense, bias could be corrected by altering the data selection. Excellent: study design can be revisit- ed if unanticipated bias effects arise. In this sense, bias could be corrected by altering the data selection. Free-text interpretation, therefore, heavily depends on contextual expert knowledge. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 3 (page number not for citation purposes) Introduction Example of expert knowledge Level and example of issue Data level Hemoglobin reference range to diagnose anemia Laboratory thresholds Body mass index Derived measurementsa Grouping of related diagnoses in a study population Diagnostic codes Same abbreviations having different meanings Jargon Glucose values are highly dependent on the time of day (eg, pre- or postprandial) Temporality Algorithm level How to handle missing data (eg, missing not at random) Methodological choices Constructing relevant derived variables from raw data (eg, torsades de pointes, Wolff-Parkinson-White syn- drome) Feature engineeringa For example, oxygen saturation of zero caused by a slipping pulse oximeter, switched leads in an electrocar- diogram Artifacts Decision support level Risk probability of 0.75 requires a warning (amber light) in a CDSb system Interpretation of model output Tuning of implantable cardioverter defibrillator Degree of autonomy Weighing a CDS system’s advice to treat while considering quality of life versus treatment burden in elderly cancer patients in a shared decision-making context Knowledge on usefulness aDerived measurements may occur at the data level but also at the algorithm level; the former being undesirable because any manipulation at the data level may result in a loss of information. bCDS: clinical decision support. aDerived measurements may occur at the data level but also at the algorithm level; the former being undesirable because any manipulation at the data level may result in a loss of information. bCDS: clinical decision support. bCDS: clinical decision support. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 2 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al Table 2. Table comparing different types of clinical data on some points important to clinical decision support systems. Structured data/electronic health record (eg, lab values and smoking status) Registry/trial data (eg, case record forms case record forms and question- naires) Electronic health record free-text/un- structured data (eg, clinical notes) Clinical decision support issues Depends on implementation. Context may be lost because of predetermined categorization. Poor: context is essentially absent as a priori interpretation is an integral part of recording data in case record forms. Excellent: contextual information can be included. Context completeness Excellent: data can be parsed or direct- ly used by computers. Good: data are uniformly formatted and can be parsed by computers. Poor: information is mostly useful for case-specific usage by humans. http://www.jmir.org/2019/3/e11732/ Introduction May require text mining/text retrieval to convert to a machine-readable format. Machine readability Good: lab values can be converted using reference values. Structured data, such as smoking and hyperten- sive status, can be reformatted for in- teroperability. Excellent: trial data are usually collect- ed using a standardized protocol, al- lowing for interoperability between institutions. Poor: free text contains jargon-specif- ic, ambiguous abbreviations (eg, PCI: percutaneous coronary interven- tion/prophylactic cranial irradiation). Translatability (between institu- tions) Good: data are often machine-derived or recorded in a standardized way. However, bias because of differences in information-recording habits among health care professionals may arise. Excellent: data are recorded in a standardized way, designed to prevent noise. Very poor: These type of data are very sensitive to interobserver noise (eg, personal abbreviations, spelling mistakes, and personal focus in recording certain types of informa- tion). Noise resistance Excellent: these type of data are readily available and can thus be used for a plethora of purposes. Limited: trials are designed and con- ducted for one specific research question. Excellent: these type of data are readily available, contain a lot of context (see Context completeness), and can thus be repurposed for a vari- ety of applications. Availability for reuse/general applicability Excellent: study design can be revisit- ed if unanticipated bias effects arise. In this sense, bias could be corrected by altering the data selection. Poor: study design is hit-or-miss. Bias cannot be corrected after the data recording process. Excellent: study design can be revisit- ed if unanticipated bias effects arise. In this sense, bias could be corrected by altering the data selection. Design flexibility Free text interpretation therefore heavily depends on contextual Structured data/electronic health record (eg, lab values and smoking status) Poor: context is essentially absent as a priori interpretation is an integral part of recording data in case record forms. Depends on implementation. Context may be lost because of predetermined categorization. Poor: information is mostly useful for case-specific usage by humans. May require text mining/text retrieval to convert to a machine-readable format. Excellent: data can be parsed or direct- ly used by computers. Excellent: data can be parsed or direct- ly used by computers. Good: lab values can be converted using reference values. Structured data, such as smoking and hyperten- sive status, can be reformatted for in- teroperability. Excellent: trial data are usually collect- ed using a standardized protocol, al- lowing for interoperability between institutions. Limited: trials are designed and con- ducted for one specific research question. Availability for reuse/general applicability Algorithm Due to this rather artificial way of collecting data as compared with clinical care, research databases and registries are unsuited for the creation of broadly applicable CDS systems using increasingly complex models. Moreover, CDS systems preferably apply information that is already available to the care provider to aid in the clinical process without impeding it by requiring the collection of various additional data. Data from EHRs contain real-world data from clinical practice. EHR-based datasets are, therefore, more suitable for CDS system development. At the same time, EHR systems were designed as a virtual patient chart and not necessarily for reuse of the data they capture. As such, turning them into valuable EHR-based datasets takes careful and skilled data processing. For example, EHR data require more data cleaning (eg, how to handle not at random missing data—also a prime example of where clinical expert knowledge plays a vital role), careful assessment of informed presence bias (ie, acknowledgment of the bias introduced by the medical process), and decisions are to be made about how certain variables are derived from often unstructured data such as free text in EHR systems or clinical notes (eg, define diabetes mellitus and define hypertension) [9]. As only health care professionals themselves know about these inherent biases of working in an EHR, expert knowledge is indispensable. Table 2 compares different types of clinical data on a number of points important to CDS development. The Utrecht Patient Oriented Database in the University Medical Center Utrecht, the Netherlands, is an example of a routinely updated EHR-based database, containing data from multiple hospital sources of about 2.3 million patients (Multimedia Appendix 1, [10]). Utrecht Patient Oriented Database is curated by clinicians who use their expert knowledge in the design of the database to counter the known biases that are inherent to EHR data. Furthermore, they assist their clinical colleagues in transforming relevant data into meaningful variables to answer clinically relevant research questions and to develop CDS systems. After the selection of the right data to develop the CDS system, the next phase is to develop a model (ie, the recipe that describes the relationship between variables and outcome in the data) by using an algorithm (a predetermined computational method to derive such a recipe from the data). Depending on the complexity of the modeling task, model development usually contains a phase of model training and phase of model validation. Algorithm In the training phase, a model that best fits the data (ie, makes the best predictions on the training data) is developed, and in the validation phase, tests are carried out to check whether the model is correct (ie, generalizes to the population). What constitutes a good prediction is dependent on the (clinical) research question (ie, identify all positive diagnoses at any cost or find a trade-off between cost and efficacy). It is common practice to test the model on a new dataset in the validation phase. This can be a previously unseen part of the total dataset or an entirely new dataset. Although modeling and algorithm development are not the natural habitat for most health care professionals, their knowledge and input are invaluable in this phase. Data Sources Widely used datasets for CDS systems include clinical trials and medical registries. Data collected within trials are of importance for a predefined research question. They are usually of high quality, may be stored in great detail, and are often richly annotated with expert knowledge (diagnostic codes and predetermined disease severity classifications). Medical registries are developed for quality control and research purposes. They are used to record a predefined limited number of variables for a specific group of patients, often focusing on particular conditions and diagnoses. Careful maintenance of research databases and registries allows for the collection of data from patients in a clean and systematic way according to protocol, preventing missingness and loss to follow-up as much as possible. However, because of their restrictive nature, research datasets and medical registries discard valuable contextual information, such as free-text notes, about included patients. Therefore, they show a limited, predefined scope of the patient’s condition. Furthermore, women and minorities are underrepresented in research datasets, and patients who are included can suffer from the Hawthorne effect (ie, altered behavior because of the fact that one is a study subject) [8]. The concept of research datasets and medical registries does not allow for flexibility in study design; the decision on what information to collect (and in what way) is single and final. http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al of data can carry value that only a health care professional is able to acknowledge, and the same holds true for outliers. Moreover, information beyond the scope cannot be added without considerable effort at a later moment (if anonymization or informed consent regulations do not prevent this at all). Simple Models In simple models, the input of expert knowledge of health care professionals is well established. As mentioned before, the simplest form of decision models is if-then rules. Examples of such models include laboratory reference values based on statistical distributions of patient measurements (eg, if fasting glucose >11 mmol/L, then the patient probably has diabetes mellitus), medical risk scores (eg, if Glasgow Coma Scale is lower than 9, then consider intubating the patient), rule-based warnings for medication (eg, if the patient has impaired kidney function, then do not allow prescription for metformin), and alarms on the intensive care when vital sign thresholds are violated (eg, sound an alarm if saturation levels drop below 95%) [11]. When building these simple models into CDS systems, the thresholds and reference values need to be provided by health care professionals. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 4 (page number not for citation purposes) Modeling Characteristics Although all models remain specific for a given question (eg, what decision to support), building a CDS system is not a static process. It often includes rounds of major and minor changes of variables included and algorithmic fine-tuning. Moreover, some algorithms are never really finished. These algorithms are called self-learning and are designed to incorporate newly acquired data over time into their modeling processes. One of the reasons one may want to adapt an algorithm is spurious associations. Agnostic analyses that do not incorporate current understanding are prone to implement clinically irrelevant or even false associations with potentially deleterious consequences [17]. Outcomes of CDS models, therefore, need critical appraisal from experts regularly [17,18]. Unsupervised learning algorithms aim to uncover regularities in data without being guided by a prelabeling of the data (ie, clustering algorithms). The scope of this technique is often to discover novel subgroups within data and populations [12,14]. This approach is useful when information on the characteristics needed to discriminate between patients and controls is not yet available, or when one aims to find starting points for more fundamental scientific research. Therefore, this approach is usually used to find novel patterns in the data instead of making predictions and is thus generally more exploratory in nature. An advantage is, thus, that it allows for hypothesis-free or agnostic detection of patterns even when expert knowledge on the difference between subgroups is missing. Nevertheless, unsupervised systems can still profit from expert knowledge in the modeling process, as for example, clustering algorithms often require the user to preset the desired number of clusters, a decision that may be based on evidence of a known pattern in the population of interest. As health care professionals are responsible for the decisions they make, they highly value transparency of a model’s decision process and its development [19]. Whether or not the respective variables are shared with the health care professionals can be up for debate, as sharing of certain variables may lead to undesirable side effects. For example, an algorithm that states that a certain keyword in a patient history carries value as a warning, one may not want the health care professional to know this keyword to prevent it from being stated just to indicate a warning. A CDS system can be transparent to a greater or lesser extent. Complex Models Before the data can be used to build a model, they need to be preprocessed. Preprocessing steps define variables from raw data that a model can use. During preprocessing steps, the expert knowledge of health care professionals is important to derive meaningful variables and values from the data. For example, disease activity variables need to be constructed because research guidelines and accompanying questionnaires are not regularly applied in clinical care. Furthermore, health care professionals may direct data scientists away from composite endpoints (eg, a patient has a 50% increased risk of pneumonia, pulmonary embolism, or chronic obstructive pulmonary disease) as they are less useful for CDS than specific endpoints that require specific actions (eg, a 50% increased risk of pneumococcal pneumonia). Moreover, the extraction of features from the data, such as differences in laboratory values over time, requires expert knowledge to determine appropriate time windows. Although the accuracy of algorithms generally increases if missing values and outliers are removed, the absence These traditional models and clinical scores are generally straightforward (Apgar score and Glasgow coma scale) to make them easily actionable, even in stressful situations. The beauty of their simplicity has ensured their broad application, but their sensitivity and specificity are unavoidably limited and usually include a substantial gray area. Moreover, most current models are based on regression or correlation measures that are less able to capture complex relationships in the data. The availability of machine learning offers novel approaches for developing medical models and risk scores. Machine learning refers to a group of statistical techniques that can be used to discern even complex patterns or regularities in data. They do so through an iterative process (in other words, the patterns are learned, hence machine learning) and produce a prediction model based on the learned patterns, which can then be incorporated in clinical support tools. [12,13]. In this complex http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al recognize patterns in complex datasets that can subsequently be used by (supervised) machine learning models, for example, using clusters of a disease as outcome variables. In some experiments, deep learning methods have been shown to have superior prediction accuracy compared with other methods [15]. With EHRs as input, deep learning can improve prediction performance in modeling tasks that can be used for CDS [16]. Complex Models Deep nets have, for example, already been proven useful in the computer-aided assessment and interpretation of medical images [4]. However, the rule of thumb garbage in is garbage out also applies when it comes to deep learning. If death is included in a model to predict readmission, it will probably come up as the most predictive variable, yet this might not be the actionable insight a health care professional is looking for. Guidance of health care professionals is needed when applying machine learning in the medical field. type of modeling, input of expert knowledge from health care professionals may seem less obvious. However, model development is not a neutral process and even the values of health care professionals may be of additional benefit [7]. Machine learning algorithms can be roughly divided into 2 classes: (1) supervised learning algorithms that make use of prior (expert) knowledge about outcomes to guide the process and (2) unsupervised learning algorithms that aim to discover data patterns irrespective of model outcome. Input data for supervised learning algorithms need to be labeled and selected manually (eg, positive/negative diagnosis, benign/malign, and concentration of inﬂammatory marker X) before modeling, and these data then constitute the outcome variable to predict for new cases. In other words, supervised learning systems rely heavily on expert knowledge [12,14]. Supervision is not only needed for the prelabeling of cases and noncases but also for statistical and methodological choices. Such choices include, for example, the choice of which algorithm to use and whether to normalize/standardize the data, and more algorithm-specific choices, such as the number of layers and nodes in a neural network or number of splits in a decision tree. Making appropriate choices on these aspects requires input from data scientists and medical scientists alike and will significantly affect the validity of the model. When the input variables in supervised models are selected by health care professionals and are based on prior knowledge and scientific evidence, supervised machine learning models may provide a safe ground for decision tools. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 5 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ Decision Support Although such experience-based recommendations remain useful, the most important evidence for the usefulness of a CDS system that will influence acceptation by health care professionals will be its ability to influence clinical outcomes. Unfortunately, the evidence for CDS systems’ frequently purported advantages over old-fashioned clinical decision in improving clinical outcome, workload, and economic cost is scarce. However, CDS systems have been shown to improve health care processes and are the best way to decrease unnecessary clinical testing [27]. Implementation and use of a CDS system consist of multiple steps, including presenting the algorithm output in a specific way, interpretation by the health care professional, and eventually, the medical decision that is made. A CDS system is not a bare model producing just an output (eg, 65%). Almost always, it contains some level of interpretation. For example, laboratory measurements are often displayed in black, red, or blue to indicate whether they fall within or are higher/lower than a reference range. Risk percentages may be accompanied by a traffic light coloring scheme, indicating risk compared with a standard disease course. These manners of presentation (eg, how to report variables and what kind of user interface) are probably the most intuitive place to integrate the expert knowledge of a health care professional. Indeed, health care professionals and patients are often included in the user experience or user interface design phases to discuss implementation. However, this is frequently perceived too late in the CDS development process, and therefore, may yield an opposite effect. Acceptation of CDS by health care professionals depends on the degree to which they feel autonomous in their decision making. Rather than choosing colors for a user interface, being part of the development process, identifying the appropriate data, discussing model design, and validation may help health care professionals to feel in control in the midst of forces that are transforming daily clinical practice. A supportive organization with inspiring leadership encouraging involvement of health care professionals in the development of CDS systems stimulates this transformation. Furthermore, model results have to be interpreted in a specific medical context before the CDS system can provide the actual tailored CDS and lead to action. This step is usually supervised by health care professionals. Modeling Characteristics A CDS system that contains an algorithm that is too complex to comprehend can result in a so-called black box situation, where it is difficult or even impossible for a human brain to understand how the prediction model works. This renders validation of these black box algorithms extremely important. Unfortunately, there is a trade-off between attainable model complexity and model interpretability [20]. The opportunities that more complex models may provide should not be underestimated. To fully benefit from complex data and incorporate it into clinical practice, health care professionals may need to accept that the ultimate goal of thoroughly clinically validated predictive models in CDS systems may not be to be fully and completely interpretable or transparent but rather to be useful to a clinical purpose and influence patient outcomes. The process of model development, including choices that are made based on expert knowledge of both health care professional and data scientist, should always be as transparent as possible for all parties involved. Transparency of the A group of supervised and unsupervised techniques that is currently the state-of-the-art in machine learning is called deep learning. These techniques often involve artificial neural networks and attempt to learn increasingly deep representations of associations in the data. Deep neural networks (deep nets) are capable of automatically determining how to represent the input data in the best way for the question at hand. Theoretically, deep nets do not even require manual data preparation. In essence, the only requirements are to standardize numeric data and to encode categorical data into a numerical format interpretable by the algorithm. Deep learning is often used to http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al basis to make them meaningful as opposed to being perceived as a nuisance has been shown to improve compliance in critical care [11]. Incorporation of the multitude of emerging CDS systems into clinical practice needs streamlining and thorough knowledge of the medical process. Health care professionals should take initiative to lead interprofessional teams, deciding how and when to report CDS. development process may ultimately be critical for acceptance of CDS systems in clinical practice. Moreover, depending on the complexity of the algorithm, internal (ie, on the same data) and external (ie, on other data) validation steps are vital before patients are exposed to the output. Decision Support For instance, a cardiologist will double check the automated interpretation of the electrocardiogram (ECG) machine, and it is very likely that an eye specialist will supervise Google’s new diabetic retinopathy algorithm before any treatment is started [4]. Supervision of interpretation does not necessarily need a thorough understanding of the algorithm itself. Rather, it is the human intervention of integrating the contextual knowledge of the health care professional and, even more importantly, the patient’s wishes, before acting upon the algorithm’s output. Modeling Characteristics To what extent CDS systems need to be assessed as medical devices, according to their intended use, is still up for debate [21,22]. If-then CDS systems and CDS systems written to combine data into a visually attractive interface may be excluded from extensive clinical research but still need quality checks, regular revisiting of the algorithm, and piloting in clinical practice to ensure the right information is displayed for the right patient (does the algorithm take the most recent value from the table and did column names change). Furthermore, scientific evidence on validity and added value of the CDS system likely increases its use by health care professionals. Fortunately, such research is currently gaining traction in the medical community [23]. Unfortunately, systematic scientific evidence outlining what requirements a useful CDS system should meet is missing [24] and has been replaced by more or less anecdotic or empirical recommendations for many years. The Ten Commandments of Clinical Decision Support [25] lists factors as speed, anticipation of information need, integration into the workflow, or general ease-of-use type of advice in alerts. Moreover, negative advice, for example, an advice not to perform or order a particular test, is rarely accepted when no alternative is suggested, and the method of alert presentation has been found to be crucial to alert compliance [26]. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 6 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ Bezemer et al Still, the decision to implement a pacemaker and monitoring and tweaking its action and settings during follow-up are a doctor’s job, including contextual expert knowledge and the patient’s own preferences into the CDS system. Therefore, clinical reasoning is unlikely to ever be completely replaced by fully automated decision making through machine learning without human intervention. Even when expert knowledge is already embedded in medical data, and when the CDS system is clinically validated, the need for health care professionals to navigate the intricacies of incorporating expert knowledge in the ultimate clinical decision must not be underestimated. This includes implicit knowledge or gut feeling for which computability is limited [29]. Moreover, the final decision (how) to use the CDS is up to the health care professional and their patient. Importantly, the decision of how to respond to a CDS system is a moral one, and moral considerations regarding when to treat or not to treat are the expertise of human beings rather than that of AI systems. A CDS system cannot decide whether primum non nocere applies to a specific situation as harm and good and quality of life depend on personal judgement, context, and preferences of human beings. Some patients may be willing to take a risk that others would not, including application of a CDS system with a black box algorithm to their specific case. This way, cultural difference may indicate the need for locally tweaked systems. People, whether patients or their loved ones, should participate in shared decision making, tailoring the usage and outcomes of CDS systems to their wishes. What is best for the patient depends on more than just the output of a CDS system. In conclusion, it is of paramount importance that health care professionals initiate and guide the development and implementation of CDS in clinical care, as opposed to waiting to be overwhelmed by current technological advancements. Most data scientists are not medical experts, and vice versa. Therefore, data scientists and health care professionals should team up in an interprofessional fashion, preferably also including patients. Data scientists who enthusiastically welcome recent innovations in AI pose a bold claim and carry the burden of proof to equip health care with suitable CDS tools. Bezemer et al regulations, health care professionals cannot easily be replaced by AI. The European General Data Protection Regulations state “The data subject shall have the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her.” This regulation appears tailored to the medical profession. Health care professionals (but basically everybody processing data) have a responsibility to guard their patients against irresponsible implementation of data-driven technologies. This especially holds true for self-learning algorithms that self-adapt to the patient population without human intervention, which may autonomously change considerably over time (so-called algorithmic drift). What is the meaning of human intervention in this sense, if the health care professional has no insight at all into the opaque model? In this light, how can health care professionals still justify their actions? How does this translate to accountability? professionals want the best for their patients, they may even develop a sense of moral obligation to embrace strategies that unravel data complexities beyond their comprehension, as relying on methods that do not use the data to their full potential leads to potentially unused value for their patients. At the same time, both data scientists and health care professionals should be alert to cognitive biases provided by pre-existing expert knowledge. Indeed, interprofessional CDS development teams should be as inclusive as possible, as the values and preferences of the people involved influence the underlying model [28]. The hallmark of biology, variability, is complex to capture in a static algorithm, and a medical decision is not based on objective single data points but on subjective, context-sensitive longitudinal observations made by health care professionals during patient contact. This has consequences for the acceptance of autonomous CDS systems. Single measurements are not likely to lead to an acceptable autonomous action by a CDS system except for when doing nothing is likely to cause more harm (eg, an automated defibrillator that decides whether electric cardioversion is indicated based on ECG input and applies the appropriate therapy itself). Other accepted autonomous CDS systems, such as pacemakers or insulin pumps, gather continuous data, and thereby, learn and improve their efficacy for the individual patient. As longitudinal data collected by sensors inside and outside the hospital are becoming more ubiquitous, the value and applicability of accepted autonomous CDS systems enriched with these data are likely to improve. Acknowledgments g TB is funded through the Netherlands CardioVascular Research Initiative CVON2017-22 ARGUS. This project is financed by the PPP Allowance made available by Top Sector Life Sciences & Health to Hartstichting to stimulate public-private partnerships. Bezemer et al Once health care professionals can be convinced of the added benefit of CDS for their patients, they may acknowledge the necessity and value of data collection, interpretation, and curation, so they may embrace their expanding role and further evolve from doctor knows best to doctor does best. Therefore, most CDS systems do not aim to replace health care professionals but are designed to support them. For example, the recent algorithms in the field of medical imaging preprocess data and take over tedious and simple tasks so that radiologists and pathologists can focus on more complex cases, acting more creatively. Given the recent developments in data protection Discussion With the rise of machine learning, and especially deep learning in CDS systems, it is perhaps tempting to let IT and data experts build CDS systems, redirecting health care professionals to merely gathering data. However, we have shown that human(e) health care professionals are still of paramount importance, as all phases of development and use of a CDS system requires the extensive expert knowledge of health care professionals. Health care professionals should not just be involved in implementing the CDS system into clinical practice but should be part of an interprofessional CDS development team from the start, initiating and guiding development through clinical demand and expert knowledge. They bring in the clinical decision they want the CDS system to support and help to understand the context in which variables are collected during routine care. Their interpretation is vital in extracting relevant variables from raw data and in avoiding the implementation of spurious associations in CDS systems. Moreover, as health care Most CDS systems do not (yet) act autonomously, so they need the attention of a health care professional to be effective. The highly technologically supported intensive care units provide ample evidence that getting the attention of a health care professional can be a challenge, as too many alarms can lead to alarm fatigue. Tweaking alert settings on an individual patient http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al g TB is funded through the Netherlands CardioVascular Research Initiative CVON2017-22 ARGUS. This project is financed by the PPP Allowance made available by Top Sector Life Sciences & Health to Hartstichting to stimulate public-private partnerships. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 7 (page number not for citation purposes) Conflicts of Interest None declared. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 7 (page number not for citation purposes) J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 7 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al y y c;104(12):e12-e22 [FREE Full text] [doi: 10.2105/AJPH.2014.302164] [Medline: 25322302] Linking laboratory and medication data: new opportunities for pharmacoepidemiological research Clin Chem Lab Med 2007;45(1):13-19 [doi: 9. Agniel D, Kohane IS, Weber GM. Biases in electronic health record data due to processes within the healthcare system: retrospective observational study. Br Med J 2018 Apr 30;361:k1479 [FREE Full text] [Medline: 29712648] 9 g , S, G p w y retrospective observational study. Br Med J 2018 Apr 30;361:k1479 [FREE Full text] [Medline: 29712648] 10. ten Berg MJ, Huisman A, van den Bemt PM, Schobben AF, Egberts AC, van Solinge WW. Linking laboratory and medication data: new opportunities for pharmacoepidemiological research. Clin Chem Lab Med 2007;45(1):13-19. [doi: 10.1515/CCLM.2007.009] [Medline: 17243908] 10. ten Berg MJ, Huisman A, van den Bemt PM, Schobben AF, Egberts AC, van Solinge WW. Linking laboratory and medication data: new opportunities for pharmacoepidemiological research. Clin Chem Lab Med 2007;45(1):13-19. [doi: 10.1515/CCLM.2007.009] [Medline: 17243908] 11. Graham KC, Cvach M. Monitor alarm fatigue: standardizing use of physiological monitors and decreasing nuisance alarms. Am J Crit Care 2010 Jan;19(1):28-34 [FREE Full text] [doi: 10.4037/ajcc2010651] [Medline: 20045845] 12. Hardin JM, Chhieng JC, Ozaydin B. Data Mining and Clinical Decision Support Systems. In: Berner E, editor. Clinical Decision Support System. New York: Springer; 2007. 13. Berner ES, editor. Clinical Decision Support Systems: Theory and Practice 3rd. New York: Spring 14. Mohri M, Rostamizadeh A, Talwalkar A. Foundations of Machine Learning. Cambridge: The MIT Press; 2012. 15. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015 May 28;521(7553):436-444. [doi: 10.1038/nature14539] [Medline: 26017442] 16. Miotto R, Li L, Kidd BA, Dudley JT. Deep patient: an unsupervised representation to predict the future of patients from the electronic health records. Sci Rep 2016 Dec 17;6:26094. [doi: 10.1038/srep26094] [Medline: 27185194] 17. Kappen TH, Vergouwe Y, van Klei WA, van Wolfswinkel L, Kalkman CJ, Moons KG. Adaptation of clinical prediction models for application in local settings. Med Decis Making 2012;32(3):E1-10 [FREE Full text] [doi: 10.1177/0272989X12439755] [Medline: 22427369] 18. Kappen TH, Peelen LM. Prediction models: the right tool for the right problem. Curr Opin Anaesthesiol 2016 Dec;29(6):717-726. [doi: 10.1097/ACO.0000000000000386] [Medline: 27536997] 19. Kappen TH, van Loon K, Kappen MA, van Wolfswinkel L, Vergouwe Y, van Klei WA, et al. Barriers and facilitators perceived by physicians when using prediction models in practice. J Clin Epidemiol 2016 Feb;70:136-145 [FREE Full text] [doi: 10.1016/j.jclinepi.2015.09.008] [Medline: 26399905] 20. Kuhn M, Johnson K. Applied Predictive Modeling. New York: Springer; 2013:600. 21. Multimedia Appendix 1 This table shows a number of indicative characteristics of the Utrecht Patient Oriented Database (UPOD) as of February 2018. This table shows a number of indicative characteristics of the Utrecht Patient Oriented Dat [PDF File (Adobe PDF File), 11KB-Multimedia Appendix 1] References 1. Moja L, Kwag KH, Lytras T, Bertizzolo L, Brandt L, Pecoraro V, et al. Effectiveness of computerized decision support systems linked to electronic health records: a systematic review and meta-analysis. Am J Public Health 2014 D 104(12) 12 22 [FREE F ll t t] [d i 10 2105/AJPH 2014 302164] [M dli 25322302] 1. Moja L, Kwag KH, Lytras T, Bertizzolo L, Brandt L, Pecoraro V, et al. Effectiveness of computerized decision support systems linked to electronic health records: a systematic review and meta-analysis. Am J Public Health 2014 Dec;104(12):e12-e22 [FREE Full text] [doi: 10.2105/AJPH.2014.302164] [Medline: 25322302] y y y Dec;104(12):e12-e22 [FREE Full text] [doi: 10.2105/AJPH.2014.302164] [Medline: 25322302] y y c;104(12):e12-e22 [FREE Full text] [doi: 10.2105/AJPH.2014.302164] [Medline: 25322302] 2. Gilbert FJ, Astley SM, Gillan MG, Agbaje OF, Wallis MG, James J, CADET II Group. Single reading with computer-aided detection for screening mammography. N Engl J Med 2008 Oct 16;359(16):1675-1684. [doi: 10.1056/NEJMoa0803545] [Medline: 18832239] 3. Cruz-Roa A, González F, Galaro J, Judkins AR, Ellison D, Baccon J, et al. A visual latent semantic approach for automatic analysis and interpretation of anaplastic medulloblastoma virtual slides. Med Image Comput Comput Assist Interv 2012;15(Pt 1):157-164. [Medline: 23285547] 4. Gulshan V, Peng L, Coram M, Stumpe MC, Wu D, Narayanaswamy A, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. J Am Med Assoc 2016 Dec 13;316(22):2402-2410. [doi: 10.1001/jama.2016.17216] [Medline: 27898976] 5. Rajkomar A, Oren E, Chen K, Dai AM, Hajaj N, Hardt M, et al. Scalable and accurate deep learning with electronic health records. npj Digital Med 2018 May 8;1(1):1-1 [FREE Full text] [doi: 10.1038/s41746-018-0029-1] 5. Rajkomar A, Oren E, Chen K, Dai AM, Hajaj N, Hardt M, et al. Scalable and accurate deep learning with electronic health records. npj Digital Med 2018 May 8;1(1):1-1 [FREE Full text] [doi: 10.1038/s41746-018-0029-1] 6. Towle A, Brown H, Hofley C, Kerston RP, Lyons H, Walsh C. The expert patient as teacher: an interprofessional health mentors programme Clin Teach 2014 Jul;11(4):301-306 [doi: 10 1111/tct 12222] [Medline: 24917101] j , , , , j j , , p g records. npj Digital Med 2018 May 8;1(1):1-1 [FREE Full text] [doi: 10.1038/s41746-018-0029-1] 6. Towle A, Brown H, Hofley C, Kerston RP, Lyons H, Walsh C. The expert patient as teacher: an interprofessional health mentors programme. Clin Teach 2014 Jul;11(4):301-306. [doi: 10.1111/tct.12222] [Medline: 24917101] 7. Mittelstadt BD, Allo P, Taddeo M, Wachter S, Floridi L. The ethics of algorithms: Mapping the debate. Big Data & Society 2016 Dec 19;3(2). [doi: 10.1177/2053951716679679] 8. Fernald DH, Coombs L, DeAlleaume L, West D, Parnes B. An assessment of the Hawthorne Effect in practice-based research. J Am Board Fam Med 2012;25(1):83-86 [FREE Full text] [doi: 10.3122/jabfm.2012.01.110019] [Medline: 22218628] 9. Agniel D, Kohane IS, Weber GM. Biases in electronic health record data due to processes within the healthcare system: retrospective observational study. Br Med J 2018 Apr 30;361:k1479 [FREE Full text] [Medline: 29712648] 10. ten Berg MJ, Huisman A, van den Bemt PM, Schobben AF, Egberts AC, van Solinge WW. JOURNAL OF MEDICAL INTERNET RESEARCH Bezemer et al 24. Lobach DF. The road to effective clinical decision support: are we there yet? B Med J 2013 Mar 13;346:f1616. [doi: 10.1136/bmj.f1616] [Medline: 23487164] 25. Bates DW, Kuperman GJ, Wang S, Gandhi T, Kittler A, Volk L, et al. Ten commandments for effective clinical decision support: making the practice of evidence-based medicine a reality. J Am Med Inform Assoc 2003;10(6):523-530 [FREE Full text] [doi: 10.1197/jamia.M1370] [Medline: 12925543] 25. Bates DW, Kuperman GJ, Wang S, Gandhi T, Kittler A, Volk L, et al. Ten commandments for effective clinical decision support: making the practice of evidence-based medicine a reality. J Am Med Inform Assoc 2003;10(6):523-530 [FREE Full text] [doi: 10.1197/jamia.M1370] [Medline: 12925543] 26. Scheepers-Hoeks AM, Grouls RJ, Neef C, Ackerman EW, Korsten EH. Physicians' responses to clinical decision support on an intensive care unit--comparison of four different alerting methods. Artif Intell Med 2013 Sep;59(1):33-38. [doi: 10.1016/j.artmed.2013.05.002] [Medline: 23746663] 26. Scheepers-Hoeks AM, Grouls RJ, Neef C, Ackerman EW, Korsten EH. Physicians' responses to clinical decision support on an intensive care unit--comparison of four different alerting methods. Artif Intell Med 2013 Sep;59(1):33-38. [doi: 10.1016/j.artmed.2013.05.002] [Medline: 23746663] 27. Bright TJ, Wong A, Dhurjati R, Bristow E, Bastian L, Coeytaux RR, et al. Effect of clinical decision-support systems: a systematic review. Ann Intern Med 2012 Jul 3;157(1):29-43. [doi: 10.7326/0003-4819-157-1-201207030-00450] [Medline: 22751758] 27. Bright TJ, Wong A, Dhurjati R, Bristow E, Bastian L, Coeytaux RR, et al. Effect of clinical decision-support systems: a systematic review. Ann Intern Med 2012 Jul 3;157(1):29-43. [doi: 10.7326/0003-4819-157-1-201207030-00450] [Medline: 22751758] ] 28. Macnish K. Unblinking eyes: the ethics of automating surveillance. Ethics Inf Technol 2012 Mar 7;14(2):151-167. [doi: 10.1007/s10676-012-9291-0] 28. Macnish K. Unblinking eyes: the ethics of automating surveillance. Ethics Inf Technol 2012 Mar 7;14(2):151-167. [doi: 10.1007/s10676-012-9291-0] 29. Morek R. The regulatory framework for online dispute resolution. U Tol L Rev 2006:163-192 [FREE Full text] y y c;104(12):e12-e22 [FREE Full text] [doi: 10.2105/AJPH.2014.302164] [Medline: 25322302] The European Parliament and the Council of the European Union. GDPR URL: https://eugdpr.org/ [accessed 2019-01-29] [WebCite Cache ID 75n8Qm2h9] Q 22. US Food & Drug Administration. 2016. Medical Device Reporting for Manufacturers (MDR) URL: http://cat.ocw.uci.edu/ media/AR01/06011/mdr.pdf[WebCite Cache ID 71DtXDpNP] 22. US Food & Drug Administration. 2016. Medical Device Reporting media/AR01/06011/mdr.pdf[WebCite Cache ID 71DtXDpNP] 23. Varghese J, Kleine M, Gessner SI, Sandmann S, Dugas M. Effects of computerized decision support system implementations on patient outcomes in inpatient care: a systematic review. J Am Med Inform Assoc 2018 May 1;25(5):593-602. [doi: 10.1093/jamia/ocx100] [Medline: 29036406] J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 8 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX JOURNAL OF MEDICAL INTERNET RESEARCH J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 9 (page number not for citation purposes) Abbreviations Please cite as: Bezemer T, de Groot MCH, Blasse E, ten Berg MJ, Kappen TH, Bredenoord AL, van Solinge WW, Hoefer IE, Haitjema S A Human(e) Factor in Clinical Decision Support Systems J Med Internet Res 2019;21(3):e11732 URL: http://www.jmir.org/2019/3/e11732/ doi: 10.2196/11732 PMID: 30888324 ©Tim Bezemer, Mark CH de Groot, Enja Blasse, Maarten J ten Berg, Teus H Kappen, Annelien L Bredenoord, Wouter W van Solinge, Imo E Hoefer, Saskia Haitjema. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 19.03.2019. This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included. J Med Internet Res 2019 \| vol. 21 \| iss. 3 \| e11732 \| p. 9 (page number not for citation purposes) http://www.jmir.org/2019/3/e11732/ XSL•FO RenderX
https://openalex.org/W2943005149	https://ojs.ifes.edu.br/index.php/ric/article/download/447/394	Portuguese	null	RECONFIGURAÇÃO INTELIGENTE DE REDES DE DISTRIBUIÇÃO DE ENERGIA ELÉTRICA CONSIDERANDO O INDICADOR FIC POR MEIO DO ALGORITMO BUSCA TABU	Revista Ifes Ciência	2,019	cc-by	6,939	SMART RECONFIGURATION OF ELECTRICAL DISTRIBUTION NETWORKS CONSIDERING THE FIC INDEX USING THE TABU SEARCH ALGORITHM Vitor Reuter Carréra Torres1 Jussara Farias Fardin2 Lucas Frizera Encarnação3 Clainer Bravin Donadel4* Rodrigo Fiorotti5 Adan Lucio Pereira6 Augusto César Rueda Medina7 Vitor Reuter Carréra Torres1 Jussara Farias Fardin2 Lucas Frizera Encarnação3 Clainer Bravin Donadel4* Rodrigo Fiorotti5 Adan Lucio Pereira6 Augusto César Rueda Medina7 1Universidade Federal do Espírito Santo. E-mail: vitoreuter@hotmail.com. 2Universidade Federal do Espírito Santo. E-mail: jussara.fardin@ufes.br. 3Universidade Federal do Espírito Santo. E-mail: lucas@ele.ufes.br. 4Instituto Federal do Espírito Santo – campus Vitória. E-mail: cdonadel@ifes.edu.br. 5Instituto Federal do Espírito Santo – campus São Mateus. E-mail: rodrigo.fiorotti@ifes.edu.br. 6Universidade Federal do Espírito Santo. E-mail: adanlucio@gmail.com. 7Universidade Federal do Espírito Santo. E-mail: augusto.rueda@ufes.br. *Autor para correspondência Artigo submetido em 25/06/2019, aceito em 19/11/2019 e publicado em 23/12/2019. Resumo: O processo de reconfiguração da rede elétrica visa atender a critérios técnico-econômicos como o número de consumidores atendidos, a redução das perdas, a manutenção dos níveis de tensão e carregamento do sistema em patamares adequados, a redução do número de manobras das chaves, dentre outros. Diversos métodos vêm sendo desenvolvidos a fim de determinar a configuração ótima para o sistema. Entretanto, diversos métodos propostos não consideram indicadores de continuidade do sistema como, por exemplo, a frequência de interrupção individual por unidade consumidora ou ponto de conexão (FIC), os quais podem acarretar multas pelo lado da distribuidora caso seus limites não sejam respeitados. Em outros casos, mesmo incluindo indicadores de continuidade tais como o FIC, os métodos propostos não conseguem evitar o desligamento repetido de unidades consumidores em caso de contingências sucessivas. Desta forma, é apresentada neste trabalho uma nova metodologia para reconfiguração das redes de distribuição de energia elétrica que atenda às restrições operacionais do sistema, incluindo o indicador de continuidade FIC, a fim de evitar que um mesmo conjunto de consumidores seja penalizado repetidamente em caso de desligamentos sucessivos. Para lidar com os múltiplos objetivos e restrições do problema de reconfiguração da rede elétrica de distribuição, a metodologia proposta foi baseada no Tabu Search Algorithm (Algoritmo Busca Tabu). O método proposto foi aplicado à rede IEEE 123 barras, rede padrão de testes estabelecida pelo IEEE (Institute of Electrical and Electronics Engineers), apresentando resultados coerentes com a proposta. 25 v.5 n.2 2019 25 v.5 n.2 2019 25 Palavras-chave: Reconfiguração das redes de distribuição de energia elétrica; sistema de distribuição; Algoritmo Busca Tabu. Palavras-chave: Reconfiguração das redes de distribuição de energia elétrica; sistema de distribuição; Algoritmo Busca Tabu. Abstract: The electrical distribution networks reconfiguration process aims to reach technical and economic criteria such as the number of served consumers, losses reduction, voltage and load adequate levels maintenance, switches changes reduction, among others. Several methods have been developed to determine the optimal system configuration. However, in several cases, the proposed methods do not consider system continuity indexes, such as the individual interruption frequency per consumer unit or connection point (FIC), which may result in penalties if their limits are not respected. In other cases, even including continuity indexes such as the FIC, the proposed methods cannot prevent repeated disconnection of consumer units in case of successive contingencies. Artigo submetido em 25/06/2019, aceito em 19/11/2019 e publicado em 23/12/2019. In this way, this paper presents a new electrical distribution networks reconfiguration methodology that consider the system operating restrictions, including the FIC continuity index, avoiding that the same consumers group is repeatedly penalized in the event of successive contingencies. To address the multiple objectives and constraints of the distribution network reconfiguration problem, the proposed methodology was based on the Tabu Search Algorithm. The proposed method was applied to the IEEE 123 buses network, a standard test network established by the IEEE (Institute of Electrical and Electronics Engineers), presenting results consistent with the proposal. Keywords: Electrical distribution networks reconfiguration; distribution system; Tabu Search Algorithm. 2 METODOLOGIA PROPOSTA Nesta seção é apresentada a metodologia proposta, incluindo os dados da rede utilizada, a descrição da função objetivo e do algoritmo de otimização utilizado. 26 v.5 n.2 2019 v.5 n.2 2019 26 forma eficiente às contingências (ABDELAZIZ, 2017; BORGES et al., 2015; GHASEMI, 2018; LÓPEZ et al., 2016; MOHAMED IMRAN et al., 2014; NGUYEN et al., 2017; NGUYEN et al., 2016; RAJARAM et al., 2015; ZHAI et al., 2018). consumidores, em situações de contingência que impossibilitem o atendimento a todos as unidades consumidoras. O restante deste artigo está organizado da seguinte forma: na Seção 2 a metodologia proposta é apresentada. Na Seção 3 são apresentados os resultados obtidos pela aplicação do método proposto, bem como é apresentada uma discussão sobre sua eficácia. Por fim, na Seção 4, são apresentadas as conclusões deste trabalho. A reconfiguração da rede de distribuição é o processo no qual o estado das chaves é alterado, visando modificar a topologia da rede de forma a respeitar as limitações impostas pelo operador (ALEMOHAMMAD et al., 2015; CHIDANANDAPPA et al., 2015; MOHAMED IMRAN et al., 2014). A alteração da topologia influencia no fluxo de potência da rede, o que impacta nas perdas do sistema, nos níveis de tensão, no carregamento da linha, dentre outros fatores. Os métodos lidam, então, com um problema de otimização não linear que pode conter vários objetivos distintos a serem tratados simultaneamente, de modo que o operador obtenha o desempenho ótimo para a rede (DUAN et al., 2015; ESMAEILIAN e FADAEINEDJAD, 2015). Estes processos de reconfiguração de redes de distribuição de energia elétrica podem ser aplicados de duas formas distintas: a primeira forma visa otimizar o funcionamento da rede elétrica em condições normais de operação, ou seja, quando toda a rede está operante e todos os consumidores estão sendo atendidos. A segunda forma visa buscar a restauração da rede elétrica de distribuição quando diante de contingências que não permitem o atendimento a todas as unidades consumidoras. Neste caso, os métodos propostos de reconfiguração da rede elétrica, mesmo incluindo indicadores de continuidade tais como a frequência de interrupção individual por unidade consumidora ou ponto de conexão (FIC), não conseguem evitar o desligamento repetido de unidades consumidores em caso de contingências sucessivas. Assim, este artigo propõe uma nova metodologia de reconfiguração de redes de distribuição de energia elétrica que permita evitar a penalização do mesmo conjunto de 1 INTRODUÇÃO distribuída (GD) que tornaram a rede ainda mais complexa e versátil, diante das possibilidades de melhorias que acrescentam ao sistema sendo, inclusive, um assunto que estará presente para estudo durante, pelo menos, a próxima década devido a sua relevância para o sistema elétrico (LÓPEZ et al., 2016). O consumo de energia elétrica per capita em qualquer país é um indicativo de seu crescimento e qualidade de vida de sua população (RAJARAM et al., 2015). À medida que o país se desenvolve, a necessidade por uma energia elétrica confiável e de qualidade vai ficando mais em evidência. Para que a energia elétrica seja considerada de boa qualidade, o sistema elétrico deve apresentar as seguintes características (BORGES et al., 2015): continuidade, ou seja, estar sempre disponível para uso; conformidade, atender aos padrões e normas técnicas especificados pelas agências reguladoras; flexibilidade, a fim de se adaptar às contínuas mudanças na estrutura topológica; manutenção, que é a capacidade da rede retornar ao modo de operação e fornecimento de energia o mais rápido possível, no caso de contingência na rede. Em geral, os sistemas de distribuição apresentam altos níveis de corrente e baixos níveis de tensão, o que acarreta perdas significativas de potência e níveis de tensão de baixa qualidade (NGUYEN et al., 2016). Paralelamente a tal situação, com o desenvolvimento das tecnologias de telecomunicação, automação e processamento de sinais digitais, chaves e dispositivos de proteção entre os alimentadores vêm evoluindo de forma a conseguir remotamente controlar e supervisionar o sistema, permitindo que este responda rapidamente às contingências (faltas) para um esquema de restauração automático (LÓPEZ et al., 2016). Diante de tal cenário, diversos métodos voltados para a reconfiguração de redes de distribuição vêm sendo desenvolvidos e aprimorados no intuito de mitigar tais perdas, melhorar os índices de qualidade da rede e responder de A rede de distribuição é o último estágio na entrega da energia elétrica. Ela é responsável por permitir o intercâmbio de energia entre o sistema de transmissão e o consumidor final. Atualmente, ela está ainda mais em evidência devido aos avanços das redes inteligentes (Smart Grids) e da geração v.5 n.2 2019 27 𝐼௦௟௔௖௞ Corrente no início do alimentador; 𝑓𝑝௦௟௔௖௞ Fator de potência no início do alimentador; 𝑉௜ Tensão nos nós de carga; 𝐼௜ Corrente nos nós de carga; 𝑓𝑝௜ Fator de potência nos nós de carga; 𝑛௖௔௥௚௔௦ Quantidade de nós de carga existentes no alimentador; 𝑃𝑟𝑒ç𝑜𝑘𝑊𝑐 Valoração das perdas de potência. 𝐼௦௟௔௖௞ Corrente no início do alimentador; 𝑓𝑝௦௟௔௖௞ Fator de potência no início do alimentador; 𝑉௜ Tensão nos nós de carga; 𝐼௜ Corrente nos nós de carga; 𝑓𝑝௜ Fator de potência nos nós de carga; 𝑛௖௔௥௚௔௦ Quantidade de nós de carga existentes no alimentador; 𝑃𝑟𝑒ç𝑜𝑘𝑊𝑐 Valoração das perdas de potência. Corrente no início do alimentador; abordar os demais objetivos não descritos no objetivo principal. Assim, os objetivos “secundários” irão implicitamente influenciar na escolha de quais elementos irão compor a solução aplicada na função objetivo principal. O enfoque mono-objetivo é um ótimo correspondente às soluções extremas (mínimas ou máximas) da função- objetivo, o que é desejado no problema em questão. A escolha da função objetivo é de suma importância para a modelagem matemática do problema mono-objetivo e, para este trabalho, consiste em minimizar o custo total do ponto de vista da distribuidora de energia elétrica, garantindo que a reconfiguração da rede mantenha a estrutura radial, atenda aos limites legais e técnicos de tensão e corrente, além de permitir considerar a influência do indicador FIC. A função objetivo a ser minimizada é apresentada pela equação (1). b) Custo das cargas não atendidas (𝐶𝑎𝑟𝑔. 𝑛𝑎): é o quanto a distribuidora de energia elétrica deixa de faturar ao rejeitar determinadas cargas, dado pela equação (4). Esse fator estimula o algoritmo a atender, caso seja possível, ao máximo de clientes possíveis. b) Custo das cargas não atendidas (𝐶𝑎𝑟𝑔. 𝑛𝑎): é o quanto a distribuidora de energia elétrica deixa de faturar ao rejeitar determinadas cargas, dado pela equação (4). Esse fator estimula o algoritmo a atender, caso seja possível, ao máximo de clientes possíveis. b) Custo das cargas não atendidas (𝐶𝑎𝑟𝑔. 𝑛𝑎): é o quanto a distribuidora de energia elétrica deixa de faturar ao rejeitar determinadas cargas, dado pela equação (4). Esse fator estimula o algoritmo a atender, caso seja possível, ao máximo de clientes possíveis. 𝑐𝑜𝑠𝑡= 𝑃. 𝑡𝑒𝑐+ 𝐶𝑎𝑟𝑔. 𝑛𝑎 + 𝐿𝑖𝑚. 𝑉+ 𝐿𝑖𝑚. 𝐼 + 𝐹𝐼𝐶 (1) 𝐶𝑎𝑟𝑔. 2.1 FUNÇÃO OBJETIVO O problema de reconfiguração de redes de distribuição de energia elétrica envolve as seguintes variáveis a serem otimizadas:  Atendimento ao maior número de consumidores possível;  Redução das perdas técnicas;  Atender aos limites do indicador FIC. E possui as seguintes restrições operacionais:  Manter a topologia radial do sistema;  Atender à capacidade de carregamento das linhas;  Atender aos limites de tensão normatizados. Desta forma, tem-se um problema multi-objetivo, onde os critérios apresentados anteriormente são conflitantes entre si. Entretanto, quando isso acontece, uma alternativa para a redução do problema é a utilização da técnica de ponderação de pesos, ou seja, vários objetivos são transformados em um único objetivo, criando uma abordagem ao problema mono- objetivo (COELLO et al., 2007). Para isso, é necessário definir uma função objetivo principal e uma função de avaliação que irá 𝑛𝑎= ෍𝑉௜𝐼௜ 𝑥 𝑃𝑟𝑒ç𝑜𝑘𝑊𝑣 ௡௔ ௜ୀଵ (4) (1) (4) Onde: Os elementos que compõem a função objetivo são: 𝑛𝑎 Cargas que não foram atendidas; 𝑛𝑎 Cargas que não foram atendidas; 𝑛𝑎 a) Perdas técnicas (𝑃. 𝑡𝑒𝑐): é a perda de potência ativa (kW) da rede de distribuição, calculada pela equação (2) e pela equação (3). 𝑃𝑟𝑒ç𝑜𝑘𝑊𝑣 Valoração da carga não atendida. c) Penalidade por infringir os limites de tensão (𝐿𝑖𝑚. 𝑉): este indicador é ativado quando a distribuidora de energia elétrica abastece a unidade consumidora; entretanto, o fornecimento de energia elétrica ocorre fora dos limites estabelecidos por regulamentação. Foram adotados, neste trabalho, os limites estabelecidos pelo PRODIST Módulo 8 (BRASIL, 2018) para redes de média tensão, apresentados no Quadro 1. 𝑃. 𝑡𝑒𝑐= 𝑃𝑒𝑟𝑑 𝑥 𝑃𝑟𝑒ç𝑜𝑘𝑊𝑐 (2) (2) 𝑃𝑒𝑟𝑑= 𝑉௦௟௔௖௞𝐼௦௟௔௖௞𝑓𝑝௦௟௔௖௞ − ෍ 𝑉௜𝐼௜𝑓𝑝௜ ௡೎ೌೝ೒ೌೞ ௜ୀଵ (3) O d (3) Onde: 𝑉௦௟௔௖௞ Tensão no nó de início do alimentador; 𝑉௦௟௔௖௞ Tensão no nó de início do alimentador; v.5 n.2 2019 28 Quadro 1: Limites de tensão para redes de distribuição de energia elétrica de média tensão Tensão de Atendimento (TA) Faixa de Variação de Tensão de Leitura (TL) em Relação à Tensão de Referência (TR) Adequada 0,93TR ≤ TL ≤ 1,05TR Precária 0,90TR ≤ TL ≤ 0,93TR Crítica TL < 0,90TR ou TL > 1,05TR Fonte: PRODIST – Módulo 8 (BRASIL, 2018). Quadro 1: Limites de tensão para redes de distribuição de energia elétrica de média t ã abaixo de 0,93TR ou acima de 1,05TR. abaixo de 0,93TR ou acima de 1,05TR. 𝑉𝑎𝑙𝑜𝑟𝑀𝑢𝑙𝑡𝑎𝑇𝑒𝑛𝑠ã𝑜 Penalidade imposta à distribuidora de energia elétrica por desrespeitar os limites de tensão. 𝑉𝑎𝑙𝑜𝑟𝑀𝑢𝑙𝑡𝑎𝑇𝑒𝑛𝑠ã𝑜 Penalidade imposta à distribuidora de energia elétrica por desrespeitar os limites de tensão. d) Penalidade por infringir o carregamento dos condutores (𝐿𝑖𝑚. 𝐼): esta é uma restrição de suma importância, uma vez que, caso seja descumprida, o condutor tende a se romper, o que poderá acarretar em prejuízos para a distribuidora de energia elétrica (por exemplo, possíveis acidentes, manutenção mais custosa e consumidores desenergizados) – equação (5). Dessa forma, na função objetivo está sendo considerada uma penalidade quando o valor fica abaixo de 93% ou acima de 105% da tensão por unidade, ou seja, quando deixa de ser adequada e passar a ser precária ou crítica conforme apresentado na Figura 1. 𝐿𝑖𝑚. 𝐼 = 𝑉𝑎𝑙𝑜𝑟𝑃𝑒𝑛𝑎𝑙𝑖𝑑𝑎𝑑𝑒𝐶𝑜𝑟𝑟𝑒𝑛𝑡𝑒 (5) Onde: 𝑉𝑎𝑙𝑜𝑟 𝑃𝑒𝑛𝑎𝑙𝑖𝑑𝑎𝑑𝑒 𝐶𝑜𝑟𝑟𝑒𝑛𝑡𝑒 Valor de penalidade atribuído a função objetivo para os casos em que houve violação dos limites de corrente. 𝐿𝑖𝑚. 𝐼 = 𝑉𝑎𝑙𝑜𝑟𝑃𝑒𝑛𝑎𝑙𝑖𝑑𝑎𝑑𝑒𝐶𝑜𝑟𝑟𝑒𝑛𝑡𝑒 (5) (5) Onde: 𝑉𝑎𝑙𝑜𝑟 𝑃𝑒𝑛𝑎𝑙𝑖𝑑𝑎𝑑𝑒 𝐶𝑜𝑟𝑟𝑒𝑛𝑡𝑒 Onde: 𝑉𝑎𝑙𝑜𝑟 Valor de penalidade atribuído a função objetivo para os casos em que houve violação dos limites de corrente. Figura 1: Fluxograma de Determinação da Penalidade por Violação dos Limites de Tensão Fonte: Autores. Figura 1: Fluxograma de Determinação da Penalidade por Violação dos Limites de Tensão Figura 1: Fluxograma de Determinação da Penalidade por Violação dos Limites de Tensão Fonte: Autores. Onde: 𝑛𝑐 Número de nós com carga; 𝑐𝑜𝑛𝑡 Quantidade de cargas que estão sendo atendidas com tensão Vale ressaltar que não há um valor de penalidade previsto na legislação para este caso. Cabe a distribuidora de energia elétrica avaliar os riscos e estabelecer o valor de penalidade correspondente. Neste trabalho, foi estabelecido o valor de 105 para tal penalidade, a fim de se evitar condutores operando em sobrecarga. e) Penalidade por infringir o indicador de continuidade FIC (𝐹𝐼𝐶): FIC representa a frequência de interrupção individual por unidade consumidora ou ponto de conexão, expressa em números de interrupção (BRASIL, 2018). Os limites para o indicador FIC são dependentes do tipo de unidade consumidora, do nível de tensão de atendimento e do valor dos indicadores de v.5 n.2 2019 29 um limite para o indicador de continuidade FIC de 2,06 ocorrências/mês. O fluxograma de aplicação da penalidade por violação do indicador de continuidade FIC é apresentado na Figura 2. conjunto (DEC ou FEC). Neste trabalho foi adotado o valor dos indicadores de conjunto igual a 5 (valor típico para a Região Metropolitana da Grande Vitória), conforme é apresentado na Tabela 1, implicando em Tabela 1: Limites de Continuidade por Unidade Consumidora - parte Faixa de variação dos Limites Anuais de Indicadores de Continuidade dos Conjuntos (DEC ou FEC) Limite de Continuidade por Unidade Consumidora ou Central Geradora Unidades Consumidoras ou Centrais Geradoras situadas em áreas urbanas com Faixa de Tensão Contratada: 1 kV < Tensão < 69 kV DIC (horas) FIC (interrupções) DMIC (horas) Anual Trim. Mensal Anual Trim. Mensal Mensal 1 11,25 5,62 2,81 6,48 3,24 1,62 2,36 2 11,68 5,84 2,92 6,93 3,46 1,73 2,39 3 12,12 6,06 3,03 7,37 3,68 1,84 2,41 4 12,55 6,27 3,13 7,82 3,91 1,95 2,44 5 12,99 6,49 3,24 8,27 4,13 2,06 2,46 ... ... ... ... ... ... ... ... Fonte: PRODIST – Módulo 8 (BRASIL, 2018). Figura 2: Fluxograma de Aplicação da Penalidade por Violação do Indicador de Continuidade FIC Fonte: Autores. Onde: 𝑉𝑎𝑙𝑜𝑟 Tabela 1: Limites de Continuidade por Unidade Consumidora - parte Faixa de variação dos Limites Anuais de Indicadores de Continuidade dos Conjuntos (DEC ou FEC) Limite de Continuidade por Unidade Consumidora ou Central Geradora Unidades Consumidoras ou Centrais Geradoras situadas em áreas urbanas com Faixa de Tensão Contratada: 1 kV < Tensão < 69 kV DIC (horas) FIC (interrupções) DMIC (horas) Anual Trim. Mensal Anual Trim. Mensal Mensal 1 11,25 5,62 2,81 6,48 3,24 1,62 2,36 2 11,68 5,84 2,92 6,93 3,46 1,73 2,39 3 12,12 6,06 3,03 7,37 3,68 1,84 2,41 4 12,55 6,27 3,13 7,82 3,91 1,95 2,44 5 12,99 6,49 3,24 8,27 4,13 2,06 2,46 ... ... ... ... ... ... ... ... Fonte: PRODIST – Módulo 8 (BRASIL, 2018). Figura 2: Fluxograma de Aplicação da Penalidade por Violação do Indicador de Continuidade FIC Fonte: Autores. Tabela 1: Limites de Continuidade por Unidade Consumidora - parte Faixa de variação dos Limites Anuais de Indicadores de Continuidade dos Conjuntos (DEC ou FEC) Limite de Continuidade por Unidade Consumidora ou Central Geradora Unidades Consumidoras ou Centrais Geradoras situadas em áreas urbanas com Faixa de Tensão Contratada: 1 kV < Tensão < 69 kV DIC (horas) FIC (interrupções) DMIC (horas) Anual Trim. Mensal Anual Trim. Mensal Mensal 1 11,25 5,62 2,81 6,48 3,24 1,62 2,36 2 11,68 5,84 2,92 6,93 3,46 1,73 2,39 3 12,12 6,06 3,03 7,37 3,68 1,84 2,41 4 12,55 6,27 3,13 7,82 3,91 1,95 2,44 5 12,99 6,49 3,24 8,27 4,13 2,06 2,46 ... ... ... ... ... ... ... ... Fonte: PRODIST – Módulo 8 (BRASIL, 2018). Tabela 1: Limites de Continuidade por Unidade Consumidora - parte Figura 2: Fluxograma de Aplicação da Penalidade por Violação do Indicador de Continuidade FIC Fonte: Autores. Figura 2: Fluxograma de Aplicação da Penalidade por Violação do Indicador de Continuidade FIC Figura 2: Fluxograma de Aplicação da Penalidade por Violação do Indicador de Continuidade FIC Fonte: Autores. Fonte: Autores. 30 v.5 n.2 2019 30 v.5 n.2 2019 30 Onde 𝐼𝑛 é uma abreviação da variável contadora 𝐼𝑛𝑡𝑒𝑟𝑟𝑢𝑝𝑡𝑖𝑜𝑛, responsável por armazenar quantas vezes cada carga foi desligada no período mensal. Caso o vetor 𝐼𝑛𝑡𝑒𝑟𝑟𝑢𝑝𝑡𝑖𝑜𝑛(𝑖) de determinado nó seja 1 (um) significa que ele já foi interrompido uma vez naquele mês, caso seja interrompido novamente irá implicar em multa. A variável 𝑉𝑎𝑙𝑜𝑟𝑀𝑢𝑙𝑡𝑎𝐹𝐼𝐶 é dada pela equação (6) e é analisada para cada consumidor que foi desligado. 730 Número médio de horas no mês; 𝑘𝑒𝑖 Coeficiente de majoração cujo valor deve ser fixado em: 15 (quinze), para unidade consumidora ou ponto de conexão atendidos em baixa tensão; 20 (vinte), para unidade consumidora ou ponto de conexão atendidos em média tensão; ou 27 (vinte e sete), para unidade consumidora ou ponto de conexão atendidos em alta tensão. 2.2 ALGORITMO DE OTIMIZAÇÃO Os algoritmos Meta-Heurísticos têm tido grande aceitação dos autores da área de sistemas elétricos de potência (KAGAN et al., 2009). Eles são métodos que se baseiam nas técnicas de inteligência artificial e possuem como vantagem a facilidade para encontrar ótimos globais, independentemente do estado inicial, uma vez que realizam suas buscas em vizinhanças dentro do espaço de pesquisa, evitando paradas prematuras em ótimos locais. Dentre os diversos algoritmos meta-heurísticos encontrados na literatura, o Algoritmo Busca Tabu (em inglês, Tabu Search Algorithm) foi escolhido para o processo de otimização. Este algoritmo é um método eficiente para respostas discretas, o qual se encaixa no perfil do problema em questão, uma vez que a resposta desejada é uma matriz de estados das chaves composta por zeros (chaves abertas) e uns (chaves fechadas). Além disso, o Algoritmo Busca Tabu possui memória dinâmica, o qual aumenta a eficiência do processo de exploração e é extremamente útil para sair de máximos ou mínimos locais. O Algoritmo Busca Tabu é um método não populacional, isto é, avalia um candidato de cada vez e possui um sistema de memória dinâmica. O diferencial do sistema de Os algoritmos Meta-Heurísticos têm tido grande aceitação dos autores da área de sistemas elétricos de potência (KAGAN et al., 2009). Eles são métodos que se baseiam nas técnicas de inteligência artificial e possuem como vantagem a facilidade para encontrar ótimos globais, 𝑉𝐴𝐿𝑂𝑅ிூ஼ = ቆ𝐹𝐼𝐶௩ 𝐹𝐼𝐶௣ − 1ቇ𝐷𝐼𝐶௣൬𝐶𝑀 730൰𝑘𝑒𝑖 (6) (6) independentemente do estado inicial, uma vez que realizam suas buscas em vizinhanças dentro do espaço de pesquisa, evitando paradas prematuras em ótimos locais. Dentre os diversos algoritmos meta-heurísticos encontrados na literatura, o Algoritmo Busca Tabu (em inglês, Tabu Search Algorithm) foi escolhido para o processo de otimização. Onde: 𝑉𝑎𝑙𝑜𝑟 Note que, caso o mesmo consumidor seja interrompido mais de uma vez dentro dos limites de FIC, quanto maior for o número de interrupções (3, 4... ao longo do mês), mais o valor da penalidade é acrescido. Isso irá garantir que o algoritmo busque não penalizar sempre a mesma unidade consumidora, mesmo estando dentro dos limites de FIC estabelecidos pela legislação. 𝐶𝑀 Média aritmética dos encargos de uso do sistema de distribuição, considerando também as demandas e energias reativas excedentes, correspondentes aos meses do período de apuração do indicador; 𝐶𝑀 Média aritmética dos encargos de uso do sistema de distribuição, considerando também as demandas e energias reativas excedentes, correspondentes aos meses do período de apuração do indicador; 730 Número médio de horas no mês; 2.3 REDE DE TESTES UTILIZADA E CONDIÇÕES GERAIS DO FLUXO DE CARGA A rede de distribuição de energia elétrica utilizada para aplicação do método proposto é a rede IEEE 123 barras, mostrada na Figura 3. Figura 3: Topologia da Rede de Distribuição IEEE 123 Barras Onde: Onde: 𝐹𝐼𝐶௩ Frequência de interrupção por unidade consumidora ou por ponto de conexão, conforme cada caso, verificada no período considerado, expressa em número de interrupções; 𝐹𝐼𝐶௣ Limite de continuidade estabelecido no período considerado para o indicador de frequência de interrupção por unidade consumidora ou por ponto de conexão, expresso em número de interrupções e centésimo do número de interrupções; p p ç Este algoritmo é um método eficiente para respostas discretas, o qual se encaixa no perfil do problema em questão, uma vez que a resposta desejada é uma matriz de estados das chaves composta por zeros (chaves abertas) e uns (chaves fechadas). Além disso, o Algoritmo Busca Tabu possui memória dinâmica, o qual aumenta a eficiência do processo de exploração e é extremamente útil para sair de máximos ou mínimos locais. O Algoritmo Busca Tabu é um método não populacional, isto é, avalia um candidato de cada vez e possui um sistema de memória dinâmica. O diferencial do sistema de 𝐷𝐼𝐶௣ Limite de continuidade estabelecido no período considerado para o indicador de duração de interrupção por unidade consumidora ou por ponto de conexão, expresso em horas e centésimos de hora; v.5 n.2 2019 31 inteligentes, devem incorporar memória adaptativa e exploração dinâmica (SOLIMANPUR e ELMI, 2013). 2.3 REDE DE TESTES UTILIZADA E CONDIÇÕES GERAIS DO FLUXO DE CARGA memória da Algoritmo Busca Tabu é que ele não mantém apenas os valores das avaliações feitas até então, como também, o itinerário das últimas soluções visitadas (COELLO et al., 2007). Além disso, passa a ser possível identificar regiões promissoras do espaço de busca. O Algoritmo Busca Tabu é uma meta- heurística para guiar heurísticas conhecidas a superar o problema da optimimalidade local. É baseado na premissa de que resoluções de problemas, para se qualificar como inteligentes, devem incorporar memória adaptativa e exploração dinâmica (SOLIMANPUR e ELMI, 2013). 3 RESULTADOS E DISCUSSÃO 3.1 SIMULAÇÃO DO CASO BASE algoritmo, pois a rede original possui somente 6 chaves.  O cabo que liga os nós 93 e 94 que é monofásico foi alterado para trifásico de modo que seja possível a utilização da chave para reconfiguração nos dois sentidos. Em sua forma original, somente o nó 54 poderia abastecer o nó 94, já que o trecho entre os nós 93 e 94 é monofásico. Não é possível uma rede monofásica abastecer outra trifásica. A definição do caso base visa estabelecer a condição operacional de referência, considerando que o alimentador consiga atender todos os nós de carga sem desrespeitar nenhuma das restrições impostas pelo algoritmo. Dessa forma, as cargas foram ajustadas para de modo que o alimentador trabalhe em seu limite operacional, porém ainda conseguindo atender a todos os nós consumidores, ou seja, sem rejeitar cargas. Os valores das cargas originais foram divididos pelo fator 1,442. Este valor foi calculado por meio da divisão entre o valor de corrente total da rede pela corrente máxima admissível do condutor existente no início do alimentador. Após o ajuste das cargas, a topologia da rede foi otimizada seguindo o método proposto neste trabalho. Naturalmente, neste caso base, não houve penalidades referentes aos 4 últimos termos da equação (1), pois não houve violação de limites operacionais nem rejeição de carga. A topologia resultante, referente ao caso base, está mostrado na Figura 4.  Por fim, as cargas foram ajustadas, a fim de fazer com que o alimentador trabalhe em sobrecarga. Desse modo, este realizará rejeição de carga ao mesmo tempo que analisa o indicador FIC do sistema. Para o cálculo do fluxo de carga, assume-se que a topologia da rede, a potência instalada (alimentador) e a potência nas cargas (ou correntes de carga) são conhecidas (KERSTING, 2002). Dessa forma, as condições operacionais da rede de distribuição de energia elétrica são conhecidas, uma vez que o fluxo de carga irá obter a tensão complexa em cada barra e os fluxos de potência (ativa e reativa) em todas as linhas ativas dessa rede e, portanto, as soluções geradas podem ser analisadas. O método utilizado foi o de varredura por soma de correntes, recomendado para cálculos em redes radiais (KAGAN et al., 2005). v.5 n.2 2019 32 Figura 3: Topologia da Rede de Distribuição IEEE 123 Barras Figura 3: Topologia da Rede de Distribuição IEEE 123 Barras Fonte: Autores. Fonte: Autores. Entretanto, algumas modificações foram feitas de modo a adequar à situação desejada, conforme listado abaixo: rejeitada. Dessa forma foram retirados três dos quatro alimentadores da rede (195, 451 e 251), bem como os reguladores de tensão localizados no meio da rede, restando apenas o alimentador localizado no nó 150 e o regulador de tensão localizado entre o mesmo e o nó 149;  Os transformadores foram simplificados, assim como o regulador de tensão localizado entre os nós 150 e 149.  A rede original apresenta 4 alimentadores que abastecem todas as cargas do sistema. Entretanto, a efetividade da metodologia proposta somente pode ser constatada em caso de sobrecarga da rede, situação na qual parte da carga deve ser Cada aresta da rede (trecho) foi avaliada como uma posição candidata a receber uma, ou seja, no processo de reconfiguração foi concedida maior liberdade ao 3 RESULTADOS E DISCUSSÃO 3.1 SIMULAÇÃO DO CASO BASE O algoritmo de fluxo de carga foi validade a partir da comparação entre os resultados obtidos para a rede IEEE 123 barras e os resultados-gabarito fornecidos pelo próprio IEEE junto às informações da rede, com diferenças inferiores à 0,1%. As simulações foram realizadas no software MATLAB 2016b. O computador utilizado foi Intel® Core ™ i5 CPU M480 2,67GHz, 3,00GB RAM e foram utilizadas para cada simulação 500 iterações e uma lista tabu de tamanho 115. Alterações feitas pelo algoritmo na topologia da rede, em relação a topologia original:  Fechamento da chave que conecta os nós 151 e 300;  Abertura da aresta que conecta os nós 101 e 97;  Fechamento da chave que conecta os nós 54 e 94;  Abertura da aresta que conecta os nós 67 e 72. Caracterização da topologia do caso  Apresentou perdas técnicas no valor de 44,83 kW;  Atendeu aos limites de carregamento dos condutores (Tabela 2). Note que a Fase A está operando próxima ao seu limite (453 A). v.5 n.2 2019 33  Manteve a topologia da rede como radial;  As tensões foram mantidas na faixa adequada.  Manteve a topologia da rede como radial;  Manteve a topologia da rede como radial;  Atendeu a todos os nós da rede;  As tensões foram mantidas na faixa adequada.  As tensões foram mantidas na faixa adequada.  As tensões foram mantidas na faixa adequada.  Atendeu a todos os nós da rede;  Atendeu a todos os nós da rede; Figura 4: Topologia da Rede referente ao Caso Base Figura 4: Topologia da Rede referente ao Caso Base Fonte: Autores. Figura 4: Topologia da Rede referente ao Caso Base Fonte: Autores. Fonte: Autores. Tabela 2: Correntes por Fase no Início do Tabela 2: Correntes por Fase no Início do Caracterização da topologia referente ao cenário de sobrecarga #1: Tabela 2: Correntes por Fase no Início do Alimentador – Caso Base Corrente total (A) Fase A Fase B Fase C 450,7 274,4 358,8 Fonte: Autores. 3.2 SIMULAÇÃO DOS CENÁRIOS DE SOBRECARGA Tabela 2: Correntes por Fase no Início do Alimentador – Caso Base  Apresentou perdas técnicas no valor de 46,05 kW;  Atendeu aos limites de carregamento dos cabos (Tabela 3). Note que a Fase A está quase atingindo seu limite (453 A).  Manteve a topologia da rede radial; 2 cenários de sobrecarga foram simulados. 3 RESULTADOS E DISCUSSÃO 3.1 SIMULAÇÃO DO CASO BASE As cargas foram ajustadas por dois fatores distintos, de modo a causar em ambas uma sobrecarga no sistema. Dessa maneira, espera-se que o algoritmo rejeite cargas para respeitas às restrições impostas na especificação do método proposto. Após a aplicação do ajuste de carga proposto referente ao cenário de sobrecarga #1 e a otimização da topologia da rede, a topologia resultante é apresentada na Figura 5.  Rejeitou as cargas dos nós 112, 113 e 114 para atender à restrição de carregamento. Vale ressaltar que o algoritmo optou por rejeitar cargas monofásicas da fase A (Tabela 4), de modo a preservar as cargas atendidas pelas outras fases, uma vez que a rede é desequilibrada com maior carregamento na fase A.  As tensões foram mantidas na faixa adequada. v.5 n.2 2019 34 Figura 5: Topologia da Rede referente ao Cenário de Sobrecarga #1 Fonte: Autores. Figura 5: Topologia da Rede referente ao Cenário de Sobrecarga #1 Fonte: Autores. Tabela 3: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #1 Corrente total (A) Fase A Fase B Fase C 440,2 280,7 365,2 Fonte: Autores. Tabela 4: Carga não Atendida – Cenário de Sobrecarga #1 Carga não atendida (kW) Fase A Fase B Fase C 56,34 0 0 Fonte: Autores. Após a aplicação do ajuste de carga Tabela 3: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #1 Tabela 3: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #1 Corrente total (A) Fase A Fase B Fase C 440,2 280,7 365,2 Fonte: Autores.  Atendeu aos limites de carregamento dos condutores (Tabela 5). Note que a Fase A está quase atingindo seu limite (453 A). p Alimentador – Cenário de Sobrecarga #1 Corrente total (A) Fase A Fase B Fase C 440,2 280,7 365,2 Fonte: Autores.  Manteve a topologia da rede radial;  Rejeitou as cargas dos nós 113 e 114 para atender à restrição de carregamento (Tabela 6). Como a sobrecarga foi menor em relação ao cenário de sobrecarga #1, foi possível atender à carga conectada ao nó 112; Tabela 4: Carga não Atendida – Cenário de Sobrecarga #1 Carga não atendida (kW) Fase A Fase B Fase C 56,34 0 0 Fonte: Autores. Tabela 4: Carga não Atendida – Cenário de Sobrecarga #1  As tensões foram mantidas na faixa adequada. 4 CONCLUSÕES Este artigo propôs uma nova metodologia de reconfiguração de redes de distribuição de energia elétrica que permita evitar a penalização do mesmo conjunto de consumidores, em situações de contingência que impossibilitem o atendimento a todos as unidades consumidoras. Este trabalho permitiu alcançar resultados promissores quanto ao processo de reconfiguração de redes de distribuição de energia elétrica, tema presente no cotidiano das distribuidoras e que interessa tanto o lado do consumidor, quanto o lado da própria distribuidora. Perdas técnicas, atendimento ao máximo de consumidores possível e limites de carregamento e tensão são elementos presentes com certa regularidade na publicação dos autores da área. Este trabalho, por sua vez, além de trabalhar com todos estes elementos, ainda adicionou métrica relacionada ao indicador de continuidade 𝐹𝐼𝐶, de forma diferenciada ao encontrado na literatura.  Manteve a topologia da rede radial;  Rejeitou, dessa vez, as cargas dos nós 70 e 71 para atender à restrição de carregamento e não ser penalizado devido à violação da parcela FIC da função objetivo, por desligar novamente as cargas conectadas aos nós 113 e 114 (Tabela 8);  As tensões foram mantidas na faixa adequada. Tabela 7: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #2 considerando Histórico de Desligamentos Tabela 7: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #2 considerando Histórico de Desligamentos Corrente total (A) Fase A Fase B Fase C 441,5 276,5 359,9 Fonte: Autores. Corrente total (A) Fase A Fase B Fase C 441,5 276,5 359,9 Fonte: Autores. O método proposto foi capaz de analisar as restrições impostas e propor uma topologia radial eficiente para a situação. Desta forma, o algoritmo proposto não penalizou o mesmo grupo de clientes ainda que fosse mantido o valor de sobrecarga no sistema. Os resultados mostraram que o desligamento do mesmo bloco de cargas quando a rede está sobrecarregada pode ser evitado. Associado a isso, os resultados também mostraram que o Algoritmo Busca Tabu, utilizado no processo de otimização, apresentou bons resultados para respostas discretas e foi essencial para a otimização da rede. Tabela 8: Carga não Atendida – Cenário de Sobrecarga #2 considerando Histórico de Desligamentos Tabela 8: Carga não Atendida – Cenário de Sobrecarga #2 considerando Histórico de Desligamentos Carga não atendida (kW) Fase A Fase B Fase C 41,67 0 0 Fonte: Autores. 3 RESULTADOS E DISCUSSÃO 3.1 SIMULAÇÃO DO CASO BASE Após a aplicação do ajuste de carga proposto referente ao cenário de sobrecarga #2 e a otimização da topologia da rede, a topologia resultante é apresentada na Figura 6. Tabela 5: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #2 Corrente total (A) Fase A Fase B Fase C 441,3 276,5 359,9 Fonte: Autores. Tabela 5: Correntes por Fase no Início do Alimentador – Cenário de Sobrecarga #2 Caracterização da topologia referente ao cenário de sobrecarga #2:  Apresentou perdas técnicas no valor de 45,58 kW; v.5 n.2 2019 35 Figura 6: Topologia da Rede referente ao Cenário de Sobrecarga #2 Fonte: Autores. Figura 6: Topologia da Rede referente ao Cenário de Sobrecarga #2 Fonte: Autores. Tabela 6: Carga não Atendida – Cenário de Sobrecarga #2 Carga não atendida (kW) Fase A Fase B Fase C 41,67 0 0 Fonte: Autores. Tabela 6: Carga não Atendida – Cenário de Sobrecarga #2 nas cargas conectadas aos nós 113 e 114. Entretanto, neste caso, poderá ocorrer penalização da função objetivo em relação a parcela FIC, uma vez que há histórico de desligamentos. Logo, o algoritmo deve avaliar se é vantajoso rejeitar cargas distintas às conectadas aos nós 113 e 144, ou rejeitar as mesmas cargas mesmo com a penalidade correspondente. Após esta nova simulação, a topologia resultante é apresentada na Figura 7. Fonte: Autores. Uma nova simulação foi realizada a partir do cenário de sobrecarga #2, agora considerando o histórico de desligamentos gura 7: Topologia da Rede referente ao Cenário de Sobrecarga #2, considerando Histórico de Desligamentos Fonte: Autores. Figura 7: Topologia da Rede referente ao Cenário de Sobrecarga #2, considerando Histórico de Desligamentos Figura 7: Topologia da Rede referente ao Cenário de Sobrecarga #2, considerando Histórico de Desligamentos Fonte: Autores. Fonte: Autores. Fonte: Autores. 36 v.5 n.2 2019 v.5 n.2 2019 36 v.5 n.2 36 v.5 n.2 2019 36 v.5 36 Caracterização da topologia referente ao cenário de sobrecarga #2, considerando o histórico de desligamentos: cargas conectadas aos nós 113 e 114 foram desconectadas inicialmente (na situação ainda sem histórico de desligamentos) por estarem mais distantes do início do alimentador, implicando em perdas técnicas maiores ao longo do percurso.  Apresentou perdas técnicas no valor de 45,88 kW;  Atendeu aos limites de carregamento dos cabos (Tabela 7). Note que a Fase A está quase atingindo seu limite (453 A): AGRADECIMENTOS DUAN, D.-L. et al. Reconfiguration of distribution network for loss reduction and reliability improvement based on an enhanced genetic algorithm. International Journal of Electrical Power & Energy Systems, v. 64, p. 88-95, 2015/01/01/ 2015. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061514004682 >. DUAN, D.-L. et al. Reconfiguration of distribution network for loss reduction and reliability improvement based on an enhanced genetic algorithm. International Journal of Electrical Power & Energy Systems, v. 64, p. 88-95, 2015/01/01/ 2015. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061514004682 >. ESMAEILIAN, H. R.; FADAEINEDJAD, R. Distribution system efficiency improvement using network reconfiguration and capacitor allocation. International Journal of Electrical Power & Energy Systems, v. 64, p. 457-468, 2015/01/01/ 2015. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061514004098 >. GHASEMI, S. Balanced and unbalanced distribution networks reconfiguration considering reliability indices. Ain Shams Engineering Journal, v. 9, n. 4, p. 1567- 1579, 2018/12/01/ 2018. ISSN 2090-4479. Di í l Os autores agradecem ao apoio institucional do Instituto Federal do Espírito Santo (Ifes) e da Universidade Federal do Espírito Santo (Ufes). 37 v.5 n.2 2019 v.5 n.2 2019 37 4 CONCLUSÕES Vale ressaltar que a soma das cargas monofásicas 113 e 114 são equivalentes às cargas 70 e 71, sendo os nós localizados no final das linhas e possuindo a mesma topologia de impedância. Entretanto, as http://www.sciencedirect.com/science/articl p KAGAN, N.; OLIVEIRA, C. C. B. D.; KAGAN, N.; OLIVEIRA, C. C. B. D.; KAGAN, N.; OLIVEIRA, C. C. B. D.; ROBBA, E. J. Introdução aos Sistemas de Distribuição de Energia Elétrica. 1ª Edição. São Paulo: Blucher, 2005. KAGAN, N.; SCHMIDT, H. P.; OLIVEIRA, C. C. B. D. Métodos de Otimização Aplicados a Sistemas Elétricos de Potência. 9ª Edição. São Paulo: Blucher, 2009. ROBBA, E. J. Introdução aos Sistemas de Distribuição de Energia Elétrica. 1ª Edição. São Paulo: Blucher, 2005. KAGAN, N.; SCHMIDT, H. P.; Qualidade da Energia Elétrica de 01 de Janeiro de 2018. Brasília 2018. Qualidade da Energia Elétrica de 01 de Janeiro de 2018. Brasília 2018. KERSTING, W. H. Distribution System Modeling and Analysis. CRC Press, 2002. KERSTING, W. H. Distribution System Modeling and Analysis. CRC Press, 2002. CHIDANANDAPPA, R.; ANANTHAPADMANABHA, T.; H.C, R. Genetic Algorithm Based Network LÓPEZ, J. C.; LAVORATO, M.; RIDER, M. J. Optimal reconfiguration of electrical distribution systems considering reliability indices improvement. International Journal of Electrical Power & Energy Systems, v. 78, p. 837-845, 2016/06/01/ 2016. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S014206151500575X >. MOHAMED IMRAN, A.; KOWSALYA, M.; KOTHARI, D. P. A novel integration technique for optimal network LÓPEZ, J. C.; LAVORATO, M.; RIDER, M. J. Optimal reconfiguration of electrical distribution systems considering reliability indices improvement. International REFERÊNCIAS ABDELAZIZ, M. Distribution network reconfiguration using a genetic algorithm with varying population size. Electric Power Systems Research, v. 142, p. 9-11, 2017/01/01/ 2017. ISSN 0378-7796. Di í l Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0378779616303273 >. GHASEMI, S. Balanced and unbalanced distribution networks reconfiguration considering reliability indices. Ain Shams Engineering Journal, v. 9, n. 4, p. 1567- 1579, 2018/12/01/ 2018. ISSN 2090-4479. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S2090447916301617 >. Disponível em: < Disponível em: < COELLO, C. C.; LAMONT, G. B.; v.5 n.2 2019 38 reconfiguration and distributed generation placement in power distribution networks. International Journal of Electrical Power & Energy Systems, v. 63, p. 461-472, 2014/12/01/ 2014. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061514003585 >. NGUYEN, T. T. et al. Multi-objective electric distribution network reconfiguration solution using runner-root algorithm. Applied Soft Computing, v. 52, p. 93-108, 2017/03/01/ 2017. ISSN 1568-4946. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S1568494616306366 >. NGUYEN, T. T.; TRUONG, A. V.; PHUNG, T. A. A novel method based on adaptive cuckoo search for optimal network reconfiguration and distributed generation allocation in distribution network. International Journal of Electrical Power & Energy Systems, v. 78, p. 801-815, 2016. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061515005827 >. reconfiguration and distributed generation placement in power distribution networks. International Journal of Electrical Power & Energy Systems, v. 63, p. 461-472, 2014/12/01/ 2014. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061514003585 >. NGUYEN, T. T. et al. Multi-objective electric distribution network reconfiguration solution using runner-root algorithm. Applied Soft Computing, v. 52, p. 93-108, 2017/03/01/ 2017. ISSN 1568-4946. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S1568494616306366 >. NGUYEN, T. T.; TRUONG, A. V.; PHUNG, T. A. A novel method based on adaptive cuckoo search for optimal network reconfiguration and distributed generation allocation in distribution network. International Journal of Electrical Power & Energy Systems, v. 78, p. 801-815, 2016. ISSN 0142-0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061515005827 >. reconfiguration and distributed generation placement in power distribution networks. International Journal of Electrical Power & Energy Systems, v. 63, p. 461-472, 2014/12/01/ 2014. ISSN 0142-0615. RAJARAM, R.; SATHISH KUMAR, K.; RAJASEKAR, N. Power system reconfiguration in a radial distribution network for reducing losses and to improve voltage profile using modified plant growth simulation algorithm with Distributed Generation (DG). Energy Reports, v. 1, p. 116-122, 2015/11/01/ 2015. ISSN 2352- 4847. Disponível em: < SOLIMANPUR, M.; ELMI, A. A tabu search approach for cell scheduling problem with makespan criterion. International Journal of Production Economics, v. 141, n. 2, p. 639-645, 2013/02/01/ 2013. ISSN 0925-5273. Disponível em: < ZHAI, H. F. et al. Dynamic reconfiguration of three-phase unbalanced distribution networks. International Journal of Electrical Power & Energy Systems, v. 99, p. 1-10, 2018/07/01/ 2018. ISSN 0142- 0615. Disponível em: < http://www.sciencedirect.com/science/articl e/pii/S0142061517321555 >.
https://openalex.org/W2939607606	https://www.frontiersin.org/articles/10.3389/fimmu.2019.00739/pdf	English	null	FcγR-TLR Cross-Talk Enhances TNF Production by Human Monocyte-Derived DCs via IRF5-Dependent Gene Transcription and Glycolytic Reprogramming	Frontiers in immunology	2,019	cc-by	8,782	FcγR-TLR Cross-Talk Enhances TNF Production by Human Monocyte-Derived DCs via IRF5-Dependent Gene Transcription and Glycolytic Reprogramming Willianne Hoepel 1,2, Melissa Newling 1,2, Lisa T. C. Vogelpoel 2, Lathees Sritharan 1,2, Ivo S. Hansen 1,2, Martien L. Kapsenberg 2, Dominique L. P. Baeten 1,2, Bart Everts 3 and Jeroen den Dunnen 1,2* Willianne Hoepel 1,2, Melissa Newling 1,2, Lisa T. C. Vogelpoel 2, Lathees Sritharan 1,2, Ivo S. Hansen 1,2, Martien L. Kapsenberg 2, Dominique L. P. Baeten 1,2, Bart Everts 3 and Jeroen den Dunnen 1,2* 1 Amsterdam Rheumatology and Immunology Center, Amsterdam, Netherlands, 2 Department of Experimental Immunology, Amsterdam UMC, Amsterdam Infection and Immunity Institute, University of Amsterdam, Amsterdam, Netherlands, 3 Department of Parasitology, Leiden University Medical Center, Leiden, Netherlands Antigen-presenting cells (APCs) such as dendritic cells (DCs) are crucial for initiation of adequate inﬂammatory responses, which critically depends on the cooperated engagement of different receptors. In addition to pattern recognition receptors (PRRs), Fc gamma receptors (FcγRs) have recently been identiﬁed to be important in induction of inﬂammation by DCs. FcγRs that recognize IgG immune complexes, which are formed upon opsonization of pathogens, induce pro-inﬂammatory cytokine production through cross-talk with PRRs such as Toll-like receptors (TLRs). While the physiological function of FcγR-TLR cross-talk is to provide protective immunity against invading pathogens, undesired activation of FcγR-TLR cross-talk, e.g., by autoantibodies, also plays a major role in the development of chronic inﬂammatory disorders such as rheumatoid arthritis (RA). Yet, the molecular mechanisms of FcγR-TLR cross-talk are still largely unknown. Here, we identiﬁed that FcγR-TLR cross-talk-induced cytokine production critically depends on activation of the transcription factor interferon regulatory factor 5 (IRF5), which results from induction of two different pathways that converge on IRF5 activation. First, TLR stimulation induced phosphorylation of TBK1/IKKε, which is required for IRF5 phosphorylation and subsequent activation. Second, FcγR stimulation induced nuclear translocation of IRF5, which is essential for gene transcription by IRF5. We identiﬁed that IRF5 activation by FcγR-TLR cross-talk ampliﬁes pro-inﬂammatory cytokine production by increasing cytokine gene transcription, but also by synergistically inducing glycolytic reprogramming, which is another essential process for induction of inﬂammatory responses by DCs. Combined, here we identiﬁed IRF5 as a pivotal component of FcγR-TLR cross-talk in human APCs. These data may provide new potential targets to suppress chronic inﬂammation in autoantibody-associated diseases that are characterized by undesired or excessive FcγR-TLR cross-talk, such as RA, systemic sclerosis, and systemic lupus erythematous. Edited by: Pierre Guermonprez, King’s College London, United Kingdom Reviewed by: Bénédicte Manoury, Institut National de la Santé et de la Recherche Médicale (INSERM), France Diana Dudziak, Universitätsklinikum Erlangen, Germany Correspondence: Jeroen den Dunnen j.dendunnen@amc.nl Specialty section: This article was submitted to Antigen Presenting Cell Biology, a section of the journal Frontiers in Immunology Received: 13 September 2018 Accepted: 19 March 2019 Published: 08 April 2019 ORIGINAL RESEARCH published: 08 April 2019 doi: 10.3389/ﬁmmu.2019.00739 ORIGINAL RESEARCH Cells and Stimulation This study was done according to the ethical guidelines of the Academic Medical Center and human material was obtained in accordance with the AMC Medical Ethics Review Committee according to the Medical Research Involving Human Subjects Act. Buﬀy coats obtained after blood donation (Sanquin blood supply) are not subjected to informed consent, which is according to the Medical Research Involving Human Subjects Act and the AMC Medical Ethics Review Committee. All samples were handled anonymously. Ethical review and approval was not required for this study in accordance with the local legislation. Monocytes were isolated from buﬀy coats by density gradient centrifugation on Lymphoprep (Nycomed) and Percoll (Pharmacia). DCs or macrophages were generated by culturing monocytes for 6 days in IMDM (Lonza) containing 5% FBS (Biowest) and 86 µg/mL gentamicin (Gibco), supplemented with 20 ng/mL GM-CSF (Invitrogen) and 2 ng/mL IL-4 (Miltenyi Biotec) for DCs or 50 ng/mL recombinant human M-CSF (BioLegend) for macrophages. At day 2 or 3, half of the medium was replaced by new medium containing cytokines. In human APCs such as DCs, the best studied cytokine- inducing FcγR is FcγRIIa. FcγRIIa has a low aﬃnity for IgG, and is therefore able to discriminate between unbound IgG and IgG immune complexes (i.e., antigen-bound). While unbound IgG, as present under homeostatic conditions, induces inhibitory signaling (5), stimulation of FcγRIIa with immune complexes, as present on opsonized pathogens, strongly enhances cytokine production induced by TLRs (1, 6). Although monocytes and macrophages are known also to express other FcRs such as FcγRI, FcγRIIa is the main IgG receptor responsible for amplifying TLR responses (2). The physiological function of FcγR-TLR cross-talk is to counteract infections with various classes of pathogens. For example, upon IgG opsonization of bacteria, the simultaneous activation of FcγRIIa and TLRs speciﬁcally ampliﬁes the production of pro-inﬂammatory cytokines TNF, IL-1β, IL-6, and IL-23 by human DCs, which in turn promote human T helper 17 (Th17) skewing, thereby tailoring immune response to counteract extracellular bacterial infections (1, 6). However, in addition to its physiological function, FcγR-TLR cross-talk can also be induced undesirably by immune complex formation of autoantibodies. This pathological role of FcγR-TLR cross-talk contributes to the pathogenesis of various autoimmune diseases including rheumatoid arthritis (RA) (7). For silencing at day 3, cells were harvest by resuspending (DCs) or by using TrypLE Select (Invitrogen) (macrophages). Citation: Hoepel W, Newling M, Vogelpoel LTC, Sritharan L, Hansen IS, Kapsenberg ML, Baeten DLP, Everts B and den Dunnen J (2019) FcγR-TLR Cross-Talk Enhances TNF Production by Human Monocyte-Derived DCs via IRF5-Dependent Gene Transcription and Glycolytic Reprogramming. Front. Immunol. 10:739. doi: 10.3389/ﬁmmu.2019.00739 Keywords: Fc gamma receptor (FcγR), interferon regulatory factor 5 (IRF5), dendritic cells, macrophages, glycolytic reprogramming, tumor necrosis factor (TNF), rheumatoid arthritis (RA), chronic inﬂammation April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 1 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. INTRODUCTION for FcγR-induced cytokine production are distinct from other FcγR-mediated functions such as phagocytosis (3). Compared to FcγR signaling, relatively more is known about the signaling pathways that are induced by individual stimulation of TLRs. TLRs signal via adaptor proteins such as MyD88 and/or TRIF to activate various transcription factors including NF-κB and MAP kinases, which are important for the transcription of pro- inﬂammatory cytokines such as TNF (9). Yet, how TLR and FcγR signaling pathways collaborate to synergistically amplify pro-inﬂammatory cytokine production is still largely unknown. Protection against diﬀerent classes of pathogens requires the activation of antigen-presenting cells (APCs) such as dendritic cells (DCs). A crucial step for shaping both innate and adaptive immunity by DCs is the production of various pro-inﬂammatory cytokines. DCs produce these cytokines upon detection of pathogens or endogenous danger signals via activation of diﬀerent families of receptors, which collectively are referred to as pattern recognition receptors (PRRs). Well-known examples of PRRs include the families of Toll-like receptors (TLRs), NOD- like receptors (NLRs), C-type lectins, and RIG-I-like receptors (RLRs). However, the list of receptor families that control cytokine production is still expanding. In this study, we identiﬁed that FcγR-TLR cross-talk-induced cytokine production critically depends on activation of the transcription factor interferon regulatory factor 5 (IRF5), which results from collaborative IRF5 activation by both FcγRs and TLRs. While TLR stimulation induced IRF5 phosphorylation, FcγR stimulation was required for IRF5 nuclear translocation. Moreover, we identiﬁed that IRF5 activation by FcγR-TLR cross- talk ampliﬁed pro-inﬂammatory cytokines production by both increasing cytokine gene transcription and by inducing glycolytic reprogramming, thereby identifying FcγRs as a new family of receptors that can induce metabolic reprogramming. In recent years, it has become clear that also the family of Fc gamma receptors (FcγRs), which are receptors for the Fc region of immunoglobulin G (IgG), play an important role in the induction of cytokines by DCs. While individual stimulation of FcγRs elicits little cytokine production, FcγRs synergize with PRRs such as TLRs to strongly but selectively amplify pro-inﬂammatory cytokine production. FcγRs synergize with TLRs that are expressed both intracellular (TLR3, TLR7/8) and extracellular (TLR2, TLR4, TLR5), as well as other receptors such as NLRs and particular cytokine receptors (1, 2). Combined, modulation of cytokine production by FcγRs thereby tailors immune responses to the immunological context (3, 4). Frontiers in Immunology \| www.frontiersin.org Cells and Stimulation Cells were microporated in the presence of 500 nM IRF5 si-RNA or control si-RNA (Dharmacon) and cultured for 3 more days in IMDM without gentamicin with supplemented cytokines. DCs were harvested at day 6 by putting the cells on ice for 30 min and macrophages were harvested at day 6 by TrypLE Select. For cIgG stimulation, 96-well high-aﬃnity Maxisorp plates (Nunc) were coated with 2 µg/mL IgG from pooled IgG (Nanogam; Sanquin Blood Supply) diluted in PBS overnight at 4◦C, followed by blocking with PBS containing 10% FBS for 1 h at 37◦C. Cells were stimulated (30,000–50,000 cells per well) with 10 µg/mL Pam3CSK4 (Invivogen). Co-stimulation experiments Remarkably, while FcγR-induced cytokine production plays an important role in both host defense and various autoimmune diseases, still very little is known about the underlying molecular mechanisms. Similar to other FcγR-mediated functions such as phagocytosis and ADCC, FcγR-induced cytokine production is dependent on the upstream kinase Syk (8). However, recent ﬁndings indicate that the downstream signaling events required April 2019 \| Volume 10 \| Article 739 2 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. round-bottom plate (Greiner Bio-One), washed in PBS, and permeabilized using Perm III buﬀer (BD Biosciences) for at least 30 min at −20◦C. Cells were then washed in PBS containing 0.5% BSA and 0.1% sodium azide and stained for 1 h at RT with a rabbit-anti-human-IRF5 antibody (1:200) (Cell Signaling) or a rabbit-anti-human-pTBK1 antibody (1:50) (Ser172; Cell Signaling), which also reacts to pIKKε, followed by a 30 min staining at room temperature with Alexaﬂuor488-labeled goat-anti-rabbit-IgG antibody (1:400) (Molecular Probes). Fluorescence was determined by ﬂow cytometry (Canto II, BD Biosciences). were performed by simultaneous exposure of the cells to cIgG and Pam3. Syk was inhibited with 1 µM R406 (Selleckchem), TBK1/IKKε was inhibited with 2 µM BX795 (Invivogen) and glycolysis was blocked using 10 mM 2-Deoxy-D-glucose (2DG; Sigma Aldrich). Cells were incubated with the inhibitor or the corresponding volume of DMSO (Sigma-Aldrich) or medium for 30 min at 37◦C before stimulation. Western Blot For analysis of IRF5 phosphorylation, DCs were stimulated as indicated in 6-well plates (1,250,000–2,000,000 cell per well) (Costar) for 30 min. Cells were gently scraped and collected in cold PBS. After washing, cells were lysed on ice for 10 min using RIPA lysis buﬀy (Cell signaling) supplemented with protease and phosphatase inhibitors (both from Roche). Lysates were brieﬂy soniﬁciated for 10 s at 30% and centrifuged for 10 min at 14,000 × g. BCA assay was performed (Thermo Scientiﬁc) and samples were boiled with 4x Laemmli Sample Buﬀer (Bio-Rad) for 15 min at 95◦C. Cell lysates were run on a 4–12% Bis-Tris protein gel (Invitrogen) using MES-running buﬀer (Invitrogen). Proteins were transferred to a PVDF membrane (GE healthcare) using transfer buﬀer (Invitrogen) and blocked with 2% milk (Bio-Rad) afterwards. Membrane was incubated in TBS Tween o/n at 4◦C with indicated antibodies: Phospho-IRF5 (Ser437) polyclonal antibody (1:1000) (Thermo Scientiﬁc), IRF5 (1:1000) (E1N9G) rabbit mAb (Cell Signaling), or Actin antibody (I-19) (1:2000) (Santa Cruz). Afterwards membrane washed with TBS Tween and incubated for 1 h at room temperature with polyclonal swine anti-rabbit immunoglobulins HRP (1:3000) (Dako). Fluorescence Microscopy py For analysis of IRF5 translocation, DCs or macrophages were stimulated as indicated in Maxisorp plates. After 2 h stimulation, cells were washed with PBS, ﬁxed with 3.7% formaldehyde (Sigma-Aldrich) for 15 min at room temperature, washed in PBS and stored in PBS containing 0.5% bovine serum albumin (BSA; PAA) and 0.1% sodium azide (Merck) at 4◦C. Cells were permeabilized with 0.2% Triton X-100 (Sigma-Aldrich) for 5 min at room temperature and blocked for 30 min in PBS containing 0.5% BSA and 0.1% sodium azide. Cells were then stained with a rabbit-anti-human-IRF5 antibody (1:400) (Cell Signaling) or rabbit-anti-human NF-kB p65 antibody (1:100) (Cell Signaling) for 45 min at room temperature, washed with PBS and stained with a Cy3-labeled goat-anti-rabbit-IgG antibody (1:50) (Jackson ImmunoResearch). Cells were again washed with PBS and nuclei were stained using 1 µg/mL Hoechst (Immunochemistry Technologies) for 1 min at room temperature. Cells were imaged using a DM IRB inverted ﬂuorescence microscope (Leica), combined with a DFC 300FX digital color camera (Leica). Metabolic Assays y Real-time analysis of the extracellular acidiﬁcation rate (ECAR) and the oxygen consumption rate (OCR) of DCs were analyzed using an XF-96 Extracellular Flux Analyzer (Seahorse Bioscience). 30,000 DCs were plated per well. To trigger FcγR on DCs XF-96 cell culture plates were coated with 4 µg/ml IgG prior to seeding of the cells. DCs were plated in glucose- free medium after which glucose was added (10 mM) to the cells during the assay to be able to determine true glycolysis- driven ECAR. Thirty minutes after glucose addition cells were stimulated with 10 µg/mL Pam3CSK4 during the essay after which OCR and glycolysis-driven ECAR were determined 30 min post stimulation. Quantitative RT-PCR For mRNA-level analysis, cells were lysed at the indicated time points, after which mRNA extraction was performed using RNeasy Mini Kit (Qiagen) and cDNA synthesis using RevertAid H Minus First Strand cDNA Synthesis Kit (Fermentas). Quantitative RT-PCR (StepOnePlus Real-Time PCR System; Thermo Fisher Scientiﬁc) was performed using Taqman Master Mix and the following Taqman primers (all from Thermo Fisher Scientiﬁc): GAPDH (4310884E), IRF5 (Hs00158114_m1), and TNF (Hs00174128_m1). mRNA levels were normalized to the geometric mean of the Ct-values of housekeeping gene GAPDH [2Ct(housekeeping)−Ct(target)], and folds were calculated compared with an unstimulated control sample (t = 0 h). ELISA For analysis of cytokine production, supernatants were harvested after overnight stimulation and stored at −20◦C. Cytokine levels in supernatants were measured by ELISA, using antibody pairs for TNF (eBioscience), IL-Iβ, IL-6, and IL-23 (U- CyTech Biosciences). Frontiers in Immunology \| www.frontiersin.org FcγR Stimulation Induces IRF5 Nuclear Translocation The transcription factor IRF5 is constitutively expressed by myeloid APCs (15), but to regulate gene transcription IRF5 needs to be translocated to the nucleus (24). Therefore, we assessed IRF5 localization in human moDCs by ﬂuorescence microscopy upon FcγR-TLR co-stimulation. IRF5 contains two nuclear localization signals (NLS) as well as a nuclear export signal (NES) and therefore continuously shuttles in and out of the nucleus (25–27). Indeed, in unstimulated moDCs IRF5 was present throughout the cell, both in the nucleus and the cytoplasm (Figure 2A). Similar to unstimulated cells, TLR2- stimulated moDCs also displayed an even distribution of IRF5 (Figure 2A). In contrast, stimulation with cIgG, either combined with TLR stimulation or not, resulted in near exclusive accumulation of IRF5 in the nucleus (Figure 2A; quantiﬁed in Figure 2B). As a control we also ascertained that (individual) TLR stimulation of moDCs results in nuclear translocation of NF-κB subunit p65 (Figures 2C,D), which is responsible for TLR-induced pro-inﬂammatory cytokine production. Very similar to moDCs, FcγR stimulation induced IRF5 nuclear translocation in human macrophages (Figures 2E,F), suggesting that nuclear translocation of IRF5 induced by FcγR stimulation is a general mechanism in myeloid APCs. To study the role of IRF5 in FcγR-TLR cross-talk, we made use of a small interfering (si)-RNA approach, which on average resulted in a 60 % reduction of IRF5 mRNA expression and a similar reduction in IRF5 protein in monocyte- derived DCs (moDCs) (Figures 1C,D). For stimulation of FcγRs and TLR2 we used plate-bound complexed IgG (cIgG) and Pam3CSK4 (Pam3), respectively. While individual stimulation with cIgG or Pam3 induced moderate amounts of TNF, combined stimulation strongly and synergistically ampliﬁed TNF production (Figure 1E). However, strikingly, silencing of IRF5 speciﬁcally reduced TNF protein production by FcγR-TLR cross- talk, without aﬀecting cytokine production induced by the individual ligands (Figure 1E). In addition, we assessed whether IRF5 is also responsible for FcγR-TLR cross talk-induced gene transcription. Indeed, (partial) silencing of IRF5 reduced TNF mRNA production upon FcγR-TLR co-stimulation (Figure 1F for kinetics of representative donor, Figure 1G for multiple donors). In contrast, TNF mRNA induced by TLR stimulation alone was not aﬀected by IRF5 silencing (Figure 1F), indicating that IRF5 speciﬁcally controls TNF transcription induced by FcγR-TLR cross-talk. Since FcγR-TLR cross-talk is known to depend on signaling through the kinase Syk, we next assessed whether Syk is required for IRF5 nuclear translocation. FcγR Stimulation Induces IRF5 Nuclear Translocation As shown in Figure 2G, Syk inhibition by therapeutic small-molecule inhibitor R406 indeed suppressed IRF5 nuclear translocation both upon individual stimulation with cIgG and upon cIgG+Pam3 co-stimulation. These data indicate that, in human moDCs and macrophages, stimulation with IgG immune complexes is responsible for nuclear translocation of IRF5. FcγR-TLR Cross-Talk in Human moDCs and Macrophages Is Dependent on IRF5 and Macrophages Is Dependent on IRF5 FcγR-TLR cross-talk plays an important role in inducing inﬂammation during both bacterial infections and autoimmune diseases (1, 2, 6, 8). As illustrated in Figures 1A,B (representative donor and multiple donors, respectively), FcγR-TLR cross-talk synergistically ampliﬁes the production of key pro-inﬂammatory cytokines TNF, IL-1β, and IL-23, while other cytokines such as IL-6 are not aﬀected. Here, we set out to identify the molecular mechanisms underlying this response, using TNF production as a main read-out for FcγR-TLR cross-talk. FcγR-TLR cross- talk is known to amplify TNF production at the level of gene transcription (1, 8). Here, we hypothesized a role for IRF5, since this transcription factor is known to be involved in enhancing TNF transcription (10–14), is highly expressed in human myeloid APCs (15), and since IRF5 polymorphisms are a known risk factor for several autoimmune diseases (16–22). Data Analysis For analysis of TBK1/IKKε phosphorylation, DCs or macrophages were stimulated as indicated in 48-well plates (Greiner Bio-One) for 30 min and ﬁxed using Lyse/Fix buﬀer (BD Biosciences) for 10 min at room temperature. For analysis of IRF5, unstimulated DCs were also lysed and transferred in a 96- well plate following the same protocol as TBK1 phosphorylation. Cells were harvested by gentle scraping, transferred to a 96-well Co-localization quantiﬁcation of the ﬂuorescence microscopy data was done using Huygens Professional software (SVI, Hilversum, The Netherlands) calculating the Manders Coeﬃcients. Western blots were analyzed using ImageJ. Data were analyzed for statistical signiﬁcance using student’s t-test with GraphPad Prism version 7 software (GraphPad Software). April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 3 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. \| FcγR-TLR cross-talk in human moDCs and macrophages is dependent on IRF5. (A,B) Human monocyte-derived dendritic cells (moDC) were stimulated CSK4 (Pam3), cIgG, or the combination for 24 h. Protein production was determined by ELISA. (A) Representative examples of experiments performed in mean + SEM). (B) Protein production of multiple donors, each pair of dots represent one donor. (C,D) IRF5 in human moDCs and macrophages was sing speciﬁc si-RNA. (C) IRF5 mRNA expression of unstimulated moDCs or macrophages (Mφ), after IRF5 silencing (si-IRF5) or non-targeted control si-C). Data shown is IRF5 mRNA expression as percentage of control of IRF5 mRNA expression in si-C conditions. Mean + SD of three (moDC) or eight (Continued) FIGURE 1 \| FcγR-TLR cross-talk in human moDCs and macrophages is dependent on IRF5. (A,B) Human monocyte-derived dendritic cells (moDC) were stimulated with Pam3CSK4 (Pam3), cIgG, or the combination for 24 h. Protein production was determined by ELISA. (A) Representative examples of experiments performed in triplicate (mean + SEM). (B) Protein production of multiple donors, each pair of dots represent one donor. (C,D) IRF5 in human moDCs and macrophages was silenced using speciﬁc si-RNA. (C) IRF5 mRNA expression of unstimulated moDCs or macrophages (Mφ), after IRF5 silencing (si-IRF5) or non-targeted control silencing (si-C). Data shown is IRF5 mRNA expression as percentage of control of IRF5 mRNA expression in si-C conditions. Mean + SD of three (moDC) or eight (Continued) April 2019 \| Volume 10 \| Article 739 4 Frontiers in Immunology \| www.frontiersin.org FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. FIGURE 1 \| experiments (Mφ). Data Analysis (D) IRF5 protein expression of unstimulated moDCs after IRF5 silencing or non-targeted control silencing measured by ﬂow cytometry. (E,H) Control or IRF5-silenced moDCs (E) and macrophages (H) were stimulated with Pam3, cIgG, or the combination for 6 h. Protein production was determined by ELISA. Data shown is protein production normalized to Pam3-induced TNF production for each experiment (set to 1), mean + SD of three (E) or six (H) experiments using different donors. (F,I) Control or IRF5-silenced moDCs (F) and macrophages (I) were stimulated with Pam3, cIgG, or in combination and TNF mRNA expression (normalized to housekeeping gene expression) was determined at indicated time points by quantitative RT-PCR. Representative examples of four experiments. (G,J) TNF mRNA expression after 6 h co-stimulation of control or IRF5-silenced moDCs (G) and Mφ (J) of multiple donors. Each pair of dots represents one donor. p < 0.05, *p < 0.01, Student’s t-test. Frontiers in Immunology \| www.frontiersin.org RESULTS Combined, these data demonstrate that the synergistic induction of TNF by FcγR-TLR cross-talk in human moDCs and macrophages is dependent on IRF5. FcγR-TLR Cross-Talk Is Dependent on TLR-Induced Phosphorylation of TBK1/IKKε and IRF5 To determine whether IRF5 is only essential for FcγR-TLR cross-talk in moDCs, or whether it is also required for FcγR- TLR cross-talk in other cell types, we assessed the eﬀect of IRF5 silencing on human macrophages, which are the main source of TNF in inﬂamed synovia of RA patients (23). Similar to moDCs, silencing of IRF5 in monocyte-derived macrophages (Figure 1C) speciﬁcally reduced TNF production induced by FcγR-TLR synergy, both on protein (Figure 1H) and mRNA (Figures 1I,J). While individual FcγR stimulation induced IRF5 translocation into the nucleus, it is not suﬃcient to induce TNF transcription (1, 6, 8). Importantly, in addition to nuclear translocation, IRF5 needs to be activated by phosphorylation in order to be transcriptionally active (14, 25, 26, 28, 29). Therefore, we determined IRF5 phosphorylation upon (co-)stimulation of human moDCs by Western blot. Interestingly, while Pam3 April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 5 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. FIGURE 2 \| FcγR stimulation induces IRF5 nuclear translocation. (A,E) Human monocyte-derived dendritic cells (moDC) (A) and macrophages (Mφ) (E) were stimulated with Pam3CSK4 (Pam3), cIgG, or the combination for 2 h and stained for IRF5 (red) and nuclei (Hoechst, blue). Representative images of three independent experiments, bar: 20 µm. (B,D,F) Quantiﬁcation of the microscopy data showing percent co-localization of three experiments (mean + SD). p < 0.05, *p < 0.01, Student’s t-test. (C) moDCs were stimulated with Pam3 and stained for p65 (red) and nuclei (Hoechst, blue). Representative image of three independent experiments. (G) Human monocyte derived macrophages were pre-incubated with the Syk inhibitor R406 and stimulated with Pam3, cIgG, or the combination for 2 h and stained for IRF5 (red) and nuclei (Hoechst, blue). Representative images of three independent experiments, bar: 20 µm. stimulation induced IRF5 phosphorylation stimulation with also needs to be phosphorylated in order to execute kinase FIGURE 2 \| FcγR stimulation induces IRF5 nuclear translocation. (A,E) Human monocyte-derived dendritic cells (moDC) (A) and macrophages (Mφ) (E) were stimulated with Pam3CSK4 (Pam3), cIgG, or the combination for 2 h and stained for IRF5 (red) and nuclei (Hoechst, blue). Representative images of three independent experiments, bar: 20 µm. (B,D,F) Quantiﬁcation of the microscopy data showing percent co-localization of three experiments (mean + SD). p < 0.05, *p < 0.01, Student’s t-test. (C) moDCs were stimulated with Pam3 and stained for p65 (red) and nuclei (Hoechst, blue). FcγR-TLR Cross-Talk Is Dependent on TLR-Induced Phosphorylation of TBK1/IKKε and IRF5 Representative image of three independent experiments. (G) Human monocyte derived macrophages were pre-incubated with the Syk inhibitor R406 and stimulated with Pam3, cIgG, or the combination for 2 h and stained for IRF5 (red) and nuclei (Hoechst, blue). Representative images of three independent experiments, bar: 20 µm. FIGURE 2 \| FcγR stimulation induces IRF5 nuclear translocation. (A,E) Human monocyte-derived dendritic cells (moDC) (A) and macrophages (Mφ) (E) were stimulated with Pam3CSK4 (Pam3), cIgG, or the combination for 2 h and stained for IRF5 (red) and nuclei (Hoechst, blue). Representative images of three independent experiments, bar: 20 µm. (B,D,F) Quantiﬁcation of the microscopy data showing percent co-localization of three experiments (mean + SD). p < 0.05, *p < 0.01, Student’s t-test. (C) moDCs were stimulated with Pam3 and stained for p65 (red) and nuclei (Hoechst, blue). Representative image of three independent experiments. (G) Human monocyte derived macrophages were pre-incubated with the Syk inhibitor R406 and stimulated with Pam3, cIgG, or the combination for 2 h and stained for IRF5 (red) and nuclei (Hoechst, blue). Representative images of three independent experiments, bar: 20 µm. stimulation induced IRF5 phosphorylation, stimulation with cIgG did not (Figure 3A, quantiﬁed as pIRF5/IRF5 ratio for multiple donors in Figure 3B). These data indicate that while FcγR stimulation induces IRF5 nuclear translocation, TLR stimulation is required for IRF5 phosphorylation. also needs to be phosphorylated in order to execute kinase activity (31), we assessed TBK1/IKKε phosphorylation by ﬂow cytometry after (co-)stimulation. Similar to IRF5 phosphorylation, we found that stimulation with Pam3 induced TBK1/IKKε phosphorylation, while stimulation with cIgG did not (Figure 3C, quantiﬁed for multiple donors in Figure 3D). IRF5 phosphorylation can be induced by TBK1, a member of the Iκ kinase (IKK) family that shares larges structural and functional similarity to IKKε (25, 26, 30). Since TBK1/IKKε To determine whether TBK1/IKKε is required for cytokine production by FcγR-TLR cross-talk, we inhibited TBK1/IKKε April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 6 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. FIGURE 3 \| FcγR-TLR cross-talk is dependent on TLR-induced phosphorylation of TBK1/IKKε and IRF5. (A) Human monocyte-derived dendritic cells (moDC) were stimulated with Pam3CSK4 (Pam3), cIgG, or the combination for 30 min. IRF5 phosphorylation at Ser437 and total IRF5 expression was assessed by Western blot. Data shown is representative example of three independent experiments. FcγR-TLR Cross-Talk Is Dependent on TLR-Induced Phosphorylation of TBK1/IKKε and IRF5 Thus, while FcγR stimulation induces nuclear translocation of IRF5, TLR stimulation induces phosphorylation of TBK1/IKKε and IRF5, which combined results in nuclear translocation of phosphorylated IRF5 to modulate cytokine gene transcription. Next, we set out to investigate whether the ampliﬁcation of the glycolytic response by FcγR-TLR cross-talk was also dependent on IRF5. While silencing of IRF5 did not aﬀect the ECAR induced by individual stimulation with cIgG or Pam3, IRF5 silencing did inhibit the increased ECAR induced upon co-stimulation (Figure 4D). These data indicate that FcγR-TLR cross-talk ampliﬁes the glycolytic response via IRF5. FcγR-TLR Cross-Talk Is Dependent on TLR-Induced Phosphorylation of TBK1/IKKε and IRF5 (B) Quantiﬁcation of the three independent Western blot experiments using ImageJ (mean + SD). First bands were corrected for actin, after which the pIRF5/IRF5 ratio was calculated. (C) Human moDCs and macrophages (Mφ) were stimulated with Pam3, cIgG, or the combination for 30 min and stained for p-TBK1/IKKε and analyzed by ﬂow cytometry (10log scale, light gray indicates background staining). Representative example of four (DCs) and three (Mφ) experiments. (D) 1MFI of pTBK1 of four (moDCs) and three (Mφ) independent experiments (mean + SD). (E) After pre-incubation with 2 µM BX795 or the corresponding volume of DMSO, moDCs, and macrophages were stimulated with Pam3, or cIgG combined with Pam3 for 24 h and TNF production was determined by ELISA. Mean + SD of four independent experiments. p < 0.05, Student’s t-test. FIGURE 3 \| FcγR-TLR cross-talk is dependent on TLR-induced phosphorylation of TBK1/IKKε and IRF5. (A) Human monocyte-derived dendritic cells (moDC) were stimulated with Pam3CSK4 (Pam3), cIgG, or the combination for 30 min. IRF5 phosphorylation at Ser437 and total IRF5 expression was assessed by Western blot. Data shown is representative example of three independent experiments. (B) Quantiﬁcation of the three independent Western blot experiments using ImageJ (mean + SD). First bands were corrected for actin, after which the pIRF5/IRF5 ratio was calculated. (C) Human moDCs and macrophages (Mφ) were stimulated with Pam3, cIgG, or the combination for 30 min and stained for p-TBK1/IKKε and analyzed by ﬂow cytometry (10log scale, light gray indicates background staining). Representative example of four (DCs) and three (Mφ) experiments. (D) 1MFI of pTBK1 of four (moDCs) and three (Mφ) independent experiments (mean + SD). (E) After pre-incubation with 2 µM BX795 or the corresponding volume of DMSO, moDCs, and macrophages were stimulated with Pam3, or cIgG combined with Pam3 for 24 h and TNF production was determined by ELISA. Mean + SD of four independent experiments. p < 0.05, Student’s t-test. using small-molecule inhibitor BX795. Indeed, BX795 abrogated FcγR-TLR cross-talk-induced TNF production (Figure 3E). for the glycolytic rate), and the rate of oxygen consumption (OCR), as a measure of oxidative phosphorylation. Notably, stimulation with cIgG indeed increased the ECAR, which was even further enhanced upon co-stimulation with cIgG and Pam3 (Figure 4B). In contrast, the OCR was not aﬀected by individual stimulation with cIgG or Pam3, and only moderately increased upon co-stimulation (Figure 4C). Frontiers in Immunology \| www.frontiersin.org FcγR-TLR Cross-Talk Induces Glycolytic Reprogramming via IRF5 Ampliﬁcation of cytokine production can be orchestrated at both the transcriptional and translational level. Interestingly, upon FcγR co-stimulation of moDCs, the fold increase in expression of TNF mRNA was lower than that fold increase at the protein level (Figure 4A), suggesting that increased translation also contributes to the ampliﬁed cytokine response. In DCs, increased cytokine mRNA translation in response to TLR stimulation has been shown to be underpinned by a rapid increase in glycolytic rate, to serve as a carbon source for de novo fatty acid synthesis to support expansion of the endoplasmic reticulum required for increased cytokine gene translation (32, 33). This, together with the recent ﬁnding that IRF5 is able to increase the glycolysis in macrophages (34), led us to hypothesize that FcγR (co-) stimulation induces a similar metabolic reprogramming via IRF5 to support increased translation. To this end, we stimulated moDCs and analyzed them for changes in rates of extracellular acidiﬁcation (ECAR), as a measure of lactate production (a proxy To assess whether the increased glycolysis by FcγR-TLR cross-talk indeed contributes to the induction of cytokine responses, we stimulated moDCs in the presence of 2- deoxyglucose (2DG), which blocks glycolysis by inhibiting hexokinase activity (35). In line with previous ﬁndings, 2DG suppressed cytokine production induced by individual TLR stimulation (Figure 4E). In addition, 2DG also strongly suppressed cytokine production upon co-stimulation with cIgG and Pam3 (Figure 4E). Interestingly, while 2DG strongly impaired FcγR-TLR cross-talk-induced TNF protein production, blocking of glycolysis had very little eﬀect on FcγR-TLR cross- talk-induced TNF gene transcription (representative donor Figure 4F, multiple donors Figure 4G). These data indicate that April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 7 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. FIGURE 4 \| FcγR-TLR cross-talk induces glycolytic reprogramming via IRF5. (A) TNF fold increase in human monocyte derived DCs (moDCs) after co-stimulation with Pam3 and cIgG for mRNA and protein. Fold increase of TNF was determined by setting Pam3 stimulation at 1 and calculating fold increase after co-stimulation at t = 3 h (mRNA) or t = 24 h (protein). Mean + SD of four (mRNA) or eight (protein) experiments. (B,C) moDCs were stimulated for 30 min with Pam3, cIgG, or the combination and extracellular acidiﬁcation rate (ECAR) (B) and oxygen consumption rate (OCR) (C) was determined. Values are normalized to unstimulated moDCs for each experiment (set to 1). FcγR-TLR Cross-Talk Induces Glycolytic Reprogramming via IRF5 Mean + SD of four experiments. (D) Control or IRF5-silenced moDCs were stimulated with Pam3, cIgG, or the combination for 30 min and ECAR was measured. Representative experiment in triplicate of three independent experiments. (E,F) After 30 min pre-incubation with 10 mM 2-Deoxy-D-glucose (2DG), moDCs were stimulated with Pam3, cIgG, or in combination. (E) TNF production after 24 h was determined by ELISA; representative example in triplicate of eight independent experiments. (F) TNF mRNA expression was determined at indicated time points (normalized to housekeeping gene expression) by quantitative RT-PCR; representative example of four independent experiments. (G) TNF fold increase after co-stimulation with Pam3 and cIgG with and without 2DG. Fold increase of TNF was determined by setting Pam3 stimulation at 1 and calculating fold increase after co-stimulation at t = 3 h. Mean + SD of four experiments. Each pair of dots represent one donor. p < 0.05, Student’s t-test. FIGURE 4 \| FcγR-TLR cross-talk induces glycolytic reprogramming via IRF5. (A) TNF fold increase in human monocyte derived DCs (moDCs) after co-stimulation with Pam3 and cIgG for mRNA and protein. Fold increase of TNF was determined by setting Pam3 stimulation at 1 and calculating fold increase after co-stimulation at t = 3 h (mRNA) or t = 24 h (protein). Mean + SD of four (mRNA) or eight (protein) experiments. (B,C) moDCs were stimulated for 30 min with Pam3, cIgG, or the combination and extracellular acidiﬁcation rate (ECAR) (B) and oxygen consumption rate (OCR) (C) was determined. Values are normalized to unstimulated moDCs for each experiment (set to 1). Mean + SD of four experiments. (D) Control or IRF5-silenced moDCs were stimulated with Pam3, cIgG, or the combination for 30 min and ECAR was measured. Representative experiment in triplicate of three independent experiments. (E,F) After 30 min pre-incubation with 10 mM 2-Deoxy-D-glucose (2DG), moDCs were stimulated with Pam3, cIgG, or in combination. (E) TNF production after 24 h was determined by ELISA; representative example in triplicate of eight independent experiments. (F) TNF mRNA expression was determined at indicated time points (normalized to housekeeping gene expression) by quantitative RT-PCR; representative example of four independent experiments. (G) TNF fold increase after co-stimulation with Pam3 and cIgG with and without 2DG. Fold increase of TNF was determined by setting Pam3 stimulation at 1 and calculating fold increase after co-stimulation at t = 3 h. Mean + SD of four experiments. FcγR-TLR Cross-Talk Induces Glycolytic Reprogramming via IRF5 Each pair of dots represent one donor. *p < 0.05, Student’s t-test. (schematically depicted in Figure 5). While TLR stimulation induces IRF5 phosphorylation, FcγR stimulation results in IRF5 nuclear translocation. In addition, we identiﬁed that during FcγR-TLR cross-talk IRF5 ampliﬁes cytokine production in at least two diﬀerent ways. First, IRF5 increases cytokine gene transcription. Second, IRF5 induces glycolytic reprogramming, which ampliﬁes cytokine production in a post-transcriptional manner. the glycolytic changes induced by FcγR-TLR cross-talk, although essential for protein production, have little eﬀect on cytokine gene transcription. Taken together, these data identify that IRF5 activation by FcγR-TLR cross-talk does not only enhance cytokine gene transcription, but also boosts translation through glycolytic reprogramming that together account for the strongly increased pro-inﬂammatory proﬁle of moDCs activated by FcγR-TLR cross-talk. IRF5 is a transcription factor that was originally identiﬁed to be involved in type I interferon (IFN) production and antiviral responses. Over the last decade, multiple additional functions of IRF5 have been identiﬁed (24). Of these, the role of IRF5 in promoting transcription of pro-inﬂammatory cytokines such as TNF is the most pronounced (10–14). In addition, IRF5 expression has been identiﬁed as a marker to discriminate between subsets of macrophages, since IRF5 expression is higher in inﬂammatory macrophage subsets (12). Although IRF5 expression levels diﬀer between diﬀerent immune cells, our data indicate that IRF5 is required for Frontiers in Immunology \| www.frontiersin.org DISCUSSION Simultaneous activation of IRF5 by TLRs and FcγRs ampliﬁes pro-inﬂammatory cytokine production in two ways. First, IRF5 increases cytokine gene transcription. Second, IRF5 increases the glycolytic rate, which ampliﬁes cytokine production in a post-transcriptional manner. FcγRs such as FcγRIIa signal through an ITAM sequence in the cytoplasmic tail, which is a common signaling module used by a variety of receptors, including B cell receptors and T cell receptors, and other members of the Fc receptor family (42, 43). Interestingly, cross-talk with TLRs has previously been described for various other Fc receptor family members, including FcαRI (33, 44) and FcεRI (45, 46). In addition, Fc receptors have been shown to not only amplify cytokine responses induced by TLRs, but also by several other receptors such as NLRs, C-type lectins, IL-1R, and IFNγR (2, 33, 44). The fact that the cross-talk of diﬀerent Fc receptors with various PRRs and cytokine receptors in diﬀerent cell types all amplify pro-inﬂammatory cytokines in a similar manner suggests that the identiﬁed pathway may be a general mechanism of synergy between ITAM signaling receptors and PRRs, analogous to the previously described collaboration between the ITAM signaling module and JAK-STAT signaling pathways (42). FcγR-TLR cross-talk in various human APCs, including DCs and macrophages. We identiﬁed that IRF5 promotes inﬂammation by both enhancing gene transcription and by inducing glycolytic reprogramming. IRF5 is known to enhance gene transcription of pro-inﬂammatory genes such as TNF by both directly binding to IFN-stimulated response element (ISRE) regions in the TNF promoter, and by forming a complex with other transcription factors, speciﬁcally NF-κB subunit p65 (13). Transcriptional activation of IRF5 is strictly regulated by diﬀerent and independent post-translational modiﬁcations, to ensure initiation of appropriate immune response and prevent unrestrained inﬂammation. On one hand, IRF5 needs to be phosphorylated, which enables dimerization that is required for DNA binding (14, 25, 26, 28, 29). On the other hand, IRF5 needs to be translocated into the nucleus, which is achieved via K63-ubiquitination of IRF5 (25, 26, 36). Hence, either phosphorylation or ubiquitination individually are generally not suﬃcient for full IRF5 activation (24–26). Based on our ﬁndings and current literature we here propose a cooperation model of IRF5-dependent gene transcription upon FcγR-TLR cross- talk (schematically depicted in Figure 5). In this model, TLR stimulation induces TBK1/IKKε-dependent phosphorylation of IRF5, which is required for IRF5 activation. Additionally, FcγR stimulation induces Syk-dependent nuclear translocation of IRF5. DISCUSSION FcγR-TLR cross-talk in human myeloid APCs is an important initiator of inﬂammation during both infection and autoimmunity (1, 2, 6, 8). However, the molecular mechanisms underlying this cross-talk are still largely unknown. Here, we identiﬁed a crucial role for IRF5, which is activated by two diﬀerent pathways during FcγR-TLR co-stimulation to synergistically amplify pro-inﬂammatory cytokine production April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 8 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. FIGURE 5 \| Model for enhanced TNF production upon FcγR-TLR cross-talk via IRF5. TLR stimulation induces TBK1/IKKε phosphorylation that leads to IRF5 phosphorylation, while FcγR signaling induces IRF5 nuclear translocation. Simultaneous activation of IRF5 by TLRs and FcγRs ampliﬁes pro-inﬂammatory cytokine production in two ways. First, IRF5 increases cytokine gene transcription. Second, IRF5 increases the glycolytic rate, which ampliﬁes cytokine production in a post-transcriptional manner. of an E3 ligase that induces K63-ubiquitination of IRF5. Interestingly, Syk has been previously coupled to IRF5 activation, which was indeed independent of IRF5 phosphorylation (24, 37). In this regard, a relevant candidate E3 ligase is TRAF6 (38, 39), which has previously been identiﬁed to K63-ubiquinate IRF5 (36). Interestingly, also TRAF6 activation by Syk has been described to be dependent on K63-linked ubiquitination (40). Another candidate is Pellino-1, which additionally provides a connection between K63-ubiquitination of IRF5 and glucose metabolism (41). In addition to increasing gene transcription, we identiﬁed that FcγR-TLR cross-talk also induces glycolytic reprogramming by IRF5. Interestingly, this ﬁnding corroborates a recent study by Hedl et al., which shows that IRF5 regulates the glycolytic rate in human and murine macrophages (34). IRF5 increases the glycolysis upon NLR stimulation via activation of the kinase Akt2, which upregulates the transcription of various glycolytic genes (34). However, remarkably, the phosphorylation of Akt2, which is essential for Akt2 activation, is independent of IRF5 phosphorylation (34), suggesting that also other posttranslational modiﬁcations of IRF5 are required for increasing glycolysis. Since FcγR stimulation induces IRF5 nuclear translocation, which is dependent on K63-ubiquitination (36, 41), the increased glycolysis by FcγR-TLR cross-talk may therefore depend on multiple posttranslational modiﬁcations of IRF5, which ultimately lead to increased Akt2 activation and glycolysis. FIGURE 5 \| Model for enhanced TNF production upon FcγR-TLR cross-talk via IRF5. TLR stimulation induces TBK1/IKKε phosphorylation that leads to IRF5 phosphorylation, while FcγR signaling induces IRF5 nuclear translocation. Frontiers in Immunology \| www.frontiersin.org REFERENCES 13. Krausgruber T, Saliba D, Ryzhakov G, Lanfrancotti A, Blazek K, Udalova IA. IRF5 is required for late-phase TNF secretion by human dendritic cells. Blood. (2010) 115:4421–30. doi: 10.1182/blood-2010-01-263020 1. den Dunnen J, Vogelpoel LT, Wypych T, Muller FJ, de Boer L, Kuijpers TW, et al. IgG opsonization of bacteria promotes Th17 responses via synergy between TLRs and FcgammaRIIa in human dendritic cells. Blood. (2012) 120:112–21. doi: 10.1182/blood-2011-12-399931 1. den Dunnen J, Vogelpoel LT, Wypych T, Muller FJ, de Boer L, Kuijpers TW, et al. IgG opsonization of bacteria promotes Th17 responses via synergy between TLRs and FcgammaRIIa in human dendritic cells. Blood. (2012) 120:112–21. doi: 10.1182/blood-2011-12-399931 14. Ren J, Chen X, Chen ZJ. IKKbeta is an IRF5 kinase that instigates inﬂammation. Proc Natl Acad Sci USA. (2014) 111:17438–43. doi: 10.1073/pnas.1418516111 2. Vogelpoel LT, Hansen IS, Visser MW, Nagelkerke SQ, Kuijpers TW, Kapsenberg ML, et al. FcgammaRIIa cross-talk with TLRs, IL-1R, and IFNgammaR selectively modulates cytokine production in human myeloid cells. Immunobiology. (2015) 220:193–9. doi: 10.1016/j.imbio.2014.07.016 15. Mancl ME, Hu G, Sangster-Guity N, Olshalsky SL, Hoops K, Fitzgerald- Bocarsly P, et al. Two discrete promoters regulate the alternatively spliced human interferon regulatory factor-5 isoforms. Multiple isoforms with distinct cell type-speciﬁc expression, localization, regulation, and function. J Biol Chem. (2005) 280:21078–90. doi: 10.1074/jbc.M5005 43200 3. Vogelpoel LT, Baeten DL, de Jong EC, den Dunnen J. Control of cytokine production by human fc gamma receptors: implications for pathogen defense and autoimmunity. Front Immunol. (2015) 6:79. doi: 10.3389/ﬁmmu.2015.00079 16. Graham RR, Kozyrev SV, Baechler EC, Reddy MV, Plenge RM, Bauer JW, et al. A common haplotype of interferon regulatory factor 5 (IRF5) regulates splicing and expression and is associated with increased risk of systemic lupus erythematosus. Nat Genet. (2006) 38:550–5. doi: 10.1038/ ng1782 4. Egmond M, Vidarsson G, Bakema JE. Cross-talk between pathogen recognizing Toll-like receptors and immunoglobulin Fc receptors in immunity. Immunol Rev. (2015) 268:311–27. doi: 10.1111/imr.12333 17. Sigurdsson S, Padyukov L, Kurreeman FA, Liljedahl U, Wiman AC, Alfredsson L, et al. Association of a haplotype in the promoter region of the interferon regulatory factor 5 gene with rheumatoid arthritis. Arthritis Rheum. (2007) 56:2202–10. doi: 10.1002/art.22704 5. Ben Mkaddem S, Hayem G, Jönsson F, Rossato E, Boedec E, Boussetta T, et al. Shifting FcγRIIA-ITAM from activation to inhibitory conﬁguration ameliorates arthritis. J Clin Investig. (2014) 124:3945–59. doi: 10.1172/JCI74572 6. REFERENCES Bakema JE, Tuk CW, van Vliet SJ, Bruijns SC, Vos JB, Letsiou S, et al. Antibody-opsonized bacteria evoke an inﬂammatory dendritic cell phenotype and polyfunctional Th cells by cross-talk between TLRs and FcRs. J Immunol. (2015) 194:1856–66. doi: 10.4049/jimmunol.1303126 18. Dieguez-Gonzalez R, Calaza M, Perez-Pampin E, de la Serna AR, Fernandez- Gutierrez B, Castaneda S, et al. Association of interferon regulatory factor 5 haplotypes, similar to that found in systemic lupus erythematosus, in a large subgroup of patients with rheumatoid arthritis. Arthritis Rheum. (2008) 58:1264–74. doi: 10.1002/art.23426 7. Choy E. Understanding the dynamics: pathways involved in the pathogenesis of rheumatoid arthritis. Rheumatology. (2012) 51(Suppl. 5):v3–11. doi: 10.1093/rheumatology/kes113 19. Feng D, Stone RC, Eloranta ML, Sangster-Guity N, Nordmark G, Sigurdsson S, et al. Genetic variants and disease-associated factors contribute to enhanced interferon regulatory factor 5 expression in blood cells of patients with systemic lupus erythematosus. Arthritis Rheum. (2010) 62:562–73. doi: 10.1002/art.27223 8. Vogelpoel LT, Hansen IS, Rispens T, Muller FJ, van Capel TM, Turina MC, et al. Fc gamma receptor-TLR cross-talk elicits pro-inﬂammatory cytokine production by human M2 macrophages. Nat Commun. (2014) 5:5444. doi: 10.1038/ncomms6444 20. Xu WD, Pan HF, Xu Y, Ye DQ. Interferon regulatory factor 5 and autoimmune lupus. Expert Rev Mol Med. (2013) 15:e6. doi: 10.1017/erm.2013.7 9. Kumar H, Kawai T, Akira S. Toll-like receptors and innate immunity. Biochem Biophys Res Commun. (2009) 388:621–5. doi: 10.1016/j.bbrc.2009.08.062 21. Tang L, Chen B, Ma B, Nie S. Association between IRF5 polymorphisms and autoimmune diseases: a meta-analysis. Genet Mol Res. (2014) 13:4473–85. doi: 10.4238/2014.June.16.6 10. Takaoka A, Yanai H, Kondo S, Duncan G, Negishi H, Mizutani T, et al. Integral role of IRF-5 in the gene induction programme activated by Toll-like receptors. Nature. (2005) 434:243–9. doi: 10.1038/nature03308 22. Almuttaqi H, Udalova IA. Advances and challenges in targeting irf5, a key regulator of inﬂammation. FEBS J. (2018). doi: 10.1111/febs.14654. [Epub ahead of print]. 11. Ouyang X, Negishi H, Takeda R, Fujita Y, Taniguchi T, Honda K. Cooperation between MyD88 and TRIF pathways in TLR synergy via IRF5 activation. Biochem Biophys Res Commun. (2007) 354:1045–51. doi: 10.1016/j.bbrc.2007.01.090 23. Brennan FM, McInnes IB. Evidence that cytokines play a role in rheumatoid arthritis. J Clin Invest. (2008) 118:3537–45. doi: 10.1172/JCI36389 12. Krausgruber T, Blazek K, Smallie T, Alzabin S, Lockstone H, Sahgal N, et al. IRF5 promotes inﬂammatory macrophage polarization and TH1-TH17 responses. Nat Immunol. (2011) 12:231–8. doi: 10.1038/ni.1990 24. FUNDING Taken together, we identiﬁed IRF5 as a key component of FcγR-TLR cross-talk in human antigen-presenting cells. Our data strengthen the concept of a powerful pro-inﬂammatory role of IRF5 through ampliﬁcation of gene transcription and metabolic reprogramming. Because undesired activation by autoantibodies contributes to the pathogenesis of various chronic inﬂammatory disorders, targeting of FcγR-TLR signaling may be a valuable tool to suppress inﬂammation in diseases such as RA, systemic lupus erythematous (SLE), and inﬂammatory bowel disease (IBD). This work was supported by grants from the Netherlands Organization of Scientiﬁc Research (NWO; VENI, project no. 91611012) and the Amsterdam UMC (AMC Fellowship 2015). Grants were given to JdD. AUTHOR CONTRIBUTIONS We thank Sonja I. Gringhuis for providing scientiﬁc input during concept development. We thank Daisy I. Picavet and Ron A. Hoebe for assistance in (analysis of) ﬂuorescence microscopy experiments. JdD: conceptualization; WH, MN, LV, MK, BE, and JdD: methodology; WH, MN, LV, LS, IH, BE, and JdD: investigation; DISCUSSION Together, these two pathways cooperate leading to activated IRF5 inside the nucleus, thereby amplifying cytokine gene transcription. FcγR-TLR cross-talk provides protective immunity against various pathogens including bacteria and viruses (1, 3, 47), but is detrimental in various autoimmune diseases, since it strongly promotes the production of pathogenic pro-inﬂammatory cytokines (6, 8). Interestingly, IRF5 activation is also tightly associated with various chronic inﬂammatory disorders (17, 18, 21, 22). In addition, disease-associated IRF5 polymorphisms have previously been shown to dramatically aﬀect cytokine production by myeloid immune cells by both increasing gene transcription and glycolysis (34, 48). Disease-associated IRF5 polymorphisms are generally associated with higher IRF5 expression, but some polymorphisms also give rise to novel IRF5 isoforms (49). For future research, it would be very interesting to determine whether disease-associated IRF5 polymorphisms also promote cytokine How FcγRIIa triggering induces IRF5 nuclear translocation is still speculative, but it may result from Syk-dependent activation April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 9 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. WH, LV, and JdD: writing the original draft; MK, DB, BE, and JdD: reviewing and editing the manuscript. production by enhancing FcγR-TLR cross-talk. In addition, targeting of IRF5, or its upstream activators such as TBK1/IKKε, may open a new avenue for therapeutic intervention (22, 49). REFERENCES Ryzhakov G, Eames HL, Udalova IA. Activation and function of interferon regulatory factor 5. J Interferon Cytokine Res. (2015) 35:71–8. doi: 10.1089/jir.2014.0023 April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 10 FcγR-TLR Cross-Talk Depends on IRF5 Hoepel et al. 39. Lin Y-C, Huang D-Y, Chu C-L, Lin Y-L, Lin W-W. The tyrosine kinase Syk diﬀerentially regulates toll-like receptor signaling downstream of the adaptor molecules TRAF6 and TRAF3. Sci Signal. (2013) 6:ra71. doi: 10.1126/scisignal.2003973 25. Lin R, Yang L, Arguello M, Penafuerte C, Hiscott J. A CRM1-dependent nuclear export pathway is involved in the regulation of IRF-5 subcellular localization. J Biol Chem. (2005) 280:3088–95. doi: 10.1074/jbc.M4084 52200 26. Chang Foreman HC, Van SS, Cheng TF, Reich NC. Activation of interferon regulatory factor 5 by site speciﬁc phosphorylation. PLoS ONE. (2012) 7:e33098. doi: 10.1371/journal.pone.0033098 40. Wu NL, Huang DY, Tsou HN, Lin YC, Lin WW. Syk mediates IL-17-induced CCL20 expression by targeting Act1-dependent K63-linked ubiquitination of TRAF6. J Invest Dermatol. (2015) 135:490–8. doi: 10.1038/jid.2014.383 27. Barnes BJ, Kellum MJ, Field AE, Pitha PM. Multiple regulatory domains of IRF-5 control activation, cellular localization, and induction of chemokines that mediate recruitment of T lymphocytes. Mol Cell Biol. (2002) 22:5721–40. doi: 10.1128/MCB.22.16.5721-5740.2002 41. Kim D, Lee H, Koh J, Ko JS, Yoon BR, Jeon YK, et al. Cytosolic pellino-1-mediated K63-linked ubiquitination of IRF5 in M1 macrophages regulates glucose intolerance in obesity. Cell Rep. (2017) 20:832–45. doi: 10.1016/j.celrep.2017.06.088 28. Chen W, Lam SS, Srinath H, Jiang Z, Correia JJ, Schiﬀer CA, et al. Insights into interferon regulatory factor activation from the crystal structure of dimeric IRF5. Nat Struct Mol Biol. (2008) 15:1213–20. doi: 10.1038/nsmb.1496 42. Bezbradica JS, Rosenstein RK, DeMarco RA, Brodsky I, Medzhitov R. A role for the ITAM signaling module in specifying cytokine-receptor functions. Nat Immunol. (2014) 15:333–42. doi: 10.1038/ni.2845 43. Nimmerjahn F, Ravetch JV. Fcgamma receptors as regulators of immune responses. Nat Rev Immunol. (2008) 8:34–47. doi: 10.1038/nri2206 29. Lopez-Pelaez M, Lamont DJ, Peggie M, Shpiro N, Gray NS, Cohen P. Protein kinase IKKbeta-catalyzed phosphorylation of IRF5 at Ser462 induces its dimerization and nuclear translocation in myeloid cells. Proc Natl Acad Sci USA. (2014) 111:17432–7. doi: 10.1073/pnas.1418399111 44. Hansen IS, Hoepel W, Zaat SAJ, Baeten DLP, den Dunnen J. Serum IgA immune complexes promote proinﬂammatory cytokine production by human macrophages, monocytes, and Kupﬀer cells through FcalphaRI-TLR Cross-Talk. J Immunol. (2017) 199:4124–31. doi: 10.4049/jimmunol.1700883 30. REFERENCES Yu T, Yi YS, Yang Y, Oh J, Jeong D, Cho JY. The pivotal role of TBK1 in inﬂammatory responses mediated by macrophages. Mediat Inﬂamm. (2012) 2012:979105. doi: 10.1155/2012/979105 45. Suurmond J, Dorjée AL, Knol EF, Huizinga TWJ, Toes REM. Diﬀerential TLR-induced cytokine production by human mast cells is ampliﬁed by FcεRI triggering. Clin Exp Allergy. (2015) 45:788–96. doi: 10.1111/cea.12509 31. Kishore N, Huynh QK, Mathialagan S, Hall T, Rouw S, Creely D, et al. IKK-i and TBK-1 are enzymatically distinct from the homologous enzyme IKK-2: comparative analysis of recombinant human IKK-i, TBK-1, and IKK-2. J Biol Chem. (2002) 277:13840–7. doi: 10.1074/jbc.M110474200 46. Suurmond J, Stoop JN, Rivellese F, Bakker AM, Huizinga TWJ, Toes REM. Activation of human basophils by combined toll-like receptor- and FcεRI- triggering can promote Th2 skewing of naive T helper cells. Eur J Immunol. (2013) 44:386–96. doi: 10.1002/eji.201343617 32. Everts B, Amiel E, Huang SC, Smith AM, Chang CH, Lam WY, et al. TLR- driven early glycolytic reprogramming via the kinases TBK1-IKKepsilon supports the anabolic demands of dendritic cell activation. Nat Immunol. (2014) 15:323–32. doi: 10.1038/ni.2833 47. Newling M, Hoepel W, Vogelpoel LTC, Heineke MH, van Burgsteden JA, Taanman-Kueter EWM, et al. Fc gamma receptor IIa suppresses type I and III interferon production by human myeloid immune cells. Eur J Immunol. (2018) 48:1796–809. doi: 10.1002/eji.201847615 33. Hansen IS, Krabbendam L, Bernink JH, Loayza-Puch F, Hoepel W, van Burgsteden JA, et al. FcαRI co-stimulation converts human intestinal CD103(+) dendritic cells into pro-inﬂammatory cells through glycolytic reprogramming. Nat Commun. (2018) 9:863. doi: 10.1038/s41467-018-03318-5 48. Hedl M, Abraham C. IRF5 risk polymorphisms contribute to interindividual variance in pattern recognition receptor-mediated cytokine secretion in human monocyte-derived cells. J Immunol. (2012) 188:5348–56. doi: 10.4049/jimmunol.1103319 34. Hedl M, Yan J, Abraham C. IRF5 and IRF5 disease-risk variants increase glycolysis and human M1 macrophage polarization by regulating proximal signaling and Akt2 activation. Cell Rep. (2016) 16:2442–55. doi: 10.1016/j.celrep.2016.07.060 49. Kaur A, Lee LH, Chow SC, Fang CM. IRF5-mediated immune responses and its implications in immunological disorders. Int Rev Immunol. (2018) 37:229–48. doi: 10.1080/08830185.2018.1469629 35. Wick AN, Drury DR, Nakada HI, Wolfe JB, Britton WttaoB, Grabowski R. Localization of the primary metabolic block produced by 2-deoxyglucose. J Biol Chem. (1957) 224:963–9. Frontiers in Immunology \| www.frontiersin.org Conﬂict of Interest Statement: DB is also an employee of Union Chimique Belge. Conﬂict of Interest Statement: DB is also an employee of Union Chimique Belge. 36. Balkhi MY, Fitzgerald KA, Pitha PM. Functional regulation of MyD88- activated interferon regulatory factor 5 by K63-linked polyubiquitination. Mol Cell Biol. (2008) 28:7296–308. doi: 10.1128/MCB.00662-08 The remaining authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. 37. del Fresno C, Soulat D, Roth S, Blazek K, Udalova I, Sancho D, et al. Interferon-β production via dectin-1-Syk-IRF5 signaling in dendritic cells is crucial for immunity to C. albicans. Immunity. (2013) 38:1176–86. doi: 10.1016/j.immuni.2013.05.010 Copyright © 2019 Hoepel, Newling, Vogelpoel, Sritharan, Hansen, Kapsenberg, Baeten, Everts and den Dunnen. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. 38. Yamada T, Fujieda S, Yanagi S, Yamamura H, Inatome R, Yamamoto H, et al. IL-1 induced chemokine production through the association of Syk with TNF receptor-associated factor-6 in nasal ﬁbroblast lines. J Immunol. (2001) 167:283. doi: 10.4049/jimmunol.167.1.283 April 2019 \| Volume 10 \| Article 739 Frontiers in Immunology \| www.frontiersin.org 11
https://openalex.org/W2557637879	https://europepmc.org/articles/pmc5356820?pdf=render	English	null	<i>KRAS</i> exon 2 codon 13 mutation is associated with a better prognosis than codon 12 mutation following lung metastasectomy in colorectal cancer	Oncotarget	2,016	cc-by	6,404	Stéphane Renaud1,2,3, Francesco Guerrera1,4, Joseph Seitlinger1, Lorena Costardi4, Mickaël Schaeffer5, Benoit Romain3,6, Claudio Mossetti4, Anne Claire-Voegeli7, Pier Luigi Filosso4, Michèle Legrain7, Enrico Ruffini4, Pierre-Emmanuel Falcoz1, Alberto Oliaro4, Gilbert Massard1 Stéphane Renaud1,2,3, Francesco Guerrera1,4, Joseph Seitlinger1, Lorena Costardi4, Mickaël Schaeffer5, Benoit Romain3,6, Claudio Mossetti4, Anne Claire-Voegeli7, Pier Luigi Filosso4, Michèle Legrain7, Enrico Ruffini4, Pierre-Emmanuel Falcoz1, Alberto Oliaro4, Gilbert Massard1 1Department of Thoracic Surgery, Strasbourg University Hospital, Strasbourg, France 2Department of Thoracic Surgery, Nancy University Hospital, Nancy, France 3Research Unit EA3430, Tumoral Progression and Micro-Environment, Epidemiological and Translational Approaches, Strasbourg University, Strasbourg, France 4Department of Thoracic Surgery, Azienda Ospedaliera Universitaria Città della Salute e della Scienza di Torino, Torino, Italy 5Department of Biostatistics, Strasbourg University Hospital, Strasbourg, France 6Department of General and Digestive Surgery, Strasbourg University Hospital, Strasbourg, France 7Department of Molecular Biology, Strasbourg University Hospital, Strasbourg, France Correspondence to: Stéphane Renaud, email: sterenaud0@gmail.com Keywords: lung metastasectomy, colorectal cancer, KRAS, surgery, codon Received: August 25, 2016 Accepted: November 21, 2016 Published: November 29, 2016 Stéphane Renaud1,2,3, Francesco Guerrera1,4, Joseph Seitlinger1, Lorena Costardi4, Mickaël Schaeffer5, Benoit Romain3,6, Claudio Mossetti4, Anne Claire-Voegeli7, Pier Luigi Filosso4, Michèle Legrain7, Enrico Ruffini4, Pierre-Emmanuel Falcoz1, Alberto Oliaro4, Gilbert Massard1 1Department of Thoracic Surgery, Strasbourg University Hospital, Strasbourg, France 2Department of Thoracic Surgery, Nancy University Hospital, Nancy, France 3Research Unit EA3430, Tumoral Progression and Micro-Environment, Epidemiological and Translational Approaches, Strasbourg University, Strasbourg, France 4Department of Thoracic Surgery, Azienda Ospedaliera Universitaria Città della Salute e della Scienza di Torino, Torino, Italy 5Department of Biostatistics, Strasbourg University Hospital, Strasbourg, France 6Department of General and Digestive Surgery, Strasbourg University Hospital, Strasbourg, France 7Department of Molecular Biology, Strasbourg University Hospital, Strasbourg, France Correspondence to: Stéphane Renaud, email: sterenaud0@gmail.com Keywords: lung metastasectomy, colorectal cancer, KRAS, surgery, codon Received: August 25, 2016 Accepted: November 21, 2016 Published: November 29, 2016 Received: August 25, 2016 Accepted: November 21, 2016 Published: Novemb www.impactjournals.com/oncotarget/ www.impactjournals.com/oncotarget/ KRAS exon 2 codon 13 mutation is associated with a better prognosis than codon 12 mutation following lung metastasectomy in colorectal cancer Stéphane Renaud1,2,3, Francesco Guerrera1,4, Joseph Seitlinger1, Lorena Costardi4, Mickaël Schaeffer5, Benoit Romain3,6, Claudio Mossetti4, Anne Claire-Voegeli7, Pier Luigi Filosso4, Michèle Legrain7, Enrico Ruffini4, Pierre-Emmanuel Falcoz1, Alberto Oliaro4, Gilbert Massard1 1Department of Thoracic Surgery, Strasbourg University Hospital, Strasbourg, France 2Department of Thoracic Surgery, Nancy University Hospital, Nancy, France 3Research Unit EA3430, Tumoral Progression and Micro-Environment, Epidemiological and Translational Approaches, Strasbourg University, Strasbourg, France 4Department of Thoracic Surgery, Azienda Ospedaliera Universitaria Città della Salute e della Scienza di Torino, Torino, Italy 5Department of Biostatistics, Strasbourg University Hospital, Strasbourg, France 6Department of General and Digestive Surgery, Strasbourg University Hospital, Strasbourg, France 7Department of Molecular Biology, Strasbourg University Hospital, Strasbourg, France Correspondence to: Stéphane Renaud, email: sterenaud0@gmail.com Keywords: lung metastasectomy, colorectal cancer, KRAS, surgery, codon Received: August 25, 2016 Accepted: November 21, 2016 Published: November 29, 2016 Survival analyses Ninety-five patients (63%) were reported to be alive, and 55 patients (37%) were dead during the follow-up period. Overall, the five-year survival rate was 42%. Survival analysis by KRAS codon mutation showed a non-significant difference between codon 12 and codon 13 mutations (median OS (mOS): 84 months vs 82 months, respectively; P = 0.167). However, bevacizumab showed a survival benefit when used in case of KRAS codon 12 mutations (mOS: Not reached (NR) vs 54 months, P < 0.001), but not in KRAS codon 13 mutations (mOS: NR vs 82 months; P = 0.48). Patients were then further analyzed with the exclusion of patients treated with bevacizumab. Hence, survival analysis in patients not treated with bevacizumab showed a significant difference between codon 12 and codon 13 mutations (mOS: 54 months vs 82 months, respectively; P = 0.009 - Figure 1). Otherwise, because among KRAS codon 12 mutations, G12D and G12V were the most frequent, other codon 12 mutations were pooled in the “other codon 12 mutations” group. There was no significant difference in mOS among KRAS codon 12 mutations (mOS: G12D 55, months (95% CI: 47.8–62) vs G12V, 55 months (95% CI 38.63–71.38) vs other codon 12 mutations, 53 months (95% CI 36.29–69.71); P = 0.78). p g g [ ] In CRC, two proto-oncogenes have been extensively studied: V-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) and V-raf murine sarcoma viral oncogene homolog B1 (BRAF). The prognostic and predictive values of mutations of these two genes in metastatic CRC is now clearly defined, particularly leading to resistance to anti- Epidermal Growth Factor Receptor (EGFR) therapies [12]. On the other hand, although a growing number of publications have focused on evaluating the prognostic value of KRAS mutations after lung metastasectomy of CRC, clinical data are still inconsistent [8–11]. Furthermore, clinical studies on lung metastases of CRC seem to provide only a fleeting glimpse of what the extent the molecular biology of cancer cells could offer in our daily practice. Particularly, the tumoral heterogeneity of KRAS mutations remains intriguing. Indeed, recently published data have highlighted the existence of two distinct groups of KRAS cells: KRAS-dependent and KRAS- independent cells [13]. RESULTS Colorectal cancer (CRC) remains one of the most diagnosed cancers and one of the leading causes of cancer- related deaths worldwide [1, 2]. Because the liver and lungs offer a favorable environment for CRC cells, up to 50% of the patients will experience metastases in these two locations [3]. Although still the subject of debate, lung metastasectomy is widely accepted by most surgical teams because, in selected populations, it leads to longer overall survival (OS) than conventional 5-fluorouracil (5-FU)- based chemotherapy regimens [4–6]. According to the selection criteria, 150 patients with lung metastasis of CRC harboring KRAS mutations were included in this study. The clinical-pathological characteristics of these patients are displayed in Table 1. Median follow-up time was 56 months (IQR: 44). Analyses of KRAS codon 12 transversions revealed 12 (8%) G12C, 39 (26%) G12V and 7 (5%) G12A. Analyses of KRAS codon 12 transitions revealed 49 (33%) G12D and 9 (6%) G12S. For KRAS codon 13, 34 (23%) cases of G13D transition were observed. Several risk factors for poor outcomes have been identified to properly select patients who would clearly benefit from surgery, both in terms of OS and disease-free survival (DFS). Moreover, recent meta-analyses have helped to define main clinical prognostic factors [7]; in the past few years, there has been an increased understanding of the molecular alterations in cancer cells, with the identification of oncogenic drivers, suggesting that clinical factors could be a reflection of only gene mutations. These observations support the perspective of a molecular classification of patients, which carries with it the possibility of a better selection process for good surgical candidates [8–11]. ABSTRACT Introduction: The utilization of molecular markers as routinely used biomarkers is steadily increasing. We aimed to evaluate the potential different prognostic values of KRAS exon 2 codons 12 and 13 after lung metastasectomy in colorectal cancer (CRC). Results: KRAS codon 12 mutations were observed in 116 patients (77%), whereas codon 13 mutations were observed in 34 patients (23%). KRAS codon 13 mutations were associated with both longer time to pulmonary recurrence (TTPR) (median TTPR: 78 months (95% CI: 50.61–82.56) vs 56 months (95% CI: 68.71–127.51), P = 0.008) and improved overall survival (OS) (median OS: 82 months vs 54 months (95% CI: 48.93–59.07), P = 0.009). Multivariate analysis confirmed that codon 13 mutations were associated with better outcomes (TTPR: HR: 0.40 (95% CI: 0.17–0.93), P = 0.033); OS: HR: 0.39 (95% CI: 0.14–1.07), P = 0.07). Otherwise, no significant difference in OS (P = 0.78) or TTPR (P = 0.72) based on the type of amino-acid substitutions was observed among KRAS codon 12 mutations. Materials and Methods: We retrospectively reviewed data from 525 patients who underwent a lung metastasectomy for CRC in two departments of thoracic surgery from 1998 to 2015 and focused on 150 patients that had KRAS exon 2 codon 12/13 mutations. Conclusions: KRAS exon 2 codon 13 mutations, compared to codon 12 mutations, seem to be associated with better outcomes following lung metastasectomy in CRC. Prospective multicenter studies are necessary to fully understand the prognostic value of KRAS mutations in the lung metastases of CRC. www.impactjournals.com/oncotarget Oncotarget 2514 Clinicopathological variables and KRAS mutations Table 1 shows the distribution of clinicopathologic variables according to KRAS codon mutation. Survival analyses Moreover, it seems that according to KRAS amino-acid substitution, different downstream signaling pathways are activated [14], likely leading to different clinical behaviors such as different degrees of aggressiveness [15], different sites of metastasis [16] and/ or different sensitivity to chemotherapy [17] and radiation therapy18. Hence, few studies on metastatic CRC tried to evaluate the prognostic significance of KRAS codon 12 and codon 13 mutations, with contradictory results [19–22]. However, to the best of our knowledge, this axis has not been yet investigated in lung metastases of CRC. Because Food and Drug Administration approval of FOLFOX and bevacizumab in CRC were both obtained in 2004, we decided to compare OS between patients who underwent a lung metastasectomy before and after 2004. The median OS of patients included before 2004 (n = 100) was not significantly different from those included thereafter (n = 50) (median OS: 75 months (95% CI: 55.25–92.68) vs 83 months (95% CI: 72.32–102.18), respectively, p = 0.22) Finally, in univariate analysis, male gender (P = 0.039), CCI (P < 0.001), WHO performance status P < 0.025), pT of CRC (P < 0.001), DFS (P = 0.006), thoracic LNI (P < 0.001) and liver metastasis P < 0.001) Because Food and Drug Administration approval of FOLFOX and bevacizumab in CRC were both obtained in 2004, we decided to compare OS between patients who underwent a lung metastasectomy before and after 2004. The median OS of patients included before 2004 (n = 100) was not significantly different from those included thereafter (n = 50) (median OS: 75 months (95% CI: 55.25–92.68) vs 83 months (95% CI: 72.32–102.18), respectively, p = 0.22) We thereby aimed to evaluate the different prognostic value of KRAS exon 2 codon 12 over codon 13 mutations in a large surgical cohort of resected lung metastases. www.impactjournals.com/oncotarget Oncotarget 2515 Oncotarget 2516 www.impactjournals.com/oncotarget Table 1: Demographic data and main covariates according to KRAS mutational status Total (n = 150) Codon 12 (n = 116) Codon 13 (n = 34) P No. % No. Col % No. Oncotarget Survival analyses CCI: Charlson Comorbidity Index, WHO-PS: World Health Organization performance status, CRC: colorectal cancer, CEA: carcinoembryonic antigen, DFS: disease free survival, LNI: lymph node involvement, IQR: inter-quartile range. number of patients for whom the corresponding data were available. CCI: Charlson Comorbidity Index, WHO-PS: World Health Organization performance status, CRC: colorectal cancer, CEA: carcinoembryonic antigen, DFS: disease free survival, LNI: lymph node involvement, IQR: inter-quartile range. showed a significant difference between codon 12 (Median TTPR (mTTPR): 60 months) and codon 13 mutations (mTTPR: NR; P = 0.041). However, bevacizumab showed a benefit on TTPR when used in the case of KRAS codon 12 mutations (mTTPR: 86 (95% CI: 57–115) vs 30 months (95% CI: 12–48), P < 0.001), but not in KRAS codon 13 mutations (mTTPR: 31 (95% CI: 15–47) vs 46 months (95% CI: 13– 79.18), P = 0.39). Patients were then further analyzed, with the exclusion of patients treated with bevacizumab. Likewise, TTPR analysis in patients not treated with bevacizumab showed a significant difference between codon 12 (mTTPR: 51 months) and codon 13 mutations (mTTPR: NR; P = 0.008 - Figure 2). Otherwise, among KRAS codon 12 mutations, even shorter in case of G12V transversion (mTTPR: 21 months, 95% CI: 0–43), there was no significant difference with G12D transition (50 months, 95% CI: 42–58) and “other codon 12 mutations” (30 months, 95% CI: 0–68, P = 0.72). were found to have a negative effect on survival. Data are shown in Table 2.i Patients who had benefited from bevacizumab were excluded from the multivariate analysis to exclude the effect of bevacizumab on the OS of KRAS codon 12 patients. Hence, the multivariate-adjusted model showed that CCI and pT of primary CRC were independent negative predictors of survival. Finally, there was a trend toward improved OS in KRAS codon 13 mutations (HR = 0.386, P = 0.068). Data are displayed in Table 2. Survival analyses Col % Bevacizumab (Yes) 37 25 27 23 10 29 0.465 Age (Mean; SD) 64 8 64 8 65 9 0.713 Sex (male) 83 55 67 58 16 47 0.328 Median follow-up time (Median, IQR) 56 44 55 43 58 47 0.89 CCI (n = 149) 0.115 0 41 28 34 29 7 21 1–2 56 38 38 33 18 55 3–4 31 21 25 22 6 18 >= 5 21 14 19 16 2 6 WHO-PS 0.708 0 112 75 85 73 27 79 1 34 23 27 23 7 21 2 3 2 3 3 3 1 1 1 1 Site of primary CRC 0.489 Colon 74 49 59 51 15 44 Rectum 76 51 57 49 19 56 pT CRC (n = 148) 0.61 I 16 11 12 11 4 12 II 21 14 14 12 7 21 III 88 59 69 61 19 56 IV 23 16 19 17 4 12 pN CRC (n = 148) 0.007 0 87 59 64 56 23 68 N1 31 21 29 25 2 6 N2 28 19 21 18 7 21 N3 2 1 2 6 CEA (n = 143) 0.584 0–5 ng/ml 94 66 71 65 23 70 > 5 ng/ml 49 34 39 35 10 30 Chemotherapy 0.062 None 39 26 29 25 10 29 Peri-operative 27 18 16 14 11 32 Post-operative 48 32 41 35 7 21 Pre-operative 36 24 30 26 6 18 Side of metastasis Bilateral 39 26 30 26 9 26 Unilateral 111 74 86 74 25 74 DFS (> 12 months) 121 81 95 82 26 76 0.481 DFS (24 months) 92 61 73 63 19 56 0.458 Number of lung metastases (> 1) 76 51 57 49 19 56 0.489 Thoracic LNI (N+; n = 148) 48 32 39 34 9 26 0.397 www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 2516 Liver Metastasis 24 16 19 16 5 15 0.815 KRAS mutation subtype G12A 7 5 7 6 G12C 12 8 12 10 G12D 49 33 49 42 G12S 9 6 9 8 G12V 39 26 39 34 G13D 34 23 34 100 number of patients for whom the corresponding data were available. CCI: Charlson Comorbidity Index, WHO-PS: World Health Organization performance status, CRC: colorectal cancer, CEA: carcinoembryonic antigen, DFS: disease free survival, LNI: lymph node involvement, IQR: inter-quartile range. *number of patients for whom the corresponding data were available. Time to pulmonary recurrence (TTPR) One hundred forty patients (93%) were available for the TTPR analysis; 68 patients (48%) experienced a lung recurrence during the follow-up period. Overall, the 5-year TTPR rate was 45%. TTPR analysis by KRAS codon mutation Figure 1: Kaplan-meier overall survival according to codon 12 or 13 mutations in patients not treated with bevacizumab. Figure 1: Kaplan-meier overall survival according to codon 12 or 13 mutations in patients not treated with bevacizumab. Figure 1: Kaplan-meier overall survival according to codon 12 or 13 mutations in patients not treated with bevacizumab. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 2517 Table 2: Uni- and multivariate analysis on overall survival (OS) Oncotarget 2518 www.impactjournals.com/oncotarget Table 2: Uni- and multivariate analysis on overall survival (OS) UNIVARIATE (N = 150) Median survival (months) P [95% Conf. Interval] MULTIVARIATE (without bevacizumab; N = 113) Hazard Ratio P [95% Conf. Oncotarget Time to pulmonary recurrence (TTPR) Interval] Mutation subtype 0.167 Codon 12 84 51 116 Codon 13 82 – – Bevacizumab < 0.001 No 56 54 82 Yes 101 84 – Mutation subtype without bevacizumab (n = 113) 0.009 Mutation subtype Codon 12 54 47 60 Codon 12 1 Codon 13 82 55 – Codon 13 0.386 0.068 0.14 1.07 Age 0.112 Age < 60 92 92 – < 60 1 >= 60 82 55 101 >= 60 1.349 0.453 0.62 2.94 Sex 0.039 Sex Female NR 84 – Female 1 Male 82 55 – Male 1.195 0.645 0.56 2.55 CCI < 0.001 CCI 0 101 84 – 0 1 1–2 NR 55 – 1–2 5.063 0.020 1.28 19.96 3–4 54 35 – 3–4 8.995 0.003 2.06 39.22 >= 5 17 17 – >= 5 23.415 0.001 3.41 160.69 WHO-PS 0.025 WHO-PS 0 92 66 – 0 1 1 82 45 – 1 0.988 0.977 0.43 2.24 2 33 33 – 2 5.602 0.163 0.49 62.89 3 NR – – 3 Site of primary CRC 0.819 Colon NR 55 – Rectum 84 60 101 pT Colon < 0.001 pT Colon T1 NR 52 – T1 1 T2 82 47 – T2 4.691 0.075 0.85 25.79 T3 84 60 – T3 1.857 0.406 0.43 7.99 T4 27 17 32 T4 9.190 0.029 1.25 67.49 pN Colon 0.066 N0 82 55 – N1 60 33 84 N2 NR – – N3 NR – – CEA 0.092 CEA 0–5 ng/m 82 55 – 0–5 ng/m 0.895 0.779 0.41 1.94 > 5 ng/ml 92 84 – > 5 ng/ml 1 Chemotherapy 0.629 www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 2518 None 82 55 – Peri-operative NR 37 – Post-operative 92 55 – Pre-operative 60 37 84 Side of surgery 0.401 Unilateral 82 60 – Bilateral 101 38 – Disease free survival 0.004 Disease free survival < 24 months 101 82 – < 24 months 1 >= 24 months 56 53 92 >= 24 months 1.561 0.269 0.71 3.44 Number of lung metastasis 0.457 1 101 60 – > 1 82 55 92 Thoracic LNI < 0.001 Thoracic LNI No 92 82 – No 1 Yes 50 37 56 Yes 2.075 0.104 0.86 5.01 Liver Metastasis < 0.001 Liver Metastasis No 84 82 – No 1 Yes 47 37 – Yes 1.509 0.379 0.61 3.77 CCI: Charlson Comorbidity Index, WHO: World Health Organization Performance Status, CRC: colorectal cancer, CEA: carcinoembryonic antigen, DFS: disease free survival, LNI: lymph node involvement. Charlson Comorbidity Index, WHO: World Health Organization Performance Status, CRC: colorectal cancer, CE gen, DFS: disease free survival, LNI: lymph node involvement. recurrence and decreased OS after lung metastasectomy in CRC [26]. These two observations suggest a higher aggressiveness of codon 12 mutations over codon 13 mutations. Although there have been no published data on the prognostic value of exon 2 codon mutations after lung metastasectomy in CRC to date, several authors have previously attempted to investigate its value in metastatic CRC. However, few data are available, and the studied populations have been heterogeneous, including all stages of CRC. In a collaborative study including 3439 CRC patients, Andreyev et al. [20] found a significant association between G12V mutation and both failure- free survival (HR: 1.3, P = 0.004) and OS (HR: 1.29, P = 0.008). However, the impact of this mutation was only observed in Duke’s C stage, but not in Duke’s B or in an advanced stage. In another work, Imamura et al. [21] confirmed in 1075 CRC patients the association between G12V mutation and worse OS (HR: 2.00 P = 0.0003), but without significant impact on prognosis of codon 13 mutations. In contrast, Samowitz et al. [22], in a large study of 1413 patients, although non-significant, found a 40% increase in short-term mortality from CRC in the case of codon 13 mutations. More recently, among 218 metastatic CRC patients, Dadduzio et al. [19] did not find any prognostic difference between codon 12 and codon 13 mutations. www.impactjournals.com/oncotarget To the best of our knowledge, our study is the first to focus on the prognostic value of these 2 codon mutations after lung metastasectomy in CRC. Consistent with in vitro studies, our work suggests a higher aggressiveness of codon 12 over codon Finally, in univariate analysis, the primary CRC site (P = 0.033), pT of CRC (P < 0.001), side of lung metastasis (P < 0.001), thoracic LNI (P = 0.015) and liver metastasis (P = 0.002) were found to have a negative effect on TTPR. Data are shown in Table 3.i Patients who had benefited from bevacizumab were excluded from the multivariate analysis to exclude the effect of bevacizumab on the mTTPR of KRAS codon 12 patients. Hence, at multivariate analysis, pT of CRC, side of lung metastasis and liver metastasis demonstrated an independent negative effect on TTPR. Moreover, an independent effect on improved TTPR in KRAS codon 13 mutations was observed (HR 0.40, P = 0.033). Data are shown in Table 3. DISCUSSION In vitro studies suggest a wide heterogeneity of KRAS mutations, suggesting different prognostic values of exon 2 codon 12 and codon 13 mutations. More specifically, exon 2 codon 12 mutations, in particular, G12D and G12V mutations, seem to be able to induce a more robust link with GTP molecules leading to higher resistance to GTPase activity compared to codon 13 mutations, in particular, G13D [23, 24]. On the other hand, it seems that KRAS codon 12 mutations, but not codon 13 mutations, are associated with a strong upregulation of vascular endothelial growth factor (VEGF) [24, 25], which is implicated in the promotion of lymphangiogenesis [26], which seems to be related to www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 2519 Uni- and multivariate analysis on time to pulmonary recurrence (TTPR) Table 3: Uni and multivariate analysis on time to pulmonary recurrence (TTPR) UNIVARIATE (N = 140) Median TTPR (months) P [95% Conf. Interval] MULTIVARIATE (without bevacizumab; N= 113) Hazard ratio P [95% Conf. Interval] Mutation subtype 0.041 Codon 12 60 51 86 Codon 13 NR 50 – Bevacizumab 0.049 No 51 50 100 Yes 92 70 98 Mutation subtype without bevacizumab (n = 113) 0.009 Mutation subtype without bevacizumab (n = 113) Codon 12 51 41 60 Codon 12 1 Codon 13 NR 50 – Codon 13 0.40 0.033 0.17 0.93 Age 0.891 < 60 86 50 – >= 60 60 51 98 Sex 0.206 Female 70 51 100 Male 57 51 – Site of primary CRC 0.033 Site of primary CRC Colon 51 49 70 Colon 1.29 0.440 0.67 2.48 Rectum 86 60 98 Rectum 1 pT Colon < 0.001 pT Colon T1 98 51 112 T1 1 T2 NR 20 – T2 8.14 0.016 1.47 45.11 T3 60 51 86 T3 5.17 0.017 1.34 19.86 T4 15 12 24 T4 24.19 < 0.001 5.11 114.65 pN Colon 0.529 N0 57 50 86 N1 NR 16 – N2 NR 51 – N3 NR – – CEA 0.283 0–5 ng/m 60 50 86 > 5 ng/ml 92 51 100 Chemotherapy 0.241 None 100 50 – Peri-operative 57 40 – Post-operative 86 51 98 Pre-operative 70 24 – www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 2520 Side of surgery < 0.001 Side of surgery Unilateral 112 60 – Unilateral 1 Bilateral 40 16 57 Bilateral 5.65 < 0.001 2.62 12.21 Disease free survival 0.489 < 24 months 60 51 98 >= 24 months 70 51 112 Number of nodules 0.409 1 70 51 98 > 1 60 50 – Thoracic LNI 0.015 Thoracic LNI No 86 60 92 No 1 Yes 50 30 98 Yes 1.86 0.101 0.89 3.89 Liver Metastasis 0.002 Liver Metastasis No 86 57 92 No 1 Yes 44 19 49 Yes 2.52 0.031 1.09 5.84 CCI: Charlson Comorbidity Index, WHO: World Health Organization Performance Status, CRC: colorectal cancer, CEA: carcinoembryonic antigen, DFS: disease free survival, LNI: lymph node involvement. CCI: Charlson Comorbidity Index, WHO: World Health Organization Performance Status, CRC: colorectal cancer, CEA: carcinoembryonic antigen, DFS: disease free survival, LNI: lymph node involvement. 13 mutations in terms of both OS and TTPR. Unlike previous authors [20, 21], we did not find any significant difference among codon 12 mutations. Indeed, even if G12V transversions were associated with decreased TTPR, there was no significant difference with all other types of codon 12 amino-acid substitutions. However, our population was small, and the absence of differences may have been related to a lack of power. Furthermore, the “other codon 12 mutations” group included different types of amino-acid substitutions; behaviors of each type Figure 2: Kaplan-meier time to pulmonary recurrence according to codon 12 or 13 mutations in patients not treated with bevacizumab. Figure 2: Kaplan-meier time to pulmonary recurrence according to codon 12 or 13 mutations in patients not treated with bevacizumab. www.impactjournals.com/oncotarget Oncotarget 2521 of amino-acid substitution could not have been studied because of the small number of patients in each group. Indeed, in vitro studies have shown that the type of amino- acid substitution may activate different downstream signaling [14]. Hence, both KRAS G12C and G12V exhibited activated Ral signaling and decreased growth factor-dependent Akt activation, although the G12D mutation exhibited activated PI3K and MEK signaling. Consequently, it would not be surprising that different amino-acid substitutions confer different behavior to cancer cells after lung metastasectomy in CRC with different prognoses as observed in other cancers. and bevacizumab introduction. Furthermore, the molecular data were obtained from the primary CRC and not from the metastatic tumors. MATERIALS AND METHODS As previously published by our team in a single- institutional study [28, 27], we confirmed in this multi- center study that bevacizumab may improve both OS and TTPR in patients with KRAS codon 12 mutations, but not with codon 13 mutations. This study was approved by the Ethics Committee of the French Society of Thoracic and Cardiovascular Surgeons (Approval Number: 2016–8-4–21–0-58-ReSt). We retrospectively reviewed the data from 525 unselected and consecutive patients with metastatic CRC who underwent a lung metastasectomy in both the Thoracic Surgery Department of Strasbourg University Hospital (France) and Torino University Hospital (Italy) from January 1998 to December 2015. Patients for whom mutational status was unknown (i.e., KRAS and BRAF) (n = 92), who exhibited a BRAF mutation (n = 26), and who did not harbor KRAS/BRAF mutations (n = 257) were excluded from this study. All included patients were considered completely cured of their primary tumor at the time of the thoracic metastasectomy, and all pulmonary metastases were metachronous. In cases of extra-thoracic metastases, only patients with hepatic metastases were included. All thoracic resections were considered R0. Finally, in our cohort known prognostic factors, namely thoracic LNI, liver metastases, DFS and pre- operative CEA did not significantly impact the OS. These observations may be related to our highly selected population, or on the other hand to the fact that mutational status is statistically more powerful to impact OS than other prognostic factors, and that these prognostic factors, as previously evoked, could be only reflections of mutational status [8]. However, our study must be interpreted with caution regarding a few limitations. First, it is a retrospective cohort study based on a relatively small sample size. Furthermore, 122 patients were from Strasbourg cohort, probably leading to a center effect. However, to the best of our knowledge, this is the largest published cohort on KRAS mutations in lung metastases of CRC, and a multicentre cohort seemed necessary to obtain enough statistical power in this highly selected population. Furthermore, our study covers a 17-year period during which there may have been changes in the management of patients; the different chemotherapy regimens used may have also influenced the survival of the study participants. However, a long study period was required to obtain enough statistical power. www.impactjournals.com/oncotarget Consequently, there is remaining doubt regarding the degree of concordance between the primary and metastatic tumors, although Cejas et al. [29] reported a concordance rate of 94%. Finally, data on extra-thoracic recurrence were not available; this would be interesting to investigate in prospective studies. In conclusion, to the best of our knowledge this is the first work on the largest published cohort of lung metastases with KRAS mutations showing that exon 2 codon 13 mutations harbor both better OS and TTPR compared to codon 12 mutations. This difference is supported by a molecular explanation because codon 12 mutations exhibit higher up-regulation of VEGF and more stable bonds between Ras and GTP. Our study adds more evidence to support that molecular biology may be helpful in our daily practice. To date, only a small portion of the large landscape of KRAS mutations have been explored. Prospective multicenter studies are necessary to understand the role of molecular biology in the proper selection of patients for lung metastasectomy of CRC. In particular, further studies are necessary to clarify the prognostic and predictive value of each KRAS amino-acid substitution. It is usually admitted that rectal cancer cells have a higher tropism to the lung than colon cancer cells. This might be partially explained by a mechanical theory, in which cancer cells can metastasize directly to the lung because of the drainage of the rectum by the rectal veins directly to the lower vena cava, meanwhile colon cancer cells must pass through the portal system and the liver, who plays the role of filter and stop cancer cells, before they can reach the lung. However, previous series have shown that CRC cells harboring KRAS mutations have a higher lung tropism than wild type or cancer cells harboring BRAF mutations [27]. This lung tropism of KRAS mutations may explain why we did not observe in our cohort a higher prevalence of rectal cancer. MATERIALS AND METHODS However, we did not observe a survival difference before and after FOLFOX Pre-operative thoracic evaluation, lung metastasectomy procedures, covariates and data collection, and molecular analysis of KRAS mutations in codons 12 and 13 were performed as previously published [8]. Statistical analysis 8. Renaud S, Romain B, Falcoz PE, Olland A, Santelmo N, Brigand C, Rohr S, Guenot D, Massard G. KRAS and BRAF mutations are prognostic biomarkers in patients undergoing lung metastasectomy of colorectal cancer. Br J Cancer. 2015; 112:720–728. Categorical data are presented as the number (percentage, %), and continuous data are presented as the mean with standard deviation (SD). Associations between Codon 12 and Codon 13 and clinical- pathological characteristics were assessed with the use of the Mann–Whitney U test for continuous variables and the chi-square test and Fisher’s exact test for categorical variables as appropriate. The prognostic influence of predictors on OS and TTPR was assessed with a log- rank test and predictors associated with P ≤ 0.2 on univariate analysis were included in the multivariate models. A multivariate Cox proportional hazard model was employed to estimate the hazard ratios (HR) and 95% CI for the possible independent predictors of OS and TTR. All the tests were two-sided, and variables were considered significant for P values < 0.05. Statistical analyses were performed using Stata 13.1 (StataCorp LP, College Station, TX, USA). 9. Ghidini M, Personeni N, Bozzarelli S, Baretti M, Basso G, Bianchi P, Tronconi MC, Pressiani T, Grizzi F, Giordano L, Malesci A, Alloisio M, Laghi L, et al. KRAS mutation in lung metastases from colorectal cancer: prognostic implications. Cancer Med. 2016; 5:256–264. 10. Pereira AA, Rego JF, Morris V, Overman MJ, Eng C, Garrett CR, Boutin AT, Ferrarotto R, Lee M, Jiang ZQ, Hoff PM, Vauthey JN, Vilar E, et al. Association between KRAS mutation and lung metastasis in advanced colorectal cancer. Br J Cancer. 2015; 112:424–428. 11. Schweiger T, Hegedüs B, Nikolowsky C, Hegedüs Z, Szirtes I, Mair R, Birner P, Döme B, Lang G, Klepetko W, Ankersmit HJ, Hoetzenecker K. EGFR, BRAF and KRAS status in patients undergoing pulmonary metastasectomy from primary colorectal carcinoma: a prospective follow- up study. Ann Surg Oncol. 2014; 21:946–954. CONFLICTS OF INTEREST 12. De Roock W, Claes B, Bernasconi D, De Schutter J, Biesmans B, Fountzilas G, Kalogeras KT, Kotoula V, Papamichael D, Laurent- Puig P, Penault-Llorca F, Rougier P, Vincenzi B, et al. Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis. The Lancet Oncology. 2010; 11:753–762. The authors have no conflicts of interests to declare. Covariates and data collection Length of DFS was calculated from the surgery of the primary CRC to the first diagnosis of a thoracic or liver metastasis by imaging. OS was calculated from the first metastasectomy until death or the last follow- www.impactjournals.com/oncotarget Oncotarget 2522 4. Pfannschmidt J, Hoffmann H, Dienemann H. Reported outcome factors for pulmonary resection in metastatic colorectal cancer. J Thorac Oncol. 2010; 5:S172–178. up. Time to pulmonary recurrence (TTPR) was defined as the time period between a thoracic metastasectomy and the first diagnosis of thoracic recurrence. Because patients were referred to our Departments of Thoracic Surgery by various oncologists from different centers, no uniform protocol for chemotherapy was performed. However, chemotherapy regimens generally consisted of fluoropyrimidines (5-fluorouracil (5-FU)), administered alone or in combination with oxaliplatin (Folfox/XelOx regimens) and/or irinotecan (Folfiri/Folfoxiri regimens). In some cases, bevacizumab was added at the discretion of the oncologist. The timing of chemotherapy was defined as follows: neo-adjuvant chemotherapy administered prior to thoracic surgery, adjuvant chemotherapy administered following lung metastasectomy, and POC was performed both in neo-adjuvant and adjuvant settings. 5. Sanoff HK, Sargent DJ, Campbell ME, Morton RF, Fuchs CS, Ramanathan RK, Williamson SK, Findlay BP, Pitot HC, Goldberg RM. Five-year data and prognostic factor analysis of oxaliplatin and irinotecan combinations for advanced colorectal cancer: N9741. J Clin Oncol. 2008; 26:5721–5727. 6. Salah S, Watanabe K, Welter S, Park JS, Park JW, Zabaleta J, Ardissone F, Kim J, Riquet M, Nojiri K, Gisabella M, Kim SY, Tanaka K, et al. Colorectal cancer pulmonary oligometastases: pooled analysis and construction of a clinical lung metastasectomy prognostic model. Ann Oncol. 2012; 23:2649–2655. 7. Gonzalez M, Gervaz P. Risk factors for survival after lung metastasectomy in colorectal cancer patients: systematic review and meta-analysis. Future Oncol. 2015; 11:31–33. REFERENCES 1. Fitzmaurice C, Dicker D, Pain A, Hamavid H, Moradi-Lakeh M, MacIntyre MF, Allen C, Hansen G, Woodbrook R, Wolfe C, Hamadeh RR, Moore A, Werdecker A, et al. The Global Burden of Cancer 2013. JAMA Oncol. 2015; 1:505–527. 1. Fitzmaurice C, Dicker D, Pain A, Hamavid H, Moradi-Lakeh M, MacIntyre MF, Allen C, Hansen G, Woodbrook R, Wolfe C, Hamadeh RR, Moore A, Werdecker A, et al. The Global Burden of Cancer 2013. JAMA Oncol. 2015; 1:505–527. 13. Singh A, Greninger P, Rhodes D, Koopman L, Violette S, Bardeesy N, Settleman J. A gene expression signature associated with “K-Ras addiction” reveals regulators of EMT and tumor cell survival. Cancer Cell. 2009; 15:489–500. 2. Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Rosso S, Coebergh JW, Comber H, Forman D, Bray F. Cancer incidence and mortality patterns in Europe: estimates for 40 countries in 2012. Eur J Cancer. 2013; 49:1374–1403. 14. Ihle NT, Byers LA, Kim ES, Saintigny P, Lee JJ, Blumenschein GR, Tsao A, Liu S, Larsen JE, Wang J, Diao L, Coombes KR, Chen L, et al. Effect of KRAS oncogene substitutions on protein behavior: implications for signaling and clinical outcome. J Natl Cancer Inst. 2012; 104:228–239. 3. August DA, Ottow RT, Sugarbaker PH. Clinical perspective of human colorectal cancer metastasis. Cancer Metastasis Rev. 1984; 3:303–324. www.impactjournals.com/oncotarget Oncotarget 2523 15. Renaud S, Falcoz PE, Schaëffer M, Guenot D, Romain B, Olland A, Reeb J, Santelmo N, Chenard MP, Legrain M, Voegeli AC, Beau-Faller M, Massard G. Prognostic value of the KRAS G12V mutation in 841 surgically resected Caucasian lung adenocarcinoma cases. Br J Cancer. 2015; 113:1206–1215. in colon cancers to tumor location, stage, and survival: a population-based study. Cancer Epidemiol Biomarkers Prev. 2000; 9:1193–1197. 23. Al-Mulla F, Milner-White EJ, Going JJ, Birnie GD. Structural differences between valine-12 and aspartate-12 Ras proteins may modify carcinoma aggression. J Pathol. 1999; 187:433–438. 16. Renaud S, Seitlinger J, Falcoz PE, Schaeffer M, Voegeli AC, Legrain M, Beau-Faller M, Massard G. Specific KRAS amino acid substitutions and EGFR mutations predict site- specific recurrence and metastasis following non-small-cell lung cancer surgery. Br J Cancer. 2016; 115:346–353. 24. Guerrero S, Casanova I, Farre L, Mazo A, Capella G, Mangues R. K-ras codon 12 mutation induces higher level of resistance to apoptosis and predisposition to anchorage- independent growth than codon 13 mutation or proto- oncogene overexpression. Cancer Res. 2000; 60:6750–6756. 17. REFERENCES Garassino MC, Marabese M, Rusconi P, Rulli E, Martelli O, Farina G, Scanni A, Broggini M. Different types of K-Ras mutations could affect drug sensitivity and tumor behavior in non-small-cell lung cancer. Annals of Oncology. 2011; 22:235–237. 25. Rak J, Mitsuhashi Y, Bayko L, Filmus J, Shirasawa S, Sasazuki T, Kerbel RS. Mutant ras oncogenes upregulate VEGF/ VPF expression: implications for induction and inhibition of tumor angiogenesis. Cancer Res. 1995; 55:4575–4580. 18. Renaud S, Schaeffer M, Voegeli AC, Legrain M, Guérin E, Meyer N, Mennecier B, Quoix E, Falcoz PE, Guénot D, Massard G, Noël G, Beau-Faller M. Impact of EGFR mutations and KRAS amino acid substitution on the response to radiotherapy for brain metastasis of non-small-cell lung cancer. Future Oncol. 2016; 12:59–70. 26. Schweiger T, Nikolowsky C, Graeter T, Seebacher G, Laufer J, Glueck O, Glogner C, Birner P, Lang G, Klepetko W, Ankersmit HJ, Hoetzenecker K. Increased lymphangiogenesis in lung metastases from colorectal cancer is associated with early lymph node recurrence and decreased overall survival. Clin Exp Metastasis. 2016; 33:133–141. 19. Dadduzio V, Basso M, Rossi S, Cenci T, Capodimonti S, Strippoli A, Orlandi A, Cerchiaro E, Schinzari G, Cassano A, Martini M, Barone C. KRAS Exon 2 Mutations as Prognostic Indicators in Advanced Colorectal Cancer in Clinical Practice: A Mono-Institutional Study. Mol Diagn Ther. 2016; 20:65–74. 27. Lipsyc M, Yager R. Impact of somatic mutations on patterns of metastasis in colorectal cancer. J Gastrointest Oncol. 2015; 6:645–649. 28. Renaud S, Schaeffer M, Falcoz PE, Seitlinger J, Romain B, Voegeli AC, Legrain M, Reeb J, Santelmo N, Rohr S, Brigand C, Olland A, Guenot D, et al. Peri- operative bevacizumab improves survival following lung metastasectomy for colorectal cancer in patients harbouring v-Ki-ras2 Kirsten rat sarcoma viral oncogen homolog exon 2 codon 12 mutations. Eur J Cardiothorac Surg. 2016. doi: 10.1093/ejcts/ezw304. 20. Andreyev HJ, Norman AR, Cunningham D, Oates J, Dix BR, Iacopetta BJ, Young J, Walsh T, Ward R, Hawkins N, Beranek M, Jandik P, Benamouzig R, et al. Kirsten ras mutations in patients with colorectal cancer: the ‘RASCAL II’ study. Br J Cancer. 2001; 85:692–696. 21. Imamura Y, Morikawa T, Liao X, Lochhead P, Kuchiba A, Yamauchi M, Qian ZR, Nishihara R, Meyerhardt JA, Haigis KM, Fuchs CS, Ogino S. Specific mutations in KRAS codons 12 and 13, and patient prognosis in 1075 BRAF wild-type colorectal cancers. Clin Cancer Res. 2012; 18:4753–4763. 29. REFERENCES Cejas P, López-Gómez M, Aguayo C, Madero R, de Castro Carpeño J, Belda-Iniesta C, Barriuso J, Moreno García V, Larrauri J, López R, Casado E, Gonzalez-Barón M, Feliu J. KRAS mutations in primary colorectal cancer tumors and related metastases: a potential role in prediction of lung metastasis. PLoS One. 2009; 4:e8199. 22. Samowitz WS, Curtin K, Schaffer D, Robertson M, Leppert M, Slattery ML. Relationship of Ki-ras mutations www.impactjournals.com/oncotarget Oncotarget 2524
https://openalex.org/W2769045464	https://europepmc.org/articles/pmc5752496?pdf=render	English	null	A novel dynamic model for predicting outcome in patients with hepatitis B virus related acute-on-chronic liver failure	Oncotarget	2,017	cc-by	7,103	ABSTRACT Aim: It is challenging to predict the outcome of patients with hepatitis B virus related acute-on-chronic liver failure (HBV-ACLF) through existing prognostic models. Our aim was to establish a novel dynamic model to improve the predictive efficiency of 30-day mortality in HBV-ACLF patients. Methods: 305 patients who were diagnosed as HBV-ACLF (derivation cohort, n=211; validation cohort, n=94) were included in this study. The HBV-ACLF dynamic (HBV-ACLFD) model was constructed based on the daily levels of predictive variables in 7 days after diagnosis combined with baseline risk factors by multivariate logistic regression analysis. The HBV-ACLFD model was compared with the Child-Turcotte- Pugh (CTP) score, end-stage liver disease (MELD) score, and MELD within corporation of serum sodium (MELD-Na) score by the area under the receiver-operating characteristic curves (AUROC). Results: The HBV-ACLFD model demonstrated excellent discrimination with AUROC of 0.848 in the derivation cohort and of 0.813 in the validation cohort (p=0.620). The performance of the HBV-ACLFD model appeared to be superior to MELD score, MELD-Na score and CTP score (P<0.0001). Conclusion: The HBV-ACLFD model can accurately predict 30-day mortality in patients with HBV-ACLF, which is helpful to select appropriate clinical procedures, so as to relieve the social and economic burden. A novel dynamic model for predicting outcome in patients with hepatitis B virus related acute-on-chronic liver failure Ran Xue1,, Zhonghui Duan1,, Haixia Liu1, Li Chen2, Hongwei Yu1, Meixin Ren1, Yueke Zhu1, Chenggang Jin3, Tao Han4, Zhiliang Gao5 and Qinghua Meng1 1Department of Critical Care Medicine of Liver Disease, Beijing You-An Hospital, Capital Medical University, Beijing, China 2Georgia Prevention Institute, Medical College of Georgia, Augusta University, Augusta, GA, USA 3The School of Social Development and Public Policy, Beijing Normal University, Beijing, China 4Department of Hepatology, Tianjin Third Central Hospital of Tianjin Medical University, Tianjin, China 5Department of Infectious Diseases, The Third Affiliated Hospital, Zhongshan University, Guangzhou, China These authors have contributed equally to this work Correspondence to: Qinghua Meng, email: meng_qh@126.com Keywords: acute-on-chronic liver failure; hepatitis B; prognostic model Received: October 06, 2017 Accepted: October 27, 2017 Published: November 14, 2017 Copyright: Xue et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Published: November 14, 2017 Copyright: Xue et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright: Xue et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. A novel dynamic model for predicting outcome in patients with hepatitis B virus related acute-on-chronic liver failure Ran Xue1,, Zhonghui Duan1,, Haixia Liu1, Li Chen2, Hongwei Yu1, Meixin Ren1, Yueke Zhu1, Chenggang Jin3, Tao Han4, Zhiliang Gao5 and Qinghua Meng1 1Department of Critical Care Medicine of Liver Disease, Beijing You-An Hospital, Capital Medical University, Beijing, China 2Georgia Prevention Institute, Medical College of Georgia, Augusta University, Augusta, GA, USA 3The School of Social Development and Public Policy, Beijing Normal University, Beijing, China 4Department of Hepatology, Tianjin Third Central Hospital of Tianjin Medical University, Tianjin, China 5Department of Infectious Diseases, The Third Affiliated Hospital, Zhongshan University, Guangzhou, China These authors have contributed equally to this work Correspondence to: Qinghua Meng, email: meng_qh@126.com Keywords: acute-on-chronic liver failure; hepatitis B; prognostic model Received: October 06, 2017 Accepted: October 27, 2017 Published: November 14, 2017 Copyright: Xue et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 65), pp: 108970-108980 Research Paper A novel dynamic model for predicting outcome in patients with hepatitis B virus related acute-on-chronic liver failure Ran Xue1,, Zhonghui Duan1,, Haixia Liu1, Li Chen2, Hongwei Yu1, Meixin Ren1, Yueke Zhu1, Chenggang Jin3, Tao Han4, Zhiliang Gao5 and Qinghua Meng1 1Department of Critical Care Medicine of Liver Disease, Beijing You-An Hospital, Capital Medical University, Beijing, China 2Georgia Prevention Institute, Medical College of Georgia, Augusta University, Augusta, GA, USA 3The School of Social Development and Public Policy, Beijing Normal University, Beijing, China 4Department of Hepatology, Tianjin Third Central Hospital of Tianjin Medical University, Tianjin, China 5Department of Infectious Diseases, The Third Affiliated Hospital, Zhongshan University, Guangzhou, China These authors have contributed equally to this work Correspondence to: Qinghua Meng, email: meng_qh@126.com Keywords: acute-on-chronic liver failure; hepatitis B; prognostic model Received: October 06, 2017 Accepted: October 27, 2017 Published: November 14, 2017 Copyright: Xue et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License 3.0 (CC BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 65), pp: 108970-108980 Research Paper www.impactjournals.com/oncotarget/ Oncotarget, 2017, Vol. 8, (No. 65), www.impactjournals.com/oncotarget/ INTRODUCTION virus (HBV) is the leading cause of chronic liver disease in the Asia-Pacific region, including China and India [4]. Liver transplantation (LT) is a feasible and beneficial treatment for patients with ACLF to achieve survival [5]. It is important to accurately distinguish the ACLF patients who need LT, and grasp the best opportunity for LT [6]. To guide and optimize therapeutic strategy for ACLF patients, an accurate prognostic scoring system is prerequisite [7]. Acute-on-chronic liver failure (ACLF) is a life-threatening syndrome with varied etiology and manifestations with a short-term mortality of 50–90% [1]. ACLF is defined as the acute decompensation of liver function in patients with either previously diagnosed or undiagnosed chronic liver disease [2–3]. Hepatitis B www.impactjournals.com/oncotarget Predictors of mortality in the derivation cohort Varied prognostic scoring systems have been developed to predict ACLF mortality and guide the decision-making of LT. The Child-Pugh-Turcotte (CPT) classification and the model of end-stage liver disease (MELD) score are the most commonly used for patient priority on the waiting list of LT [8–9]. MELD Serum sodium (MELD-Na) score, as a modified MELD score, presented promising value for predicting mortality among patients on the LT waiting list [10–11]. 30-day mortality was applied as the end-point in the multivariate logistic regression analysis, and all variables in univariate analyses were imported into the model. In this model, TBiL, albumin, HE, INR, blood neutrophils percentage count, and suspicion of infection were independent risk factors for death. The β-coefficient, OR and 95% CI for independent predictors were presented in Table 2. However, each of the models above cannot apply reasonably to evaluate the clinical outcome in hepatitis B virus related acute-on-chronic liver failure (HBV- ACLF), mainly due to differences in patient background queues. There exist great differences between Eastern and Western ACLF in definition and diagnostic criteria, involving the basis of chronic liver disease, the concept of organ failure and the diagnostic criteria for organ failure [12–13]. Furthermore, HBV-ACLF is a dynamic process in which the variables at the time of hospitalization are predicted to vary over time, accompanied with the clinical processes and outcomes change accordingly. The existing prognostic models were established based on static baseline, which cannot logically evaluate the predictive outcome [14]. Therefore, it is urgently needed to derivation and validation of a novel dynamic model for predicting outcome in patients with HBV-ACLF. www.impactjournals.com/oncotarget Oncotarget 108970 108970 Baseline characteristics of patients 305 patients who were diagnosed as HBV- ACLF (derivation cohort, n=211; validation cohort, n=94) were included in this study. The comparisons of patients’ characteristics in the derivation and validation cohorts were shown in Table 1. There was no significant difference in gender distribution, died numbers within 30-day, HE, ascites and laboratory parameters at baseline between derivation cohort and validation cohort. The pre- existing chronic liver disease (p<0.001), infection (p< 0.001) and PT (p=0.007) were significantly less in the validation cohort than those in the derivation cohort. The age (p=0.016) and suspicion of infection (p=0.005) were higher in the validation cohort than those in the derivation cohort. The development of prognostic model The prognostic model derived from the derivation cohort (Table 2). The initial model was established based on the independent predictors of admission mortality. The model passed the Hosmer-Lemeshow goodness-of- fit test (p=0.535); the area under the Receiver-operating characteristic (AUROC) curve was also used. However, its discriminative ability was only modest (AUROC 0.745, 95% CI: 0.667 to 0.823). Ultimately, the HBV- ACLF dynamic (HBV-ACLFD) model was constructed based on the independent predictors of mortality and the daily changes within 7 days after diagnosis (Table 3). The Hosmer-Lemeshow goodness-of-fit test for HBV-ACLFD model was also performed (p=0.288). The HBV-ACLFD model was discriminated (AUROC=0.848, 95% CI: 0.793 to 0.902; Figure 1A). The specificity, sensitivity, negative predictive value and positive predictive value of the HBV-ACLFD model were 93.63%, 44.44%, 83.05% and70.59%, respectively. Our aim was to establish a prognostic model according to early changes of independent predictive variables at admission in patients with HBV-ACLF; to identify whether the model was superior to the existing prognostic models such as CPT scores, MELD scores, and MELD-Na scores; and ultimately to validate the model by a cohort of 94 patients with HBV-ACLF from three different geographical spread medical centers, so as to confirm the potential value of model for clinical treatment decision making. The validation of the predicted model The HBV-ACLFD model retained a good discrimination when applied to a validation cohort (AUROC=0.813; 95% CI, 0.720 to 0.905; Figure 1B). The Hosmer-Lemeshow goodness-of-fit test for HBV-ACLFD model was also performed (p=0.612). The AUROC from three different medical centers were provided, respectively (Figure 2). Comparison with alternative predicted models By the 30-day mortality and c-statistic as the endpoint, we identified four post-admission prognostic models of AUROC. The overall 30-day mortality was 25.59% (n=54 of 211) (Supplementary Figure 1). On admission, the AUROC was highest for the HBV-ACLFD model (0.848), followed by the MELD (0.696; 95% CI, 0.609 to 0.784), MELD-Na (0.686; 95% CI, 0.597 to 0.776), and CTP (0.566; 95% CI, 0.471 to 0.660) scores (Figure 3). There were significant differences between all pairs of scores for the dynamic changes (p <0.05). Comparison with alternative predicted models www.impactjournals.com/oncotarget Oncotarget 108971 Table 1: Clinical profiles of patients with HBV-ACLF in the derivation and validation cohorts : Clinical profiles of patients with HBV-ACLF in the derivation and validation cohorts Table 1: Clinical profiles of patients with HBV-ACLF in the derivation and validation cohorts Parameters Derivation cohort (n=211) Validation cohort (n=94) P values Character Age (years), mean±SD 43.94±12.63 47.15±10.60 0.016 Male, n (%) 176 (83.41) 81(86.17) 0.541 Died within 30 days, n (%) 54 (25.59) 33 (35.11) 0.089 HE a None HE, n (%) 152 (72.04) 61 (64.89) 0.429 I-II HE, n (%) 41 (19.43) 24 (25.53) III-IV HE, n (%) 18 (8.53) 9 (9.57) Pre-existing chronic liver diseaseb Chronic hepatitis B, n (%) 167(79.15) 49(52.13) <0.001* Compensated liver cirrhosis, n (%) 44 (20.85) 45 (47.87) In-hospital complication Ascites None, n (%) 67 (31.75) 26 (27.66) 0.184 Mild, n (%) 78 (36.97) 27 (28.72) Moderate, n (%) 47 (22.27) 27 (28.72) Severe, n (%) 19 (9.00) 14 (14.89) Infection, n (%) 12 (5.68) 28 (29.78) <0.001* Suspicion of infection, n(%)c 77 (36.49) 19 (20.21) 0.005* Laboratory parameters at baseline ALT (U/L), median (range) 163.50 (13.70 -2550.00) 197.75 (19.00 -7822.60) 0.227 AST (U/L), median (range) 161.35 (31.50 -5715.00) 188.35 (21.00 -5688.20) 0.118 TBil (μmol/L), median (range) 347.90 (171.93 -966.40) 325.75 (172.50 -780.30) 0.446 INR, median (range) 2.55 (1.08 -5.04) 2.37 (1.32 -9.00) 0.127 Na (mmol/L), median (range) 135.00 (111.00 -145.40) 135.00 (113.50 -143.70) 0.840 Albumin(g/L), mean±SD 31.00±4.73 31.25±4.51 0.667 WBC (×109/L), median (range) 7.01 (1.10 - 29.41) 7.32 (1.11 -35.64) 0.489 Blood neutrophils percentage count, median (range) 72.90 (26.60 -94.00) 70.90 (46.90 -92.50) 0.438 Hemoglobin (g/L), mean±SD 119.63±23.74 121.04±25.99 0.642 PLT (×109/L), median (range) 87.00 (9.00 -298.00) 96.50 (26.00-297.00) 0.077 PTA, median (range) 29.50 (8.00 -39.50) 32.00 (9.00 -40.00) 0.081 a Mild to moderate hepatic encephalopathy (HE) was defined as grade I or II HE, severe HE was defined as grade III or IV HE, according to the West Haven classification. b The pre-existing chronic liver disease patients were divided into two groups, counter balanced between liver cirrhosis and chronic hepatitis B. p c Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or ≥50% increase with respect to baseline with a final value >8,000/mm3; more than 5% of band forms; and/or temperature >37.5°C. * P<0.05. a Mild to moderate hepatic encephalopathy (HE) was defined as grade I or II HE, severe HE was defined as grade III or IV HE, according to the West Haven classification. a Mild to moderate hepatic encephalopathy (HE) was defined as grade I or II HE, severe HE was defined as grade III or IV HE, according to the West Haven classification. b The pre-existing chronic liver disease patients were divided into two groups, counter balanced between liver cirrhosis and chronic hepatitis B. c Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or ≥50% increase with respect to baseline with a final value >8,000/mm3; more than 5% of band forms; and/or temperature >37.5°C. * P<0.05. p c Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or ≥50% increase with respect to baseline with a final value >8,000/mm3; more than 5% of band forms; and/or temperature >37.5°C. * P<0 05 a Mild to moderate hepatic encephalopathy (HE) was defined as grade I or II HE, severe HE was defined as grade III or IV HE, according to the West Haven classification. b The pre-existing chronic liver disease patients were divided into two groups, counter balanced between liver cirrhosis and chronic hepatitis B. c Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or ≥50% increase with respect to b li i h fi l l 8 000/ 3 h 5% f b d f d/ 37 5°C www.impactjournals.com/oncotarget Table 3: The HBV-ACLFD model development based on predictors of mortality at baseline and their daily changes (Δ) within 7 days after diagnosis Variables OR CI1 CI2 β-coefficient P values Ln (TBiL, μmol/L) 3.02 0.99 9.22 1.11 0.052 ΔLn (TBiL, μmol/L) 3928.33 1.71 9.02e+06 8.28 0.036* Ln (Albumin, g/L) 0.93 0.84 1.02 -0.08 0.110 ΔLn (Albumin, g/L) 0.69 0.46 1.02 -0.38 0.063 Ln (INR) 2.25 0.56 9.05 0.81 0.255 ΔLn (INR) 9910.49 23.74 4.14e+06 9.20 0.003* Ln (Blood neutrophils percentage count) 1.06 1.02 1.10 0.06 0.001* ΔLn (Blood neutrophils percentage count) 1.38 1.13 1.68 0.32 0.002* HE I-II 0.97 0.38 2.52 -0.03 0.957 III-IV 3.47 1.00 12.09 1.24 0.051 Suspicion of infection 0.37 0.15 0.87 -1.00 0.022* Constant 0.00 0.00 0.02 -10.27 0.002* p<0.05. Table 3: The HBV-ACLFD model development based on predictors of mortality at baseline and their daily changes (Δ) within 7 days after diagnosis LFD model development based on predictors of mortality at baseline and their daily changes diagnosis Table 3: The HBV-ACLFD model development based on predictors of mortality at baseline an Δ) within 7 days after diagnosis scoring system is a precondition for guiding and optimizing therapeutic strategy for HBV-ACLF patients. The results of our research showed that the HBV-ACLFD model exhibited excellent discrimination and almost the same performance in two cohorts (AUROC 0.848 in the derivation cohort and 0.813 in the validation cohort). The performance of the HBV-ACLFD model appeared to be DISCUSSION HBV-ACLF can cause irreversible liver failure, leading to severe liver function damage. If LT is not available, it may lead to an at least 70% mortality rate [15]. Therefore, it is urgent to guide and optimize therapeutic strategy for HBV-ACLF patients. An accurate prognostic www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 108972 Table 2: The initial model was based on independent predictors of mortality at admission. Variables OR CI1 CI2 β-coefficient P values Ln (TBiL, μmol/L) 3.07 1.23 7.65 1.12 0.016 Ln (Albumin, g/L) 0.98 0.91 1.06 -0.02 0.667 Ln (INR) 1.55 0.45 5.32 0.44 0.487 Ln (Blood neutrophils percentage count) 1.04 1.01 1.08 0.04 0.012* HE I-II 1.16 0.49 2.75 0.15 0.735 III-IV 5.29 1.73 16.19 1.66 0.004* Suspicion of infectiona 0.30 0.14 0.68 -1.19 0.004* Constant 0.00 0.00 0.01 -10.33 <0.001* a Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or≥50% increase with respect to baseline with a final value >8,000/mm3; more than 5% of band forms; and/or temperature >37.5 °C. P<0.05. Table 2: The initial model was based on independent predictors of mortality at admission. a Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or≥50% increase with respect to baseline with a final value >8,000/mm3; more than 5% of band forms; and/or temperature >37.5 °C. P<0.05. Oncotarget 108973 www.impactjournals.com/oncotarget DISCUSSION HBV-ACLF can cause irreversible liver failure, leading to severe liver function damage. If LT is not available, it may lead to an at least 70% mortality rate [15]. Therefore, it is urgent to guide and optimize therapeutic strategy for HBV-ACLF patients. An accurate prognostic scoring system is a precondition for guiding and optimizing therapeutic strategy for HBV-ACLF patients. The results of our research showed that the HBV-ACLFD model exhibited excellent discrimination and almost the same performance in two cohorts (AUROC 0.848 in the derivation cohort and 0.813 in the validation cohort). The performance of the HBV-ACLFD model appeared to be Table 3: The HBV-ACLFD model development based on predictors of mortality at baseline and their daily changes (Δ) within 7 days after diagnosis Variables OR CI1 CI2 β-coefficient P values Ln (TBiL, μmol/L) 3.02 0.99 9.22 1.11 0.052 ΔLn (TBiL, μmol/L) 3928.33 1.71 9.02e+06 8.28 0.036* Ln (Albumin, g/L) 0.93 0.84 1.02 -0.08 0.110 ΔLn (Albumin, g/L) 0.69 0.46 1.02 -0.38 0.063 Ln (INR) 2.25 0.56 9.05 0.81 0.255 ΔLn (INR) 9910.49 23.74 4.14e+06 9.20 0.003* Ln (Blood neutrophils percentage count) 1.06 1.02 1.10 0.06 0.001* ΔLn (Blood neutrophils percentage count) 1.38 1.13 1.68 0.32 0.002* HE I-II 0.97 0.38 2.52 -0.03 0.957 III-IV 3.47 1.00 12.09 1.24 0.051 Suspicion of infection 0.37 0.15 0.87 -1.00 0.022* Constant 0.00 0.00 0.02 -10.27 0.002* *p<0.05. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 108973 in Eastern and Western countries [12]. In China, chronic HBV infection contributes to the ACLF. The early diagnosis and prompt treatment play a core role in the liver failure therapeutic strategies. In the western countries, ACLF is defined on the basis of compensated or decompensated cirrhosis [3]. Acute liver function decompensation mainly caused by alcohol abuse and bacterial infection. The therapeutic strategies of western countries focus on the multiple organ dysfunction syndrome (MODS), to distinguish the patients with superior to MELD score, MELD-Na score and CTP score (P<0.0001). It is indicated that the HBV-ACLFD model can accurately predict 30-day mortality in patients with HBV-ACLF, which is helpful to select appropriate clinical procedures for HBV-ACLF patients, so as to relieve the social and economic burden. A number of definitions of ACLF have been put forward, based on advice from experts rather than evidence-based data. The definitions of heterogeneity show the differences from the etiology of liver disease O t 108974 pactjournals com/oncotarget 1: (A) Receiver operating characteristic curves of the HBV-ACLFD model in predicting mortality in the tion cohort. The HBV-ACLFD model had discrimination (AUROC=0.848; 95% CI: 0.793 to 0.902). (B) Receiver operating eristic curves of the HBV-ACLFD model in predicting mortality in the validation cohort. The HBV-ACLFD model retained a good nation when applied to a validation cohort (AUROC=0.813; 95% CI: 0.720 to 0.905). The Hosmer-Lemeshow goodness-of-fit tes V-ACLFD model has been performed (p=0.612). Figure 1: (A) Receiver operating characteristic curves of the HBV-ACLFD model in predicting mortality in the derivation cohort. The HBV-ACLFD model had discrimination (AUROC=0.848; 95% CI: 0.793 to 0.902). (B) Receiver operating characteristic curves of the HBV-ACLFD model in predicting mortality in the validation cohort. The HBV-ACLFD model retained a good discrimination when applied to a validation cohort (AUROC=0.813; 95% CI: 0.720 to 0.905). The Hosmer-Lemeshow goodness-of-fit test for HBV-ACLFD model has been performed (p=0.612). www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 108974 high mortality risk who need to be treated in the ICU or moved to the waiting list of LT [16]. In addition, there are many differences of pathophysiology, clinical manifestation and prognosis between HBV related liver disease and alcoholic liver disease. Actually, in the Asia- Pacific region, including China and India, ACLF is mainly induced by the HBV infection. Antiviral therapy is also the special treatment strategy in the comprehensive treatment of HBV related ACLF [12]. Therefore, it is necessary to build up an accurate prognostic scoring system based on HBV-ACLF patients specifically. derivation cohort and a validation cohort. The data of patients with HBV-ACLF, including demographics, clinical, laboratory variables, underlying chronic liver disease, complications during the hospital course, as well as in-hospital special treatment, were collected. The derivation cohort was used to identify predictors of 30- day mortality and construct the HBV-ACLFD prognostic model. Δbiomarker within 7 days after diagnosis was calculated and constructed into the model together with baseline risk factors based on logistic regression. The mortality rates in validation cohorts and derivation cohort with HBV-ACLF at 30-day after diagnosis were 25.59% and 35.11%, respectively, which are consistent with the results reported by Xia et al [20]. This study focuses on patients with HBV-ACLF from China. HBV-ACLF is one of the most lethal, costly, and pervasive diseases in China. Early predictors are essentially required to distinguish patients with ACLF who need orthotropic LT from those that can survive by only intensive medical care [17]. HBV-ACLF is a dynamic process in which the variables at the time of hospitalization are predicted to vary over time, accompanied with the clinical processes and outcomes change accordingly. Meanwhile, prognosis predictions fluctuate over different clinical treatment. A study of acute liver failure (ALF) has shown that the model based on the early change of dynamic variables is better for the prediction than the model based on static baseline variables [21]. The high absolute values of AFP cannot predict the prognosis well, but the uptrend of AFP Currently, it remains difficult to identify appropriate indicators of poor outcome in HBV-ACLF [18]. The established prognostic models are mostly based on static baseline variables. However, patients’ responses to the treatment could also affect the outcomes [19]. Thus, we retrospectively reviewed patients diagnosed as HBV- ACLF. Patients with HBV-ACLF were divided into a Figure 2: The AUROC from the three different medical centers. www.impactjournals.com/oncotarget 1: Beijing You-an Hospital (derivation cohort), 2: the Third Affiliated Hospital of Sun Yat-Sen University, 3: Tianjin Third Central Hospital, 4: Beijing You-an Hospital (validation cohort). Figure 2: The AUROC from the three different medical centers. 1: Beijing You-an Hospital (derivation cohort), 2: the Third Affiliated Hospital of Sun Yat-Sen University, 3: Tianjin Third Central Hospital, 4: Beijing You-an Hospital (validation cohort). www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 108975 Figure 3: Comparison of the predictive accuracy for 30-day mortality between MELD, MELD-Na, CTP, and the dynamic prognostic model. On admission, the AUROC was highest for the HBV-ACLFD model (0.848), followed by the MELD (0.696; 95% CI, 0.609 to 0.784), MELD-Na (0.686; 95% CI, 0.597 to 0.776), and CTP (0.566; 95% CI, 0.471 to 0.660) scores. There were significant differences between all pair of scores for the dynamic changes (p <0.05). Figure 3: Comparison of the predictive accuracy for 30-day mortality between MELD, MELD-Na, CTP, and the dynamic prognostic model. On admission, the AUROC was highest for the HBV-ACLFD model (0.848), followed by the MELD (0.696; 95% CI, 0.609 to 0.784), MELD-Na (0.686; 95% CI, 0.597 to 0.776), and CTP (0.566; 95% CI, 0.471 to 0.660) scores. There were significant differences between all pair of scores for the dynamic changes (p <0.05). Figure 3: Comparison of the predictive accuracy for 30-day mortality between MELD, MELD-Na, CTP, and dynamic prognostic model. On admission, the AUROC was highest for the HBV-ACLFD model (0.848), followed by the M (0 696 95% CI 0 609 t 0 784) MELD N (0 686 95% CI 0 597 t 0 776) d CTP (0 566 95% CI 0 471 t 0 660) Th Figure 3: Comparison of the predictive accuracy for 30-day mortality between MELD, MELD-Na, CTP, and the dynamic prognostic model. On admission, the AUROC was highest for the HBV-ACLFD model (0.848), followed by the MELD (0.696; 95% CI, 0.609 to 0.784), MELD-Na (0.686; 95% CI, 0.597 to 0.776), and CTP (0.566; 95% CI, 0.471 to 0.660) scores. There were significant differences between all pair of scores for the dynamic changes (p <0.05). Figure 3: Comparison of the predictive accuracy for 30-day mortality between MELD, MELD-Na, CTP, and the dynamic prognostic model. On admission, the AUROC was highest for the HBV-ACLFD model (0.848), followed by the MELD (0.696; 95% CI, 0.609 to 0.784), MELD-Na (0.686; 95% CI, 0.597 to 0.776), and CTP (0.566; 95% CI, 0.471 to 0.660) scores. There were significant differences between all pair of scores for the dynamic changes (p <0.05). Figure 4: Study flow: diagram showing the process of study selection and exclusion in of HBV-ACLF patients. flow: diagram showing the process of study selection and exclusion in of HBV-ACLF patients. Figure 4: Study flow: diagram showing the process of study selection and exclusion in of HBV-ACLF patients. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 108976 over the first 3 days of hospitalization can reflect the survival rate of ALF [22]. Another study reported that ΔMELD is superior to initial MELD and that CTP scores are reliable in patients with advanced cirrhosis [14]. In our study, the initial model was constructed based on the independent predictors of admission mortality. However, the discriminative ability was just moderate (AUROC = 0.7451). When Δbiomarker was calculated and constructed into the model together with baseline risk factors, this model had a great discrimination (AUROC=0.8475). Taken together, it is suggested that continuous changes in predictive variables can better predict mortality than using static variables. mildly elevated body temperature [26]. The suspicion of infection is correlated with a good prognosis [29]. Rapid initiation of appropriate antibiotic therapy is critical to the management of patients suspected of being infected. Delayed and inappropriate treatments are correlated with increased mortality [30]. Moreover, the validation cohort data were collected from three different medical centers of geographic location in China, which means our model is validated in different regions of the population and in different areas of HBV-ACLF lesions. It also improves the accuracy and credibility of our HBV-ACLFD model, which is another principal strength of our model. Because this model is based on retrospective data on a great number of patients with HBV-ACLF from three different medical centers, it also means that homogeneous cohort managed with similar treatment regimens cannot be secured. It is indicated that this model has a broad spectrum of practicality. On admission, the AUROC for MELD, MELD-Na, and CTP scores were 0.6962, 0.6862, 0.5656, respectively, which is consistent with the results reported [23]. However, the MELD, MELD-Na and CTP scoring systems are used to predict mortality risk in untreated cirrhotic patients [24–25]. These models did not consider the impact of biomarker changes after diagnosis, HBV-ACLF related complications and in-hospital special treatment on the prognosis. When HBV-ACLF related complications and biomarker (Δbiomarker) changes daily within 7 days after diagnosis were established with baseline risk factors according to logistic regression, the HBV-ACLFD model (AUROC=0.8475) showed better potential than the established prognostic models like MELD, MELD-Na and CTP scores in a cohort of patients with HBV-ACLF for predicting the 30-day mortality. By the AUROC, these results support that Δbiomarker and HE can improve its accuracy in predicting mortality. Thus, the HBV-ACLFD model showed good discrimination in the derivation cohort. www.impactjournals.com/oncotarget When applied to the separate validation cohort of patients with HBV-ACLF, the new model retained good discrimination, accurately distinguishing the ACLF patients who need LT, and grasping the best opportunity for their transplantation. However, novel clinical strategies, including artificial liver support system and stem cell transplantation, apply to treat HBV-ACLF now. Our model cannot assess the effect of these novel clinical strategies for accurately predict outcome in patients with HBV-ACLF. Meanwhile, the comparison between the CLIF Consortium ACLF score (CLIF-C ACLFs) and the HBV-ACLFD model should be explored, despite the fact that the CLIF-C ACLFs is helpful to predicting short-term mortality in ACLF patients in Western countries [31–32], where the most common etiology of ACLF is alcoholic liver disease. In conclusion, we retrospectively deduced and validated the dynamic models of predictive outcomes in patients with HBV-ACLF. This model may be helpful in clinical decision making and risk stratification for patients with HBV-ACLF. MATERIALS AND METHODS We also evaluated biochemical and clinical variables via multivariate logistic regression. HE, suspicion of infection, baseline and average daily changes of serum TBiL, INR, serum albumin and blood neutrophils percentage were independent prognostic factors for 30-day mortality. Meanwhile, on multivariate analysis, suspicion of infection is associated with mortality. The hazard ratio for mortality of patients with HBV-ACLF was 30.37 during the first 30-day. It is recommended that patients with high risk of complications and cirrhosis are highly suspected bacterial infection [26]. All procedures and methods related to this research were accorded morally with current laws as well as the creeds of the Declaration of Helsinki. The research was permitted by the Ethical Committee of Beijing You-An Hospital, Capital Medical University. Procedures The derivation cohort was applied to determine the predictors of mortality and thus established a prognostic model. Δbiomarker was calculated and constructed into the model together with baseline risk factors based on logistic regression, which is our dynamic prognostic model for HBV-ACLF, named the HBV-ACLF dynamic (HBV-ACLFD) model. AUROC were used as a control to compare the predictive values in the HBV-ACLFD model, i.e. MELD score, MELD-Na score and CTP score. The results of blood tests performed on the day of diagnosis and within 7 days after diagnosis were recorded. The blood tests consisted of white blood count (WBC), platelet count (PLT), hemoglobin level, blood neutrophils percentage count (NEUT), international normalized ratio (INR), prothrombin time (PT), creatinine level, aspartate transaminase (AST) level, aspartate alanine transaminase (ALT) level, TBil level (total bilirubin), direct bilirubin (DBil) level, cholinesterase, glucose, total cholesterol level, albumin level, triglyceride level, serum sodium level, ammonia level, serum chloride level, serum potassium level, serum magnesium level and total serum calcium level. Study design and patients selection A total of 445 patients who were diagnosed as HBV-ACLF from January 2005 to February 2014 were included in this research. The minimum follow-up period for enrolled patients was 30-day after diagnosis. The diagnosis of cirrhosis was according to a composite of clinical signs and findings provided through laboratory test results, radiologic imaging, endoscopy and liver biopsy. The systemic inflammatory response syndrome (SIRS) occurs in patients with advanced cirrhosis and is correlated with poor prognosis [27]. The definition of SIRS and sepsis are very difficult because of the following findings [28]: elevated baseline heart rate for hyperdynamic circulatory syndrome; reduced baseline cell count for hypersplenism; hepatic encephalopathy leading to excessive ventilation; cirrhotic patients with The entry criteria comprised the following:HBV- ACLF is defined as ACLF with previously diagnosed or undiagnosed HBV. All enrolled patients met the criteria for ACLF from the consensus recommendations of the Asian www.impactjournals.com/oncotarget Oncotarget 108977 Pacific Association for the Study of the Liver (APASL) [33]. All treatments were performed based on the criteria of diagnostic and treatment guidelines for ACLF adopted by the Chinese Medical Association [34]. Differences of biomarker (Δ biomarker) levels within 7 days after diagnosis were calculated. For example, a patient with HBV-ACLF with an ALT of 100U/L on the day of admission and 380U/L on day 7 after admission, had ΔALT= (380-100) /7=40. In the case of variceal hemorrhage or plasma exchange, the results of blood tests after variceal hemorrhage or plasma exchange were chosen after more than 2 days. The exclusion criteria were the following: other factors induce severe liver injury, such as alcohol, drugs, hepatoviruses other than HBV, autoimmunity and pregnancy, as well as genetic and metabolic disorders. HBV-ACLF patients with hepatocellular carcinoma, known decompensated cirrhosis prior to onset of acute hepatic insult, age less than 18 years, jaundice induced by hemolytic jaundice and obstructive jaundice, absence of any chronic liver disease on investigations, and prolonged prothrombin time induced by blood system diseases were also excluded. The flowchart for the selection of HBV- ACLF patients was shown in Figure 4. Calculation of the CTP, MELD and MELD-Na The MELD equation was applied to calculate the score of severity: 9.57×ln(creatinine, mg/ dl)+3.78×ln(bilirubin, mg/dl)+11.2×ln(INR)+6.43, in which the minimal values were forced to 1.0 for calculation purposes [29]. The MELD-Na equation was constructed based on the Na and MELD, MELD+1.59×(135–Na), with minimum and maximum Na values as 120 and 135 mEq/L, respectively [35]. The CTP classification was assessed according to the standard criteria [36]. Observed parameters Data on patient demographics, clinical and laboratory variables, evaluation of the underlying chronic liver disease, complications during the medication course, in-hospital special treatment (antiviral therapy, plasma exchange, corticosteroid exposure) were retrospectively collected. All enrolled patients were followed up for a minimum of 30-day by clinic visits or telephone. The outcome (death or survival) of each patient with HBV- ACLF was documented. Suspicion of infection based on at least one of the following: WBC count >10,000/mm3 or ≥50% increase with respect to baseline with a final value >8,000/mm3; more than 5% of band forms; and/or temperature >37.5°C. Management protocol Patients who were positive for HBV-DNA at presentation underwent antiviral therapy (lamivudine, telbivudine, or entecavir) with informed consent. In order to prevent/treat complications, comprehensive medical interventions were applied, including absolute bed rest, intravenous drop infusion of albumin or plasma, energy supplements, maintenance of acid-base equilibrium or electrolyte, plasma exchange the use of adenosylmethionine, glutathione or branched-chain amino acids to nourish liver cells, as well as antibiotics for infection. Patients with decompensation requiring organ support (such as variceal hemorrhage, HE, hepatorenal syndrome, mechanical ventilationor multiorgan failure) were admitted to ICU. Derivation cohort data were screened from Beijing You-an Hospital from January 2005 to January 2013, while validation cohort data was collected from three different medical centers from January 2013 to February 2014. Specifically, 30 cases, 24 cases and 40 cases of HBV-ACLF as validation cohort data were collected from Beijing You-an Hospital, Tianjin Third Central Hospital and the Third Affiliated Hospital of Sun Yat- Sen University, respectively. The derivation cohort was applied to determine the predictors of mortality and thus established a prognostic model. REFERENCES 14. Huo TI, Wu JC, Lin HC, Lee FY, Hou MC, Lee PC, Chang FY, Lee SD. Evaluation of the increase in model for end- stage liver disease (DeltaMELD) score over time as a prognostic predictor in patients with advanced cirrhosis: risk factor analysis and comparison with initial MELD and Child-Turcotte-Pugh score. J Hepatol. 2005; 42:826-32. 1. Liu Q, Liu Z, Wang T, Wang Q, Shi X, Dao W. Characteristics of acute and sub-acute liver failure in China: nomination, classification and interval. J Gastroenterol Hepatol. 2007;22:2101-6. 2. Olson JC, Kamath PS. Acute-on-chronic liver failure: what are the implications? Curr Gastroenterol Rep. 2012;14:63-6. 15. Mas A, Rodeg J. Fulminant hepatitis failure. Lancet. 1997; 349:1081-5. 3. Sarin SK, Choudhury A. Acute-on-chronic liver failure: terminology, mechanisms and management. Nat Rev Gastroenterol Hepatol. 2016;13:131-49. 16. Arroyo V, Moreau R, Jalan R, Ginès P; EASL-CLIF Consortium CANONIC Study. Acute-on-chronic liver failure: a new syndrome that will re-classify cirrhosis. J Hepatol. 2015;62:S131-43. 4. Bernal W, Jalan R, Quaglia A, Simpson K, Wendon J, Burroughs A. Acute-on-chronic liver failure. Lancet 2015;386:1576-87. 17. Wlodzimirow KA, Eslami S, Abu-Hanna A, Nieuwoudt M, Chamuleau RA. A systematic review on prognostic indicators of acute on chronic liver failure and their predictive value for mortality. Liver Int. 2013;33:40-52. 5. Asrani SK, Simonetto DA, Kamath PS. Acute-on-chronic liver failure. Clin Gastroenterol Hepatol. 2015; 13:2128-39. 6. Shalimar, Kumar D, Vadiraja PK, Nayak B, Thakur B, Das P, Datta Gupta S, Panda SK, Acharya SK. Acute on chronic liver failure because of acute hepatic insults: etiologies, course, extrahepatic organ failure and predictors of mortality. J Gastroenterol Hepatol. 2016;31:856-64. 18. Anand AC, Dhiman RK. Acute on chronic liver failure-what is in a 'definition'?. J Clin Exp Hepatol. 2016;6:233-40. 19. Yamagishi Y, Saito H, Ebinuma H, Kikuchi M, Ojiro K, Kanamori H, Tada S, Horie Y, Kato S, Hibi T. A new prognostic formula for adult acute liver failure using computer tomography-derived hepatic volumetric analysis. J Gastroenterol. 2009;44:615-23. 7. Zheng YX, Zhong X, Li YJ, Fan XG. Performance of scoring systems to predict mortality of patients with acute-on-chronic liver failure: a systematic review and meta-analysis. J Gastroenterol Hepatol. 2017. https://doi. org/10.1111/jgh.13786. 20. Xia Q, Dai X, Zhang Y, Guo Y, Xu X, Yang Q, Du W, Liu X, Chen Y, Huang J, Li L. A modified MELD model for Chinese pre-ACLF and ACLF patients and it reveals poor prognosis in pre-ACLF patients. PLoS One. 2013;8:e64379. 8. FUNDING 11. Kim WR, Biggins SW, Kremers WK, Wiesner RH, Kamath PS, Benson JT, Edwards E, Therneau TM. Hyponatremia and mortality among patients on the liver-transplant waiting list. N Engl J Med. 2008;359:1018-26. The study was supported by the National Natural Science Foundation of China (No. 81470877), the Foundation of Capital Science Committee of China (H020920020890), and Capital Health Research and Development of Special (2011-2018-09). 12. Hernaez R, Solà E, Moreau R, Ginès P. Acute-on-chronic liver failure: an update. Gut. 2017;66:541-53. 13. Jalan R, Yurdaydin C, Bajaj JS, Ginès P. Toward an improved definition of acute-on-chronic liver failure. Gastroenterology. 2014;147:4-10. CONFLICTS OF INTEREST No potential conflicts (financial, professional, or personal) is relevant to the manuscript. Statistical analysis All statistical analyses were conducted by STATA version 13.1. Univariate analyses were applied by appropriate tests to identify the variables which were significantly different in patients who died or survived in the derivation cohort. Multivariable logistic regression www.impactjournals.com/oncotarget Oncotarget 108978 end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology. 2003;124:91-6. model was used by taking these predictor variables with the outcome (survived vs. death) through a stepwise forward selection procedure and thereby establish the dynamic prognostic model. The validity of this model was constructed by concordance (c) statistics, which is equivalent to the area under the AUROC curve. C-value >0.7 was recommended useful, and the C - value >0.8 would be considered excellent. Then, the performance of this model was finally verified in an independent cohort. 9. Hong G, Lee KW, Suh S, Yoo T, Kim H, Park MS, Choi Y, Yi NJ, Suh KS. The model for end-stage liver disease score- based systempredicts short term mortality better than the current Child-Turcotte-Pugh score-based allocation system during waiting for deceased liver transplantation. J Korean Med Sci. 2013;28:1207-12. 10. Marroni CP, de Mello Brandão AB, Hennigen AW, Marroni C, Zanotelli ML, Cantisani G, Fuchs SC; Liver Transplantation Group. MELD scores with incorporation of serum sodium and death prediction in cirrhotic patients on the waiting list for liver transplantation: a single center experience in southern Brazil. Clin Transplant. 2012;26:E395-401. REFERENCES Wiesner R, Edwards E, Freeman R, Harper A, Kim R, Kamath P, Kremers W, Lake J, Howard T, Merion RM, Wolfe RA, Krom R; United Network for Organ Sharing Liver Disease Severity Score Committee. Model for 21. Kumar R, Shalimar, Sharma H, Goyal R, Kumar A, Khanal S, Prakash S, Gupta SD, Panda SK, Acharya SK. Prospective derivation and validation of early dynamic www.impactjournals.com/oncotarget Oncotarget 108979 30. European Association for the Study of the Liver. EASL clinical practice guidelines on the management of ascites, spontaneous bacterial peritonitis, and hepatorenal syndrome in cirrhosis. J Hepatol. 2010;53:397-417. model for predicting outcome in patients with acute liver failure. Gut. 2012;61:1068-75. 22. Schiødt FV, Ostapowicz G, Murray N, Satyanarana R, Zaman A, Munoz S, Lee WM. Alpha-fetoprotein and prognosis in acute liver failure. Liver Transpl. 2006;12:1776-81. 31. Sy E, Ronco JJ, Searle R, Karvellas CJ. Prognostication of critically ill patients with acute-on-chronic liver failure using the Chronic Liver Failure-Sequential Organ Failure Assessment: a Canadian retrospective study. J Crit Care. 2016;36:234-9. 23. Xun YH, Shi JP, Li CQ, Li D, Shi WZ, Pan QC, Guo JC, Zang GQ. Prognostic performance of a series of model for end-stage liver disease and respective Delta scores in patients with hepatitis B acute-on-chronic liver failure. Mol Med Rep. 2014;9:1559-68. 32. Lee M, Lee JH, Oh S, Jang Y, Lee W, Lee HJ, Yoo JJ, Choi WM, Cho YY, Cho Y, Lee DH, Lee YB, Yu SJ, et al. CLIF-SOFA scoring system accurately predicts short-term mortality in acutely decompensated patients with alcoholic cirrhosis: a retrospective analysis. Liver Int. 2015;35:46-57. 24. Zhang QB, Chen YT, Lian GD, Qian CC, Chen SJ, Huang KH. A combination of models for end-stage liver disease and cirrhosis-related complications to predict the prognosis of liver cirrhosis. Clin Res Hepatol Gastroenterol. 2012;36:583-91. 33. Sarin SK, Kedarisetty CK, Abbas Z, Amarapurkar D, Bihari C, Chan AC, Chawla YK, Dokmeci AK, Garg H, Ghazinyan H, Hamid S, Kim DJ, Komolmit P. Acute-on-chronic liver failure: consensus recommendations of the Asian Pacific Association for the Study of the Liver (APASL) 2014. Hepatol. Int. 2014;8:453-71. 25. Peng Y, Qi X, Tang S, Deng H, Li J, Ning Z, Dai J, Hou F, Zhao J, Wang R, Guo X. Child-Pugh, MELD, and ALBI scores for predicting the in-hospital mortality in cirrhotic patients with acute-on-chronic liver failure. Expert Rev Gastroenterol Hepatol. 2016;10:971-80. 34. REFERENCES Liver Failure and Artificial Liver Group, Chinese Society of Infectious Diseases, Chinese Medical Association; Severe Liver Diseases and Artificial Liver Group, Chinese Society of Hepatology, Chinese Medical Association. [Diagnostic and treatment guidelines for liver failure (2012 version)]. [Article in Chinese]. Zhonghua Gan Zang Bing Za Zhi. 2013;21:177-83. 26. Fagiuoli S, Colli A, Bruno R, Burra P, Craxì A, Gaeta GB, Grossi P, Mondelli MU, Puoti M, Sagnelli E, Stefani S, Toniutto P. Management of infections in cirrhotic patients: report of a consensus conference. Dig Liver Dis. 2014;46:204-12. 27. Abdel-Khalek EE, El-Fakhry A, Helaly M, Hamed M, Elbaz O. Systemic inflammatory response syndrome in patients with liver cirrhosis. Arab J Gastroenterol. 2011;12:173-7. 35. Biggins SW, Kim WR, Terrault NA, Saab S, Balan V, Schiano T, Benson J, Therneau T, Kremers W, Wiesner R, Kamath P, Klintmalm G. Evidence-based incorporation of serum sodium concentration into MELD. Gastroenterology. 2006;130:1652-60. 28. Cazzaniga M, Dionigi E, Gobbo G, Fioretti A, Monti V, Salerno F. The systemic inflammatory response syndrome in cirrhotic patients: relationship with their in-hospital outcome. J Hepatol. 2009;51:475-82. 36. Pugh RN, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg. 1973;60:646-9. 29. Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. A model to predict poor survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology. 2000;31:864-71. www.impactjournals.com/oncotarget www.impactjournals.com/oncotarget Oncotarget 108980
https://openalex.org/W2136222411	https://europepmc.org/articles/pmc4127204?pdf=render	English	null	An Ant Colony Optimization Based Feature Selection for Web Page Classification	The scientific world journal/TheScientificWorldjournal	2,014	cc-by	13,758	Hindawi Publishing Corporation e Scientiﬁc World Journal Volume 2014, Article ID 649260, 16 pages http://dx.doi.org/10.1155/2014/649260 Hindawi Publishing Corporation e Scientiﬁc World Journal Volume 2014, Article ID 649260, 16 pages http://dx.doi.org/10.1155/2014/649260 Hindawi Publishing Corporation e Scientiﬁc World Journal Volume 2014, Article ID 649260, 16 pages http://dx.doi.org/10.1155/2014/649260 1. Introduction manual classification is time consuming and needs a lot of human effort, which makes it unscalable with respect to the high growing speed of the web. Therefore, there has been great need for automated web page classification systems [4]. A major problem of the web page classification is the high dimensionality of the feature space. We need to select “good” subsets of features from the original feature space to reduce the dimensionality and to improve the efficiency and run time performance of the classification process [5]. Several approaches such as document frequency, information gain, mutual information, chi square analysis, and term strength have been applied to select proper features for text categoriza- tion. According to Yang and Pedersen [6] chi square analysis and information gain are more effective for optimizing classi- fication results, and document frequency is less effective but it is scalable and affordable. In Aghdam et al. [7] nature inspired search and optimization algorithms have been applied for feature selection of text classification problem. According to [7], ant colony optimization and genetic algorithms can choose better features than the information gain and chi square analysis, and performance of ant colony optimization manual classification is time consuming and needs a lot of human effort, which makes it unscalable with respect to the high growing speed of the web. Therefore, there has been great need for automated web page classification systems [4].i The aim of text classification is to categorize documents into a certain number of predefined classes by using document features. Text classification plays a crucial role in many retrieval and management tasks such as information retrieval, information extraction, document filtering, and building hierarchical directories [1]. When text classification focuses on web pages it is named web classification or web page classification. However, web pages are different from text, and they contain a lot of additional information, such as URLs, links, HTML tags, which are not supported by text documents. Because of this property of web pages, web classification is different from traditional text classification [1].ii A major problem of the web page classification is the high dimensionality of the feature space. We need to select “good” subsets of features from the original feature space to reduce the dimensionality and to improve the efficiency and run time performance of the classification process [5]. Esra Saraç and Selma AyGe Özel Department of Computer Engineering, C¸ukurova University, Balcali, Sarıc¸am, 01330 Adana, Turkey Correspondence should be addressed to Selma Ays¸e ¨Ozel; saozel@cu.edu.tr Received 25 April 2014; Revised 20 June 2014; Accepted 22 June 2014; Published 17 July 2014 Academic Editor: T. O. Ting Copyright © 2014 E. Sarac¸ and S. A. ¨Ozel. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The increased popularity of the web has caused the inclusion of huge amount of information to the web, and as a result of this explosive information growth, automated web page classification systems are needed to improve search engines’ performance. Web pages have a large number of features such as HTML/XML tags, URLs, hyperlinks, and text contents that should be considered during an automated classification process. The aim of this study is to reduce the number of features to be used to improve runtime and accuracy of the classification of web pages. In this study, we used an ant colony optimization (ACO) algorithm to select the best features, and then we applied the well-known C4.5, naive Bayes, and k nearest neighbor classifiers to assign class labels to web pages. We used the WebKB and Conference datasets in our experiments, and we showed that using the ACO for feature selection improves both accuracy and runtime performance of classification. We also showed that the proposed ACO based algorithm can select better features with respect to the well-known information gain and chi square feature selection methods. Research Article An Ant Colony Optimization Based Feature Selection for Web Page Classification Esra Saraç and Selma AyGe Özel Department of Computer Engineering, C¸ukurova University, Balcali, Sarıc¸am, 01330 Adana, Turkey Correspondence should be addressed to Selma Ays¸e ¨Ozel; saozel@cu.edu.tr Received 25 April 2014; Revised 20 June 2014; Accepted 22 June 2014; Published 17 July 2014 2. Related Work In this section, we give a brief review of web page classifica- tion, feature selection, and ant colony optimization for feature selection. 2.1. Web Page Classification. Classification is the process of assigning predefined class labels to some unseen or test data. For this purpose, a set of labeled data is used to train a classifier which is then used for labeling unseen data. This classification process is also defined as a supervised learning [11]. The process is not different in web page classification such that there is one or more predefined class labels and a classification model assigns class labels to web pages which are in fact hypertext and have many features such as textual tokens, markup tags, URLs, and host names in URLs that are meaningful for classifiers. As web pages have additional properties, their classification has several differences from traditional text classification [1].ii i Web page classification has some subfields like subject classification and functional classification [1]. In subject classification, classifier is concerned with the content of a web page and tries to determine the “subject” of the web page. For example, categories of online newspapers like finance, sport, and technology are instances of subject classification. Functional classification on the other hand deals with the function or type of the web page. For example, determining whether a web page is a “personal homepage” or a “course page” is an instance of a functional classification. Subject and functional classification are the most popular classification types [1].ii In this study, a new ACO based feature selection algo- rithm, which selects best features from web pages, has been proposed. The contributions of this study are summarized as follows. (i) In the earlier studies, only the bag of terms approach is used for feature extraction and ACO is applied to select features among the small number of terms. However, in our study each term in the URLs and in the HTML tags is taken as a different feature. Therefore, a feature is represented as <URL><term>, or <tag><term> pair which we call “tagged terms” representation that yields thousands of features to be extracted from web pages. According to our research in the literature, there exists no ACO based study which works on such large scale feature space. y Classification can be divided into binary classification and multiclass classification according to the number of classes [1]. 1. Introduction Several approaches such as document frequency, information gain, mutual information, chi square analysis, and term strength have been applied to select proper features for text categoriza- tion. According to Yang and Pedersen [6] chi square analysis and information gain are more effective for optimizing classi- fication results, and document frequency is less effective but it is scalable and affordable. In Aghdam et al. [7] nature inspired search and optimization algorithms have been applied for feature selection of text classification problem. According to [7], ant colony optimization and genetic algorithms can choose better features than the information gain and chi square analysis, and performance of ant colony optimization On the web, classification is used for topic-specific web link selection, analysis of the topical structure of the web, development of web directories, and focused crawling [1]. Previously, people manually constructed some web directo- ries such as Yahoo! [2] and the Open Directory Project [3] and manually assigned class labels to web documents. However, 2 The Scientific World Journal classification. Section 3 describes our ACO-based feature selection algorithm. Section 4 includes the datasets used, the experiments performed, their results, and discussions. Section 5 concludes the study and gives some future work. is better than the genetic algorithm. For this reason, in this study we applied an ant colony optimization, which was originally developed to solve optimization problems, to select the best features from the web pages for accurate and time efficient classification. fii Ant colony optimization (ACO) was inspired from the behavior of real ant colonies, and it is used to solve discrete optimization problems. The first ACO system was introduced by Marco Dorigo in his Ph.D. thesis [8] and was called the ant system (AS). The AS is the result of a research on computa- tional intelligence approaches to combinatorial optimization [8]. The AS was initially applied to the travelling salesman problem [9] and then to other hard problems. The original AS is motivated by the natural phenomenon that ants deposit pheromone on the ground in order to mark some favorable path that should be followed by other members of the colony. The aim of the colony is to find the shortest path between a food source and the nest. 1. Introduction The behavior of an ant colony is a good example of a self-organized system such that it is based on positive feedback (i.e., the deposit of pheromone) and negative feedback (i.e., the evaporation of pheromone). Theoretically, if the quantity of pheromone remains the same over time on all edges, no route is chosen. However, because of feedback, a slight variation on an edge allows the edge to be chosen. The algorithm moves from an unstable state in which no edge is stronger than another, to a stable state where the route is composed of the strongest edges. Ant colony optimization has a wide application domain; for example, Liu et al. [10] have used ACO for continuous domains. 2. Related Work In binary classification, there is only one class label. Classifier looks for an instance and assigns it to the specific class or not. Instances of the specific class are called relevant, and the others are named nonrelevant. If there is more than one class, this type of classification is called multiclass classification [1]. The classifier also assigns an instance to one of the multiple classes. In our study, we focus on functional classification of web pages, and we make binary classification since it is the basis of the focused crawlers [12, 13] or topical crawlers [14] of the search engines. The techniques developed in this study can also be used for subject classification and/or multiclass classification of web pages. (ii) In earlier studies, each ant selects features one by one; however, in our study each ant selects a set of features at a time since our feature space is too high, and selecting features one by one increases running time of ACO sharply. (iii) In earlier studies, only the effect of using features from bag of terms approach has been studied. In this study, the effect of using features only from URLs, from <title> tags, and tagged terms, as well as bag of terms approach, is investigated. We also study the effects of HTML tags on classification performance. (iii) In earlier studies, only the effect of using features from bag of terms approach has been studied. In this study, the effect of using features only from URLs, from <title> tags, and tagged terms, as well as bag of terms approach, is investigated. We also study the effects of HTML tags on classification performance. 2.2. Feature Selection for Web Page/Text Classification. Fea- ture selection is the one of the most important steps in classification systems. Web pages are generally in HTML format. This means that web pages are semistructured data, as they contain HTML tags and hyperlinks in addition to pure text. Because of this property of web pages, feature selection in web page classification is different than traditional text 2.2. Feature Selection for Web Page/Text Classification. Fea- ture selection is the one of the most important steps in classification systems. Web pages are generally in HTML format. This means that web pages are semistructured data, as they contain HTML tags and hyperlinks in addition to pure text. 2. Related Work They have studied the UCI dataset with C4.5 classifier and achieved 98.2% average accuracy for classification.i Mutual information (MI) was first introduced by Shan- non [15] in the context of digital communications between discrete random variables and was generalized to contin- uous random variables. Mutual information is considered as an acceptable measure of relevance between two ran- dom variables [17]. Mutual information is a probabilistic method which measures how much information the pres- ence/absence of a term contributes to making the correct classification decision on a class [18]. i Huang [24] has used classification accuracy and feature weights of the constructed SVM classifier to design the pheromone update in ACO based feature selection. In this study the UCI and simulated datasets are used and 94.65% average accuracy value is obtained. Document frequency (DF) is the number of documents in which a term occurs in a dataset. It is the simplest criterion for term selection and easily scales to a large dataset with linear computational complexity. It is a simple but effective feature selection method for text categorization [6]. Sivagaminathan and Ramakrishnan [25] have proposed a hybrid approach which is a combination of the neural networks and ACO for feature selection. Neural networks are used for error prediction and classification. In the exper- iments 3, 5, 8, 10, and 12 ants are used for the medical diagnosis dataset which contains 21 to 34 features, and 77.50% average accuracy on breast cancer (Wisconsin prognostic) and 98.22% average accuracy on thyroid disease datasets are achieved. Term strength (TS) has been proposed and evaluated by Wilbur and Sirotkin [16] for vocabulary reduction in text retrieval. Term strength is also used in text categorization [19, 20], such that it predicts term importance based on how commonly a term is likely to appear in “closely-related” documents. TS uses training set of documents to derive document pairs whose measured similarity according to the cosine value of the two document vectors is above a threshold. Then, “term strength” is computed based on the predicted conditional probability that a term occurs in the second half of a pair of related documents given that it occurs in the first half. The above methods namely the IG, the DF, the MI, and the TS have been compared by Yang and Pedersen [6] by using the kNN classifier on the Reuters [21] corpus. 2. Related Work Because of this property of web pages, feature selection in web page classification is different than traditional text This paper is organized as follows. In Section 2, we give related work on ACO based feature selection and web page 3 The Scientific World Journal which is a combination of fuzzy rough set and ACO to select features from the Water Treatment Plant database. Fuzzy rough set dependency measure is used for features which are more informative in the currently given selected subset. The number of features is reduced from 38 to 10 with fuzzy rough sets, and then it is further reduced from 10 to 9.5 with ACO. Dependency measure is used as the stopping criteria, and C4.5 is employed for classification. Instead of accuracy and F-measure values only the training and testing errors are presented. The error rates in training with no feature reduction, fuzzy rough set reduction, and ant fuzzy rough set reduction are 1.5%, 10.8%, and 6.5%, respectively. They have observed 19.1%, 25.2%, and 22.1% testing errors with no feature reduction, fuzzy rough set reduction, and ant fuzzy rough set reduction, respectively. classification. Feature selection is generally used to reduce dimension of data with tens or hundreds of thousands of features which would be impossible to process further. A major problem of web page classification is the high dimensionality of the feature space. The best feature subset contains the least number of features that most contribute to classification accuracy and efficiency.i ifi To improve the performance of web page classification, several approaches that are imported from feature selection for text classification have been applied. Information gain [11], mutual information [15], document frequency [6], and term strength [16] are the most popular traditional feature selection techniques. Information gain (IG) measures the amount of information in bits about the class prediction, if the only information available is the presence of a feature and the corresponding class distribution. Concretely, it measures the expected reduction in entropy [11].i Chen et al. [23] have proposed a rough set approach for feature selection based on ACO. Mutual information is used as heuristic information. Feature selection is started with a feature core rather than a random feature which causes complete graph to become in a smaller form. Feature reduction is finished after the core has been found. 2.3. Ant Colony Optimization for General Classification Prob- lems. Jensen and Shen [22] have proposed a hybrid approach 2. Related Work According to [6], IG is the most effective method with 98% feature reduction; DF is the simplest method with the lowest cost in computation and it can be credibly used instead of IG if computation of this measure is too expensive. Vieira et al. [26] have proposed two separate ant colonies combined with fuzzy models for feature selection; one colony is used for minimizing number of features, and the other one is employed to minimize classification error. In this study, the first colony determines the number of features and the second one selects the features. A fuzzy model is used as a classifier and 96.4% average accuracy on breast cancer (Wisconsin prognostic) dataset is observed. Nemati and Basiri [27] have proposed an ACO based method for a speaker verification system. Gaussian mixture model universal background model (GMM-UBM) is used as a classifier over the TIMIT corpora and equal error rate (EER) of the classification is taken as evaluation criteria. 4.56%, 2.634%, and 3.679% EER values are observed with GMM- UBM, ACO, and GA methods, respectively. In addition to traditional feature selection methods, swarm intelligence techniques are also popular to be used for feature selection. In this study, we review the application of ACO for feature selection in general classification problems and web/text classification domains. Kabir et al. [28] have combined ACO with neural net- works where neural networks are used as classifier. The proposed study contains both wrapper and filter methods such that a probabilistic formula for random selection and determination of subset size is used. Eight well-known cancer datasets are used in the experiments and 98.91% average accuracy is achieved. 2.3. Ant Colony Optimization for General Classification Prob- lems. Jensen and Shen [22] have proposed a hybrid approach 4 The Scientific World Journal the web page classification domain. Holden and Freitas [36] have been inspired by AntMiner [35] and used the ant colony paradigm to find a set of rules that classify the web pages into several categories. They have no prior assumptions about which words in the web pages to be classified can be used as potential discriminators. To reduce data rarity, they use stemming which is a technique in which different grammatical forms of a root word are considered as equivalent such that help, helping, and helped are taken as help. 2. Related Work Holden and Freitas [36] have also gathered sets of words if they are closely related in the WordNet electronic thesaurus. They have compared their AntMiner with the rule inference algorithms C4.5 and CN2. They have found that AntMiner is comparable in accuracy and forms simpler rules with respect to C4.5 and CN2. The best result of AntMiner is 81.0% classification accuracy and this result is obtained when WordNet generalization is used with Title features. Akarsu and Karahoca [29] have used ACO for clustering and feature selection. Ant colony clustering technique is used to segment breast cancer dataset. To remove irrelevant or redundant features from the dataset, sequential backward search technique is applied. Feature selection and clustering algorithms are incorporated as a wrapper. The results showed that the accuracy of the FS-ACO clustering approach is better than the filter approaches. They have compared their study with the clustering algorithm in Weka with respect to the sum of squared error value. According to the experimental evaluation, the sum of squared error is 732 for Weka and 758 for the proposed method. p p Al-Ani [30] has proposed an ACO based subset search procedure for speech classification problem. Local impor- tance of a given feature is measured by using the mutual information evaluation function, and only the best k subsets are used to update the pheromone. First iteration starts with 𝑚features. The second and following steps start with 𝑚−𝑝 features that are selected randomly from the previous k-best subsets. 𝑝is a numeric value which changes from 1 to 𝑚−1. They have studied TIMIT corpora with ANN classifier. The average classification accuracy of ACO, GA, and SFS over all the cases is 84.22%, 83.49%, and 83.19%, respectively. Aghdam et al. [7] have proposed an ACO based fea- ture selection algorithm for text classification. The features selected by an ant are evaluated according to the classifier performance and the feature subset length. In the exper- iments, it is assumed that classifier performance is more important than subset length, so that they assign 80% and 20% weights to classifier performance and the subset length, respectively. To measure classification performance, a simple k nearest neighbor classifier is used in the experiments. The performance of the proposed algorithm is compared with the performance of a genetic algorithm, information gain, and chi square analysis on the task of feature selection in Reuters-21578 dataset [21]. 2. Related Work Their experimental evaluation showed the superiority of the ACO based feature section over genetic algorithms, information gain, and chi square analysis methods. They studied only bag of terms feature extraction method and observed 89.08% microaverage F-measure value for Reuters-21578 dataset. Wang et al. [31] have developed an ACO based feature selection, which employs SVM classifier, to find the best feature subset for the UCI dataset. The experiments are performed with 5, 10, 15, 20, and 25 ants, and the best feature subset is found with 15 ants. In the experimental evaluation 94.83% average accuracy for Wine dataset and 79.57% average accuracy for Image Segment dataset are observed.i Jain and Singh [32] have modified probability function of ANT with exponential function for feature selection. Two modified ant algorithms are applied to a number of problems and the results are compared with those obtained by applying the original ACO and genetic algorithm, and the same results but with better execution time are observed. Jensen and Shen [37] have proposed an ACO enhanced fuzzy rough feature selection for web page classification. Terms extracted from web pages are weighted according to the 𝑡𝑓∗𝑖𝑑𝑓weighting scheme. In the proposed ACO based feature selection algorithm, each subset selected by each ant is evaluated by a fuzzy-rough measure of the selected subset. The pheromone values are updated according to this measure and the length of the selected subset. After selecting the best feature set, the web pages are then classified. The experiments are performed on a small dataset which contains 280 web pages collected from Arts & Humanities, Entertainment, Computers & Internet, Health, Business & Economy categories of Yahoo directory and it is observed that ACO based feature selection performs the highest degree of reduction in the feature space with minimal loss of information. Abd-Alsabour and Randall [33] have proposed a wrapper based system for feature selection. In the experiments, UCI dataset and SVM classifier are used. In the proposed feature selection algorithm, the number of selected features is not fixed, so the length of each ant’s feature subset may differ. By using the features selected by ants, 1 and 0.8584 classification accuracy values are observed for Wine and Vehicle datasets, respectively. y Rasmy et al. [34] have proposed a hybrid approach for feature selection. They have used ACO for selecting features and ANTMiner for classification. 2.4. Ant Colony Optimization for Text and Web Classification. AntMiner [35] is the first study that uses the ACO in 2. Related Work (v) We have also investigated which tags are more impor- tant for web page classification. (v) We have also investigated which tags are more impor- tant for web page classification. 2. Related Work In their ACO, ants start choosing a node (i.e., feature) randomly, after that classifier performance and length of selected feature vectors are adopted as heuristic information for ACO. In this study, UCI dataset having 150 to 560 features is used and the number of features is reduced to 9 and 70 for Diabetes and Analcatdata classes, respectively. After feature selection, 94.4% and 85% average classification accuracy is achieved for Sonar and Diabetes datasets, respectively. Janaki Meena et al. [38] have used ACO for feature selection and na¨ıve Bayes for classification over the 20 Newsgroup dataset. The ratio between observed frequency and expected frequency of the term is applied as a heuristic measure to the features extracted according to the bag of terms method. Map reduce is used for parallelization. Experiments are performed with 500 ants and 150 iterations. 2.4. Ant Colony Optimization for Text and Web Classification. AntMiner [35] is the first study that uses the ACO in The Scientific World Journal 5 aining ature traction Feature selection Classification The best features New (test) Preprocessing Classification Results ACO based feature selection b pages web pages Figure 1: Architecture of the proposed system. For talk.politics.mideast dataset, recall and precision values are 0.94 and 0.68, respectively. Training web pages Mangai et al. [39] have studied the WebKB dataset. Fea- tures are selected with Ward’s minimum variance measure, information gain, and 𝑡𝑓∗𝑖𝑑𝑓methods. Ward’s minimum variance measure is first used to identify clusters of redundant features in a web page. In each cluster, the best represen- tative features are retained and the others are eliminated. Removing such redundant features helps to minimize the resource utilization during classification. After clustering process, features are selected from these clusters, then kNN, SVM, na¨ıve Bayes, and C4.5 classifiers are used with 10- fold cross validation for classification. Course web pages are used as positive instances and student web pages are used as negative instances. The proposed method of feature selection is compared with other common feature selection methods. Experiments showed that the proposed method performs better than most of the other feature selection methods in terms of reducing the number of features and the classifier training time. 95.00% and 95.65% accuracy values are achieved with kNN and SVM classifiers, respectively.f New (test) web pages Feature extraction Results i Our proposed system is different from the above studies in the following respects. Figure 1: Architecture of the proposed system. 2. Related Work (i) We use ACO for selecting a predefined number of features from the large feature space extracted from the web pages instead of selecting features one by one. Our feature selection method chooses features as feature groups. Since using a single feature to determine the class of a web page is not enough, also, including features one by one to the selected feature list of each ant increases run time of ACO because all the computations for pheromone update must be repeated for each selection process. However, when we choose a set of features, we perform the necessary computations just once for each selected set. features are extracted, after that the best subset of features is selected by our ACO algorithm and then, by using the selected best features, new web pages are classified with Weka [40] data mining software implemented in Java. The components of our proposed system are explained in detail in the following subsections. 3.1. Feature Extraction. In the feature extraction phase all terms from the <title>, <h1>, <h2>, <h3>, <a>, <b>, <i>, <em>, <strong>, <p>, and <li> tags which denote title, header at level 1, header at level 2, header at level 3, anchor, bold, italic, emphasize, strong, paragraph, and list item, and additionally URL addresses of web pages are used. According to the experimental results of the earlier studies [41–43], these tags have useful information and should be used during feature extraction. To extract features, all the terms from each of the above mentioned tags and URL addresses of the relevant pages in the training set are taken. After term extraction, stopwords are removed and the remaining terms are stemmed by using the Porter’s stemmer [44]. The stemmed terms and their corresponding tags form our whole feature set. The details of the feature extraction step are presented in Section 4.3. (ii) We adopt ACO pheromone update formula for web page classification such that our ants are not blind as in the original ACO’s ants; we feed them with 𝑡𝑓∗𝑖𝑑𝑓 value of each term. So, they have an idea about terms before selecting features. (iii) We have investigated effects of using features from URLs, <title> tags, tagged terms, and bag of terms on the classification performance. (iv) We have used larger datasets with larger feature spaces. (iv) We have used larger datasets with larger feature spaces. 3. Ant Colony Optimization for Feature Selection 3.2. Feature Selection. After the feature extraction step, the (sub)optimum subset of features is selected with our ACO based feature selection algorithm. The flowchart of our proposed ACO algorithm is presented in Figure 2. This section includes our ACO based feature selection and classification system. The main structure of the proposed system is shown in Figure 1. Our system consists of feature extraction, ACO based feature selection, and classification components. The training dataset is prepared according to binary class classification problem. From the training dataset, In our proposed method, each feature represents a node, and all nodes are independent of each other. Nodes (i.e. features) are selected according to their selection probability 6 The Scientific World Journal 6 h Start Iteration count = 0 Iteration count ++ Select next feature Compute document vectors for each web page and form the arff file Classify web pages by using Weka Update pheromone values Iteration count > N Return the best feature set Stop For each ant Yes Yes No No \|Features\| < n Figure 2: Flow chart of the proposed ACO algorithm. Start Iteration count = 0 Select next feature Yes \|Features\| < n No Compute document vectors for each web page and form the arff file Classify web pages by using Weka Classify web pages by using Weka Update pheromone values No Iteration count > N Return the best feature set Figure 2: Flow chart of the proposed ACO algorithm. 𝑃𝑘(𝑖) which is given in (1). Initially, all nodes have the same selection probability: pheromone information respectively. The pheromone values, and parameters 𝛼and 𝛽are initialized according to [45] which have shown that 1 is the best value for 𝛼and 𝛽, and 10 is suitable for initial pheromone trail value. After all the ants have built a complete tour, the pheromone trail is updated according to the global update rule which is given in (2) as 𝑃𝑘(𝑖) = [𝜏(𝑖)]𝛼[𝜂(𝑖)]𝛽 ∑𝑙∈𝑁𝑘 𝑖[𝜏(𝑙)]𝛼[𝜂(𝑙)]𝛽, (1) (1) where 𝜂(𝑖) is equal to the document frequency of feature 𝑖which is the number of documents in the training set that contains feature 𝑖and represents heuristic information available to the ants. 𝑁𝑘 𝑖is the “feasible” neighborhood of ant 𝑘, that is, all features as yet unvisited by ant 𝑘. 𝜏(𝑖) is the pheromone trail value of feature 𝑖. 4. Experimental Evaluation and Results This section includes the datasets, namely, the WebKB and the Conference that were used in this study, the experiments performed, and their results. The proposed ACO based feature selection algorithm works as follows. Initially all features have the same selection probability 𝑃𝑘(𝑖) value. According to 𝑃𝑘(𝑖) value of each feature 𝑖, an ant 𝑘chooses 𝑛features. A roulette wheel selection algorithm [46] is used to select each of the 𝑛features. When an ant 𝑘chooses 𝑛features, the web pages in the training dataset are classified with respect to the selected 𝑛 features by using the C4.5 classifier (i.e., the J48 classifier) of Weka data mining tool. To classify the web pages in the training dataset, 𝑡𝑓∗𝑖𝑑𝑓values where 𝑡𝑓is the term frequency and 𝑖𝑑𝑓is the inverse document frequency of the selected features for each web page are taken as the feature values.hi The proposed ACO based feature selection algorithm works as follows. Initially all features have the same selection probability 𝑃𝑘(𝑖) value. According to 𝑃𝑘(𝑖) value of each feature 𝑖, an ant 𝑘chooses 𝑛features. A roulette wheel selection algorithm [46] is used to select each of the 𝑛features. 4.1. WebKB Dataset. The WebKB dataset [49] is a set of web pages collected by the World Wide Knowledge Base (Web- >KB) project of the CMU [50] text learning group and has been downloaded from The 4 Universities Dataset Homepage [51]. These pages are collected from computer science depart- ments of various universities in 1997 and manually classified into seven different classes, namely, student, faculty, staff, department, course, project, and others. For each class, the collection contains web pages from four universities which are Cornell, Texas, Washington, Wisconsin universities, and other miscellaneous pages collected from other universities.hi When an ant 𝑘chooses 𝑛features, the web pages in the training dataset are classified with respect to the selected 𝑛 features by using the C4.5 classifier (i.e., the J48 classifier) of Weka data mining tool. To classify the web pages in the training dataset, 𝑡𝑓∗𝑖𝑑𝑓values where 𝑡𝑓is the term frequency and 𝑖𝑑𝑓is the inverse document frequency of the selected features for each web page are taken as the feature values.hi The J48 classifier is an open source Java implementation of the C4.5 algorithm in the Weka data mining tool. C4.5 is a well-known decision tree algorithm, and it is an extension of Quinlan’s earlier ID3 algorithm [47]. 3. Ant Colony Optimization for Feature Selection Parameters 𝛼and 𝛽determine the relative influence of heuristic and 𝜏(𝑖) = 𝜌𝜏(𝑖) + 𝑛 ∑ 𝑘=1 Δ𝜏𝑘(𝑖) , (2) (2) where 𝜌denotes pheromone evaporation parameter which decays the pheromone trail, and 𝑛is the number of ants. According to [45], 𝜌value is selected as 0.2. The specific The Scientific World Journal 7 amount of pheromone, Δ𝜏𝑘(𝑖), that each ant k deposits on the trail is given by (3) as Table 1: Train/test distribution of WebKB dataset for binary class classification. Class Train relevant/nonrelevant Test relevant/nonrelevant Course 846/2822 86/942 Project 840/2822 26/942 Student 1485/2822 43/942 Faculty 1084/2822 42/942 Δ𝜏𝑘(𝑖) = { { { { { 2𝐵𝑘𝐿𝑘 if node 𝑖is used by elithist ant 𝑘 𝐵𝑘𝐿𝑘 if node 𝑖is used by any ant 𝑘 0 otherwise. (3) Δ𝜏𝑘(𝑖) = { { { { { 2𝐵𝑘𝐿𝑘 if node 𝑖is used by elithist ant 𝑘 𝐵𝑘𝐿𝑘 if node 𝑖is used by any ant 𝑘 0 otherwise. (3) Δ𝜏𝑘(𝑖) = { { { { { 2𝐵𝑘𝐿𝑘 if node 𝑖is used by elithist ant 𝑘 𝐵𝑘𝐿𝑘 if node 𝑖is used by any ant 𝑘 0 otherwise. (3) In (3), 𝐿𝑘is the F-measure value of ant k’s feature subset, and 𝐵𝑘is the unit pheromone value. This means that the higher the F-measure of the ant’s selected subset, the more the pheromone deposited on the features used in the subset, and these features are more likely to be selected in the next iteration.h 4. Experimental Evaluation and Results It builds decision trees from a set of training data using an extension of information gain known as gain ratio. After making classification by using the selected features, the classification performance is measured with respect to F-measure [48] value which is given in (4) as The 8,282 web pages are manually classified into the seven categories such that the student category has 1641 pages, faculty has 1124, staff has 137, department has 182, course has 930, project has 504, and other contains 3764 pages. The class other is a collection of pages that are not deemed as the “main page” and are not representing an instance of the previous six classes. The WebKB dataset includes 867 web pages from Cornell University, 827 pages from Texas University, 1205 pages from Washington University, 1263 pages from Wisconsin University, and finally 4120 miscellaneous pages from other universities. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 proposed method was tested under Microsoft Windows 7 operating system. The hardware used in the experiments has 16 GB of RAM and Intel Xenon E5-2643 3.30 GHz processor. Our feature selection method is tested on the Conference and the WebKB datasets. The proposed ACO based feature selection method is run for 250 iterations, since after 250 iterations we observed that there is no improvement on the classification performance. We have determined the number of ants as 30 experimentally, since we observed that 30 ants give satisfactory results for our study. pages are labeled as positive documents in the dataset. To complete the dataset, the short names of the conferences were queried using the Google search engine [53] manually, and the irrelevant pages in the result set were taken as negative documents. The dataset consists of 824 relevant and 1545 irrelevant pages which are approximately 2 times of the relevant pages. In our ACO based feature selection algorithm, each ant chooses a predefined number (i.e., 𝑛) of features. However, in classical ACO based systems, each ant choses features one by one. The reasons for choosing 𝑛features are that (i) our classification problem has thousands of features and inclusion or removal of one single feature does not make a considerable effect on the classification performance; (ii) choosing features one by one among the thousands of features increases time complexity of the ACO. In the experiments the number 𝑛is taken as 10, 100, and 500. In the Conference dataset, approximately 75% of both relevant and irrelevant pages are taken as the training set, and the remaining 25% of the relevant and irrelevant pages are included into the test set. The number of web pages in the train and test part of the Conference dataset is given in Table 2. In our ACO based feature selection system, we need a classifier to evaluate the fitness of each set of features selected by each ant. For this purpose, we employed C4.5 classifier since in our previous study [54] we compared classification performance of navie Bayes, kNN (i.e., IBk) and C4.5 (i.e., J48) algorithms of Weka data mining tool, and we observed that C4.5 is the best performer for our datasets.t 4.3. Feature Extraction. In the feature extraction phase all <title>, <h1>, <h2>, <h3>, <a>, <b>, <i>, <em>, <strong>, <li>, and <p> tags, text content, and URL addresses of web pages are used. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 All the terms from each of the above mentioned tags and URL addresses of the relevant web pages in the training set are taken. After term extraction, stopword removal and Porter’s stemming algorithm [44] are applied. Each stemmed term and its corresponding tag or URL pair forms a feature. For example, a term “program” in a <title> tag, in a <li> tag, or in a URL is taken as a different feature and this feature extraction method is called “tagged terms” method. Terms from similar HTML tags, for example, <strong>, <b>, <em>, and <i>, are grouped together to reduce the feature space.f After determining the best set of features by using our ACO based feature selection algorithm, we use the selected features to classify the web pages in the test datasets. We repeat this process (i.e., ACO based feature selection and then classification of test datasets) 5 times, and the best, worst, and average F-measure values of these 5 runs are presented in the following sections. In this study, features are selected from four different feature sets for each class. In the first set, features are extracted only from the URL addresses of web pages. Secondly, only <title> tags are used for feature extraction. In the third feature extraction method, all terms that appear in the web pages regardless of their HTML tag are used as features. In other words, a term which appears in the document regardless of its position is taken as a feature. This feature extraction approach is called “bag-of-terms” method. Finally, all terms that appear in each of the above listed HTML tags are used as features. In other words, a term which appears in different HTML tags is taken as a different feature (i.e., tagged terms).h 4.5. Experiment 1: Selecting Features Only from URLs of Web Pages. Performance of the proposed method for selecting features from only URL addresses of web pages is considered in this experiment. For this purpose, features are extracted only from the URL addresses of web pages in the training datasets. For all classes, 𝑛features are selected with our ACO based feature selection algorithm, and then test (unseen) web pages are classified with respect to these selected 𝑛features. To classify test web pages, J48 classifier is used. 𝐹-measure = 2 ∗recall ∗precision recall + precision . (4) (4) From the WebKB dataset Project, Faculty, Student, and Course classes are used in this study. As Staff and Department classes have small number of positive examples, they are not considered. Training and test datasets are constructed as described in the WebKB project website [51]. For each class, training set includes relevant pages which belong to randomly chosen three universities and others class of the dataset. The fourth university’s pages are used in the test phase. Approximately 75% of the irrelevant pages from others class are included in the training set and the remaining 25% of them are included in the test set. The number of web pages in the train and test part of the WebKB dataset, which is used in this study, is given in Table 1. For example, the Course class includes 846 relevant and 2822 irrelevant pages for the training phase and 86 relevant and 942 irrelevant pages for the test phase. F-measure is a combination of precision and recall such that recall is the proportion of web pages which are classified as class 𝐶𝑖, among all web pages which truly have class 𝐶𝑖, and precision is the proportion of the web pages which truly have class 𝐶𝑖among all those which are classified as class 𝐶𝑖. In earlier studies, researchers measured performance with respect to F-measure value. To comply with the standards on this issue, F-measure value is chosen as the performance metric in this study. The above feature selection and F- measure value computations are repeated for all ants. After these computations, an ant is chosen as an elitist ant which has the highest F-measure value. Then, the pheromone values are updated based on (2) and (3). This process is repeated a predetermined number of times (i.e., N). Finally, the feature subset having the best F-measure value is chosen as the best feature set which can then be used for classifying new (unseen) web pages. 4.2. Conference Dataset. The Conference dataset consists of the computer science related conference homepages that were obtained from the DBLP website [52]. The conference web The Scientific World Journal 8 8 Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 — 1 1 1 0.794 Worst — 1 1 1 0.745 Table 5: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from ⟨title⟩tags for all classes. rst, and average 𝐹-measure values of classification using ACO selected 𝑛features from URLs for all classes. Table 4: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from URLs for all classes. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 1 1 1 1 0.782 Avg. 1 0.911 0.993 1 0.664 Worst 1 0.822 0.986 1 0.545 100 Best 1 1 1 1 0.883 Avg. 1 1 1 1 0.745 Worst 1 1 1 1 0.606 500 Best — 1 1 1 0.842 Avg. — 1 1 1 0.794 Worst — 1 1 1 0.745 Table 5: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from ⟨title⟩tags for all classes. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.983 0.92 0.891 0.947 0.736 Avg. 0.966 0.903 0.883 0.901 0.712 Worst 0.948 0.885 0.875 0.854 0.687 100 Best 0.983 0.92 0.883 0.939 0.739 Avg. 0.976 0.916 0.879 0.930 0.713 Worst 0.980 0.911 0.869 0.921 0.687 500 Best — 0.92 0.883 0.94 0.741 Avg. — 0.915 0.879 0.933 0.678 Worst — 0.91 0.869 0.922 0.710 Table 5: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from ⟨title⟩tags for all classes. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.983 0.92 0.891 0.947 0.736 Avg. 0.966 0.903 0.883 0.901 0.712 Worst 0.948 0.885 0.875 0.854 0.687 100 Best 0.983 0.92 0.883 0.939 0.739 Avg. 0.976 0.916 0.879 0.930 0.713 Worst 0.980 0.911 0.869 0.921 0.687 500 Best — 0.92 0.883 0.94 0.741 Avg. — 0.915 0.879 0.933 0.678 Worst — 0.91 0.869 0.922 0.710 , and average 𝐹-measure values of classification using ACO selected 𝑛features from ⟨title⟩tags for all classes. the best F-measure value obtained is 0.883. So, we can say that the Conference dataset has less meaningful URL addresses with respect to the WebKB dataset. There is no considerable difference between varying numbers of features for the WebKB dataset. For the Conference dataset, on the other hand, reducing the number of features also reduces the average F-measure values of classification of the test web pages. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 This process is repeated 5 times and the best, worst, and average classification performance of these 5 runs in F-measure values for the selected 𝑛features for all datasets is given in Table 4. As Course dataset has less than 500 features extracted from URLs (see Table 3), we only select 100 and 10 features from URLs for Course class. f The number of features varies according to the dataset and the feature extraction method used. Numbers of features for each class of all datasets with respect to the feature extraction method used are shown in Table 3. As an example, 33519 features are extracted when tagged terms method are used for the Course class. When only the <title> tag is considered, the number of features extracted reduces to 305 for this class. In Table 4, the best F-measure values are written in bold face for all classes. According to the experimental results presented in Table 4, we observed that the WebKB dataset includes meaningful URL addresses so that, for all 𝑛values, web pages from the Course, Project, Student, and Faculty classes are classified at 100% accuracy by using the ACO selected features from the URLs of the web pages. However, this observation is not true for the Conference dataset, as 4.4. Experimental Setup. In this study, Perl script language was used for the feature extraction phase, and our ACO based feature selection algorithm was implemented in Java programming language under Eclipse environment. The The Scientific World Journal 9 Table 3: Number of features for all classes with respect to feature extraction methods. Class Feature extraction method Tagged terms Bag of terms ⟨title⟩tag URL Course 33519 16344 305 479 Project 30856 15307 596 686 Student 49452 22245 1987 1557 Faculty 47376 24641 1502 1208 Conference 34952 18572 890 1115 Table 3: Number of features for all classes with respect to feature extraction methods. Table 4: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from URLs for all classes. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 1 1 1 1 0.782 Avg. 1 0.911 0.993 1 0.664 Worst 1 0.822 0.986 1 0.545 100 Best 1 1 1 1 0.883 Avg. 1 1 1 1 0.745 Worst 1 1 1 1 0.606 500 Best — 1 1 1 0.842 Avg. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 g When the results of these four experiments are compared, we observed that the WebKB dataset (Course, Project, Stu- dent, and Faculty classes) is classified with 100% accuracy (i.e., F measure values are 1.0), and the classification per- formance of the Conference dataset is also satisfactory (i.e., F-measure values are up to 0.952). For the WebKB dataset, when the number of selected features is small (i.e., 𝑛= 10), the best feature extraction methods are URL only and tagged terms method. For the Conference dataset, on the other hand, the best feature extraction method is bag of terms and then tagged terms method. This result has occurred because of the fact that the URLs of the WebKB dataset have class specific terms; for the Conference dataset on the other hand, URLs do not have class specific terms. Since tagged terms feature extraction method is successful in general cases, we use this feature extraction method in the rest of the experiments. 4.7. Experiment 3: Selecting Features from Bag of Terms Method. In this experiment, features are selected among terms which are extracted by using bag of terms method. In this method, only the terms regardless of their position or tag are taken as features. Our ACO based feature selection algorithm runs 5 times for each dataset, and we obtain 5 sets of best features. Then by using the selected set of features, we classify the test web pages with J48 classifier. The best, worst, and average classification performance in F-measure value of these 5 runs for classifying the test web pages for all datasets is given in Table 6. In Table 6, the best F-measure values are written in bold face for all classes. When we compare Tables 4, 5, and 6, we observed that text contents in web pages are more meaningful for classification than URL addresses and titles for the Conference dataset. However, in the WebKB dataset, URL addresses have better features for classification than page contents and titles of web pages. As in the previous experiments, F-measure values change with the number of features and the average classification performance decreases when the number of features is decreased. 4.9. Experiment 5: Using Other Classifiers for the Test Phase. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 the test dataset are classified with respect to these selected 𝑛 features by using J48 classifier. This process is repeated 5 times and the best, worst, and average classification performance for these 5 runs in F-measure value for the selected 𝑛features for all datasets is given in Table 5. As Course dataset has less than 500 features extracted from <title> tags (see Table 3), we only select 100 and 10 features from <title> tags for Course class. In Table 5, the best F-measure values are written in bold face for all classes. According to the results presented in Table 5, the F-measure values of the classification of the test web pages by using the ACO selected 𝑛features are also high for the WebKB dataset, which implies that the WebKB dataset includes meaningful title declarations. However, the title declarations of the Conference dataset are not meaningful as the WebKB dataset. In this experiment, the reduced number of features affects average F-measure values negatively for the Course, Student, and Faculty classes; however, for the 4.6. Experiment 2: Selecting Features Only from <title> Tags. Performance of the proposed method using only <title> tags of Web pages is considered in this experiment. For this purpose, features are extracted only from the <title> tags of web pages in the training datasets. Similar to the first experiment, each ant selects 𝑛features from the whole feature set that are obtained from <title> tags, and our ACO algorithm returns the best 𝑛features. After that web pages in The Scientific World Journal 10 Table 6: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from bag of terms method for all classes. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.926 0.896 0.95 0.862 0.933 Avg. 0.880 0.780 0.944 0.836 0.926 Worst 0.834 0.665 0.947 0.811 0.92 100 Best 0.895 0.840 0.934 0.928 0.948 Avg. 0.873 0.821 0.903 0.918 0.940 Worst 0.851 0.802 0.872 0.908 0.932 500 Best 0.960 0.973 0.981 0.921 0.952 Avg. 0.955 0.969 0.964 0.918 0.949 Worst 0.950 0.966 0.947 0.915 0.946 Project and Conference classes the average F-measure values are improved when the number of features is reduced. Project and Conference classes the average F-measure values are improved when the number of features is reduced. classification performance is the best for the medium and high 𝑛values (i.e., 100 and 500). Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 Table 7: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from tagged terms method for all classes. h , , gi g gg classes. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.963 0.94 0.954 0.965 0.921 Avg. 0.957 0.938 0.95 0.952 0.904 Worst 0.954 0.937 0.941 0.946 0.877 100 Best 0.902 0.82 0.923 0.928 0.94 Avg. 0.881 0.808 0.887 0.914 0.932 Worst 0.865 0.79 0.845 0.869 0.911 500 Best 0.9 0.35 0.91 0.356 0.936 Avg. 0.87 0.345 0.896 0.355 0.932 Worst 0.885 0.345 0.884 0.354 0.929 Table 8: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from tagged terms for all classes when IBk classifier is used in the test phase. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.948 0.929 0.937 0.948 0.906 Avg 0.943 0.914 0.92 0.936 0.888 Worst 0.929 0.909 0.896 0.915 0.842 100 Best 0.95 0.887 0.918 0.965 0.866 Avg 0.94 0.872 0.906 0.951 0.861 Worst 0.932 0.858 0.896 0.941 0.855 500 Best 0.951 0.902 0.918 0.94 0.844 Avg 0.944 0.899 0.911 0.938 0.84 Worst 0.937 0.898 0.908 0.936 0.837 Table 8: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from ta IBk classifier is used in the test phase. Table 8: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from tagged terms for all classes when IBk classifier is used in the test phase. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.948 0.929 0.937 0.948 0.906 Avg 0.943 0.914 0.92 0.936 0.888 Worst 0.929 0.909 0.896 0.915 0.842 100 Best 0.95 0.887 0.918 0.965 0.866 Avg 0.94 0.872 0.906 0.951 0.861 Worst 0.932 0.858 0.896 0.941 0.855 500 Best 0.951 0.902 0.918 0.94 0.844 Avg 0.944 0.899 0.911 0.938 0.84 Worst 0.937 0.898 0.908 0.936 0.837 According to Table 8, as the number of features selected decreases, average classification performance increases for the Project, Student, and Conference datasets when IBk classifier is used in the test phase. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 For the Faculty dataset, the best classification performance is obtained when 100 features are selected.i According to Figures 3, 4, 5, 6, and 7, when the number of selected features is small (i.e., 10 to 100) the most dis- criminative features are obtained from h1, title, and anchor tags. As the number of selected features increases (i.e., 500), features extracted from anchor tag and body text become more dominating. According to our experimental results we observed that using small number of features (10 to 100) is enough to make a good classification for most of our datasets, so instead of using all the features extracted from the web pages, it is enough to use features extracted from URL, h1, title, and anchor tags. For the na¨ıve Bayes classifier, as the number of selected features decreases, the classification performance increases as shown in Table 9. According to Tables 7, 8, and 9 the features selected by our ACO based feature selection algorithm which uses J48 classifier give satisfactory classification performance when these features are used to classify new web pages with IBk and na¨ıve Bayes classifiers. However, when the results presented in Tables 7, 8, and 9 are compared, we can say that using J48 in the test phase yields more satisfactory classification, since we choose features in ACO according to the J48 classifier. 4.11. Experiment 7: Effect of ACO Based Feature Selection on Classification Performance in Terms of Accuracy and Running Time. In this section, effect of the proposed ACO based feature selection algorithm on the classification performance has been investigated. For this purpose, F-measure values and running time of the classification of the test web pages using the C4.5 classifier with and without the ACO based feature selection are compared. The results of this experiment are presented in Table 10 and Figure 8. In Table 10, F-measure values of classification of test web pages with ACO based feature selection and without making any feature selection (i.e., by using all features) are compared for features extracted from tagged terms, URLs, and <title> tags. 4.11. Experiment 7: Effect of ACO Based Feature Selection on Classification Performance in Terms of Accuracy and Running Time. In this section, effect of the proposed ACO based feature selection algorithm on the classification performance has been investigated. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 C4.5 decision tree classifier, which is named J48 in Weka, is employed in our ACO based feature selection system to compute the fitness of the selected features and therefore to update pheromone values and selection probabilities of features. In the test phase of the previous experiments, we also applied J48 classifier to classify the test (unseen) web pages. In this experiment, our aim is to show whether our ACO based selected features give good classification performance with other well-known classifiers, namely, the kNN and the na¨ıve Bayes. For this purpose, we employ kNN, which is named IBk in Weka, and na¨ıve Bayes classifiers in the test phase. In the ACO-based feature selection phase, we employed J48 classifier. It is also possible to employ kNN or na¨ıve Bayes classifiers in the ACO-based feature selection algorithm; however, according to our previous study [54], J48 has better classification performance with respect to kNN and na¨ıve Bayes for our datasets and that is why we prefer J48 in the feature selection process. Using kNN or na¨ıve Bayes in the ACO based feature selection may be considered for the future work. The best, worst, and average classification performance in F-measure values of the 5 runs is presented in Tables 8 and 9. 4.8. Experiment 4: Selecting Features from Tagged Terms Method. In this experiment, features are selected among terms which are extracted by using tagged terms method such that each term in each HTML tag is taken as a different feature. As URLs have very discriminating terms for the WebKB dataset, terms from URLs are not taken in this experiment. Our ACO based feature selection algorithm is run 5 times and we obtain 5 sets of best features for each case. Then by using the selected set of features, we classify the test web pages by using the J48 classifier. The best, worst, and average classification performance in F-measure value of these 5 runs for ACO selected 𝑛features is presented in Table 7. In Table 7, the best F-measure values are written in bold face. According to Table 7, as 𝑛decreases classification performance increases for the Course, Student, Project, and Faculty classes. However, for the Conference dataset, average 11 The Scientific World Journal The Scientific World Journal Table 7: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from tagged terms method for all classes. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 0.858 0.354 0.901 0.762 0.695 Worst 0.829 0.324 0.889 0.602 0.645 500 Best 0.543 0.54 0.502 0.527 0.826 Avg. 0.51 0.539 0.484 0.51 0.821 Worst 0.478 0.53 0.472 0.501 0.816 Table 10: 𝐹-measure values of the C4 5 classifier with and without making any feature selection Table 10: 𝐹-measure values of the C4.5 classifier with and without making any feature selection. Dataset Feature extraction methods Tagged terms Title URL ACO selected features Without feature selection ACO selected features Without feature selection ACO selected features Without feature selection Course 0.963 0.909 0.983 0.898 1 1 Faculty 0.965 0.911 0.947 0.920 1 1 Project 0.954 0.355 0.891 0.985 1 1 Student 0.940 0.337 0.920 0.679 1 1 Conference 0.940 0.930 0.741 0.372 0.883 0.972 Table 10: 𝐹-measure values of the C4.5 classifier with and without making any feature selection. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 350 Figure 3: Distribution of the ACO selected tags for course class. When we compared the results presented in Table 10, we can say that classification performance in terms of F-measure increases when we used features selected by our ACO based feature selection method from tagged terms. For the features extracted from title tags, classification performance increases with our ACO based feature selection method for the Course, Student, Faculty, and Conference datasets, but F-measure value decreases for the Project dataset. For the features extracted from URLs both using all features and making ACO based feature selection give satisfactory classification performance. In Figure 8, time required to classify test web pages for all classes when C4.5 classifier is used are displayed. As it can be easily seen from the figure, making feature selection reduces the time required to classify new (unseen) web pages sharply without making reduction in classification accuracy (Table 10). Figure 3: Distribution of the ACO selected tags for course class. 4.12. Experiment 8: Comparison of the Proposed Method with the Well-Known Feature Selection Methods. In this experi- ment, we compared our ACO based feature selection method with two well-known feature selection methods that are information gain and chi square analysis. To accomplish this, we select 10, 100, and 500 features extracted from tagged terms method with information gain (IG) and chi square (Chi) feature selection methods. After that, we classify test datasets by using the selected features. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 For this purpose, F-measure values and running time of the classification of the test web pages using the C4.5 classifier with and without the ACO based feature selection are compared. The results of this experiment are presented in Table 10 and Figure 8. In Table 10, F-measure values of classification of test web pages with ACO based feature selection and without making any feature selection (i.e., by using all features) are compared for features extracted from tagged terms, URLs, and <title> tags. 4.10. Experiment 6: Distribution of Tags in the Selected Feature Subsets. In this section, we have investigated the distribution of tags in the ACO selected subset of features from tagged terms method for each class. Figures 3, 4, 5, 6, and 7 show tag distributions for the ACO selected features for these five classes. Since URLs are very dominant features for the WebKB dataset as it can be seen from Table 4, we did not include features from URLs in this experiment. The Scientific World Journal 12 Table 9: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from tagged terms for all classes when na¨ıve Bayes classifier is used in the test phase. Table 9: The best, worst, and average 𝐹-measure values of classification using ACO selected 𝑛features from tagged terms for all classes when na¨ıve Bayes classifier is used in the test phase. No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.933 0.864 0.917 0.936 0.897 Avg. 0.856 0.863 0.88 0.926 0.855 Worst 0.618 0.859 0.809 0.909 0.838 100 Best 0.868 0.38 0.912 0.875 0.797 Avg. 0.858 0.354 0.901 0.762 0.695 Worst 0.829 0.324 0.889 0.602 0.645 500 Best 0.543 0.54 0.502 0.527 0.826 Avg. 0.51 0.539 0.484 0.51 0.821 Worst 0.478 0.53 0.472 0.501 0.816 Table 10: 𝐹-measure values of the C4.5 classifier with and without making any feature selection. Dataset Feature extraction methods Tagged terms Title URL ACO selected features Without feature selection ACO selected features Without feature selection ACO selected features Without feature selection Course 0.963 0.909 0.983 0.898 1 1 Faculty 0.965 0 911 0.947 0 920 1 1 No. of features selected (𝑛) 𝐹-measure Dataset Course Student Project Faculty Conference 10 Best 0.933 0.864 0.917 0.936 0.897 Avg. 0.856 0.863 0.88 0.926 0.855 Worst 0.618 0.859 0.809 0.909 0.838 100 Best 0.868 0.38 0.912 0.875 0.797 Avg. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 The classification performance in F-measure values of our ACO based method, chi Square, and information gain methods is presented in Table 11.h The best F-measure values are written in bold face for all cases in Table 11. According to the results given in Table 11, our ACO based feature selection algorithm can select better features with respect to chi square and information gain especially when the number of selected feature is small (i.e., less than 500). As shown in Table 11, when the number of selected features increases, the classification performance of 13 The Scientific World Journal The Scientific World Journal 13 Table 11: Comparison of the proposed method with well-known feature selection methods. Datasets No. of selected features 10 100 500 IG Chi ACO IG Chi ACO IG Chi ACO Faculty 0.89 0.93 0.96 0.90 0.91 0.92 0.90 0.90 0.35 Course 0.91 0.88 0.96 0.36 0.36 0.90 0.90 0.91 0.90 Student 0.80 0.79 0.94 0.42 0.42 0.82 0.34 0.34 0.35 Project 0.83 0.90 0.95 0.39 0.39 0.92 0.35 0.35 0.91 Conference 0.93 0.57 0.92 0.94 0.94 0.94 0.93 0.93 0.93 Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 Figure 4: Distribution of the ACO selected tags for project class. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 350 Figure 5: Distribution of the ACO selected tags for faculty class. test data decreases, so it is better to use small number of features in terms of classification accuracy and running time (i.e., Figure 8). 4.13. Comparison of the Proposed Method with Earlier Studies. Several ACO based feature selection algorithms have been Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 Figure 6: Distribution of the ACO selected tags for student class. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 Figure 7: Distribution of the ACO selected tags for conference class. proposed for the UCI dataset. The authors of [22–26, 28, 29, 31, 33, 34] have reported that their proposed ACO based algorithms increase the performance of classification. According to our experimental results, we also observed that ACO based feature selection algorithm improves the 13 Table 11: Comparison of the proposed method with well-known feature selection methods. 5. Conclusion In this study we have developed an ACO-based feature selection system for the web page classification problem. We have used four kinds of feature extraction methods that are, namely, URL only, <title> tag only, bag of terms, and tagged terms. After the feature extraction step, in our ACO based feature selection algorithm, for each ant we select a predefined number of features, and then we evaluate the performance of the selection made by each ant by using the C4.5 classifier. According to the performance of the selected features, we update pheromone values and selection probabilities of the features. This process is repeated until the best features are selected. After selecting the best features, classification of the new (unseen) web pages is made by using the selected feature set. Experimental evaluation shows that using tagged terms as feature extraction method gives good classification performance on the average cases with respect to using bag-of-terms, URL alone, or <title> tag alone methods. Our ACO based feature selection system is able to select better features with respect to well-known information gain and chi square selection methods. The proposed system is effective in reducing the number of features so that it is suitable for classification of high dimensional data. By reducing the feature space, our system also reduces the time required to classify new web pages sharply without loss of accuracy in classification. As future work, performance of our ACO-based feature selection system may be evaluated for the multiclass case. Figure 8: Comparison of time required to classify test pages. Figure 8: Comparison of time required to classify test pages. classification performance in terms of classification accuracy and time for the WebKB and Conference datasets in general. In [55], sequential n-grams are used to extract features from the URLs of web pages. The selected features are then classified with maximum entropy and support vector machine separately. The average F-measure value is 0.525 for the multiclass classification of the WebKB dataset. Our average F-measure value is 1.0 for WebKB dataset when features are selected from URLs and binary classification is made. Based on these results, we can say that our ACO based feature selection with binary class classification has better classification performance for the WebKB dataset. i ¨Ozel [56] has used tagged terms as features with a GA based classifier. URL addresses are not used in feature extraction step. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 Table 11: Comparison of the proposed method with well-known feature selection methods. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 Figure 6: Distribution of the ACO selected tags for student class. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 Figure 4: Distribution of the ACO selected tags for project class. Figure 4: Distribution of the ACO selected tags for project class. Figure 6: Distribution of the ACO selected tags for student class. Figure 4: Distribution of the ACO selected tags for project class. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 350 Figure 5: Distribution of the ACO selected tags for faculty class. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 300 350 Figure 5: Distribution of the ACO selected tags for faculty class. Title h1 Anchor Bold Text li 500 features 100 features 10 features 0 50 100 150 200 250 Figure 7: Distribution of the ACO selected tags for conference class. Figure 5: Distribution of the ACO selected tags for faculty class. Figure 7: Distribution of the ACO selected tags for conference class. test data decreases, so it is better to use small number of features in terms of classification accuracy and running time (i.e., Figure 8). proposed for the UCI dataset. The authors of [22–26, 28, 29, 31, 33, 34] have reported that their proposed ACO based algorithms increase the performance of classification. According to our experimental results, we also observed that ACO based feature selection algorithm improves the 4.13. Comparison of the Proposed Method with Earlier Studies. Several ACO based feature selection algorithms have been 14 The Scientific World Journal The Scientific World Journal 500 features 100 features 10 features Conference Course Faculty Project Student All features 0 10 20 30 40 50 60 70 Figure 8: Comparison of time required to classify test pages. Mangai et al. [39] have used Ward’s minimum variance measure and information gain for feature selection from the WebKB dataset. 95% accuracy is achieved for Couse dataset with a kNN classifier. 5. Conclusion Average F-measure values are 0.9 and 0.7 for the Course and the Student classes of the WebKB dataset. In our proposed method, on the other hand, average F-measure value is increased up to 1.0 for the Course and Student classes. This comparison shows that features from URLs affect the classification performance positively. In addition to this, using ACO based feature selection and then applying the C4.5 classifier perform better than a GA based classifier.i Conflict of Interests ii Jiang [57] has proposed a text classification algorithm that combines a k-means clustering scheme with a variation of expectation maximization (EM), and the proposed method can learn from a very small number of labeled samples and a large quantity of unlabeled data. Experimental results show that the average F-measure value is 0.7 for WebKB dataset in multiclass classification [57]. This result shows that our ACO- based algorithm with binary classification performs better since it yields higher F-measure value. The authors declare that there is no conflict of interests regarding the publication of this paper. Acknowledgments This work was supported by C¸ukurova University Academic Research Project Unit under Grant no MMF2013D10 and by The Scientific and Technological Research Council of Turkey (T ¨UB˙ITAK) scholarship 2211-C. Joachims [58] has used transductive support vector machines on WebKB dataset with binary class classification and for feature extraction. Bag-of-terms method is used. According to the experimental results of this study, average F-measure values are reported as 0.938, 0.537, 0.184, and 0.838 for the Course, Faculty, Project, and Student classes, respectively. Also, these results show that our proposed algorithm has better performance with respect to the SVM algorithm. Table 2: Train/test distribution of the conference dataset. Train relevant/nonrelevant Test relevant/nonrelevant Conference 618/1159 206/386 However, when we use our ACO-based feature selection algorithm and then apply a kNN classifier to test web pages from the Course dataset, we also obtained 0.951 F-measure value. When we apply C4.5 classifier after our ACO-based feature selection algorithm, the F-measure value increases up to 1. References [1] X. Qi and B. D. Davison, “Web page classification: features and algorithms,” ACM Computing Surveys, vol. 41, no. 2, article 12, 2009. [2] Yahoo!, https://maktoob.yahoo.com/?p=us. [2] Yahoo!, https://maktoob.yahoo.com/?p=us. The Scientific World Journal 15 [24] C. L. Huang, “ACO-based hybrid classification system with feature subset selection and model parameters optimization,” Neurocomputing, vol. 73, no. 1–3, pp. 438–448, 2009. [3] Open direct Project, http://www.dmoz.org/. [4] C. Chen, H. Lee, and C. Tan, “An intelligent web-page classifier with fair feature-subset selection,” Engineering Applications of Artificial Intelligence, vol. 19, no. 8, pp. 967–978, 2006. [25] R. K. Sivagaminathan and S. Ramakrishnan, “A hybrid approach for feature subset selection using neural networks and ant colony optimization,” Expert Systems with Applications, vol. 33, no. 1, pp. 49–60, 2007. i [5] W. Shang, H. Huang, H. Zhu, Y. Lin, Y. Qu, and Z. Wang, “A novel feature selection algorithm for text categorization,” Expert Systems with Applications, vol. 33, no. 1, pp. 1–5, 2007. [26] S. M. Vieira, J. M. C. Sousa, and T. A. Runkler, “Two cooperative ant colonies for feature selection using fuzzy models,” Expert Systems with Applications, vol. 37, no. 4, pp. 2714–2723, 2010. [6] Y. Yang and J. O. Pedersen, “A comparative study on feature selection in text categorization,” in Proceedings of the 14th International Conference on Machine Learning (ICML ’97), pp. 412–420, Nashville, Tenn, USA, July 1997. [27] S. Nemati and M. E. Basiri, “Text-independent speaker verifi- cation using ant colony optimization-based selected features,” Expert Systems with Applications, vol. 38, no. 1, pp. 620–630, 2011. [7] M. H. Aghdam, N. Ghasem-Aghaee, and M. E. Basiri, “Text feature selection using ant colony optimization,” Expert Systems with Applications, vol. 36, no. 3, pp. 6843–6853, 2009. [8] M. Dorigo, Optimization, learning and natural algorithms [Ph. D. thesis], Politecnico di Milano, Milan, Italy, 1992. [28] M. M. Kabir, M. Shahjahan, and K. Murase, “A new hybrid ant colony optimization algorithm for feature selection,” Expert Systems with Applications, vol. 39, no. 3, pp. 3747–3763, 2012. [9] M. Dorigo, V. Maniezzo, and A. Colorni, “Ant system: optimiza- tion by a colony of cooperating agents,” IEEE Transactions on Systems, Man, and Cybernetics B: Cybernetics, vol. 26, no. 1, pp. 29–41, 1996. [29] E. Akarsu and A. Karahoca, “Simultaneous feature selection and ant colony clustering,” Procedia Computer Science, vol. 3, pp. 1432–1438, 2011. [10] L. Liu, Y. Dai, and J. References Gao, “Ant colony optimization algorithm for continuous domains based on position distribution model of ant colony foraging,” The Scientific World Journal, vol. 2014, Article ID 428539, 9 pages, 2014. [30] A. Al-Ani, “Ant colony optimization for feature subset selec- tion,” World Academy of Science, Engineering and Technology, vol. 1, no. 4, 2007. [31] W. Wang, Y. Jiang, and S. W. Chen, “Feature subset selection based on ant colony optimization and support vector machine,” in Proceedings of the 7th WSEAS International Conference of Signal Processing, Computational Geometry & Artificial Vision, p. 182, Athens, Greece, August 2007. [11] T. M. Mitchell, Machine Learning, McGraw-Hill, New York, NY, USA, 1st edition, 1997. [12] S. Chakrabarti, M. van den Berg, and B. Dom, “Focused crawl- ing: a new approach to topic-specific Web resource discovery,” Computer Networks, vol. 31, no. 11–16, pp. 1623–1640, 1999. [32] N. Jain and J. P. Singh, “Modification of ant algorithm for feature selection,” in Proceedings of the International Conference on Control Automation, Communication and Energy Conservation (INCACEC ’09), June 2009. [13] S. A. ¨Ozel and E. Sarac¸, “Focused crawler for finding profes- sional events based on user interests,” in Proceedings of the 23rd International Symposium on Computer and Information Sciences (ISCIS ’08), pp. 441–444, Istanbul, Turkey, October 2008. [33] N. Abd-Alsabour and M. Randall, “Feature selection for clas- sification using an ant colony system,” in Proceedings of the 6th IEEE International Conference on e-Science Workshops (e- ScienceW ’10), pp. 86–91, Brisbane, Australia, December 2010. [14] F. Menczer and R. K. Belew, “Adaptive information agents in distributed textual environments,” in Proceedings of the 2nd International Conference on Autonomous Agents (AGENTS ’98), pp. 157–164, Minneapolis, Minn, USA, May 1998.h [34] M. H. Rasmy, M. El-Beltagy, M. Saleh, and B. Mostafa, “A hybridized approach for feature selection using ant colony optimization and ant-miner for classification,” in Proceeding of the 8th International Conference on Informatics and Systems (INFOS '12), pp. BIO-211–BIO-219, Cairo, Egypt, May 2012. [15] C. E. Shannon, “A mathematical theory of communication,” The Bell System Technical Journal, vol. 27, pp. 379–423, 1948.hi [16] W. J. Wilbur and K. Sirotkin, “The automatic identification of stop words,” Journal of Information Science, vol. 18, no. 1, pp. 45– 55, 1992. [35] R. S. Parpinelli, H. S. Lopes, and A. Freitas, “An ant colony algorithm for classification rule discovery,” IEEE Transactions on Evolutionary Computation, vol. 6, no. 4, pp. 321–332, 2002. [17] T. M. References Cover and J. A. Thomas, Elements of Information Theory, Wiley-Interscience Press, New York, NY, USA, 1991.h [18] S. Guiasu, Information Theory with Applications, McGraw-Hill Press, New York, NY, USA, 1st edition, 1977. [36] N. Holden and A. A. Freitas, “Web page classification with an ant colony algorithm,” in Parallel Problem Solving from Nature— PPSN VIII, vol. 3242 of Lecture Notes in Computer Science, pp. 1092–1102, Springer, Berlin, Germany, 2004. [19] Y. Yang, “Noise reduction in a statistical approach to text cate- gorization,” in Proceedings of the 18th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 256–263, July 1995. [37] R. Jensen and Q. Shen, “Web page classification with aco- enhanced fuzzy-rough fetaure selection,” vol. 4259 of Lecture Notes in Artificial Intelligence, pp. 147–156, 2006. [20] Y. Yang and W. J. Wilbur, “Using corpus statistics to remove redundant words in text categorization,” Journal of the American Society for Information Science, vol. 47, no. 5, pp. 357–369, 1996. [38] M. Janaki Meena, K. R. Chandran, A. Karthik, and A. Vijay Samuel, “An enhanced ACO algorithm to select features for text categorization and its parallelization,” Expert Systems with Applications, vol. 39, no. 5, pp. 5861–5871, 2012. [21] Reuters Dataset, http://kdd.ics.uci.edu/databases/reuters21578/ reuters21578.html. [22] R. Jensen and Q. Shen, “Fuzzy-rough data reduction with ant colony optimization,” Fuzzy Sets and Systems, vol. 149, no. 1, pp. 5–20, 2005. [39] J. A. Mangai, V. S. Kumar, and S. A. Balamurugan, “A novel fea- ture selection framework for automatic web page classification,” International Journal of Automation and Computing, vol. 9, no. 4, pp. 442–448, 2012. [23] Y. Chen, D. Miao, and R. Wang, “A rough set approach to feature selection based on ant colony optimization,” Pattern Recognition Letters, vol. 31, no. 3, pp. 226–233, 2010. [40] WEKA, http://www.cs.waikato.ac.nz/∼ml/weka. The Scientific World Journal 16 [41] S. Kim and B. Zhang, “Genetic mining of HTML structures for effective web-document retrieval,” Applied Intelligence, vol. 18, no. 3, pp. 243–256, 2003. [42] A. Ribeiro, V. Fresno, M. C. Garcia-Alegre, and D. Guinea, “Web page classification: a soft computing approach,” Lecture Notes in Artificial Intelligence, vol. 2663, pp. 103–112, 2003. [43] A. Trotman, “Choosing document structure weights,” Informa- tion Processing & Management, vol. 41, no. 2, pp. 243–264, 2005. [44] M. F. Porter, “An algorithm for suffix stripping,” Program, vol. 14, no. 3, pp. 130–137, 1980. [45] M. Dorigo and T. References St¨utzle, Ant Colony Optimization, The MIT Press, 2004. [46] T. B¨ack, Evolutionary Algorithms in Theory and Practice, Oxford University Press, New York, NY, USA, 1996. [47] J. R. Quinlan, C4.5: Programs for Machine Learning, Morgan Kaufmann, San Mateo, Calif, USA, 1st edition, 1993. [48] C. J. van Rijsbergen, Information Retrieval, Butterworth-Heine- mann, London, UK, 2nd edition, 1979. [49] M. Craven, D. DiPasquo, D. Freitag et al., “Learning to extract symbolic knowledge from the World Wide Web,” in Proceedings of the 15th National Conference on Artificial Intelligence (AAAI ’98), pp. 509–516, AAAI Press, July 1998. [50] CMU, http://www.cs.cmu.edu/. [51] WebKB, http://www.cs.cmu.edu/∼webkb/. [52] DBLP, http://www.informatik.uni-trier.de/∼ley/db/. [53] Google, http://www.google.com. [54] E. Sarac¸ and S. A. ¨Ozel, “URL tabanlı web sayfası sınıflandırma,” in Akıllı Sistemlerde Yenilikler ve. Uygulamaları Sempozyumu (ASYU ’10), pp. 13–18, 2010. [55] M.-Y. Kan and H. O. N. Thi, “Fast webpage classification using URL features,” in Proceedings of the 14th ACM International Conference on Information and Knowledge Management (CIKM ’05), pp. 325–326, New York, NY, USA, November 2005.i [56] S. A. ¨Ozel, “A Web page classification system based on a genetic algorithm using tagged-terms as features,” Expert Systems with Applications, vol. 38, no. 4, pp. 3407–3415, 2011. [57] E. P. Jiang, “Learning to integrate unlabeled data in text classifi- cation,” in Proceedings of the 3rd IEEE International Conference on Computer Science and Information Technology (ICCSIT '10), pp. 82–86, Chengdu, China, July 2010. [58] T. Joachims, “Transductive inference for text classification using support vector machines,” in Proceedings of the 16th International Conference on Machine Learning, pp. 200–209, 1999.
https://openalex.org/W3192178871	https://www.researchsquare.com/article/rs-673213/latest.pdf	English	null	Transversely Isotropic Creep Characteristics and Damage Mechanism of Layered Phyllite Under Uniaxial Compression Creep Test	Research Square (Research Square)	2,021	cc-by	12,537	Transversely Isotropic Creep Characteristics and Damage Mechanism of Layered Phyllite Under Uniaxial Compression Creep Test Jiabing Zhang Guangxi University Xiaohu Zhang Guangxi University Zhen Huang (  hzcslg@163.com ) Guangxi University Helin Fu Central South University Jiabing Zhang 1, Xiaohu Zhang 2, Zhen Huang 3, and Helin Fu 4 Jiabing Zhang 1, Xiaohu Zhang 2, Zhen Huang 3, and Helin Fu 4 1 Assistant professor, College of Civil Engineering and Architecture, Guangxi University, Nanning 53000 China (E-mail: zhang.j.b@gxu.edu.cn) 2 Postgraduate student, College of Civil Engineering and Architecture, Guangxi University, Nanning 530004, China (E-mail: zhang_xh2020@163.com) 3 Assistant professor, College of Civil Engineering and Architecture, Guangxi University, Nanning 530004, China (E-mail: hzcslg@163.com) (Corresponding author) 3 Assistant professor, College of Civil Engineering and Architecture, Guangxi University, Nanning 530004, China (E-mail: hzcslg@163.com) (Corresponding author) 4 Professor, School of Civil Engineering, Central South University, Changsha 410075, China (E-mail: fu.h.l@csu.edu.cn) 4 Professor, School of Civil Engineering, Central South University, Changsha 410075, China (E-mail: fu.h.l@csu.edu.cn) Abstract The layered surrounding rocks of deep tunnels undergo large creep deformation due to the presence of planes of weakness and the presence of prolonged high in-situ stress, thereby the deformation severely endangers the safety of tunnels. This study conducts uniaxial compression creep tests to experimentally investigate the transversely isotropic creep characteristics and the damage mechanism of layered phyllite samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90°. The results indicate that the creep deformation of the specimens takes place in four stages: the instantaneous elastic deformation stage, the deceleration creep stage, the steady-state creep stage, and the accelerated creep stage. The cumulative creep deformation and the creep time during the steady-state creep stage of the specimens initially decrease and then increase as the bedding angle changes from 0° to 90°, thereby, corresponding to the initial increase and subsequent decrease in creep rate during the deceleration creep stage. Based on the existing viscoelastic-plastic damage creep model, the creep parameters E1, E2, η2, and η3 are observed to initially decrease and then increase with the increase in bedding angle, hence demonstrating that the creep characteristics and damage mechanism of the layered rock mass are controlled by the effect of the natural weakness planes and show significant transversely isotropic characteristics. Keywords Layered phyllite; creep characteristics; damage mechanism; bedding angle; uniaxial compressive creep test Research Article Posted Date: August 4th, 2021 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License License:   This work is licensed under a Creative Commons Attribution 4.0 International License. R d F ll Li Transversely isotropic creep characteristics and damage mechanism o layered phyllite under uniaxial compression creep test Transversely isotropic creep characteristics and damage mechanism of layered phyllite under uniaxial compression creep test Jiabing Zhang 1, Xiaohu Zhang 2, Zhen Huang 3, and Helin Fu 4 1 Assistant professor, College of Civil Engineering and Architecture, Guangxi University, Nanning 530004, China (E-mail: zhang.j.b@gxu.edu.cn) 2 Postgraduate student, College of Civil Engineering and Architecture, Guangxi University, Nanning 530004, China (E-mail: zhang_xh2020@163.com) 3 Assistant professor, College of Civil Engineering and Architecture, Guangxi University, Nanning 530004, China (E-mail: hzcslg@163.com) (Corresponding author) 4 Professor, School of Civil Engineering, Central South University, Changsha 410075, China (E-mail: fu.h.l@csu.edu.cn) 1. Introduction With the continued advancement of "the Belt and Road" project and the comprehensive launch of the With the continued advancement of "the Belt and Road" project and the comprehensive launch of th With the continued advancement of "the Belt and Road" project and the comprehensive launch of the Sichuan-Tibet railway project in China, the execution of deep rock mass engineering projects has become increasingly common. The long-term safety of such engineering projects is primarily controlled by the rock mass creep characteristics, and the influence of rock mass creep characteristics gets more pronounced in layered rock masses (Chen et al. 2019). The previous studies have shown that the creep characteristics of layered surrounding rocks of deep tunnels deteriorate significantly due to the presence of natural weakness planes and of the long-term high in-situ stress, which may lead to large deformations and subsequent failure of structures (Fu et al. 2018; Chen et al. 2019; Xu et al. 2020) (as shown in Fig. 1). Therefore, it is essential to examine the transversely isotropic creep characteristics and the damage mechanism of layered rock masses to enhance the long-term stability and safety of deep rock mass engineering structures. a) b) c) Fig. 1 Large deformation failure characteristics of layered surrounding rock in a tunnel. a) bedding angle of 0°–30°; b) bedding angle of 30°–75°; c) bedding angle of 75°–90°. 拱顶弯曲折断破坏仰拱弯曲折断破坏边墙顺层滑移破坏边墙弯曲折断破坏拱顶顺层滑移破坏边墙弯曲折断破坏拱顶顺层滑移破坏 Fig. 1 Large deformation failure characteristics of layered surrounding rock in a tunnel. a) bedding angle of 0°–30°; b) bedding angle of 30°–75°; c) bedding angle of 75°–90°. Fig. 1 Large deformation failure characteristics of layered surrounding rock in a tunnel. a) bedding angle of 0°–30°; b) bedding angle of 30°–75°; c) bedding angle of 75°–90°. Considerable efforts have been made to investigate the strength and the deformation characteristics of layered rock masses using the uniaxial and triaxial compression tests and the shear tests (Tien et al. 2006; Lee et al. 2015; Hu et al. 2017; Zhang et al. 2019; Shen et al. 2021; Yang et al. 2021). A detailed understanding of the effect of the orientations of the bedding plane on the mechanical properties of layered rock masses has been established. For studying the creep characteristics of rock masses, Xie et al. 1. Introduction (2011) conducted a series of hydrostatic and triaxial compression tests on chemically degraded samples and reported the influence of ionic concentration and pH of aqueous solutions on the creep characteristics of limestone. Nadimi et al. (2011) studied the rheological characteristics and long-term strength of surrounding rock containing deeply buried tunnels by triaxial compression creep tests. The results show that the internal structure of the surrounding rock is constantly adjusted and reorganized due to the presence of prolonged high in-situ stress, continuously developing the creep deformation of the surrounding rock and decreasing the bearing capacity. Chen et al. (2017) investigated the effect of temperature and stress on the creep characteristics of Beishan granite. They found that the peak strength and peak strain of rock initially increases and then decreases with increasing temperature. Moreover, Hadiseh et al. (2018); Reza Taheri et al. (2020); Wang et al. (2021), and many others have studied the creep characteristics of rock masses through multi-step and mono-step creep tests. In addition, the creep damage mechanism of rock masses has been extensively studied by various researchers, and some creep constitutive models have been developed (e.g., Zhao et al. 2009; Zhou et al. 2011; Yang et al. 2015; Li et al. 2017; Wu et al. 2020). These studies primarily focused on the creep characteristics and damage mechanism of isotropic rocks; however, there is still a lack of systematic understanding of transversely isotropic creep characteristics and the damage mechanism of layered rock masses considering the weakening effect of macroscopic bedding planes, which needs to be further supplemented and improved. In this paper, a series of uniaxial compression creep tests were conducted on layered phyllite samples having different bedding angles to obtain the creep curve and the creep rate curve. The effect of the angle of bedding plane at each creep stage and the damage mechanism of the evolution of creep failure morphology of the layered phyllites had been discussed in detail. Finally, the variation of rock creep parameters with the change in bedding angles was determined based on the existing viscoelastic-plastic damage creep model, which provides a reference for further understanding the transverse isotropic creep damage mechanism of the layered rock mass. 2.1 Sample preparation In this experiment, few representative phyllite rock blocks were collected from a typical supporting project site. The layered phyllite samples had a height of 100 mm and a diameter of 50 mm, which is in strict accordance with the International Society for Rock Mechanics ISMR (ISMR, 2007). It was ensured that the error in the sample preparation process does not exceed 0.3 mm, and the non-parallelism at both ends remains lower than 0.05 mm. Five sets of a total fifteen drill cores were drilled out from the phyllite blocks; each set containing three samples having an angle of 0°, 22.5°, 45°, 67.5°, and 90° between the loading direction and the bedding surface (as shown in Fig. 2). After processing, the samples with higher integrity were selected for acoustic measurement, and the samples with greater homogeneity were further screened for testing by measuring the wave speed of the samples. Among the three samples of each bedding angle, two were selected for the conventional uniaxial compression test and one for the uniaxial compression creep test. p p Fig. 2 Schematic diagram of sample preparation. 50 mm 100 mm 50 mm 100 mm 90° 67.5° 45° 22.5° 0° 50 mm 100 mm 50 mm 100 mm 50 mm 100 mm 67.5° 50 mm 100 mm 90° 50 mm 100 mm 45° 50 mm 100 mm 50 mm 100 mm 50 mm 100 mm 90° 67.5° 22.5° 50 mm 100 mm 50 mm 100 mm 5 100 mm 100 mm 100 mm 100 mm 50 mm 50 mm 100 mm 22.5° 50 mm 100 mm 50 mm 100 mm 22.5° 0° 50 mm 100 mm 0° 100 mm 50 mm Fig. 2 Schematic diagram of sample preparation. 2.2 Experimental procedures The conventional uniaxial compression test and uniaxial compression creep test were performed using the RMT-301 rock and concrete mechanics testing machine (as shown in Fig. 3). The maximum test force of the equipment is 1500 kN; the accuracy is  0.5%; the deformation rate range is 0.0001 to 1 mm/s; and the loading rate range is 0.01–90 kN/s. In the uniaxial compression test, displacement is used to control the loading, the loading rate is 0.005 mm/s, and the axial deformation of the sample is measured using a linear variable differential transducer (LVDT). The uniaxial peak strength of a rock sample at a given angle to bedding σc is the average value of the uniaxial peak strength of the rock sample under the same bedding angle. The uniaxial compression creep test was conducted using stress-controlled loading at a loading rate of 0.01 MPa/s. After each load level was added to the predetermined value and was kept stable for 48 hours, ensuring that the displacement rate of the rock sample after 48 hours was less than 0.01 mm/h, the next loading level was applied in a step-wise fashion until the sample failed (as shown in Fig. 4). A high-speed camera was used to record the entire deformational process and the failure of rock samples. The failure strength, failure mode, and fragmentation distribution characteristics for each bedding direction were analyzed. y Fig. 3 Equipment used in the uniaxial compression creep tests. Fig. 4 Loading time control schematic diagram for each uniaxial compression creep test. 48 96 144 192 240 288 0 Axial strain /10-3 Time /h Time /h Axial strain /10-3 Rock sample Linear variable differential transducer Control system Load frame, upper beam low-stress levels Force sensor high-stress level Load frame platen Fig. 4 Loading time control schematic diagram for each uniaxial compression creep test. 48 96 144 192 240 288 0 Axial strain /10-3 Time /h Time /h Axial strain /10-3 low-stress levels high-stress level Fig. 3 Equipment used in the uniaxial compression creep tests. Rock sample Linear variable differential transducer Control system Load frame, upper beam Force sensor Load frame platen Load frame, upper beam high-stress level Force sensor Axial strain /10-3 Linear variable differential transduce low-stress levels Fig. 4 Loading time control schematic diagram for each uniaxial compression creep test. Fig. 3 Equipment used in the uniaxial compression creep tests. 3.1 Uniaxial compression test results and analysis Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 15 30 45 60 75 90 105 120 135 150 Axial stress, σ (MPa) Axial strain, ε (10 -3) Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 15 30 45 60 75 90 105 120 135 150 Axial stress, σ (MPa) Axial strain, ε (10 -3) Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 15 30 45 60 75 90 105 120 135 150 Axial stress, σ (MPa) Axial strain, ε (10 -3) A i l (MP ) b) The second group 5 Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 15 30 45 60 75 90 105 120 135 150 Axial stress, σ (MPa) Axial strain, ε (10 -3) Axial stress, σ (MPa) Axial stress, σ (MPa) ) The first group Axial strain, ε (10 -3) b) The second group Axial strain, ε (10 -3) ) The first group Axial strain, ε (10 -3) b) The second group , ( ) a) The first group a) The first group b) The second group Fig. 5 Stress-strain relationship curves of layered phyllite under uniaxial compression. Fig. 6 Relationships between the bedding angle and the uniaxial compressive strength and peak strain of layered phyllite. 3.2 Uniaxial compression creep test results and analysis 0.0 22.5 45.0 67.5 90.0 0 20 40 60 80 100 120 140 160 UCS, σc: Peak strain, εc: Group one Group one Group two Group two Average Average Bedding orientation, β (°) Uniaxial compressive strength, σc (MPa) 2.0 2.5 3.0 3.5 4.0 4.5 Peak strain, εc (10 -3) Fig. 5 Stress-strain relationship curves of layered phyllite under uniaxial compression. 3.1 Uniaxial compression test results and analysis Fig. 5 shows the stress-strain relationship curves of two sets of rock samples having different bedding angles. The stress-strain curves of rock samples having different bedding angles are similar and can be divided into four stages: the compaction stage, the linear elastic stage, the plastic yielding stage, and the failure stage. The relationships between the bedding angle and the uniaxial compressive strength and peak strain of layered phyllite are illustrated in Fig. 6. Notably, the bedding angle has a significant influence on the uniaxial compressive strength and peak strain of layered phyllite, and the uniaxial compressive strength and peak strain of layered phyllite initially decrease and then increase as the bedding angle changes from 0° to 90°, forming a U-shaped curve. The change trend is similar to the test results of phyllite from Xu et al. (2018). When the bedding angle is 90°, the uniaxial compressive strength of the rock sample is the largest, and the average compressive strength is 141.80 MPa. When the bedding angle is 22.5°, the uniaxial compressive strength of the rock sample is the smallest, and the average compressive strength is 50.28 MPa. It is concluded that the rock samples used in the test have little dispersion, which is helpful for further exploring the creep characteristics of layered phyllite. a) The first group b) The second group Fig. 5 Stress-strain relationship curves of layered phyllite under uniaxial compression. Fig. 6 Relationships between the bedding angle and the uniaxial compressive strength and peak strain of layered phyllite. 3.1 Uniaxial compression test results and analysis 3.2 Uniaxial compression creep test results and analysis Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 15 30 45 60 75 90 105 120 135 150 Axial stress, σ (MPa) Axial strain, ε (10 -3) Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 15 30 45 60 75 90 105 120 135 150 Axial stress, σ (MPa) Axial strain, ε (10 -3) 0.0 22.5 45.0 67.5 90.0 0 20 40 60 80 100 120 140 160 UCS, σc: Peak strain, εc: Group one Group one Group two Group two Average Average Bedding orientation, β (°) Uniaxial compressive strength, σc (MPa) 2.0 2.5 3.0 3.5 4.0 4.5 Peak strain, εc (10 -3) a) The first group b) The second group Fig. 5 Stress-strain relationship curves of layered phyllite under uniaxial compression. 3.1 Uniaxial compression test results and analysis The following conclusions can be drawn: (1) When the axial stress on the rock sample is at the first four stress levels, the rock sample first undergoes instantaneous elastic deformation after each level of load is applied and then enters the deceleration creep stage. The creep rate gradually decreases with increasing time, at which point the slope of the creep curve becomes approximately 0, and the creep tends to be in the steady-state creep stage. Because the loading stress at this stage is less than the long-term strength of the rock, the particles inside the rock are not destroyed, but the internal pores and cracks are gradually compressed and they ultimately close, which leads to an increase in the directional flow resistance inside the rock as well as a gradual increase in rock creep. (2) When the axial stress on the rock sample is at the fifth stress level, the rock will enter the steady-state creep stage after a short deceleration creep stage. The creep strain value of the rock in the steady-state creep stage has a linear relationship with time. Since the loading stress at this stage is close to the long-term rock strength; as the loading time increases, the rock sample develops shear cracks along the bedding direction. Cracks within the rock sample continue to develop; however, the failure does not occur. (3) When the axial stress on the rock sample is at the last stress level, the stable creep rate accelerates and quickly enters the accelerated creep stage. In this stage, the rock creep rate increases rapidly with time, and the development of non-linear accelerating creep occurs, which finally reaches the ultimate compressive strain and leads to the failure of rock sample. (3) When the axial stress on the rock sample is at the last stress level, the stable creep rate accelerates and quickly enters the accelerated creep stage. In this stage, the rock creep rate increases rapidly with time, and the development of non-linear accelerating creep occurs, which finally reaches the ultimate compressive strain and leads to the failure of rock sample. (4) When the axial stress on the rock sample is less than the long-term strength of the specimens, the creep strain of the rock mass is composed of three parts: the instantaneous elastic deformation stage, the deceleration creep stage, and the steady-state creep stage. 3.1 Uniaxial compression test results and analysis 0.0 22.5 45.0 67.5 90.0 0 20 40 60 80 100 120 140 160 UCS, σc: Peak strain, εc: Group one Group one Group two Group two Average Average Bedding orientation, β (°) Uniaxial compressive strength, σc (MPa) 2.0 2.5 3.0 3.5 4.0 4.5 Peak strain, εc (10 -3) Fig. 6 Relationships between the bedding angle and the uniaxial compressive strength and peak strain of layered phyllite. 3.2 Uniaxial compression creep test results and analysis The uniaxial compression creep test was conducted using the graded incremental loading method, and the axial loading was divided into 6 loading levels from low-stress level to high-stress level according to the long-term rock strength σs (the long-term rock strength takes 75% of the uniaxial peak strength σc of the rock sample). It is assumed that the first 5 loading levels are low-stress levels σm (σm<σs), and the last loading level is a high-stress level σn (σn>σs). For example, the long-term strength of a rock sample with a bedding angle of 90° is 106.35 MPa. The loading level of the rock sample was divided into 6 levels (86, 91, 96, 101, 106, and 111 MPa), the rock is at a low-stress level when the first 5 loading levels (86, 91, 96, 101, and 106 MPa) are applied, and the rock is at a high-stress level when the last loading level (111 MPa) is applied. The uniaxial compression creep test schemes of rock samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90° are listed in Table 1. Table 1 Load scheme of uniaxial compression creep test. Table 1 Load scheme of uniaxial compression creep test. Bedding angle, β (°) Uniaxial peak strength, σc (MPa) Long-term rock strength, σs (MPa) Stress levels of various loading levels (MPa) 1 2 3 4 5 6 0 61.49 46.12 25 30 35 40 45 50 22.5 50.28 37.71 20 24 28 32 36 40 45 54.48 40.86 24 28 32 36 40 44 67.5 67.76 50.82 29 34 39 44 49 54 90 141.80 106.35 86 91 96 101 106 111 The strain-time relationship curves of layered phyllite samples under various loading levels are presented in Fig. 7. The following conclusions can be drawn: The strain-time relationship curves of layered phyllite samples under various loading levels are presented in Fig. 7. 3.1 Uniaxial compression test results and analysis When the axial stress on the rock sample is greater than the long-term strength of the specimens, the creep strain of rock mass is composed of the instantaneous elastic deformation stage, the deceleration creep stage, the steady-state creep stage, and the accelerated creep stage. a) Bedding angle of 0 b) Bedding angle of 22.5 c) Bedding angle of 45 d) Bedding angle of 67.5 e) Bedding angle of 90 f) The last-stage loading Fig. 7 Creep curves of layered phyllite samples under various loading conditions. 3.1 Uniaxial compression test results and analysis a) the first five-stage loading creep curve of the rock sample with a bedding angle of 0°; b) the first five-stage loading creep curve of the rock sample with a bedding angle of 22.5°; c) the first five-stage loading creep curve of the rock sample with a bedding angle of 45°; d) the first five-stage loading creep curve of the rock sample with a bedding angle of 67.5°; e) the first five-stage loading creep curve of the rock sample with a bedding angle of 90°; f) the last-stage loading creep curve of rock samples having bedding angles of 0° 22 5° 45° 67 5° d 90° 0 5 10 15 20 25 30 35 40 45 50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Axial compression 25 MPa Axial compression 30 MPa Axial compression 35 MPa Axial compression 40 MPa Axial compression 45 MPa Axial strain / 10 -3 Time / h 0 5 10 15 20 25 30 35 40 45 50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Axial strain / 10 -3 Time / h Axial compression 20 MPa Axial compression 24 MPa Axial compression 28 MPa Axial compression 32 MPa Axial compression 36 MPa 0 5 10 15 20 25 30 35 40 45 50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Axial strain / 10 -3 Time / h Axial compression 24 MPa Axial compression 28 MPa Axial compression 32 MPa Axial compression 36 MPa Axial compression 40 MPa 0 5 10 15 20 25 30 35 40 45 50 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Axial strain / 10 -3 Time / h Axial compression 29 MPa Axial compression 34 MPa Axial compression 39 MPa Axial compression 44 MPa Axial compression 49 MPa Axial compression 86 MPa Axial compression 91 MPa Axial compression 96 MPa Axial compression 101 MPa Axial compression 106 MPa 0 5 10 15 20 25 30 35 40 45 50 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Axial strain / 10 -3 Time / h 0 1 2 3 4 5 6 7 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 Axial strain / 10 -3 Time / h Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° 0 0 5 10 15 20 25 30 35 40 45 50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Axial strain / 10 -3 Time / h Axial compression 20 MPa Axial compression 24 MPa Axial compression 28 MPa Axial compression 32 MPa Axial compression 36 MPa 0 5 10 15 20 25 30 35 40 45 50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Axial compression 25 MPa Axial compression 30 MPa Axial compression 35 MPa Axial compression 40 MPa Axial compression 45 MPa Axial strain / 10 -3 Time / h Axial strain / 10 -3 Time / h Time / h a) Bedding angle of 0 Time / h 0 5 10 15 20 25 30 35 40 45 50 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Axial strain / 10 -3 Time / h Axial compression 24 MPa Axial compression 28 MPa Axial compression 32 MPa Axial compression 36 MPa Axial compression 40 MPa b) Bedding angle of 22.5 0 5 10 15 20 25 30 35 40 45 50 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Axial strain / 10 -3 Time / h Axial compression 29 MPa Axial compression 34 MPa Axial compression 39 MPa Axial compression 44 MPa Axial compression 49 MPa Axial strain / 10 -3 d) Bedding angle of 67.5 f) The last-stage loading Time / h 0 1 2 3 4 5 6 7 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 Axial strain / 10 -3 Time / h Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° c) Bedding angle of 45 e) Bedding angle of 90 Time / h Axial compression 86 MPa Axial compression 91 MPa Axial compression 96 MPa Axial compression 101 MPa Axial compression 106 MPa 0 5 10 15 20 25 30 35 40 45 50 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 Axial strain / 10 -3 Time / h 3 Axial strain / 10 -3 Axial strain / 10 -3 f) The last-stage loading Fig. 3.1 Uniaxial compression test results and analysis 7 Creep curves of layered phyllite samples under various loading conditions. a) the first five-stage loading creep curve f th k l ith b ddi l f 0° b) th fi t fi t l di f th k l ith b ddi Fig. 7 Creep curves of layered phyllite samples under various loading conditions. a) the first five-stage loading creep curve of the rock sample with a bedding angle of 0°; b) the first five-stage loading creep curve of the rock sample with a bedding angle of 22.5°; c) the first five-stage loading creep curve of the rock sample with a bedding angle of 45°; d) the first five-stage loading creep curve of the rock sample with a bedding angle of 67.5°; e) the first five-stage loading creep curve of the rock sample with a bedding angle of 90°; f) the last-stage loading creep curve of rock samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90°. strain-time relationship curves of layered phyllite samples under the last-stage loading are shown The strain-time relationship curves of layered phyllite samples under the last-stage loading are show in Fig. 7f . The creep time of the rock sample in the steady-state creep stage first decreases and then increases as the bedding angle changes from 0° to 90°. It is primarily because of the reason that with the increase in bedding angle, the shear stress on the bedding surface first increases and then decreases, and the growth rate of shear cracks first accelerates and then decelerates. Figs. 8 and 9 show the cumulative creep deformation curves and the creep rate curves of the deceleration creep stage of the specimens under various loading levels, respectively. For the same bedding angle, the cumulative creep deformation of the rock sample and the axial displacement rate of the deceleration creep stage both show concave growth roughly as the loading level increases. Under the same stress loading level, as the bedding angle changes from 0° to 90°, the cumulative creep deformation of the rock sample and creep rate of the deceleration creep stage are different. When the bedding angle is 22.5°, the cumulative creep deformation of the rock sample is the smallest, and the creep rate corresponding to the deceleration creep stage is the largest. The effect of the bedding angle on the creep characteristics of layered phyllite is clearly visible under the high-stress loading level. Fig. 3.1 Uniaxial compression test results and analysis 8 Cumulative creep deformation curves of the specimens under various loading conditions. Fig. 9 Creep rate curves of the specimens during the deceleration creep stage under various loading conditions. 48 96 144 192 240 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° Accumulated strain /10 -3 Time /h 48 96 144 192 240 288 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 Axial displacement rate /mm·h -1 Time /h Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° Accumulated strain /10 -3 Fig. 8 Cumulative creep deformation curves of the Fig. 9 Creep rate curves of the specimens during the g p specimens under various loading conditions. deceleration creep stage under various loading conditions. The typical failure characteristics of layered phyllite samples having different bedding angles under uniaxial compression creep test are presented in Fig. 10. When the bedding angle is 0°, the rock sample undergoes shear failure along a certain angle relative to the vertical axis, and the long-term strength of the rock is relatively high. When the bedding angle is 22.5° – 67.5°, because the bedding surface has a weakening effect on the rock sample, shear cracks along the bedding direction form near the bedding surface, causing the rock sample to undergo shear slip failure at the bedding surface. When the bedding angle is 90°, a typical vertical splitting failure occurs parallel to the bedding plane. a) b) c) d) e) Fig. 10 Typical failure characteristics of layered phyllite samples under uniaxial compression creep for a: a) bedding angle of 0°; b) bedding angle of 22.5°; c) bedding of 45°; d) bedding angle of 67.5°; e) bedding angle of 90°. d) e) b) a) d) c) Fig. 10 Typical failure characteristics of layered phyllite samples under uniaxial compression creep for a: a) bedding angle of 0°; b) bedding angle of 22.5°; c) bedding of 45°; d) bedding angle of 67.5°; e) bedding angle of 90°. 4.1 Damage creep model of the layered rock masses To further explore the transversely isotropic creep damage mechanism of layered rock masses, a damage creep model that considers the weakening effect of bedding planes has been constructed based on the existing viscoelastic-plastic damage creep model, and the influence of bedding angle on the transversely isotropic creep damage mechanism of layered rock masses has been quantitatively discussed. The results of the laboratory uniaxial compression creep test of layered phyllite show the development of instantaneous strain in the rock sample in the initial loading stage. Since the loading time at this stage is shorter than the later creep time, the elastic strain is considered to be completed instantaneous at this stage, and its constitutive relations are described by elastomers. Then it enters the deceleration creep stage. When the level of the axial stress on the rock sample is small, the strain rate gradually decreases and tends towards 0. The strain curve at this stage shows significant nonlinear characteristics, and its constitutive relationship can be described as a nonlinear Kelvin body. When the stress level of the creep test is close to the long-term rock strength, the rock strain has a linear relationship with time after the deceleration creep stage. The constitutive relationship at this stage can be described as a viscous body. When the stress level exceeds the long-term strength of the rock itself, the rock undergoes the deceleration and steady-state creep stage, and quickly enters the accelerated creep stage. At this time, the internal fissures within the rock sample continue to develop and gradually form macroscopic cracks. In this process, the creep damage increases sharply with time, resulting in failure of the rock sample (as shown in Fig. 11). Fig. 11 Creep curves of layered rock masses. Fig. 12 Damage creep model of layered rock masses. εe ε εve εv η4 σ 0 t η3 8 (t ) 0, σ0 σs (t ) 0, σ0 σs εvp η2 E1 E2 σs σ0 8 E1 η2 η3 σs εve εv εvp εe E2 η4, D Fig. 12 Damage creep model of layered rock masses. E1 η2 η3 σs εve εv εvp εe E2 η4, D Fig. 11 Creep curves of layered rock masses. Fig. 12 Damage creep model of layered rock masses. 4.1 Damage creep model of the layered rock masses Based on the creep curves of layered rock masses under different stress levels, this study connects the following classical rock mechanics elements in a series: elastic body, nonlinear Kelvin body, viscous body, and damage viscoplastic body. By considering the non-stationarity component of the element parameters under the action of bedding surface weakening and aging damage, a damage creep model of the whole rock creep process considering the weakening effect of the macro-bedding surface and aging damage has been constructed (as shown in Fig. 12). 4.2 One-dimensional equation of the damage creep model of the layered rock masses According to the model series and parallel stress-strain relationship: According to the model series and parallel stress-strain relationship:            vp v ve e vp v ve e           (1) (1) (1) where ε is the total strain; εe, εve, εv, and εvp are the strain of the elastic body, nonlinear Kelvin body, viscous body, and damage viscoplastic body, respectively; σe, σve, σv, and σvp are the stress of the elastic body, nonlinear Kelvin body, viscous body, and damage viscoplastic body, respectively. where ε is the total strain; εe, εve, εv, and εvp are the strain of the elastic body, nonlinear Kelvin body, viscous body, and damage viscoplastic body, respectively; σe, σve, σv, and σvp are the stress of the elastic body, nonlinear Kelvin body, viscous body, and damage viscoplastic body, respectively. The one-dimensional equations of the elastic body, nonlinear Kelvin body, and viscous body are given as follows (Zhao et al. 2017; Liu et al. 2017): 1 1 E E e e      (2) )] exp( 1[ 2 2 2 t η E E ve     (3) t v 3    (4) (2) (3) (4) where E1 and E2 are the elastic modulus of the elastic body and nonlinear Kelvin body, respectively; η2 and η3 are the viscosity coefficient of the nonlinear Kelvin body and viscous body, respectively. where E1 and E2 are the elastic modulus of the elastic body and nonlinear Kelvin body, respectively; η2 and η3 are the viscosity coefficient of the nonlinear Kelvin body and viscous body, respectively. The differential constitutive equation of the damage viscoplastic body by introducing the damage variable D (Kachanvo 1992; Cao et al. 2016) is as follows: ) (1 4 4 D ~ vp          (5) λ t t D     1 1 1 ） 1 （ 1 (6) (5) (6) where ~  is the effective stress; η4 is the viscosity coefficient of the damage viscoplastic body; D is the damage variable; C and λ are material parameters; and t1 is the time at which the rock undergoes creep failure. Substituting Eq. (6) into Eq. 4.2 One-dimensional equation of the damage creep model of the layered rock masses (5), a one-dimensional creep constitutive equation of the damage viscoplastic body can be obtained as follows:          1 1 1 4 1 （） t t vp (7) (7) By integrating Eq. (7) with the initial conditions of t = 0 and ε = 0 and replacing σ with σ－σs, an one-dimensional equation of the damage viscoplastic body can be obtained as follows: By integrating Eq. (7) with the initial conditions of t = 0 and ε = 0 and replacing σ with σ－σs, an one-dimensional equation of the damage viscoplastic body can be obtained as follows: ] ) 1( 1[) 1( 1 4        D t s vp      (8) (8) where σs is the long-term strength, which can be obtained from the uniaxial compression creep test. According to Eqs. (1)–(8), the one-dimensional equation of the damage creep model of layered rock masses can be expressed as following: 2 0 1 2 2 2 0 1 2 2 3 2 1 2 2 1 1 0 4 1 [1 exp( )] ( ( ) 0 ) [1 exp( )] ( ( ) 0 ) [1 exp( )] (1 )[1 1 ] ( ) s s 3 λ λ s s E σ σ t t σ σ E E η E σ σ σ t t t σ σ E E η η ε E σ σ t t E E η η σ σ t λ t σ σ η λ t                                                ，， (9) (9) 4.3 Identification and sensitivity analysis of creep parameters 2.17 0.024      ) 0.13(1 1.34 0.056 t e      ) 0.15(1 1.54 0.042 t e      ) 0.16(1 1.73 0.039 t e      ) 0.18(1 1.93 0.032 t e      Axial compression 24MPa Axial compression 28MPa Axial compression 32MPa Axial compression 36MPa Axial compression 40MPa Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 2.4 2.8 3.2 3.6 4.0 4.4 ) 0.13(1 2.79 0.059 t e      ) 0.14(1 2.95 0.050 t e      ) 0.15(1 3.20 0.044 t e      t e t 0.020 ) 0.15(1 3.46 0.039       ) (1 3 0.1 2.56 0.071 t e      Axial compression 86MPa Axial compression 91MPa Axial compression 96MPa Axial compression 101MPa Axial compression 106MPa Axial strain / 10 -3 Time / h b) Bedding angle of 22.5 d) Bedding angle of 67.5 f) The last-stage loading 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.2 1.5 1.8 2.1 2.4 2.7 ) 0.14(1 1.09 ε 0.066 t e     ) 0.16(1 1.30 ε 0.049 t e     ) 0.17(1 1.52 ε 0.044 t e     ) 0.19(1 1.71 ε 0.036 t e     t e t 0.020 ) 0.22(1 1.96 ε 0.027      Axial compression 20MPa Axial compression 24MPa Axial compression 28MPa Axial compression 32MPa Axial compression 36MPa Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.5 1.8 2.1 2.4 2.7 3.0 3.3 3.6 ) 0.18(1 1.86 0.057 t e      ) 0.19(1 2.13 0.047 t e      ) 0.22(1 2.36 0.039 t e      t e t 4 0.01 ) 0.24(1 2.61 0.028       ) 0.16(1 1.68 0.064 t e      Axial compression 29MPa Axial compression 34MPa Axial compression 39MPa Axial compression 44MPa Axial compression 49MPa Axial strain / 10 -3 Time / h a a 0 1 2 3 4 5 6 7 8 3 4 5 6 7 8 9 0.29 0.023 ) 1.95 1.06(1 0.20 ) 0.31(1 8 .2 4 t t e t         4 0.3 0.024 ) 2.89 1.19(1 0.145 ) 0.28(1 56 .4 t t e t         9 0.3 0.030 ) 4.17 1.60(1 0.138 ) 0.35(1 55 .5 t t e t         2 0.4 0.031 ) 5.23 1.56(1 3 0.1 ) 0.29(1 87 .5 t t e t         5 0.4 0.032 ) 7.12 2.07(1 2 0.1 ) 0.37(1 59 .6 t t e t         Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° Axial strain / 10 -3 Time / h a) Bedding angle of 0 b) Bedding angle of 22.5 c) Bedding angle of 45 d) Bedding angle of 67.5 e) Bedding angle of 90 f) The last-stage loading Fig 13 Comparison of creep test data and the theoretical curve of layered phyllite samples a) the first five-stage loading Axial compression 25MPa Axial compression 30MPa Axial compression 35MPa Axial compression 40MPa Axial compression 45MPa 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 ) 0.13(1 1.38 ε 0.068 t e     ) 0.14(1 1.64 ε 0.057 t e     ) 0.16(1 1.91 ε 0.048 t e     ) 0.18(1 2.17 ε 0.040 t e     t e t 0.013 ) 0.21(1 2.44 ε 0.032      Axial strain /10 -3 Time /h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.2 1.5 1.8 2.1 2.4 2.7 ) 0.14(1 1.09 ε 0.066 t e     ) 0.16(1 1.30 ε 0.049 t e     ) 0.17(1 1.52 ε 0.044 t e     ) 0.19(1 1.71 ε 0.036 t e     t e t 0.020 ) 0.22(1 1.96 ε 0.027      Axial compression 20MPa Axial compression 24MPa Axial compression 28MPa Axial compression 32MPa Axial compression 36MPa Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 t e ε t 5 0.01 ) (1 20 0. 4.3 Identification and sensitivity analysis of creep parameters a) the first five-sta creep curve of the rock sample with a bedding angle of 0°; b) the first five-stage loading creep curve of the rock Fig. 13 Comparison of creep test data and the theoretical curve of layered phyllite samples. a) the first five-stage loading creep curve of the rock sample with a bedding angle of 0°; b) the first five-stage loading creep curve of the rock sample with a bedding angle of 22 5°; c) the first five stage loading creep curve of the rock sample with a bedding angle of 45°; d) the Fig. 13 Comparison of creep test data and the theoretical curve of layered phyllite samples. a) the first five-stage loading creep curve of the rock sample with a bedding angle of 0°; b) the first five-stage loading creep curve of the rock sample with a bedding angle of 22.5°; c) the first five-stage loading creep curve of the rock sample with a bedding angle of 45°; d) the first five-stage loading creep curve of the rock sample with a bedding angle of 67.5°; e) the first five-stage loading creep curve of the rock sample with a bedding angle of 90°; f) the last-stage loading creep curve of rock samples having bedding angles of 0° 22 5° 45° 67 5° and 90° creep curve of the rock sample with a bedding angle of 0 ; b) the first five stage loading creep curve of the rock sample with a bedding angle of 22.5°; c) the first five-stage loading creep curve of the rock sample with a bedding angle of 45°; d) the first five-stage loading creep curve of the rock sample with a bedding angle of 67.5°; e) the first five-stage loading creep curve of the rock sample with a bedding angle of 90°; f) the last-stage loading creep curve of rock samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90°. dding angle of 22.5°; c) the first five-stage loading creep curve of the rock sample with a bedding angle of 45°; d) the five-stage loading creep curve of the rock sample with a bedding angle of 67.5°; e) the first five-stage loading creep e of the rock sample with a bedding angle of 90°; f) the last-stage loading creep curve of rock samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90°. 4.3 Identification and sensitivity analysis of creep parameters 2.17 0.024      ) 0.13(1 1.34 0.056 t e      ) 0.15(1 1.54 0.042 t e      ) 0.16(1 1.73 0.039 t e      ) 0.18(1 1.93 0.032 t e      Axial strain / 10 -3 Time / h Axial strain / 10 -3 c) Bedding angle of 45 e) Bedding angle of 90 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 t e ε t 5 0.01 ) (1 20 0. 4.3 Identification and sensitivity analysis of creep parameters 2.17 0.024      ) 0.13(1 1.34 0.056 t e      ) 0.15(1 1.54 0.042 t e      ) 0.16(1 1.73 0.039 t e      ) 0.18(1 1.93 0.032 t e      Axial compression 24MPa Axial compression 28MPa Axial compression 32MPa Axial compression 36MPa Axial compression 40MPa Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.5 1.8 2.1 2.4 2.7 3.0 3.3 3.6 ) 0.18(1 1.86 0.057 t e      ) 0.19(1 2.13 0.047 t e      ) 0.22(1 2.36 0.039 t e      t e t 4 0.01 ) 0.24(1 2.61 0.028       ) 0.16(1 1.68 0.064 t e      Axial compression 29MPa Axial compression 34MPa Axial compression 39MPa Axial compression 44MPa Axial compression 49MPa Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 2.4 2.8 3.2 3.6 4.0 4.4 ) 0.13(1 2.79 0.059 t e      ) 0.14(1 2.95 0.050 t e      ) 0.15(1 3.20 0.044 t e      t e t 0.020 ) 0.15(1 3.46 0.039       ) (1 3 0.1 2.56 0.071 t e      Axial compression 86MPa Axial compression 91MPa Axial compression 96MPa Axial compression 101MPa Axial compression 106MPa Axial strain / 10 -3 Time / h 0 1 2 3 4 5 6 7 8 3 4 5 6 7 8 9 0.29 0.023 ) 1.95 1.06(1 0.20 ) 0.31(1 8 .2 4 t t e t         4 0.3 0.024 ) 2.89 1.19(1 0.145 ) 0.28(1 56 .4 t t e t         9 0.3 0.030 ) 4.17 1.60(1 0.138 ) 0.35(1 55 .5 t t e t         2 0.4 0.031 ) 5.23 1.56(1 3 0.1 ) 0.29(1 87 .5 t t e t         5 0.4 0.032 ) 7.12 2.07(1 2 0.1 ) 0.37(1 59 .6 t t e t         Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° Axial strain / 10 -3 Time / h b) ddi l f 22 5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.2 1.5 1.8 2.1 2.4 2.7 ) 0.14(1 1.09 ε 0.066 t e     ) 0.16(1 1.30 ε 0.049 t e     ) 0.17(1 1.52 ε 0.044 t e     ) 0.19(1 1.71 ε 0.036 t e     t e t 0.020 ) 0.22(1 1.96 ε 0.027      Axial compression 20MPa Axial compression 24MPa Axial compression 28MPa Axial compression 32MPa Axial compression 36MPa Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 ) 0.13(1 1.38 ε 0.068 t e     ) 0.14(1 1.64 ε 0.057 t e     ) 0.16(1 1.91 ε 0.048 t e     ) 0.18(1 2.17 ε 0.040 t e     t e t 0.013 ) 0.21(1 2.44 ε 0.032      Axial strain /10 -3 Ti /h Axial strain / 10 -3 b) Bedding angle of 22.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.5 1.8 2.1 2.4 2.7 3.0 3.3 3.6 ) 0.18(1 1.86 0.057 t e      ) 0.19(1 2.13 0.047 t e      ) 0.22(1 2.36 0.039 t e      t e t 4 0.01 ) 0.24(1 2.61 0.028       ) 0.16(1 1.68 0.064 t e      Axial compression 29MPa Axial compression 34MPa Axial compression 39MPa Axial compression 44MPa Axial compression 49MPa Axial strain / 10 -3 Time / h b) Bedding angle of 22.5 3.3 3.6 Axial compression 29MPa Axial compression 34MPa Axial compression 39MPa Axial compression 44MPa Axial compression 49MPa 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 t e ε t 5 0.01 ) (1 20 0. 4.3 Identification and sensitivity analysis of creep parameters 2.17 0.024      ) 0.13(1 1.34 0.056 t e      ) 0.15(1 1.54 0.042 t e      ) 0.16(1 1.73 0.039 t e      ) 0.18(1 1.93 0.032 t e      Axial strain / 10 -3 Time / h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 2.4 2.8 3.2 3.6 4.0 4.4 ) 0.13(1 2.79 0.059 t e      ) 0.14(1 2.95 0.050 t e      ) 0.15(1 3.20 0.044 t e      t e t 0.020 ) 0.15(1 3.46 0.039       ) (1 3 0.1 2.56 0.071 t e      Axial compression 86MPa Axial compression 91MPa Axial compression 96MPa Axial compression 101MPa Axial compression 106MPa Axial strain / 10 -3 Time / h Time / h d) Bedding angle of 67.5 f) Th l t t l di Time / h 0 1 2 3 4 5 6 7 8 3 4 5 6 7 8 9 0.29 0.023 ) 1.95 1.06(1 0.20 ) 0.31(1 8 .2 4 t t e t         4 0.3 0.024 ) 2.89 1.19(1 0.145 ) 0.28(1 56 .4 t t e t         9 0.3 0.030 ) 4.17 1.60(1 0.138 ) 0.35(1 55 .5 t t e t         2 0.4 0.031 ) 5.23 1.56(1 3 0.1 ) 0.29(1 87 .5 t t e t         5 0.4 0.032 ) 7.12 2.07(1 2 0.1 ) 0.37(1 59 .6 t t e t         Bedding angle of 0° Bedding angle of 22.5° Bedding angle of 45° Bedding angle of 67.5° Bedding angle of 90° Axial strain / 10 -3 Time / h Axial strain / 10 -3 f) The last-stage loading Time / h e) Bedding angle of 90 Time / h Time / h f) The last-stage loading f) The last-stage loading Fig. 13 Comparison of creep test data and the theoretical curve of layered phyllite samples. 4.3 Identification and sensitivity analysis of creep parameters 4.3 Identification and sensitivity analysis of creep parameters The nonlinear regression method was used on the uniaxial creep test data of layered phyllite to identify the parameters of the damage creep model of layered rock masses and to obtain the elastic moduli E1, E2, viscosity coefficient η2, η3, η4, and the value of the material parameter λ. The time t1 for the creep failure of the rock can be obtained from the uniaxial compression creep test. Based on the model parameters corresponding to the layered phyllite creep tests on samples having different bedding angles, the theoretical curve of the model is established with the test time as a consistent independent variable, and the curve is compared with the test data under the same conditions (as shown in Fig. 13). a) Bedding angle of 0 c) Bedding angle of 45 e) Bedding angle of 90 Axial compression 25MPa Axial compression 30MPa Axial compression 35MPa Axial compression 40MPa Axial compression 45MPa 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 ) 0.13(1 1.38 ε 0.068 t e     ) 0.14(1 1.64 ε 0.057 t e     ) 0.16(1 1.91 ε 0.048 t e     ) 0.18(1 2.17 ε 0.040 t e     t e t 0.013 ) 0.21(1 2.44 ε 0.032      Axial strain /10 -3 Time /h 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 t e ε t 5 0.01 ) (1 20 0. 4.3 Identification and sensitivity analysis of creep parameters 13f), and this is consistent with the experimental results. Under high-stress levels, the directional flow speed of the microelements inside the rock is accelerated, and the internal mechanical properties continue to deteriorate. Compared with the first five stress loading levels the creep parameters E1 E2 η2 and η3 of the rock E2 as well as the viscosity coefficients η2 and η3 initially decrease and then increase with the increase in bedding angle under low stress levels. Moreover, the changes in the creep parameters causes the cumulative creep deformation and creep time during the steady-state creep stage of the specimens to initially decrease and then increase as the bedding angle changes from 0° to 90° (as shown in Fig. 13f), and this is consistent with the experimental results. Under high-stress levels, the directional flow speed of the microelements inside the rock is accelerated, and the internal mechanical properties continue to deteriorate. Compared with the first five stress loading levels, the creep parameters E1, E2, η2, and η3 of the rock sample are small in the final loading stage, which demonstrates that the internal viscoelastic and viscoplastic deformations of the rock sample at this time are both large; the rock mass elastic modulus E1, E2, viscosity coefficient η2, η3, η4, and the time required for creep failure t1, all decrease first and then increase with increase in the bedding angle (as shown in Fig. 14). The influence of the transverse isotropic Compared with the first five stress loading levels, the creep parameters E1, E2, η2, and η3 of the rock sample are small in the final loading stage, which demonstrates that the internal viscoelastic and viscoplastic deformations of the rock sample at this time are both large; the rock mass elastic modulus E1, E2, viscosity coefficient η2, η3, η4, and the time required for creep failure t1, all decrease first and then increase with increase in the bedding angle (as shown in Fig. 14). The influence of the transverse isotropic characteristics of rocks on the rheological characteristics of layered phyllite cannot be ignored. a) Relationship between the bedding angle and elastic modulus E1. b) Relationship between the bedding angle and elastic modulus E2. c) Relationship between the bedding angle and viscosity coefficients η2. d) Relationship between the bedding angle and viscosity coefficients η3. 4.3 Identification and sensitivity analysis of creep parameters Table 2 shows the fitting parameters of layered phyllite samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90° under the creep failure stress. The experimental results show that the elastic modulus E2 as well as the viscosity coefficients η2 and η3 initially decrease and then increase with the increase in bedding angle under low stress levels. Moreover, the changes in the creep parameters causes the cumulative creep deformation and creep time during the steady-state creep stage of the specimens to initially decrease and then increase as the bedding angle changes from 0° to 90° (as shown in Fig. 13f), and this is consistent with the experimental results. Under high-stress levels, the directional flow speed of the microelements inside the rock is accelerated, and the internal mechanical properties continue to deteriorate. Compared with the first five stress loading levels, the creep parameters E1, E2, η2, and η3 of the rock sample are small in the final loading stage, which demonstrates that the internal viscoelastic and viscoplastic deformations of the rock sample at this time are both large; the rock mass elastic modulus E1, E2, viscosity coefficient η2, η3, η4, and the time required for creep failure t1, all decrease first and then increase with increase in the bedding angle (as shown in Fig. 14). The influence of the transverse isotropic characteristics of rocks on the rheological characteristics of layered phyllite cannot be ignored. E2 as well as the viscosity coefficients η2 and η3 initially decrease and then increase with the increase in bedding angle under low stress levels. Moreover, the changes in the creep parameters causes the cumulative creep deformation and creep time during the steady-state creep stage of the specimens to initially decrease and then increase as the bedding angle changes from 0° to 90° (as shown in Fig. 13f), and this is consistent with the experimental results. Under high-stress levels, the directional flow speed of the microelements inside the rock is accelerated, and the internal mechanical properties continue to deteriorate. E2 as well as the viscosity coefficients η2 and η3 initially decrease and then increase with the increase in bedding angle under low stress levels. Moreover, the changes in the creep parameters causes the cumulative creep deformation and creep time during the steady-state creep stage of the specimens to initially decrease and then increase as the bedding angle changes from 0° to 90° (as shown in Fig. 4.3 Identification and sensitivity analysis of creep parameters 8.52 9.35 9.65 9.73 16.84 0.0 22.5 45.0 67.5 90.0 0 2 4 6 8 10 12 14 16 18 20 E1 /GPa Bedding angle /° 172.41 128.62 157.14 158.82 299.19 0.0 22.5 45.0 67.5 90.0 0 50 100 150 200 250 300 350 E2 /GPa Bedding angle /° 5.34 2.97 3.77 4.76 9.57 0.0 22.5 45.0 67.5 90.0 0 2 4 6 8 10 12 η2 /(GPa·h) Bedding angle /° 384.62 199.01 303.45 398.55 895.16 0.0 22.5 45.0 67.5 90.0 0 200 400 600 800 1000 1200 η3 /(GPa·h) Bedding angle /° b) Relationship between the bedding angle and elastic modulus E2. 172.41 128.62 157.14 158.82 299.19 0.0 22.5 45.0 67.5 90.0 0 50 100 150 200 250 300 350 E2 /GPa Bedding angle /° 8.52 9.35 9.65 9.73 16.84 0.0 22.5 45.0 67.5 90.0 0 2 4 6 8 10 12 14 16 18 20 E1 /GPa Bedding angle /° b) Relationship between the bedding angle and elastic modulus E2. a) Relationship between the bedding angle and elastic modulus E1. 2 384.62 199.01 303.45 398.55 895.16 0.0 22.5 45.0 67.5 90.0 0 200 400 600 800 1000 1200 η3 /(GPa·h) Bedding angle /° 5.34 2.97 3.77 4.76 9.57 0.0 22.5 45.0 67.5 90.0 0 2 4 6 8 10 12 η2 /(GPa·h) Bedding angle /° η3 /(GPa·h) Bedding angle /° c) Relationship between the bedding angle and viscosity coefficients η2. d) Relationship between the bedding angle and viscosity coefficients η3. d) Relationship between the bedding angle and viscosity coefficients η3. e) Relationship between the bedding angle and viscosity coefficients η4. f) Relationship between the bedding angle and creep failure time t1. Fig. 14 Relationship between the bedding angle and creep parameters. 31.31 14.52 22.64 23.58 35.49 0.0 22.5 45.0 67.5 90.0 0 5 10 15 20 25 30 35 40 45 50 η4 /(GPa·h) Bedding angle /° 5.23 1.95 2.89 4.17 7.13 0.0 22.5 45.0 67.5 90.0 0 1 2 3 4 5 6 7 8 9 10 t1 /h Bedding angle /° 31.31 14.52 22.64 23.58 35.49 0.0 22.5 45.0 67.5 90.0 0 5 10 15 20 25 30 35 40 45 50 η4 /(GPa·h) Bedding angle /° 5.23 1.95 2.89 4.17 7.13 0.0 22.5 45.0 67.5 90.0 0 1 2 3 4 5 6 7 8 9 10 t1 /h Bedding angle /° e) Relationship between the bedding angle and viscosity coefficients η4. 4.3 Identification and sensitivity analysis of creep parameters f) Relationship between the bedding angle and creep failure time t1. Fig. 14 Relationship between the bedding angle and creep parameters. Table 2 Creep parameters of layered phyllite samples having different bedding angles. Bedding angle, β (°) Axial compression (MPa) Creep parameters E1 (GPa) E2(GPa) η2 (GPa·h) η3 (GPa·h) η4 (GPa·h) t1 (h) 0 25 18.06 192.30 13.08 - - - 30 18.29 214.29 12.21 - - - 35 18.32 218.75 10.50 - - - 40 18.43 222.22 8.89 - - - 45 18.52 224.17 7.19 3461.54 - - 50 8.52 172.41 5.34 384.62 31.31 5.23 22.5 20 18.39 142.86 9.43 - - - 24 18.46 150.00 7.35 - - - 28 18.54 164.71 7.24 - - - 32 18.60 168.42 6.06 - - - 36 18.62 171.42 4.63 1800.05 - - 40 9.35 128.62 2.97 199.01 14.52 1.95 45 24 17.91 179.10 10.03 - - - 28 18.18 186.66 7.84 - - - 32 18.49 191.62 7.47 - - - 36 18.65 200.02 6.40 - - - 40 18.96 203.42 4.88 2758.62 - - 44 9.65 157.14 3.77 303.45 22.64 2.89 67.5 29 17.26 181.25 11.60 - - - 34 18.28 188.89 10.75 - - - 39 18.31 198.98 9.35 - - - 44 18.64 209.53 8.17 - - - 49 18.77 213.04 5.96 3450.70 - - 54 9.73 158.82 4.76 398.55 23.58 4.17 Table 2 Creep parameters of layered phyllite samples having different bedding angles. 90 86 32.45 661.54 46.97 - - - 91 32.61 684.21 40.37 - - - 96 32.64 685.71 34.28 - - - 101 32.65 687.07 30.23 - - - 106 32.77 706.66 27.55 5273.63 - - 111 16.84 299.19 9.57 895.16 35.49 7.13 5. Discussion 5. Discussion The long-term load damage parameters D and λ were introduced into the damage creep model of layered rock masses, and these parameters play a key role in describing the damage characteristics of layered rock masses during the accelerated creep stage. Therefore, this section focuses on the effect of the long-term load damage parameters on the transversely isotropic creep characteristics and the damage mechanism of rock samples having different bedding angles during the accelerated creep stage. Fig. 15 shows the variation of damage variable with time for rock samples having different bedding angles during the accelerated creep stage. As the long-term load damage parameter λ of the specimens decreases, the nonlinear change trend of the curve becomes more significant and the shorter the time required for the damage variable D of the rock sample to reach 1. These observations indicate that the damage creep model can describe well the accelerated creep characteristics of layered rock masses. Fig. 15 Damage variable vs. time for rock samples having different bedding angles during the accelerated creep stage. Fig. 16 Creep curves of rock samples having different bedding angles during the accelerated creep stage. 0 1 2 3 4 5 6 7 8 9 10 0.0 0.2 0.4 0.6 0.8 1.0 Damage variable，D Time / h Bedding angle of 0°, λ=0.71 Bedding angle of 22.5°, λ=0.41 Bedding angle of 45°, λ=0.51 Bedding angle of 67.5°, λ=0.63 Bedding angle of 90°, λ=0.82 0 1 2 3 4 5 6 7 8 0.0 0.5 1.0 1.5 2.0 Axial strain / 10 -3 Time / h Bedding angle of 0°, λ=0.71 Bedding angle of 22.5°, λ=0.41 Bedding angle of 45°, λ=0.51 Bedding angle of 67.5°, λ=0.63 Bedding angle of 90°, λ=0.82 0 1 2 3 4 5 6 7 8 9 10 0.0 0.2 0.4 0.6 0.8 1.0 Damage variable，D Time / h Bedding angle of 0°, λ=0.71 Bedding angle of 22.5°, λ=0.41 Bedding angle of 45°, λ=0.51 Bedding angle of 67.5°, λ=0.63 Bedding angle of 90°, λ=0.82 0 1 2 3 4 5 6 7 8 0.0 0.5 1.0 1.5 2.0 Axial strain / 10 -3 Time / h Bedding angle of 0°, λ=0.71 Bedding angle of 22.5°, λ=0.41 Bedding angle of 45°, λ=0.51 Bedding angle of 67.5°, λ=0.63 Bedding angle of 90°, λ=0.82 Damage variable，D Axial strain / 10 -3 Fig. 16 Creep curves of rock samples having different bedding angles during the accelerated creep stage. 5. Discussion Fig. 16 Creep curves of rock samples having different Fig. 15 Damage variable vs. time for rock samples having different bedding angles during the accelerated creep stage. Fig. 16 Creep curves of rock samples having different bedding angles during the accelerated creep stage. Fig. 16 shows the creep curves of rock samples having different bedding angles during the accelerated creep stage. When the bedding angle is 22.5°, 45°, and 67.5°, the creep deformation rate of the rock sample increases, the curve becomes steeper, and the long-term load damage parameter λ of the rock sample is 0.41, 0.51, and 0.63, respectively. When the bedding angles are 0° and 90°, the accelerated creep stage of the rock sample is longer, the creep deformation rate is smaller, the curve is smoother, and the long-term load damage parameters λ of the rock sample are 0.71 and 0.82, respectively. Results show that as the bedding angle changes from 0° to 90°, the long-term load damage parameter λ of the rock sample first decreases and then increases. In addition, the time required for the accumulation of microscopic cracks in the corresponding rock sample during the accelerated creep stage to the generation of the macroscopic section first decreases and then increases. The damage creep model effectively reflects and predicts the transversely isotropic creep damage mechanism of layered rock masses. 6. Summary and conclusions To develop a comprehensive understanding of the creep characteristics and damage mechanism of layered rock masses, a series of uniaxial compression creep experiments were conducted on layered phyllite samples having bedding angles of 0°, 22.5°, 45°, 67.5°, and 90°. The laboratory test results show that the creep deformation of layered phyllite can be classified into four stages: the instantaneous elastic deformation stage, the deceleration creep stage, the steady-state creep stage, and the accelerated creep stage. The creep characteristics of layered phyllites are significantly affected by the weak bedding planes; specifically, the cumulative creep deformation and the creep time during the steady-state creep stage of the specimens initially decrease and then increase as the bedding angle changes from 0° to 90°, thereby, corresponding to the initial increase and subsequent decrease in creep rate during the deceleration creep stage. Based on the existing viscoelastic-plastic damage creep model, the creep parameters were identified, and the variation trend of creep parameters was studied. At low-stress levels, the elastic modulus E2 as well as the viscosity coefficients η2 and η3 all first decrease and then increase as the bedding angle changes from 0° to 90°. Under high-stress levels, the elastic moduli E1, E2, viscosity coefficient η2, η3, η4, and the time required for creep failure t1 all first decrease and then increase with increase in the bedding angle. Acknowledgements This study was financially supported by the National Natural Science Foundation of China (Grant Nos. 51978668 and 51578550), the Research Foundation of Education Bureau of Hunan Province (Grant No. 19K100), and the Innovation Project of Guangxi Graduate Education (Grant No. YCSW2021021). References Cao P, Wen YD, Wang YX, Yuan HP, Yuan BX (2016) Study on nonlinear damage creep constitutive model for high–stress soft rock. Environmental Earth Sciences 75(10): 900, DOI: 10.1007/s12665-016-5699-x Chen DF, Xu DP, Ren GF, Jiang Q, Liu GF, Wan LP, Li N (2019) Simulation of cross–correlated non–Gaussian random fields for layered rock mass mechanical parameters. Computers and Geotechnics 112: 104–119, DOI: 10.1016/j.compgeo.2019.04.012 Chen L, Wang CP, Liu JF, Li Y, Liu J, Wang J (2017) Effects of temperature and stress on the time–dependent behavior of Beishan granite. International Journal of Rock Mechanics and Mining Sciences 93: 316–323, DOI: 10.1016/j.ijrmms.2016.11.007 Chen ZQ, He C, Xu GW, MA GY, Wu D (2019) A case study on the asymmetric deformation characteristics and mechanical behavior of deep–buried tunnel in phyllite. Rock Mechanics and Rock Engineering 52: 4527–4545, DOI: 10.1007/s00603-019-01836-2 Fu HL, Zhang JB, Huang Z, Shi Y, Chen W (2018) A statistical model for predicting the triaxial compressive strength of transversely isotropic rocks subjected to freeze–thaw cycling. Cold Regions Science and Technology 145: 237–248, DOI: 10.1016/j.coldregions.2017.11.003 Hadiseh M, Ajalloeian R (2018) Mechanical behavior of salt rock under uniaxial compression and creep tests. International Journal of Rock Mechanics and Mining Sciences 110: 19–27, DOI: 10.1016/j.ijrmms.2018.07.006 Hu SC, Tan YL, Zhou H, Guo WY, Hu DW, Meng FZ, Liu ZG (2017) Impact of bedding planes on mechanical properties of sandstone. Rock Mechanics and Rock Engineering 50(8): 2243–2251, DOI: 10.1007/s00603-017-1239-6 ISRM (2007) The complete ISRM suggested methods for rock characterization, testing and monitoring: 1974–2006. In: Ulusay, R., Hudson, J.A. (Eds.), Suggested Methods Prepared by the Commission on Testing Methods. International Society for Rock Mechanics, Compilation Arranged by the ISRM Turkish National Group, Ankara, Turkey, pp. 628. Kachanov M (1992) Effective elastic properties of cracked solids: critical review of basic concepts. Applied Mechanics Reviews 45(8): 304–335, DOI: 10.1115/1.3119761 Lee YK, Pietruszczak S (2015) Tensile failure criterion for transversely isotropic rocks. International Journal of Rock Mechanics and Mining Sciences 79: 205–215, DOI: 10.1016/j.ijrmms.2015.08.019 Li XC, Yang CL, Ren T, Nie BS, Zhao CH, Liu SW, Jiang T (2017) Creep behaviour and constitutive model of coal filled with gas. International Journal of Mining Science and Technology 27: 847–851, DOI: 10.1016/j.ijmst.2017.07.017 Liu HZ, Xie HQ, He JD, Xiao ML, Zhuo L (2017) Nonlinear creep damage constitutive model for soft rocks. References Mechanics of Time-dependent Materials 21(1): 73–96, DOI: 10.1007/s11043-016-9319-7 Nadimi S, Shahriar K, Sharifzadeh M, Moarefvand P (2011) Triaxial creep tests and back analysis of time–dependent behavior of Siah Bisheh cavern by 3–Dimensional Distinct Element Method. Tunnelling and Underground Space Technology 26: 155–162, DOI: 10.1016/j.tust.2010.09.002 Reza Taheri S, Pak A, Shad S, Mehrgini B, Razifar M (2020) Investigation of rock salt layer creep and its effects on casing collapse. International Journal of Mining Science and Technology 30: 357–365, DOI: 10.1016/j.ijmst.2020.02.001 Shen PW, Tang HM, Zhang BC, Ning YB, He C (2021) Investigation on the fracture and mechanical behaviors of simulated transversely isotropic rock made of two interbedded materials. Engineering Geology 286: 106058, DOI: 10.1016/j.enggeo.2021.106058 Tien YM, Kuo MC, Juang CH (2006) An experimental investigation of the failure mechanism of simulated transversely isotropic rocks. International Journal of Rock Mechanics and Mining Sciences 43: 1163–1181, DOI: 10.1016/j.ijrmms.2006.03.011 Wang JB, Zhang Q, Song ZP, Zhang YW (2021) Experimental study on creep properties of salt rock under long–period cyclic loading. International Journal of Fatigue 143: 106009, DOI: 10.1016/j.ijfatigue.2020.106009 Wu F, Zhang H, Zou QL, Li CB, Chen J, Gao RB (2020) Viscoelastic–plastic damage creep model for salt rock based on fractional derivative theory. Mechanics of Materials 150: 103600, DOI: 10.1016/j.mechmat.2020.103600 Xie SY, Shao JF, Xu WY (2011) Influences of chemical degradation on mechanical behaviour of a limestone. International Journal of Rock Mechanics and Mining Sciences 48: 741–747, DOI: 10.1016/j.ijrmms.2011.04.015 Xu GW, He C, Su Ang, Chen ZQ (2018) Experimental investigation of the anisotropic mechanical behavior of phyllite under triaxial compression. International Journal of Rock Mechanics and Mining Sciences 104: 100–112, DOI: 10.1016/j.ijrmms.2018.02.017 Xu GW, He C, Chen ZQ, Yang QH (2020) Transversely isotropic creep behavior of phyllite and its influence on the long–term safety of the secondary lining of tunnels. Engineering Geology 278: 105834, DOI: 10.1016/j.enggeo.2020.105834 Yang WD, Zhang QY, Li SC, Wang SG (2014) Time–dependent behavior of diabase and a nonlinear creep model. Rock Mechanics and Rock Engineering 47: 1211–1224, DOI: 10.1007/s00603-013-0478-4 Yang SQ, Xu P, Ranjith PG (2015) Damage model of coal under creep and triaxial compression. International Journal of Rock Mechanics and Mining Sciences 80: 337–345, DOI: 10.1016/j.ijrmms.2015.10.006 Yang SQ, Yin PF, Huang YH, Cheng JL (2021) Strength, deformability and X-ray micro-CT observations of transversely isotropic composite rock under different confining pressures. References Engineering Fracture Mechanics 241: 107389, DOI: 10.1016/j.engfracmech.2019.04.030 Zhang JB, Fu HL, Huang Z, Wu YM, Chen W, Shi Y (2019) Experimental study on the tensile strength and failure characteristics of transversely isotropic rocks after freeze–thaw cycles. Cold Regions Science and Technology 163: 68–77, DOI: 10.1016/j.coldregions.2019.04.006 Zhao YL, Cao P, Wang WJ, Wan W, Liu YK (2009) Viscoelasto-plastic rheological experiment under circular increment step load and unload and nonlinear creep model of soft rocks. Journal of Central South University of Technology 16: 0488–0494, DOI: 10.1007/s11771-009-0082-7 Zhao YL, Wang YX, Wang WJ, Wan W, Tang JZ (2017) Modeling of non-linear rheological behavior of hard rock using triaxial rheological experiment. International Journal of Rock Mechanics and Mining Sciences 93: 66–75, DOI: 10.1016/j.ijrmms.2017.01.004 Zhao YL, Wang YX, Wang WJ, Wan W, Tang JZ (2017) Modeling of non-linear rheological behavior of hard rock using triaxial rheological experiment. International Journal of Rock Mechanics and Mining Sciences 93: 66–75, DOI: 10.1016/j.ijrmms.2017.01.004 Zhou HW, Wang CP, Han BB, Duan ZQ (2011) A creep constitutive model for salt rock based on fractional derivatives. International Journal of Rock Mechanics and Mining Sciences 48(1): 116–121, DOI: 10.1016/j.ijrmms.2010.11.004 Zhou HW, Wang CP, Han BB, Duan ZQ (2011) A creep constitutive model for salt rock based on fractional derivatives. International Journal of Rock Mechanics and Mining Sciences 48(1): 116–121, DOI: 10.1016/j.ijrmms.2010.11.004 Zhou HW, Wang CP, Han BB, Duan ZQ (2011) A creep constitutive model for salt rock based on fractional derivatives. International Journal of Rock Mechanics and Mining Sciences 48(1): 116–121, DOI: 10.1016/j.ijrmms.2010.11.004
https://openalex.org/W2107797757	https://europepmc.org/articles/pmc3553030?pdf=render	English	null	Prognostic value of stromal and epithelial periostin expression in human prostate cancer: correlation with clinical pathological features and the risk of biochemical relapse or death	BMC cancer	2,012	cc-by	6,845	RESEARCH ARTICLE Open Access * Correspondence: fboccardo@unige.it 1IRCCS San Martino University Hospital – IST National Cancer Research Institute and the University of Genoa, Academic Unit of Medical Oncology (UOC Oncologia Medica B), Largo Rosanna Benzi 10, 16132 Genoa, Italy 3University of Genoa, Department of Internal Medicine, School of Medicine, Genoa, Italy Full list of author information is available at the end of the article © 2012 Nuzzo et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Prognostic value of stromal and epithelial periostin expression in human prostate cancer: correlation with clinical pathological features and the risk of biochemical relapse or death Pier Vitale Nuzzo1,3, Alessandra Rubagotti1,3, Linda Zinoli1,3, Francesco Ricci3, Sandra Salvi2, Simona Boccardo2 and Francesco Boccardo1,3* Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Patient selection Since we originally planned to study POSTN overexpres- sion in cryopreserved material by immune blotting techni- ques, a patient cohort we had previously utilized for studies on NM proteins was selected [7]. This cohort is made up of 90 patients who underwent radical prostatec- tomy for biopsy proven PCa between October 1995 and October 2003, and who were subsequently referred to our Unit for treatment or follow-up. Before surgery, all patients had provided consent allowing tumor tissue specimens to be collected for proteomic analysis. This research project was approved by the Ethical Committee of the National Cancer Research Institute of Genoa, Italy. Unfortunately, most of the stored material had been used for previous studies on NM proteins and therefore large enough sam- ples for the present evaluations were no longer available. This prompted us to retrieve corresponding archival mater- ial in order to allow us to carry out immune histochemical studies. We were able to follow-up most of our cohort patients at regular intervals. However, over time, a relevant number of patients failed to attend clinical examinations, so the vital status of these patients had to be checked by phone, or, when this was not possible, by contacting the local tumor registry or the registry office of the patient’s place of residence. The main characteristics of the patients making up the study cohort are summarized in Table 1. Many of the cellular abnormalities that are present in most solid tumors are structural in nature and involve either the nuclear matrix (NM) or the extracellular matrix (ECM), both of which are regarded as a promising source of new markers [6-8]. Among the components of ECM, increasing interest has been shown in Periostin (POSTN), a protein produced by fibroblasts, as a major putative player in human carcinogenesis [8]. During embryogenesis, this protein is preferentially expressed in the periosteum and periodontal ligaments where it acts as a critical regulator of bone and tooth formation and maintenance [9]. However, it was also shown to play an important role in cardiac devel- opment [10]. In adults, POSTN is up-regulated by mechan- ical stress and contributes to tissue repair and regeneration [11,12]. It has recently been suggested that POSTN might also play a relevant role in human carcinogenesis. Abstract Background: The purpose of the present study was to evaluate the prognostic value of POSTN expression following prostatectomy. Methods: Periostin (POSTN) expression in prostate cancer (PCa) and in normal specimens was evaluated in 90 patients by an immuno-reactive score(IRS) based on the intensity of immunostaining and on the quantity of stained cells. The t-test was applied to compare IRS values in cancer specimens to values in normal specimens. Pearson’s test was used to correlate POSTN expression to clinical pathologic features. PSA progression-free and survival curves were constructed by the Kaplan–Meier method and compared using the log-rank test. Multi-parametric models were constructed according to the Cox technique adding all the covariates predicting for either PSA progression or death into the models after univariate analysis. Results: Both stromal and epithelial POSTN expression were significantly increased in tumor tissues. In particular, we found stromal expression to be significantly higher than epithelial expression as compared to normal tissues (p<0.000 and p=0.001).A significant correlation between POSTN epithelial expression and extra-prostatic extension was found (p=0.03). While high stromal expression was significantly associated with shorter survival (p=0.008), a low epithelial score significantly correlated with shorter PSA-free survival (p=0.04), suggesting that POSTN plays an apparently opposing biological role depending on its compartmentalization.Regardless of the mechanism that is involved, patients showing both high stromal and low epithelial expression made up a subgroup with a very bleak prognosis. Conclusions: Although requiring further validation through larger studies, our findings show that POSTN might represent a novel prognostic marker for PCa. rds: POSTN protein, Human, Prostatic neoplasms, Extracellular matrix proteins, Prognosis, Biomarker © 2012 Nuzzo et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 2 of 12 Background overexpression in PCa [15,16]. In one study, increased epithelial expression was found during the early stages of PCa, whereas stromal POSTN expression prevailed in advanced stages [15]. In the other study, which also showed POSTN to be far more overexpressed in tumor tissues than in peritumoral tissues, POSTN appeared to be overexpressed both by the epithelial and by the stromal cells [16]. In this study, a strong association between epithelial expression and local tumor stage was observed, while stromal overexpression appeared to be correlated mainly with a high Gleason score and an increased risk of biochemical failure [16]. Prostate cancer (PCa) has become the most common ma- lignancy among men in most Western countries [1]. Even when this tumor is apparently confined to the prostate, it encompasses a broad spectrum of diseases, some of which are characterized by extremely indolent behavior and others by very poor outcome. Therefore, an important clinical question is how aggressively to treat patients diagnosed at this stage. Among patients who are treated with radical prostatectomy, the most commonly used parameters for defining prognosis and choosing the right candidates for adjuvant local irradiation or systemic treatments include tumor volume and pathological grade and status of surgical margins, seminal vesicles and pelvic nodes [2]. However, there is no widely accepted method for quantifying tumor volume [3]. Moreover, tumor grade scoring methods can result in significant inter-observer variations, particularly when defining intermediate tumor grades [4,5]. This applies specifically to the old Gleason scoring method. New prog- nostic markers are therefore required. In the present study we investigated POSTN expres- sion in PCa tissue specimens and in normal peritumoral specimens in order to confirm previous findings and to evaluate the putative prognostic value of POSTN also as a function of its compartmentalization. Patient selection In fact, this protein interacts with multiple cell-surface receptors, most notably integrins, as well as with signals mainly via the PI3-K/Akt and other pathways, thus promoting cancer cell survival, epithelial–mesenchymal transition (EMT), in- vasion, and metastasis [13]. Though it is currently not clear whether the production and secretion of POSTN is directly mediated by tumor epithelial cells or by stromal cells, or by both, overexpression of POSTN in cancer stroma and/or in the epithelium is usually associated with the most ma- lignant phenotypes and with poor clinical outcome [8]. In bladder cancer however, protein down-regulation was shown to be associated with poorer prognostic features [14], suggesting that POSTN can act either as a tumor promoter or as a tumor suppressor gene, most likely depending on several variables, including the protein isoform and /or the interactor involved in the process. Immunohistochemistry (IHC) IHC analysis was carried out using 3-μm sections of paraf- fin embedded prostate tissue using the POSTN (OSF-2) Polyclonal Antibody (Acris Antibodies, Germany; Host/ Isotype:Rabbit), suitable for various isoforms of POSTN. The antibody was diluted at 1:500. The most representative tumor and normal peritu- moral tissue sections were immunostained using the To the best of our knowledge, to date only two studies have investigated the clinical relevance of POSTN Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 3 of 12 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Table 1 Main characteristic of study patients Table 1 Main characteristic of study patients Total patients N=90(%) Median age in years(range) 64(48–77) Median PSA level at surgery in ng/ml(range) 11.0(1.7-167.4) Extra-prostatic extension 46(51.1) Pelvic nodes involved 16(17.8) Surgical margins involved 34(37.8) Seminal vesicles involved 24(26.7) Gleason score<7 32(35.6) =7 30(33.3) >7 28(31.1) A negative and a positive control were used for each stain- ing run. The negative control consisted of performing the entire IHC procedure on adjacent sections in the absence of the primary antibody. Then the sections were counter- stained with Gill's modified hematoxylin, cover-slipped and evaluated by two different observers using an Olympus multi-headed light microscope using 10X, 40X and 63X magnifications. Evaluation of staining To evaluate epithelial and stromal POSTN expression, we used the immuno-reactive score (IRS) as previously implemented by Tischler et al. [16], based on the inten- sity of immune staining and the quantity of stained cells. The intensity of staining was arbitrarily graded as: absent (0), weak (1+), moderate (2+), strong (3+). The percen- tage of stained cells was used to quantify the reaction as negative (0% of positive cells), 1+ (<10% positive cells); 2+ (10-50% of positive cells); 3+ (51-80% of positive cells); 4+ (>80% of positive cells). The final value of the analysis of each tissue sample was then expressed as an Benchmark XT automatic stainer (Ventana Medical Systems, SA Stasbourg, France). Slides were deparaffi- nized, and after adding high Ph, heat induced, standard cit- rate buffer (30 min), the antibody-antigen complex was relieved using the polymeric detection system (Ventana Medical System Ultraview Universal DAB Detection Kit). Figure 1 POSTN expression in tumor stroma: tumor specimens graded 0 to 3+ according to arbitrary scoring (see text) are shown. Figure 1 POSTN expression in tumor stroma: tumor specimens graded 0 to 3+ according to arbitrary scoring (see text) are shown. Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 4 of 12 Figure 2 POSTN expression in tumor epithelium: tumor specimens graded 0 to 3+ according to arbitrary scoring (see text) are shown. Figure 2 POSTN expression in tumor epithelium: tumor specimens graded 0 to 3+ according to arbitrary scoring (see text) are shown. sion in tumor epithelium: tumor specimens graded 0 to 3+ according to arbitrary scoring (see text) are shown. igure 2 POSTN expression in tumor epithelium: tumor specimens graded 0 to 3+ according to arbitrary scoring (se absolute value through the obtained score by multiplying the two individual scores (i.e., intensity of staining score times the percentage of stained cells score). Examples of scoring according to staining intensity and the percen- tage of stained cells are shown in Figures 1 and 2. was defined by any PSA serum level of 0.4 ng/ml or more following prostatectomy, provided that this value had been confirmed at least once, and at least 4 weeks apart. PSA progression-free survival was thus defined by the amount of time that elapsed from the date the patient underwent prostatectomy to the date of the documented PSA progres- sion as defined above. Evaluation of staining Overall survival was calculated as the amount of time between the date of prostatectomy and the date of death, regardless of the cause. Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Main analysis endpoints PSA progression-free survival and overall survival were the main end-points of the present analysis. PSA progression Table 2 POSTN expression in epithelium or stroma of tumor or normal tissue specimens Epithelium Stroma Mean value(SE) p≤ Mean value(SE) p≤ Tumor tissue:Intensity 0.56(0.08) 2.38(0.08) Normal tissue:Intensity 0.26(0.05) 0.002 1.49(0.09) 0.000 Tumor tissue:%Cells stained 0.89(0.13) 2.46(0.11) Normal tissue:%Cells stained 0.41(0.09) 0.003 1.52(0.10) 0.000 Tumor tissue:IRS 1.24(0.21) 6.37(0.39) Normal tissue:IRS 0.46(0.10) 0.001 2.80(0.27) 0.000 Abbreviations:SE,Standard Error;IRS, Immuno-reactive score. Table 2 POSTN expression in epithelium or stroma of tumor or normal tissue specimens Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 Page 5 of 12 Figure 3 PSA progression-free Survival (A) and Overall Survival (B) as a function of stromal IRS (see text for explanations). Figure 3 PSA progression-free Survival (A) and Overall Survival (B) as a function of stromal IRS (see text for explanations). Statistical analysis survival curves were constructed by the Kaplan–Meier method [17] and compared using the log-rank test [18]. To evaluate the role of prognostic variables, a series of Cox proportional hazards models were fitted to PSA progression-free and overall survival data [19]. The follow- ing covariates were included in all models: pre-surgery PSA levels (<=10 ng/ml, >10 ng/ml); extra-prostatic exten- sion (Yes, No); involvement of surgical margins (No, Yes); involvement of seminal vesicles (No, Yes); Gleason score The t-test was applied to compare mean values (Standard Error: SE) of stromal or epithelial IRS in PCa tissue spe- cimens with those calculated in normal peritumoral tissue specimens. Pearson’s correlation was used to correlate either epithelial or stromal POSTN expression with all of the following: PSA, Gleason score, extra-prostatic exten- sion, lymph node status, involvement of surgical margins or of seminal vesicles. PSA progression-free and overall Figure 4 PSA progression-free (A) and Overall Survival (B) as a function of epithelial IRS (see text for explanations). Figure 4 PSA progression-free (A) and Overall Survival (B) as a function of epithelial IRS (see text for explanations). Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 6 of 12 Table 3 Multivariate analysis PSA Progression-Free Survival Model 1 Model 2 Model 3 HR (95%CI) p= HR (95%CI) p= HR (95%CI) p= PSA,ng/ml ≤10ng/ml 1.0 1.0 1.0 >10ng/ml 1.41 (0.67-2.97) 0.4 1.26 (0.59-2.68) 0.5 1.30 (0.61-2.79) 0.5 Extra-prostatic extension Yes 1.0 1.0 1.0 No 3.33 (1.31-8.49) 0.01 3.21 (1.28-8.07) 0.01 3.21 (1.27-8.09) 0.01 Pelvic nodes involved N0 1.0 1.0 1.0 N+ 1.56 (0.69-3.56) 0.3 1.38 (0.60-3.14) 0.4 1.40 (0.61-3.18) 0.4 Surgical margins involved No 1.0 1.0 1.0 Yes 0.75 (0.31-1.79) 0.5 0.72 (0.31-1.67) 0.4 0.68 (0.29-1.62) 0.4 Seminal vesicles involved No 1.0 1.0 1.0 Yes 0.77 (0.35-1.68) 0.5 0.64 (0.28-1.43) 0.3 0.64 (0.28-1.42) 0.3 Gleason score <7 1.0 ≥7 2.03 (0.91-4.55) 0.08 Gleason score ≤7 1.0 >7 3.29 (1.63-6.65) 0.001 Gleason score <7 1.0 0.003 =7 1.33 (0.54-3.26) 0.5 >7 3.90 (1.59-9.55) 0.003 Epithelial IRS >2 1.0 1.0 1.0 ≤2 6.20 (1.30-29.44) 0.02 4.85 (0.98-24.09) 0.05 4.86 (0.98-24.05) 0.05 Stroma IRS <12 1.0 1.0 1.0 =12 2.06 (0.87-4.88) 0.1 1.90 (0.77-4.65) 0.1 1.84 (0.76-4.50) 0.6 Abbreviations:HR, Hazard Ratio;CI,Confidence Interval;PSA,Prostate-specific antigen;IRS,Immuno-reactive score. Table 3 Multivariate analysis (HR) estimates and their 95% Confidence Intervals (CIs) were also calculated [18]. All P values were two-tailed. The IBM software Statistical Package for Social Sciences (SPSS) version 19.0 for Windows (SPSS Inc. Statistical analysis Chicago, Illinois, USA) was used for data analysis. (Model I: <7, >=7), (Model II: <=7, >7), (Model III: <7, =7, >7); involvement of pelvic nodes (N0, N+); epithelial IRS; stromal IRS and stromal plus epithelial IRS. Relative to IRS, the value corresponding to the 75th percentile was used as an arbitrary cut-off (75th percentile value of epithe- lial IRS=2; 75th percentile value of stromal IRS=12). The 75th percentile value was infact the value which better dis- criminated the patient-cohort according to the main clin- ical outcome endpoints on study. (Model I: <7, >=7), (Model II: <=7, >7), (Model III: <7, =7, >7); involvement of pelvic nodes (N0, N+); epithelial IRS; stromal IRS and stromal plus epithelial IRS. Relative to IRS, the value corresponding to the 75th percentile was used as an arbitrary cut-off (75th percentile value of epithe- lial IRS=2; 75th percentile value of stromal IRS=12). The 75th percentile value was infact the value which better dis- criminated the patient-cohort according to the main clin- ical outcome endpoints on study. POSTN expression in the epithelium and stroma of prostate tissues A stepwise procedure was used with a significance level of p=0.05 to retain variables in the model. Hazard Ratio Distinct stromal and epithelial staining characteristics allowed for absolutely certain evaluation of POSTN Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 7 of 12 Page 7 of 12 Table 4 Multivariate analysis Overall survival Model 1 Model 2 Model 3 HR (95%CI) p= HR (95%CI) p= HR (95%CI) p= PSA,ng/ml ≤10ng/ml 1.0 1.0 1.0 >10ng/ml 0.54 (0.15-2.03) 0.4 0.48 (0.12-1.86) 0.3 0.48 (0.12-1.87) 0.3 Extra-prostatic extension Yes 1.0 1.0 (0.76-17.35) 1.0 No 3.43 (0.73-16.12) 0.1 3.63 0.1 3.48 (0.73-16.54) 0.1 Pelvic nodes involved N0 1.0 1.0 1.0 N+ 1.71 (0.50-5.79) 0.4 1.44 (0.43-4.83) 0.5 1.40 (0.42-4.72) 0.6 Surgical margins involved No 1.0 1.0 1.0 Yes 1.37 (0.45-4.21) 0.6 1.38 (0.47-4.01) 0.6 1.33 (0.45-3.92) 0.6 Seminal vesicles involved No 1.0 1.0 1.0 Yes 1.39 (0.49-3.94) 0.5 1.08 (0.37-3.10) 0.9 1.11 (0.38-3.25) 0.8 Gleason score <7 1.0 ≥7 2.13 (0.55-8.19) 0.3 Gleason score ≤7 1.0 >7 2.80 (0.93-8.47) 0.07 Gleason score <7 1.0 0.1 =7 1.43 (0.32-6.40) 0.6 >7 3.52 (0.79-15.61) 0.1 Epithelial IRS >2 1.0 1.0 1.0 ≤2 3.22 (0.56-18.48) 0.2 3.09 (0.51-18.63) 0.2 3.19 (0.52-19.47) 0.2 Stroma IRS <12 1.0 1.0 1.0 =12 5.85 (1.82-18.77) 0.003 5.76 (1.84-18.07) 0.003 5.99 (1.87-19.22) 0.003 Abbreviations:HR,Hazard Ratio;CI Confidence Interval;PSA,Prostate-specific antigen;IRS Immuno-reactive score. Table 4 Multivariate analysis correlated with epithelial POSTN expression (p<0.000). As shown in Table 2, both stromal and epithelial POSTN expressions were significantly increased in tumor tissues as compared to normal adjacent tissues (p<0.000 and p=0.001, respectively). Nevertheless, POSTN expression in the stromal component of nor- mal tissues was about twice as high as what was observed in the epithelial component of prostate can- cer tissues (p=0.003), indicating that stromal cells mostly contribute to POSTN secretion both in normal and in neoplastic conditions. correlated with epithelial POSTN expression (p<0.000). As shown in Table 2, both stromal and epithelial POSTN expressions were significantly increased in tumor tissues as compared to normal adjacent tissues (p<0.000 and p=0.001, respectively). Nevertheless, POSTN expression in the stromal component of nor- mal tissues was about twice as high as what was observed in the epithelial component of prostate can- cer tissues (p=0.003), indicating that stromal cells mostly contribute to POSTN secretion both in normal and in neoplastic conditions. staining in the selected tumor and peritumor tissue spe- cimens. POSTN expression in the epithelium and stroma of prostate tissues Of the 90 prostate cancers, 79 (87.8%) dis- played moderate (N°=32) or strong (N°=47) stromal POSTN expression. The mean (SE) overall IRS value was 6.37 (0.39). The expression of POSTN by tumor epithelial cells was significantly lower than what was observed in stromal cells (p=0.003). In fact, only 36 out of 90 (40%) cancer specimens showed weak (N°=25) or moderate (N°=11) staining intensity. The mean (ES) overall IRS in this case was 1.24 (0.21). Notably, stromal POSTN expression significantly Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 Page 8 of 12 http://www.biomedcentral.com/1471-2407/12/625 Figure 5 PSA progression-free (A) and Overall Survival (B) as a function of both stromal and epithelial IRS (see text for explanations). on-free (A) and Overall Survival (B) as a function of both stromal and epithelial IRS (see text for explanations). Figure 5 PSA progression-free (A) and Overall Survival (B) as a function of both stromal and epithelial IRS (see t Correlation with clinical-pathological variables corresponding to the 75th percentile) was significantly correlated with longer PSA-free survival (p=0.04) (Figure 4). Since stromal and epithelial expression appeared to have different correlations despite being dir- ectly correlated to each other as previously mentioned, the two variables (i.e., stromal and epithelial POSTN scores) were both included in the same multiparametric models together with the other variables that predicted both PSA–free and overall survival in univariate analysis. Differ- ent models were created on the basis of the Gleason score (Tables 3 and 4). Multivariate analysis confirmed that epi- thelial expression independently correlated with the risk of PSA progression, regardless of how the Gleason score was analyzed. Low expression did in fact imply an increase in the risk of PSA failure, which became statistically sig- nificant when the Gleason score was analyzed by arbitrar- ily grouping patients with a score >7 with those showing a score equal to 7. Multivariate analysis also confirmed that stromal IRS did not predict the risk of PSA failure, regard- less of which Gleason score variable was used. By contrast, stromal IRS was the only independent predictor of the risk of death, regardless of how the Gleason score variable was analyzed. Epithelial IRS was not predictive of the risk of death, similarly to all the other variables we considered, including extra-prostatic extension and Gleason score. However, the latter variables were predictors of the risk of PSA failure. POSTN expression in the epithelium and stroma of prostate tissues Noteworthy, the patients showing both high stromal expression (IRS=12) and low epithelial expression (IRS<=2) made up a subgroup with a very bleak prognosis, showing both the shortest PSA-free (p=0.005) and the shortest overall survival (p=0.02) (Figure 5).This trend was confirmed by multivariate analysis (Tables 5 and 6). We found no specific correlation between epithelial or stromal POSTN expression and any of the clinical- pathological parameters (PSA preoperative level; Gleason score; extra-prostatic extension; lymph node status; in- volvement of surgical margins or seminal vesicles) that we evaluated in the present analysis except for a weak positive correlation between POSTN stromal expression and Gleason score (p=0.08), and a significant correlation be- tween epithelial expression of the protein and extra- prostatic extension (p=0.03). Correlation of POSTN expression with PSA-free and overall survival Besides the number of patients and disease stage, these studies differ in some methodological aspects. In our own, as well as in Tischler’s study [16], POSTN detection was performed using the same rabbit polyclonal antibody capable of recognizing all the different POSTN isoforms. A polyclonal anti-POSTN antibody was also used by the Japanese investigators for their IHC determinations [15]. However, there were major Correlation of POSTN expression with PSA-free and overall survival After a median follow-up time of 134.5 months (range, 33.7–178.2), 44 patients progressed and 19 died. Median time to PSA progression for the whole cohort was 94.5 months (range 3–169.6), while median time to death has not been reached yet. POSTN expression in the stromal and epithelial compartments of the tumor both correlated with PSA progression-free survival or overall survival, al- beit in a different manner. Stromal POSTN expression was significantly associated with overall survival, for an IRS value of 12, corresponding to the 75th percentile. The group of patients with an IRS=12 (strong staining intensity and 80% of positive cells) showed significantly shorter sur- vival than patients with an IRS<12 (p=0.008). These patients also showed a trend for shorter PSA-free survival (Figure 3). While no significant correlation was found be- tween epithelial IRS and patients' survival, in contrast to previous findings, a higher epithelial score (IRS>2, again Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 9 of 12 Page 9 of 12 Table 5 Multivariate analysis PSA Progression-Free Survival Model 1 Model 2 Model 3 HR (95%CI) p= HR (95%CI) p= HR (95%CI) p= PSA,ng/ml ≤10ng/ml 1.0 1.0 1.0 >10ng/ml 1.39 (0.66-2.96) 0.4 1.25 (0.58-2.68) 0.6 1.29 (0.60-2.78) 0.5 Extra-prostatic extension Yes 1.0 1.0 1.0 No 3.30 (1.29-8.47) 0.01 3.21 (1.27-8.07) 0.01 3.20 (1.27-8.09) 0.01 Pelvic nodes involved N0 1.0 1.0 1.0 N+ 1.55 (0.68-3.55) 0.3 1.37 (0.60-3.14) 0.4 1.39 (0.61-3.18) 0.4 Surgical margins involved No 1.0 1.0 1.0 Yes 0.75 (0.31-1.81) 0.5 0.72 (0.31-1.67) 0.4 0.68 (0.29-1.62) 0.4 Seminal vesicles involved No 1.0 1.0 1.0 Yes 0.77 (0.35-1.70) 0.5 0.64 (0.28-1.44) 0.3 0.64 (0.28-1.43) 0.3 Gleason score <7 1.0 ≥7 2.02 (0.90-4.54) 0.09 Gleason score ≤7 1.0 >7 3.29 (1.62-6.65) 0.001 Gleason score <7 1.0 0.03 =7 1.32 (0.54-3.26) 0.5 >7 3.89 (1.58-9.54) 0.003 Stromal -Epithelial IRS Score =12/≤2 1.0 0.08 1.0 0.2 1.0 0.2 <12/≤2 0.47 (0.19-1.17) 0.1 0.52 (0.20-1.33) 0.2 0.53 (0.21-1.37) 0.2 =12/>2 0.14 (0.01-1.33) 0.09 0.19 (0.02-1.89) 0.2 0.19 (0.02-1.89) 0.2 <12/>2 0.09 (0.01-0.77) 0.03 0.11 (0.01-1.10) 0.06 0.12 (0.01-1.11) 0.06 Abbreviations:HR,Hazard Ratio;CI ,Confidence Interval; PSA,Prostate-specific antigen;IRS, Immuno-reactive score. Table 5 Multivariate analysis found higher POSTN expression in normal epithelial cells. However, in this study, POSTN expression was also higher in the epithelium than in the stroma of cancerous tissues [15]. It is not easy to explain the differences that were observed in the three studies. Discussion Comparably to previous findings by Tischler et al. [16], we have demonstrated that POSTN is far more highly expressed in cancer tissues than in normal tissues. In our study, POSTN appears to be expressed mainly in the stromal compartment, both in normal and in cancerous tissues. In Tischler’s study, epithelial expression was higher in normal tissues, while in cancerous tissues it was higher in the larger test cohort, but it was lower in the smaller training cohort [16]. Tischler's findings on normal prostate gland tissues are comparable with those previously reported by Tsunoda et al. [15] who also Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 10 of 12 Table 6 Multivariate analysis Overall survival Model 1 Model 2 Model 3 HR (95%CI) p= HR (95%CI) p= HR (95%CI) p= PSA,ng/ml ≤10ng/ml 1.0 1.0 1.0 >10ng/ml 0.49 (0.13-1.83) 0.3 0.45 (0.12-1.74) 0.2 0.45 (0.12-1.74) 0.2 Extra-prostatic extension Yes 1.0 1.0 1.0 No 3.34 (0.71-15.66) 0.1 3.75 (0.79-17.88) 0.1 3.52 (0.74-16.77) 0.1 Pelvic nodes involved N0 1.0 1.0 1.0 N+ 1.72 (0.50-5.88) 0.4 1.45 (0.43-4.93) 0.5 1.41 (0.41-4.79) 0.6 Surgical margins involved No 1.0 1.0 1.0 Yes 1.42 (0.46-4.38) 0.5 1.36 (0.46-3.99) 0.6 1.31 (0.44-3.89) 0.6 Seminal vesicles involved No 1.0 1.0 1.0 Yes 1.41 (0.50-3.97) 0.5 1.06 (0.37-3.05) 0.9 1.12 (0.38-3.20) 0.8 Gleason score <7 1.0 ≥7 2.19 (0.56-8.62) 0.3 Gleason score ≤7 1.0 >7 2.75 (0.90-8.42) 0.08 Gleason score <7 1.0 0.2 =7 1.49 (0.33-6.81) 0.6 >7 3.56 (0.78-16.20) 0.1 Stromal -Epithelial IRS Score =12/≤2 1.0 0.01 1.0 0.01 1.0 0.01 <12/≤2 0.13 (0.04-0.44) 0.001 0.14 (0.04-0.46) 0.001 0.14 (0.04-0.45) 0.001 =12/>2 0.19 (0.02-1.37) 0.1 0.21 (0.03-1.58) 0.1 0.20 (0.02-1.53) 0.1 <12/>2 0.12 (0.01-1.32) 0.08 0.12 (0.01-1.43) 0.09 0.11 (0.01-1.37) 0.09 Abbreviations:HR,Hazard Ratio;CI ,Confidence Interval; PSA, Prostate-specific antigen;IRS,Immuno-reactive score. Table 6 Multivariate analysis differences concerning staining evaluation. In fact, we and Tischler used an immune score (IRS) obtained by mul- tiplying the intensity of staining by the percentage of stained cells. Notably, comparable median IRS values were obtained by us and by the Swiss colleagues. However, the results obtained in our two studies are not comparable since patients were analyzed after arbitrarily grouping them in different manners. Authors’ contributions NPV participated in the design of the study, in data interpretation and in drafting the manuscript. RA participated in the design of the study and performed the statistical analysis. ZL made substantial contributions in updating patients histories and helped in statistical analysis. RF participated in study conception and acquisition of data. SS carried out the immunohistochemistry analysis and evaluation of staining. BS carried out the immunohistochemistry analysis and evaluation of staining. BF: made substantial contributions to conception, design, analysis and interpretation of data. He has been involved in drafting the manuscript and revising it critically for important intellectual content. He gave final approval of the version to be published. No degree of relationship do exhist between Boccardo S and Boccardo F, as they simply share an homonymy of family name. All authors read and approved the final manuscript. We also observed a direct relationship between stromal POSTN IRS and PSA-free survival, however the difference was not statistically significant, exactly as reported by Tischler et al. [16] in their validation set which was much larger than our own set but showed a comparable clinical outcome. However, low epithelial POSTN expression was associated with shorter PSA-free survival in our study, and epithelial expression was not predictive of patient sur- vival. By contrast, stromal expression was highly predictive of the risk of death, while it was only a weak predictor of PSA progression. These findings suggest that POSTN might play a different biological role in tumor progression, depending on its compartmentalization. In fact, while POSTN down-regulation in PCa epithelium appears to be correlated with extra-prostatic extension and biochemical failure, both of which represent early events in the natural history of the disease, POSTN overexpression in the stroma appears to be highly predictive of the risk of death, a late event that usually follows distant spreading and the loss of hormone dependency. This differential effect has been confirmed by multivariate analysis and suggests that it may be possible to identify different tumor phenotypes which are characterized by an increasing risk of PSA fai- lure and death. In this regard, it is certainly intriguing that the prognostic value of POSTN overexpression in stroma is especially evident in patients whose tumors down- regulate POSTN expression in the epithelial compart- ment, but not in those whose tumors also overexpress the protein in the epithelial component (Figure 5). Discussion In fact, in the Swiss study, median IRS values were used as cut-off points to dichotomize the tumors into a “POSTN low” and “POSTN high” population, while we found that the cut off score that best defined patient risk was the 75th per- centile. In Tsunoda's study [15], IHC analysis only took into consideration the quantitative expression of POSTN (positive: at least >5% of staining cells), without evaluating staining intensity. differences concerning staining evaluation. In fact, we and Tischler used an immune score (IRS) obtained by mul- tiplying the intensity of staining by the percentage of stained cells. Notably, comparable median IRS values were obtained by us and by the Swiss colleagues. However, the results obtained in our two studies are not comparable since patients were analyzed after arbitrarily grouping them in different manners. In fact, in the Swiss study, median IRS values were used as cut-off points to dichotomize the tumors into a “POSTN low” and “POSTN high” population, while we found that the cut off score that best defined patient risk was the 75th per- centile. In Tsunoda's study [15], IHC analysis only took The differences in how patients were grouped, i.e. according to their IRS, may have been particularly relevant when POSTN expression was correlated with clinical outcome. Tischler’s study [16] evaluated the correlation between POSTN expression and PSA relapse- free survival. They showed that higher stromal POSTN was significantly associated with shorter PSA-free survival both in the training cohort and in the test cohort set. However, the difference between low and high POSTN subgroups was Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Page 11 of 12 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Author details 1 RCCS S 1IRCCS San Martino University Hospital – IST National Cancer Research Institute and the University of Genoa, Academic Unit of Medical Oncology (UOC Oncologia Medica B), Largo Rosanna Benzi 10, 16132 Genoa, Italy. 2IRCCS San Martino University Hospital – IST National Cancer Research Institute, Pathology Unit, Genoa, Italy. 3University of Genoa, Department of Internal Medicine, School of Medicine, Genoa, Italy. Received: 5 July 2012 Accepted: 17 December 2012 Published: 28 December 2012 Authors’ contributions The phenotype characterized by low POSTN expression in the epithelium and high protein expression in the stroma showed a bleak prognosis, both in terms of PSA-free and overall survival. Acknowledgments Thi k i l This work was partially supported by grants from: the Italian Health Ministry (Grant number 2005 conv. 93; 2005 conv. 65); Italian Education, University and Research Ministry (Grant number: 9806118184 and 2008YFRLC8-004); Liguria Region, Department of Health (Grant number: 563/2009); “Compagnia di San Paolo”, Italy (Grant number: 2009.1271). The authors are indebted to Dr Claudia Casella (Descriptive Epidemiology Unit: Tumor Registry, IRCCS San Martino University Hospital –IST National Cancer Research Institute, Genoa, Italy) for helping in data collection. The authors would also like to thank Prof. Mauro Truini (Pathology Unit, Director; IRCCS San Martino University Hospital - IST National Cancer Research Institute, Genoa, Italy;) and Prof. Giorgio Carmignani (Department of Urology “Luciano Giuliani”, Director; IRCCS San Martino University Hospital – IST National Cancer Research Institute, Genoa, Italy) for their kind and greatly appreciated co-operation. Finally they are indebted to Ms Valerie Perricone for English editing. This work was partially supported by grants from: the Italian Health Ministry (Grant number 2005 conv. 93; 2005 conv. 65); Italian Education, University and Research Ministry (Grant number: 9806118184 and 2008YFRLC8-004); Liguria Region, Department of Health (Grant number: 563/2009); “Compagnia di San Paolo”, Italy (Grant number: 2009.1271). y The authors are indebted to Dr Claudia Casella (Descriptive Epidemiology Unit: Tumor Registry, IRCCS San Martino University Hospital –IST National Cancer Research Institute, Genoa, Italy) for helping in data collection. The authors would also like to thank Prof. Mauro Truini (Pathology Unit, Director; IRCCS San Martino University Hospital - IST National Cancer Research Institute, Genoa, Italy;) and Prof. Giorgio Carmignani (Department of Urology “Luciano Giuliani”, Director; IRCCS San Martino University Hospital – IST National Cancer Research Institute, Genoa, Italy) for their kind and greatly appreciated co-operation. Finally they are indebted to Ms Valerie Perricone for English editing. Conclusions statistically significant only in the training set. This might reflect the different size of the two cohorts and/or the length of follow-up (45 vs 72 months, respectively). However, it might also be associated with the different characteristics of the two study populations. In fact, 46% of the patients making up the training set showed bio- chemical failure compared with 20% of those forming the validation set. No relationship between PSA-free sur- vival and epithelial POSTN expression was reported in this study. Unfortunately, these investigators did not explain the criteria that were adopted to define PSA pro- gression, nor did they attempt to correlate POSTN expression with patients' survival. Our findings should be regarded as merely exploratory and, as such, they should be evaluated with caution. Nonethe- less, they warrant IHC methodological standardization and further validation of the potential usefulness of POSTN as a prognostic marker in larger prospective series. Competing interest The authors declare no competing interest. Competing interest The authors declare no competing interest. Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 6. Boccardo F, Rubagotti A, Carmignani G, Romagnoli A, Nicolò G, Barboro P, Parodi S, Patrone E, Balbi C: Nuclear matrix proteins changes in cancerous prostate tissues and their prognostic value in clinically localized prostate cancer. Prostate 2003, 55:259–64. 6. Boccardo F, Rubagotti A, Carmignani G, Romagnoli A, Nicolò G, Barboro P, Parodi S, Patrone E, Balbi C: Nuclear matrix proteins changes in cancerous prostate tissues and their prognostic value in clinically localized prostate cancer. Prostate 2003, 55:259–64. 7. Ricci F, Rubagotti A, Zinoli L, Mangerii R, Nuzzo PV, Carmignani G, Simonato A, Barboro P, Balbi C, Boccardo F: Prognostic value of nuclear matrix protein expression in localized prostate cancer. J Cancer Res Clin Oncol 2012, 138(8):1379–84. 8. Ruan K, Bao S, Ouyang G: The multifaceted role of periostin in tumorigenesis. Cell Mol Life Sci 2009, 66(14):2219–30. 8. Ruan K, Bao S, Ouyang G: The multifaceted role of periostin in tumorigenesis. Cell Mol Life Sci 2009, 66(14):2219–30. 9. Oshima A, Tanabe H, Yan T, Lowe GN, Glackin CA, Kudo A: A novel mechanism for the regulation of osteoblast differentiation: transcription of periostin, a member of the fasciclin I family, is regulated by the bHLH transcription factor, twist. J Cell Biochem 2002, 86(4):792–804. 9. Oshima A, Tanabe H, Yan T, Lowe GN, Glackin CA, Kudo A: A novel mechanism for the regulation of osteoblast differentiation: transcription of periostin, a member of the fasciclin I family, is regulated by the bHLH transcription factor, twist. J Cell Biochem 2002, 86(4):792–804. 10. Shimazaki M, Nakamura K, Kii I, Kashima T, Amizuka N, Li M, Saito M, Fukuda K, Nishiyama T, Kitajima S, et al: Periostin is essential for cardiac healing after acute myocardial infarction. J Exp Med 2008, 205(2):295–303. 10. Shimazaki M, Nakamura K, Kii I, Kashima T, Amizuka N, Li M, Saito M, Fukuda K, Nishiyama T, Kitajima S, et al: Periostin is essential for cardiac healing after acute myocardial infarction. J Exp Med 2008, 205(2):295–303. 11. Nakazawa T, Nakajima A, Seki N, Okawa A, Kato M, Moriya H, Amizuka N, Einhorn TA, Yamazaki M: Gene expression of periostin in the early stage of fracture healing detected by cDNA microarray analysis. J Orthop Res 2004, 22(3):520–525. 12. Li P, Oparil S, Feng W, Chen YF: Hypoxia-responsive growth factors upregulate periostin and osteopontin expression via distinct signaling pathways in rat pulmonary arterial smooth muscle cells. References 1. Jemal A, Siegel R, Xu J, Ward E: Cancer statistics, 2010. CA Cancer J Clin 2010, 60(5):277–300. 2. Bostwick DG, Grignon DJ, Hammond ME, et al: Prognostic factors in prostate cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000, 124(7):995–1000. 2. Bostwick DG, Grignon DJ, Hammond ME, et al: Prognostic factors in prostate cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 2000, 124(7):995–1000. 3. Humphrey PA, Vollmer RT: Percentage of carcinoma as a measure of prostatic tumor size in radical prostatectomy tissues. Mod Pathol 1997, 10:326–33. Moreover, multi-parametric models showed that the proteomic signature based on the epithelial and stromal expression of POSTN indeed added to the prognostic information provided by the currently available variables, including the Gleason score. It should be stated in this regard that the old Gleason grading system was used and that caution is therefore warranted in data interpretation. 4. Tefilli MV, Gheiler EL, Tiguert R, et al: Should Gleason score 7 prostate cancer be considered a unique grade category? Urology 1999, 53:372–377. 5. Montironi R, Mazzuccheli R, Scarpelli M, Lopez-Beltran A, Fellegara G, Algaba F: Gleason grading of prostate cancer in needle biopsies or radical prostatectomy specimens: contemporary approach, current clinical significance and sources of pathology discrepancies. BJU Int 2005, 95(8):1146–52. Page 12 of 12 Page 12 of 12 Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 J Appl Physiol 2004, 97(4):1550–1558. discussion 1549. 13. Morra L, Moch H: Periostin expression and epithelial-mesenchymal transition in cancer: a review and an update. Virchows Arch 2011, 459(5):465–75. 14. Kim CJ, Yoshioka N, Tambe Y, Kushima R, Okada Y, Inoue H: Periostin is down-regulated in high grade human bladder cancers and suppresses in vitro cell invasiveness and in vivo metastasis of cancer cells. Int J Cancer 2005, 117(1):51–8. 14. Kim CJ, Yoshioka N, Tambe Y, Kushima R, Okada Y, Inoue H: Periostin is down-regulated in high grade human bladder cancers and suppresses in vitro cell invasiveness and in vivo metastasis of cancer cells. Int J Cancer 2005, 117(1):51–8. 15. Tsunoda T, Furusato B, Takashima Y, Ravulapalli S, Dobi A, Srivastava S, McLeod DG, Sesterhenn IA, Ornstein DK, Shirasawa S: The increased expression of periostin during early stages of prostate cancer and advanced stages of cancer stroma. Prostate 2009, 69(13):1398–403. 15. Tsunoda T, Furusato B, Takashima Y, Ravulapalli S, Dobi A, Srivastava S, McLeod DG, Sesterhenn IA, Ornstein DK, Shirasawa S: The increased expression of periostin during early stages of prostate cancer and advanced stages of cancer stroma. Prostate 2009, 69(13):1398–403. 16. Tischler V, Fritzsche FR, Wild PJ, Stephan C, Seifert HH, Riener MO, Hermanns T, Mortezavi A, Gerhardt J, Schraml P, Jung K, Moch H, Soltermann A, Kristiansen G: Periostin is up-regulated in high grade and high stage prostate cancer. BMC Cancer 2010, 10:273. 17. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958, 53:457–481. 18. Peto R, Pike MC, Armitage P, et al: Design and analysis of randomised clinical trials requiring prolonged observation of each patient: I. Introduction and design. Br J Cancer 1976, 34:585–612. Analysis and examples. Br J Cancer 1977;35:1–39. 19. Cox DR: Regression models and life-tables. J R Stat Soc Ser B 1972, 34:187–220. doi:10.1186/1471-2407-12-625 Cite this article as: Nuzzo et al.: Prognostic value of stromal and epithelial periostin expression in human prostate cancer: correlation with clinical pathological features and the risk of biochemical relapse or death. BMC Cancer 2012 12:625. doi:10.1186/1471-2407-12-625 Cite this article as: Nuzzo et al.: Prognostic value of stromal and epithelial periostin expression in human prostate cancer: correlation with clinical pathological features and the risk of biochemical relapse or death. BMC Cancer 2012 12:625. Nuzzo et al. BMC Cancer 2012, 12:625 http://www.biomedcentral.com/1471-2407/12/625 Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color ﬁgure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Submit your next manuscript to BioMed Central and take full advantage of: Submit your next manuscript to BioMed Central and take full advantage of: • Convenient online submission
https://openalex.org/W2063085129	https://europepmc.org/articles/pmc3978045?pdf=render	English	null	An Elite Controller of Picornavirus Infection Targets an Epitope That Is Resistant to Immune Escape	PloS one	2,014	cc-by	11,176	Introduction the oral-fecal route and normal transmission is most often asymptomatic. Occasionally, through unknown mechanisms, virions gain access to the central nervous system (CNS) where the virus readily infects neurons and causes acute encephalitis [8]. Intracerebral infection of mice with the picornavirus TMEV is accompanied by an acute inflammatory response that is not cleared by most strains of mice. Infection often leads to persistence and pathology that mimics the disease multiple sclerosis in humans [8,9]. The major determinant regulating persistence is MHC, specifically the H-2D gene cluster of the MHC class I locus [10]. Among the H-2D alleles, only H-2Db clearly confers resistance to virus infection regardless of genetic background [11,12]. This phenotype is dependent on the generation of an immunodominant CD8+ T-cell response to the viral protein 2 (VP2), specifically the peptide sequence VP2121-130 bound to H-2Db. Although responses to other peptides derived from TMEV have been documented [13], responses to the 10 amino acid sequence VP2121-130 have been shown to comprise up to 70% of the CD8+ T-cell response after infection and is the primary response needed for virus clearance [14]. The quality of the immunodominant response observed after TMEV infection is rare among known CD8+ T-cell responses in that it is sufficient to control virus infection. The presence of only a rare CD8+ T-cell response suggests that TMEV has evolved to avoid detection by the CD8+ T-cell response, Interaction with pathogens has been proposed to be the main driver of human immunity as well as the evolution of infectious diseases and the agents that cause them [1,2,3]. Infection with disease causing pathogens has been important for shaping the genetics of human populations and the most prominent element is the diversity of major histocompatibility genes and their role in recognizing emerging infectious diseases. Occasionally, pathogens emerge which are not effectively targeted by the immune response of most individuals, leading to overwhelming persistent infection often with significant pathologic consequences. Rare individuals are able to control these infections [4,5,6]. Fortuitously, the vast diversity of MHC genes in the population favors the presence of rare alleles that are capable of recognizing and clearing these emerging pathogens. Therefore, understanding the mechanisms leading to this super resistance phenotype is critically important. We have been examining the underlying basis of super resistance in a mouse model of picornavirus infection [7]. Michael P. Bell1, Danielle N. Renner3, Aaron J. Johnson1,2, Kevin D. Pavelko1* 1 Department of Immunology, Mayo Clinic, Rochester, Minnesota, United States of America, 2 Department of Neurology, Mayo Graduate School, Mayo Clinic, Rochester, Minnesota, United States of America, 3 Neurobiology of Disease Program, Mayo Clinic, Rochester, Minnesota, United States of America Abstract This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Copyright: 2014 Bell et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by grants from the National Institutes of Health (5R01CA104996-09). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: pavelko.kevin@mayo.edu Abstract The emergence of novel viral pathogens can lead to devastating consequences in the infected population. However, on occasion, rare hyper-responsive elite controllers are able to mount a protective primary response to infection and clear the new pathogen. Factors distinguishing elite controllers from other members of the population are not completely understood. We have been using Theiler’s murine encephalomyelitis as a model of primary infection in mice and clearance of the virus is limited to one MHC genotype capable of generating a protective response to a single viral peptide VP2121-130. The genetics of host susceptibility to TMEV, a natural mouse pathogen, has been studied extensively and non-protective CD8 responses to other peptides have been documented, however, little is known why the protective response to infection focuses on the VP2121-130 peptide. To study this question, we have generated TMEV mutants that encode for mutations within the VP2121-130 peptide. We find that very few of mutants are able to assemble and infect in vitro. These mutations are not related to virus RNA structure since non-coding mutations do not interfere with assembly. In the rare event when functional VP2121-130 mutant viruses did emerge, they were attenuated to some level or retained the ability to develop an immune response to the wild-type VP2121-130 sequence, demonstrating that the virus is incapable of escaping the protective response. These findings advance our understanding of how characteristics of the host immune response and an infectious agent can interact to lead to the appearance of rare super controllers in a population. Furthermore, the immutable nature of the viral antigen highlights the importance of choosing appropriate vaccine antigens and has implications for the development of agents that are able to generate protective CD8 T-cell responses. Citation: Bell MP, Renner DN, Johnson AJ, Pavelko KD (2014) An Elite Controller of Picornavirus Infection Targets an Epitope That Is Resistant to Immune Escape. PLoS ONE 9(4): e94332. doi:10.1371/journal.pone.0094332 Editor: Ralph Tripp University of Georgia United States of America Citation: Bell MP, Renner DN, Johnson AJ, Pavelko KD (2014) An Elite Controller of Picornavirus Infection Targets an Epitope That Is Resistant to Immune Escape. PLoS ONE 9(4): e94332. doi:10.1371/journal.pone.0094332 Editor: Ralph Tripp, University of Georgia, United States of America Received January 22, 2014; Accepted March 13, 2014; Published April 7, 2014 Copyright: 2014 Bell et al. Introduction These elite controllers represent a unique population that is able to recognize and eliminate an emerging pathogen while allowing survivors to pass on genes necessary for disease resistance. The mouse pathogen Theiler’s murine encephalomyelitis virus (TMEV) is a natural pathogen in mice that is transmitted through PLOS ONE \| www.plosone.org 1 April 2014 \| Volume 9 \| Issue 4 \| e94332 Picornavirus Epitope Resistant to Immune Escape Table 1. VP2121-130 mutants. Model Virus PDB ID VP2 Mutation VP2121-130 Inf. Sequence Theiler’s murine encephalomyelitis virus 1TME TMEV VP2121-130 F H A G S L L V F M + TTT CAC GCC GGC TCT CTT CTT GTT TTC ATG Seneca Valley Virus-001 3CJI SVV126-135 F H Q G A L L V A M 2 TTT CAC CAA GGC GCT CTT CTT GTT GCC ATG A123Q F H Q G S L L V F M 2 TTT CAC CAA GGC TCT CTT CTT GTT TTC ATG S125A F H A G A L L V F M + TTT CAC GCC GGC GCC CTT CTT GTT TTC ATG L129A F H A G S L L V A M 2 TTT CAC GCC GGC TCT CTT CTT GTT GCC CTA Coxsackievirus A21 1Z7S CSV117-126 F H Q G A L G V F L 2 TTT CAC CAA GGC GCC CTT GGT GTT TTC CTA L127G F H A G S L G V F L 2 TTT CAC GCC GGC TCT CTT GGT GTT TTC CTA M130L F H A G S L L V F L + TTT CAC GCC GGC TCT CTT CTT GTT TTC CTA Seneca Valley Virus-001/Coxsackie A21 S125A/M130L F H A G A L L V F L 2 TTT CAC GCC GGC GCC CTT CTT GTT TTC CTA Human Rhinovirus 16 1AYN S125T F H A G T L L V F M 2 TTT CAC GCC GGC ACT CTT CTT GTT TTC ATG Echovirus 1 1EV1 S125C F H A G C L L V F M 2 TTT CAC GCC GGC TGT CTT CTT GTT TTC ATG TMEV VP2121-30 Alt. Introduction F H A G S L L V F M + TTC CAT GCA GGA AGC TTA TTG GTC TTT ATG doi:10 1371/journal pone 0094332 t001 doi:10.1371/journal.pone.0094332.t001 several amino acid residues of the VP2121-130 epitope of TMEV to identify modifications that could potentially escape the immuno- dominant response normally observed in H-2Db mice. Our hypothesis is that modifications to the VP2121-130 sequence would evade immunodominant T-cell responses and allow virus persis- tence. However, we find that rarely do any modifications within this region generate functional virions. Through homology mapping of related picornavirus species we have identified two TMEV mutations within the VP2121-130 region that has a diminished immunodominant response. This diminished response is not sufficient to promote virus persistence, demonstrating that a diminution of the VP2121-130 CD8+ T-cell response continues to support viral clearance. These results provide important details regarding the characteristics of viral antigens that drive protective CD8+ T-cell responses to primary infections and will provide insight for the design of vaccines that are able to elicit these desired responses. through mutations that promote immune escape and detection by most MHC class I alleles. However, a protective response does exist in the mouse population, suggesting that virus escape from detection by the H-2Db allele may not be possible and that this sequence serves a critical function for the virus and is consequently maintained in the TMEV genome. Several potential outcomes to a change in a viral epitope recognized by an immunodominant CD8+ T-cell response could be realized. The acquisition of novel mutations within a virus epitope may ultimately result in a virus that is no longer functional or is unable to assemble due to structural or functional constraints. Consequently, the immune system will not recognize or detect these viruses due to low fidelity and minimized antigen load. Alternatively, novel mutations will be tolerated and generate functional virions. These new viruses may acquire mutations that alter MHC binding by either increasing or decreasing their affinity for MHC or alter interactions with CD8+ T-cells and MHC bound to peptide. Nonetheless, alteration of a single amino acid has the potential to either dampen a response or enhance the response to viral peptides. However, as an immune evasion strategy mutations that dampen the immune response are more likely to ensure virus survival and persistence. PLOS ONE \| www.plosone.org Silent RNA mutations at the coding sequence for VP2121-130 do not affect virus assembly Although we were able to generate two VP2121-130 mutant viruses using this approach a majority of the virus vectors we synthesized failed to generate infectious virions. To rule out the role of RNA structure in this region, we generated a codon alternative version of TMEV. This mutant contained 13 silent nucleotide changes within the 30 nucleotide coding sequence for the VP2121-130 epitope (Figure 2A) and changed 9 of the 10 codons for this epitope. Three replicates of wild-type and VP2121-130 codon alternative viruses were generated. Viruses recovered using this approach were sequenced to determine whether the mutations were maintained in the genome. The sequences at VP2121-130 shared complete identity to the plasmid used to generate both of these viruses (Figure 2B). Further, both viruses could be detected by Western blot in supernatants derived from cells transfected 7 days prior to collection (Figure 2C). Although modifications to picornavirus RNA at other sites within VP2 can influence the efficiency of translation and replication [17] silent mutations that encode the VP2121-130 do not affect virus assembly or replications, demonstrating that the inability to generate mutations in this region is due to the amino acid structure rather than the encoding RNA. To verify the production of viral proteins and proper cleavage and assembly of the VP2 mutant viruses we transfected mutant vectors into cells and then analyzed whole cell lysates and supernatants for the presence of viral capsid proteins after 3 and 7 days post-transfection by Western blot using polyclonal serum that recognizes both VP1 and VP2 protein. Using wild-type TMEV, viral proteins are typically detected in the whole cell lysates or supernatant by day 7 (Figure 1B). After 3 days of transfection only the S125A and M130L mutant had detectable VP1 and VP2 proteins demonstrating that the viral polyprotein was processed and cleaved to the appropriate size and that a functionally mature viral protease was generated (Figure 1C). Further, the larger VP0 protein was only detected with the whole cell lysate at 3 days demonstrating only immature virus is associated with cell lysate consistent with other picornavirus strains [15]. By day 7, only mature viral capsids were detected with both mutants killing the remainder of the cells as was observed by the cytopathic effect (CPE) seen in these cultures (Figure 1C). Silent RNA mutations at the coding sequence for VP2121-130 do not affect virus assembly Further, only superna- tants derived from the wild-type TMEV, S125A and M130L viruses could induce CPE on susceptible cell lines as was demonstrated through the generation of high titer virus superna- tants after serial passage. Mutants S125A and M130L generate infectious virus Mutants S125A and M130L generate infectious virus The vectors encoding mutations within the VP2121-130 epitope were transfected into cells that support the generation and propagation of TMEV. RNA isolated 24 hours after transfection revealed that all of the vectors expressed TMEV encoding RNA as well as the vector encoded neomycin resistance gene. We found that only two of the mutant TMEV strains had particularly high expression of VP2 RNA suggesting that these mutants were able to generate additional TMEV RNA, potentially through use of their own functioning viral polymerase, suggesting viral reconstitution and assembly using this method (Figure 1A). Further, co- expressing both of these mutations did not lead to the enhanced RNA expression observed with the individual mutations, demon- strating that intermolecular interactions among amino acids in this area may be important for the generation of infectious virions. Of interest was the S125A mutant, since its central location within the peptide was most likely to influence interactions with the T-cell receptor when bound to H-2Db, we identified two other homologous virus sequences with changes at this position. We found that that human rhinovirus 16 and echovirus 1 encode threonine and cysteine at position 125 of this peptide. Vectors encoding these mutations (Table 1) however did not generate high RNA levels like that seen with wild-type TMEV (Figure 1A). Results Development of TMEV VP2121-130 mutants To identify mutations that can be introduced into the immunodominant TMEV sequence FHAGSLLVFM at positions 121-130 of VP2, we used homology mapping to identify similar linear and conformational epitopes derived from related viruses. In this manuscript, we sought to identify TMEV mutations that could modulate CD8+ T-cell immunodominance through manip- ulation of viral genetics rather than host genetics. We mutagenized April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org 2 Picornavirus Epitope Resistant to Immune Escape prediction tool I-mutant 2.0 [16] to predict mutations within this region that would promote stabilization of the folded VP2 protein. The use of this algorithm would have predicted the stability of the M130L mutation that we found using the homologous sequence from CSV. The predicted DDG most likely to stabilize the virus structure was leucine at position 130 with a net change in Gibbs free energy of -0.29 (Table 2). Using -0.29 as a threshold, we identified the most likely stable changes throughout this peptide fragment. We found that the serine at position 125 was predicted to be most amenable to change, with 10 different amino acids potentially able to stabilize this structure including the alanine substitution used previously (Table 2). We generated the remain- ing 9 mutants and transfected them into BHK cells to determine whether new variants would emerge from plasmids expressing these mutations. Only the wild-type and S125A mutant generated detectable VP1 and VP2 protein as assessed by western blot. The remaining mutants did not generate detectable virus proteins after transfection in vitro for 7 days (Figure 1D), demonstrating the inability to generate assembled virus products using cDNA encoding these mutations. Further, we generated the remaining amino acid substitutions at VP2125 and no productive virus assembly was observed (data not shown), demonstrating that interactions beyond those employed by this algorhithm govern the ability to generate infectious virus. Our search identified 51 known homologous structures from several picornavirus species including two strains of TMEV and of Mengo virus, which were identical at VP2121-130. The fourth hit on our search was Seneca Valley Virus-001 (SVV). SVV exhibited a 44% sequence identity when comparing chain C of SVV to VP2 of TMEV. Three amino acid differences were observed in the sequence homologous to TMEV-VP2121-130, a glutamine at position 123, an alanine at position 125 and an alanine at position 129 (Table 1). Results The next virus with the highest sequence identity to TMEV-VP2 was coxsackievirus A21 (CSV) (36%). Four differ- ences within the VP2121-130 sequence were identified, including two unique amino acids, a glycine at position 127 and a leucine at position 130. Having identified five changes to the homologous VP2 protein, we generated TMEV mutant vectors that encode the sequences identified in SVV and CSV en masse or as individual amino acid changes (Table 1). Our search identified 51 known homologous structures from several picornavirus species including two strains of TMEV and of Mengo virus, which were identical at VP2121-130. The fourth hit on our search was Seneca Valley Virus-001 (SVV). SVV exhibited a 44% sequence identity when comparing chain C of SVV to VP2 of TMEV. Three amino acid differences were observed in the sequence homologous to TMEV-VP2121-130, a glutamine at position 123, an alanine at position 125 and an alanine at position 129 (Table 1). The next virus with the highest sequence identity to TMEV-VP2 was coxsackievirus A21 (CSV) (36%). Four differ- ences within the VP2121-130 sequence were identified, including two unique amino acids, a glycine at position 127 and a leucine at position 130. Having identified five changes to the homologous VP2 protein, we generated TMEV mutant vectors that encode the sequences identified in SVV and CSV en masse or as individual amino acid changes (Table 1). In vitro and in vivo replication of VP2 S125A and VP2 M130L mutant viruses After gating on CD45+ lymphocytes, we found that mutant viruses induced a varied CD8+ T-cell infiltration (Figure 4A), as a percentage of total CD45+ cells the S125A mutant was equivalent to the wild-type TMEV and the M130L mutant had a reduced proportion of CD8+ T-cells compared to the S125A mutant and the wild-type virus. However, quantita- tively the absolute numbers of CD8+ T-cells that infiltrate the CNS after infection were only reduced with the S125A mutant compared to wild-type TMEV (Figure 4A). CD8+ T-cells using the S125A, M130L and wild-type TMEV viruses. After gating on CD45+ lymphocytes, we found that mutant viruses induced a varied CD8+ T-cell infiltration (Figure 4A), as a percentage of total CD45+ cells the S125A mutant was equivalent to the wild-type TMEV and the M130L mutant had a reduced proportion of CD8+ T-cells compared to the S125A mutant and the wild-type virus. However, quantita- tively the absolute numbers of CD8+ T-cells that infiltrate the CNS after infection were only reduced with the S125A mutant compared to wild-type TMEV (Figure 4A). immunodeficient mice. Similar to its growth in vitro, the S125A virus had reduced virulence as determined by weight loss (Figure 3C) and survival (Figure 3D). Further, infected CNS tissues were recovered as mice succumbed to infection and were assessed by plaque assay to determine viral load. We found that virus titers at death for both of the VP2 mutant viruses were significantly lower than that reached by wild-type TMEV (Figure 3E). This is in contrast to the in vitro response, the RAG-/- survival after M130L infection induced morbidity at a lower titer than the wild-type virus, suggesting that the lower level of recovered infectious virus does not necessarily correlate with morbidity at these high levels of infection. Since immune escape is predicated on the avoidance of an immunodominant CD8+ T-cell response we asked whether infection with VP2 mutant viruses would induce an immune response that overlaps with the wild type VP2121-130 response. Six days after intracranial infection with the S125A, M130L and wild- type TMEV viruses, lymphocytes isolated from the central nervous system were stained with tetramer to detect H-2Db/VP2121-130 specific CD8+ T-cells. In vitro and in vivo replication of VP2 S125A and VP2 M130L mutant viruses This response was reduced by 48% with the S125A mutant and 46% by the M130L mutant compared to wild- type TMEV (Figure 4B), demonstrating that the VP2 mutations altered the quality and quantity of the CD8+ T-cell response after infection and that the response induced with VP2 mutants overlaps with that induced with wild-type sequence. In vitro and in vivo replication of VP2 S125A and VP2 M130L mutant viruses We found that two mutations within the immunodominant viral epitope from TMEV could generate functional virus particles. To further characterize these viruses we infected BHK cells to further understand the virus’ ability to lyse and kill in vitro. Virus stocks of the S125A and M130L viruses both grew to high titers using in vitro culture techniques. However, we found that the plaque size of the S125A mutant virus was significantly smaller compared to the wild-type virus or to the M130L virus (Figure 3A) consistent with its reduced ability to kill cells in vitro (Figure 3B). Since these viruses were able to replicate in vitro, we infected RAG deficient mice with these viruses to determine whether the mutations alter in vivo growth characteristics or lethality toward To further investigate the potential for changes at positions within the VP2121-130 sequence we used the protein stability PLOS ONE \| www.plosone.org April 2014 \| Volume 9 \| Issue 4 \| e94332 April 2014 \| Volume 9 \| Issue 4 \| e94332 3 Picornavirus Epitope Resistant to Immune Escape PLOS ONE \| www.plosone.org 4 April 2014 \| Volume 9 \| Issue 4 \| e94332 Picornavirus Epitope Resistant to Immune Escape Picornavirus Epitope Resistant to Immune Escape Figure 1. Development of TMEV VP2121-130 mutant viruses. (A) Real-time RT-PCR expression analysis of viral VP2 and plasmid neomycin phosphotransferase from BHK cells transfected with TMEV VP2121-130 mutant cDNA. (B) Western blot analysis of whole cell lysates and viral supernatants for TMEV viral proteins from cells transfected with wild-type TMEV-DA cDNA for 3 and 7 days. (C) Western blot of whole cell lysates and supernatants from cells transfected with TMEV VP2121-130 mutant plasmid cDNA (D) Western blot of supernatants derived from cells transfected with TMEV VP2125 mutant plasmid cDNA. doi:10.1371/journal.pone.0094332.g001 Figure 1. Development of TMEV VP2121-130 mutant viruses. (A) Real-time RT-PCR expression analysis of viral VP2 and plasmid neomycin phosphotransferase from BHK cells transfected with TMEV VP2121-130 mutant cDNA. (B) Western blot analysis of whole cell lysates and viral supernatants for TMEV viral proteins from cells transfected with wild-type TMEV-DA cDNA for 3 and 7 days. (C) Western blot of whole cell lysates and supernatants from cells transfected with TMEV VP2121-130 mutant plasmid cDNA (D) Western blot of supernatants derived from cells transfected with TMEV VP2125 mutant plasmid cDNA. doi:10.1371/journal.pone.0094332.g001 CD8+ T-cells using the S125A, M130L and wild-type TMEV viruses. April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org 5 April 2014 \| Volume 9 \| Issue 4 \| e94332 Figure 2. Altered VP2121-130 codon structure does not influence virus fidelity. (A) Thirteen silent nucleotide substitutions representing changes to 9 of the 10 codons of VP2121-130 were introduced into the TMEV-DA plasmid cDNA by site-directed mutagenesis. (B) Sequence verification of codon alternate VP2121-130 virus recovered from infected BHK cells. (C) Western blot analysis of virus supernatant recovered from cells infected with codon alternate TMEV-VP2121-130. doi:10.1371/journal.pone.0094332.g002 Figure 2. Altered VP2121-130 codon structure does not influence virus fidelity. (A) Thirteen silent nucleotide substitutions representing changes to 9 of the 10 codons of VP2121-130 were introduced into the TMEV-DA plasmid cDNA by site-directed mutagenesis. (B) Sequence verification of codon alternate VP2121-130 virus recovered from infected BHK cells. (C) Western blot analysis of virus supernatant recovered from cells infected with codon alternate TMEV-VP2121-130. doi:10.1371/journal.pone.0094332.g002 infected mice that normally fail to control infection with wild-type TMEV, S125A and M130L virus and assessed virus transcript levels after 30 days of infection. As suspected wild-type virus was detected in the CNS with the highest levels being detected in the spinal cord rather than the brain, consistent with susceptibility to chronic infection and demyelinating disease. Both VP2 mutant strains were detectable in the CNS after infection; however the M130L mutant was detected at lower levels compared to the S125A mutant or to wild-type TMEV, demonstrating that the ability of this mutant to persist was attenuated in FVB mice (Figure 4D). similarity in virus replication observed in vitro and in vivo. Attempts to determine the precise mechanism were impeded by our inability to generate properly folded VP2-M130L/H-2Db tetramers after multiple attempts (data not shown), suggesting that the M130L mutant peptide could not sufficiently stabilize H-2Db in a tetramer folding reaction. In contrast, the VP2-S125A peptide with a mutation outside of the major anchor residues was able to properly fold in vitro. This reagent revealed detectable VP2- S125A specific CNS infiltrating CD8+ T-cells after infection with the S125A mutant virus (2468%, n = 4), suggesting that the S125A mutant peptide stabilizes MHC I in vivo and activates CD8+ T-cells specific to this peptide, although to a lesser degree than wild-type infection. Since our semi-quantitative RT-PCR strategy could detect all three strains of virus, even potential VP2121-130 revertants, we amplified and sequenced VP2 segments corresponding to the regions containing the VP2 mutations. Altered CD8 T-cell immunity and viral persistence after infection with VP2121-130 mutants Previously, we had shown that a perturbation of the VP2121-130 immunodominant response through genetic deletion decreased the number of CD8+ T-cells that entered the brain after TMEV infection [18]. Since we were able to identify two VP2121-130 mutant viruses that are able to replicate in mice we asked whether viral mutants would generate an altered CD8+ T-cell response in immunocompetent mice that normally control infection. Six days after intracranial infection, lymphocytes isolated from the central nervous system were analyzed for the presence and quantity of We found that the CD8+ T-cell response to the M130L mutant was not as robust as the response to wild-type virus, in spite of the Table 2. I-Mutant predicted DDG values for TMEV VP2121-130 amino acid substitutions. Table 2. I-Mutant predicted DDG values for TMEV VP2121-130 amino acid substitutions. 121 122 123 124 125 126 127 128 129 130 Amino Acid F H A G S L L V F M V 22.02 20.49 0.01 21.14 1.03 21.09 20.71 - 21.71 20.91 L 21.47 20.20 20.16 20.15 0.97 - - 21.12 21.24 20.29 I 20.70 20.88 20.18 20.42 1.00 20.86 20.38 20.54 20.80 20.71 M 20.97 20.96 21.04 21.06 0.33 20.95 20.48 22.16 21.55 - F - 20.54 20.26 21.00 0.50 20.76 20.42 22.58 - 20.88 W 0.19 20.43 20.33 20.19 1.11 20.67 20.21 22.03 21.40 20.82 Y 20.16 20.17 20.90 21.24 20.46 21.91 21.02 23.41 21.77 21.24 G 22.07 21.58 22.14 2 21.56 24.31 23.36 25.38 24.29 23.48 A 21.95 21.75 - 21.76 20.20 23.49 22.56 24.05 23.79 22.71 P 21.98 21.38 21.44 22.00 20.69 22.21 21.72 23.37 22.46 21.60 S 21.36 21.14 20.56 22.00 - 22.79 22.02 24.04 23.16 22.03 T 21.07 21.03 20.91 21.87 0.25 23.03 22.59 23.23 23.23 22.02 C 20.80 0.14 20.97 21.38 20.45 22.37 22.16 22.13 22.50 21.61 H 21.51 - 21.45 22.06 20.87 22.63 22.26 23.73 23.28 22.06 R 21.23 21.55 21.45 21.72 20.54 23.01 22.14 23.75 22.83 21.99 K 21.70 21.79 22.02 22.25 20.57 22.86 21.96 24.02 22.70 21.93 Q 21.20 21.36 21.21 21.46 20.20 22.39 21.79 23.19 22.52 21.83 E 20.68 20.97 21.02 20.94 20.17 22.37 21.61 23.09 22.18 21.51 N 21.63 21.77 21.21 21.91 20.44 22.69 21.98 23.99 22.67 21.69 D 21.01 21.36 20.67 21.75 20.50 22.98 22.20 23.42 22.66 21.72 doi:10.1371/journal.pone.0094332.t002 April 2014 \| Volume 9 \| Issue 4 \| e94332 5 Picornavirus Epitope Resistant to Immune Escape Picornavirus Epitope Resistant to Immune Escape All RNA’s detected in spinal cord tissues demonstrated maintenance of the mutant VP2 strains after 30 days of infection (data not shown), demonstrating the stability of these mutants in vivo. Since the immunodominant response is critical for determining viral control after acute infection, we analyzed spinal cord and brain homogenates from 30 day infected mice for the presence of persisting virus to determine whether these mice could escape the immunodominant CD8+ T-cell response and convert to a susceptible non-controlling phenotype [19]. Mice infected with S125A, M130L or with wild-type TMEV all exhibited low levels of viral transcripts in both the spinal cord and brain, suggesting that the mutations in the VP2121-130 epitope did not dramatically affect the ability of the host to clear these viruses (Figure 4C). Clearance of the VP2 mutant viruses by resistant mice was similar to the wild type, however the mechanism of clearance may be unique since attenuating mutations within the VP2 region may promote clearance exclusively through innate immune mechanisms rather than through cytotoxic T-cell responses [20]. To test this we We analyzed the TMEV specific IgG in C57BL/6 and FVB mice from these animals to determine whether the VP2 mutations altered the ability to mount an antibody response against the wild- type virus. We found that infection with mutant viruses induced anti-TMEV responses and neutralizing titers against wild-type virus that were equivalent to the response induced with wild-type (Figure 4C and D), suggesting that the mutations within the VP2121-130 region did not alter the neutralizing antibody response to TMEV. This evidence supports the hypothesis that the modification of residues within the VP2121-130 region does not April 2014 \| Volume 9 \| Issue 4 \| e94332 April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org 6 Picornavirus Epitope Resistant to Immune Escape Picornavirus Epitope Resistant to Immune Escape weight after infection with TMEV-DA-wt, S125A and M130L (Aver- ge+STD) (* significant by Two-way ANOVA). (D) Survival analysis of RAG deficient mice from 3C after infection with wild-type and mutant viruses (* significant by Kaplan-Meier Log-Rank test). (E) Virus titer of viruses recovered from CNS tissues at the completion of analyses in C and D (* significant by ANOVA excluding animals that survived viral infection). Figure 3. In vitro and in vivo growth of VP2-SD125A and VP2- MD130L mutant viruses. (A) Plaque diameter of TMEV-DA wild-type compared to S125A and M130L mutant virus (* significant by ANOVA). (B) Virus killing of BHK cells as measured by MTT assay at titrating concentrations of virus (* significant by Two-way ANOVA). (C) Body mass of RAG-deficient mice at specified times after infection with TMEV- wt, S125A and M130L viruses. (C bottom) Percent change in body appreciably modulate the structures that are recognized by anti- TMEV antibodies. appreciably modulate the structures that are recognized by anti- TMEV antibodies. The in vivo immunodominant response to wild-type TMEV is reactive to the mutant VP2 peptides S125A and M130L We have previously used VP2121-130 peptide depletion in H-2Db mice to block the generation of the immunodominant CD8+ T- cell response to TMEV [19]. To determine whether the response to wild-type TMEV induces a CD8+ T-cell response that overlaps with the putative MHC class I epitopes derived from the S125A or the M130L mutant viruses we used this strategy to probe the lymphocyte response to TMEV CNS infection. After peptide administration of a control peptide the immunodominant CD8+ T cell response to VP2121-130 comprises 50% of the total CD8+ T- cell response in the CNS and wild-type VP2121-130 peptide depletes this response by 92% (Figure 5A). The use of the S125A peptide reduces the response to wild-type virus by 30% and the M130L peptide depletes the response by 22% (Figure 5A and 5B) and decreases the overall number of VP2121 reactive CD8+ T-cells when compared to the no depletion control (Figure 5B). Statistical significance was not reached using One-way ANOVA when comparing absolute numbers of VP2121-130 specific CD8+ T-cells to E749 specific depletion; therefore due to procedural rules comparisons between the individual VP2 mutant peptides were not made. Nonetheless, each of the mutant VP2 peptides is able to modulate the quality of the VP2121 specific CD8+ T-cell response to intracerebral infection with wild-type virus. Picornavirus Epitope Resistant to Immune Escape Picornavirus Epitope Resistant to Immune Escape Figure 4. CD8+ T-cell response and virus RNA levels in resistant and susceptible mice after intracranial infection with TMEV-wt, S125A and M130L viruses. (A) CD45+ cells isolated from the central nervous system of VP2 mutant infected C57BL/6 mice were analyzed for the presence of CD8+ cells. (a significant compared to M130L by ANOVA, * significant compared to TMEV-wt by ANOVA). Figure is an experiment replicated 3 times with 3-5 animals per group. (B) CNS infiltrating lymphocytes were stained with H-2Db/VP2121 or H-2Db/E749 tetramers and analyzed by flow cytometry to determine the percent of the CD8+ T-cell population that is positive for the immunodominant VP2121 epitope (* significant compared to S125A and M130L by ANOVA). A representative example of 2 experiments using 3-5 animals per group. (C) Semi-quantitative RT-PCR analysis of RNA isolated by C57BL/6 mice infected with TMEV-wt, S125A or M130L mutants (n = 3/group). No significant differences in viral transcripts were detected between the groups. IgG specific responses to wild- type TMEV in C57BL/6 mice infected for 30 days. No significant differences were detected by ELISA between S125, M130L or pciDA (wild-type TMEV) (D) The same RT-PCR analysis in C on the susceptible strain FVB (* significant compared to TMEV-wild type by ANOVA, n = 5/ group). TMEV-specific IgG response and antibody neutralization to wild- type TMEV in FVB mice infected for 30 days. No significant differences were detected by ELISA or neutralization assay between S125, M130L or pciDA (wild-type TMEV). vaccine potential by preventing the emergence of antigen loss variants. The goal of this work was to identify virus mutations that could be used to escape a CD8+ T-cell response and to better understand the characteristics of strong virus antigens that are Figure 4. CD8+ T-cell response and virus RNA levels in resistant and susceptible mice after intracranial infection with TMEV-wt, S125A and M130L viruses. (A) CD45+ cells isolated from the central nervous system of VP2 mutant infected C57BL/6 mice were analyzed for the presence of CD8+ cells. (a significant compared to M130L by ANOVA, * significant compared to TMEV-wt by ANOVA). Figure is an experiment replicated 3 times with 3-5 animals per group. (B) CNS infiltrating lymphocytes were stained with H-2Db/VP2121 or H-2Db/E749 tetramers and analyzed by flow cytometry to determine the percent of the CD8+ T-cell population that is positive for the immunodominant ANOVA). Discussion Viral interactions with strong CD8+ T-cell responses have been shown to drive the diversity of the RNA viruses HIV-1 and hepatitis C virus [21,22]. Of interest, the immunodominant CD8+ T-cell response to the picornavirus TMEV targets known sequences within the VP2 capsid protein, however no CD8+ T- cell escape mutations have been documented to this response. This protein has higher sequence homology among members of the cardiovirus genus when compared to capsid protein VP1 [23], which is less conserved and has more documented functional mutations [24,25,26], suggesting it has a greater potential for immune escape. However, the forces that drive these mutations may differ from those that drive variation in other viral proteins. The B-cell response targets the outer surfaces of most viruses, consistent with this, the primary neutralizing antibody response against TMEV targets surface exposed residues of VP1 [27,28]. In contrast, the protective CD8+ T cell response targets a sequence in VP2 that contains only a partial surface exposed loop and is the major constituent of a b strand buried within the structure of the VP2 protein [29], demonstrating that the CD8+ T-cell response targets a sequence that is likely to be important for protein-protein interactions that affect virus assembly. Although prediction algorithms for the selection of MHC binding peptides are based on the binding of peptide fragments to MHC, our findings suggest that the context of the peptide within the unprocessed antigen or within an assembled virus should also be considered and that highly conserved amino acid stretches may provide enhanced Figure 3. In vitro and in vivo growth of VP2-SD125A and VP2- MD130L mutant viruses. (A) Plaque diameter of TMEV-DA wild-type compared to S125A and M130L mutant virus (* significant by ANOVA). (B) Virus killing of BHK cells as measured by MTT assay at titrating concentrations of virus (* significant by Two-way ANOVA). (C) Body mass of RAG-deficient mice at specified times after infection with TMEV- wt, S125A and M130L viruses. (C bottom) Percent change in body PLOS ONE \| www.plosone.org April 2014 \| Volume 9 \| Issue 4 \| e94332 7 Picornavirus Epitope Resistant to Immune Escape A representative example of 2 experiments using 3-5 animals per group. (C) Semi-quantitative RT-PCR analysis of RNA isolated by C57BL/6 mice infected with TMEV-wt, S125A or M130L mutants (n = 3/group). No significant differences in viral transcripts were detected between the groups. IgG specific responses to wild- type TMEV in C57BL/6 mice infected for 30 days. No significant differences were detected by ELISA between S125, M130L or pciDA (wild-type TMEV) (D) The same RT-PCR analysis in C on the susceptible strain FVB (* significant compared to TMEV-wild type by ANOVA, n = 5/ group). TMEV-specific IgG response and antibody neutralization to wild- type TMEV in FVB mice infected for 30 days. No significant differences were detected by ELISA or neutralization assay between S125, M130L or pciDA (wild-type TMEV). doi:10.1371/journal.pone.0094332.g004 Figure 5. Peptide depletion reveals an overlap in the wild-type VP2121-130 CD8+ T cell response after infection with S125A and M130L mutant viruses. (A) Mice were pre-depleted with control E749, wild-type VP2121-130, VP2-S125A and VP2-M130L peptide prior to intracranial infection with wild-type virus. CNS infiltrating lymphocytes were analyzed by flow cytometry to determine the percentage and number of CD8+ T-cells specific for the immunodominant epitope after peptide depletion. Data are representative examples of three individual mice per group. (B) Percent and absolute number of H-2Db/VP2121+ CD8 T-cells derived from infected CNS tissue including a no peptide group (* significant by ANOVA). doi:10.1371/journal.pone.0094332.g005 \| \| \| \| y doi:10.1371/journal.pone.0094332.g004 vaccine potential by preventing the emergence of antigen loss variants. The goal of this work was to identify virus mutations that could be used to escape a CD8+ T-cell response and to better understand the characteristics of strong virus antigens that are Figure 5. Peptide depletion reveals an overlap in the wild-type VP2121-130 CD8+ T cell response after infection with S125A and M130L mutant viruses. (A) Mice were pre-depleted with control E749, wild-type VP2121-130, VP2-S125A and VP2-M130L peptide prior to intracranial infection with wild-type virus. CNS infiltrating lymphocytes were analyzed by flow cytometry to determine the percentage and number of CD8+ T-cells specific for the immunodominant epitope after peptide depletion. Data are representative examples of three individual mice per group. (B) Percent and absolute number of H-2Db/VP2121+ CD8 T-cells derived from infected CNS tissue including a no peptide group (* significant by ANOVA). doi:10.1371/journal.pone.0094332.g005 Figure 4. Picornavirus Epitope Resistant to Immune Escape Further, the use of picornavirus vectors may provide an advantage in that their relatively small size may allow for the rational design of escape vectors that can be used to efficiently target desired immune responses. Dominant CD8+ T-cell responses occur after infection or immunization with a variety of antigens, including pathogens that vary considerably in size, nucleic acid composition as well as cellular tropism, yet the immune system tends to focus on a short amino acid sequence derived from these antigens [39]. Although peptide binding and stabilization of MHC class I is necessary for the development of this response, it is not necessarily a predictor of an individual’s ability to control infection, since dominant CD8+ T-cell responses can occur in the absence of viral clearance [40,41,42]. The distinction between immunodominance and the ability to clear an infection is important, since targeting virus antigens subjects them to immune selection and pressures to evade detection by T-cells, which could lead to evasion of on immunodominant response without consequence or evasion of a super response that allows viral persistence. In spite of the potential for immune escape, certain individuals within popula- tions retain the ability to target virus sequences that lead to viral clearance via the CD8+ T-cell response, suggesting that the response is impervious to change and can broadly recognize variation in virus derived MHC class I bound peptides [43] or that the sequence provides a necessary function for the virus and resists change. In spite of the numerous VP2121-130 mutations that promote an attenuated virus phenotype shown here and by others [31], the M130L mutant was the only one that replicated as efficiently as the wild-type virus after infection in vitro and in RAG deficient animals and similar to wild-type TMEV its clearance was promoted in H-2b mice. In contrast to wild-type, the M130L mutant was cleared more efficiently in FVB H-2q mice. A strain that is susceptible to chronic infection and demyelination after infection with wild-type TMEV [11]. Although a precise mechanism for this difference was not identified, we hypothesize that the emergence of viruses with comparable fitness to wild-type virus may consequently be recognized by a unique MHC within the population, thus promoting viral clearance. In this case, the H- 2Dq, H-2Kq or H-2Lq alleles of FVB mice may have promoted viral clearance through the emergence of new epitopes within the M130L virus. Picornavirus Epitope Resistant to Immune Escape Picornavirus Epitope Resistant to Immune Escape targeted after a primary virus infection. The disruption of CD8+ T-cell activation using this approach can be accomplished using several strategies. The approaches used can either be designed to inhibit binding of the specific virus epitope to MHC class I through modification of MHC binding residues or to introduce mutations that inhibit interactions with the H-2Db/VP2121-130 responsive T-cells. Although attempts to characterize these peptide residues as functionally distinct have been documented [30] it is unclear whether residues that interact with the T-cell receptor are mutually exclusive from those that bind to MHC. Previously, Myoung et. al. [31] specifically targeted the carboxy-terminal H- 2Db anchor residue at position 130 of VP2 as a means of destabilizing interactions of the antigenic peptide with MHC I. Using this strategy, this group found that only rare mutations at position 130 of VP2 generated infectious particles and that only one infected equivalently in vivo, the M130L mutant. In addition to the M130L mutation, we found that only one substitution at position 125 yielded a virus that could assemble in vitro. Although the contribution of this amino acid position to the induction of the immunodominant VP2121-130 response is not known, an absent canonical asparagine in the fifth position of nonamer H-2Db binding peptides is thought to be an anchor residue [30]. The non- canonical nature of the decamer VP2121-130 peptide calls into question the precise contribution amino acids at position 125 play in binding to MHC or to interactions with the T-cell receptor [32]. We find that VP2-S125A mutant peptides can assemble with H- 2Db molecules, suggesting that this peptide can stabilize H-2Db and induce responses that overlap with the wild-type peptide. The response to this peptide was decreased however, suggesting that this change primarily altered T-cell interactions with the peptide which may not be as robust as those involving the wild-type peptide. found that antigen encoding TMEV vaccines can effectively be used as immunotherapy for the treatment of melanoma and breast cancer in models of these diseases [20,36]. However, as with other live vaccine vectors the endogenous response to the virus often competes with the desired response against the vaccine antigen, limiting the usefulness of live virus vaccines [37,38]. Identifying strategies for evading these responses in a variety of vectors will be useful for improving the effective immune response against the target. Picornavirus Epitope Resistant to Immune Escape Of interest, Vilyuisk virus was previously identified as a divergent TMEV strain [33] and contains the M130L mutation within its VP2 capsid, providing a natural variant to TMEV within the VP2121-130 region. Humans have been identified as a host for Vilyuisk virus [34], however this virus may represent a natural CD8+ T-cell escape variant of TMEV since replication of this virus in other rodent species has been observed [33,35]. Although these viruses are similar, genomic differences exist outside of the VP2121-130 region, suggesting that additional factors outside of CD8+ T-cell immunity may also drive their evolution. Additional controlled studies using recombinant TMEV viruses and transgenic mice may more precisely determine the role CD8+ T-cells and other factors play in driving virus diversity. The findings in this report provide further detail into how superior virus controllers can occur in infected populations and how both host and pathogen genetics play a role in determining the outcome of infection with emerging pathogens. Previously, we have shown that a virus controlling phenotype depends on an underappreciated function of MHC that goes beyond its ability to present peptides. We found that host regulation of MHC class I genes can have a profound impact on the ability to generate this response [7]. Here, we explored the contribution that virus genetics plays in the generation of this response. We find that the VP2121-130 epitope is crucial to virus fidelity and that this amino acid sequence is critical to virus assembly. Further, when rare virus mutants emerged, they were unable to avoid detection by CD8+ T-cells activated with wild-type virus, revealing that an immune escape virus can be controlled by a cross-reactive response or be forced to extinction through a reduction in its overall fidelity. These findings provide further insight into how an effective cellular immune response is generated and will have consequences for better understanding the antigens targeted by CD8+ T cells that effectively control virus infection. Picornavirus Epitope Resistant to Immune Escape CD8+ T-cell response and virus RNA levels in resistant and susceptible mice after intracranial infection with TMEV-wt, S125A and M130L viruses. (A) CD45+ cells isolated from the central nervous system of VP2 mutant infected C57BL/6 mice were analyzed for the presence of CD8+ cells. (a significant compared to M130L by ANOVA, * significant compared to TMEV-wt by ANOVA). Figure is an experiment replicated 3 times with 3-5 animals per group. (B) CNS infiltrating lymphocytes were stained with H-2Db/VP2121 or H-2Db/E749 tetramers and analyzed by flow cytometry to determine the percent of Figure 5. Peptide depletion reveals an overlap in the wild-type VP2121-130 CD8+ T cell response after infection with S125A and M130L mutant viruses. (A) Mice were pre-depleted with control E749, wild-type VP2121-130, VP2-S125A and VP2-M130L peptide prior to intracranial infection with wild-type virus. CNS infiltrating lymphocytes were analyzed by flow cytometry to determine the percentage and number of CD8+ T-cells specific for the immunodominant epitope after peptide depletion. Data are representative examples of three individual mice per group. (B) Percent and absolute number of H-2Db/VP2121+ CD8 T-cells derived from infected CNS tissue including a no peptide group (* significant by ANOVA). doi:10.1371/journal.pone.0094332.g005 Figure 4. CD8+ T-cell response and virus RNA levels in resistant and susceptible mice after intracranial infection with TMEV-wt, S125A and M130L viruses. (A) CD45+ cells isolated from the central nervous system of VP2 mutant infected C57BL/6 mice were analyzed for the presence of CD8+ cells. (a significant compared to M130L by ANOVA, * significant compared to TMEV-wt by ANOVA). Figure is an experiment replicated 3 times with 3-5 animals per group. (B) CNS infiltrating lymphocytes were stained with H-2Db/VP2121 or H-2Db/E749 tetramers and analyzed by flow cytometry to determine the percent of the CD8+ T-cell population that is positive for the immunodominant April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org 8 Mice, cell lines and reagents PE labeled tetramers for H-2Db/ E749 were either purchased from Beckman Coulter (Brea, CA). PE labeled H-2Db/VP2121 tetramers were kindly provided by the NIH Tetramer Core at Emory University (Atlanta, GA). APC labeled H-2Db/VP2121 tetramers were provided by Dr. Aaron Johnson. Rabbit anti-TMEV polyclonal serum used for Western blots was provided by Dr. Moses Rodriguez (Mayo Clinic, Rochester, MN). PerCP labeled anti-CD45 and PE-Cy7 or allophycocyanin (APC) labeled anti-mouse CD8 were purchased from BD Biosciences (San Diego, CA). PE labeled tetramers for H-2Db/ E749 were either purchased from Beckman Coulter (Brea, CA). PE labeled H-2Db/VP2121 tetramers were kindly provided by the NIH Tetramer Core at Emory University (Atlanta, GA). APC labeled H-2Db/VP2121 tetramers were provided by Dr. Aaron Johnson. Rabbit anti-TMEV polyclonal serum used for Western blots was provided by Dr. Moses Rodriguez (Mayo Clinic, Rochester, MN). Virology y Plaque assays of viral supernatants or of homogenized CNS tissues were performed on L929 cells in 12 well plates as described previously [48]. Plates were scanned using an Epson Perfection V600 flatbed scanner (Long Beach, CA), images were analyzed using the measuring tool in the ImageJ software [49] to assess the mean diameter of the mutant and wild-type viruses. Virus RNA isolated from the brains and spinal cords of infected mice were assessed by semi-quantitative RT-PCR as described previously [20]. RNA isolated from FVB mice infected with wild-type TMEV or with VP2 mutant viruses for 30 days was used to verify the presence of the infecting virus by sequence analysis of RT-PCR amplicons across the mutation site. TMEV specific ELISA and neutralization TMEV specific ELISA was performed as d TMEV specific ELISA was performed as described previously [18]. TMEV neutralization assays were performed as described previously [50]. Briefly, TMEV immune serum was incubated with at least 100 PFU of wild-type TMEV for one hour prior to plating on L929 cell monolayers for plaque assay. Percent neutralization was calculated using the number of plaques remaining compared to non-immune serum ((1 – (PFU immune serum/PFU non-immune serum)) x 100%). To identify potential mutations that could be introduced into the VP2121-130 region of TMEV, we used I-Mutant2.0, a web server for the automatic prediction of protein stability change upon single-site mutation (http://gpcr2.biocomp.unibo.it/I- Mutant.htm). This tool was trained on data derived from ProTherm [16], a database of experimental data on protein mutations. We used the known crystal structure data from TMEV (1TME) as input for predicting whether mutations within this region are stabilizing or destabilizing. The results are expressed as the predicted DDG, the unfolding Gibbs free energy of the mutated protein minus the unfolding Gibbs free energy of the wild-type (Kcal/mol). Mice, cell lines and reagents C57BL/6 mice were purchased from Jackson Laboratory (Bar Harbor, ME). FVB and FVB RAG-/- mice were kindly provided by Dr. Moses Rodriguez (Mayo Clinic, Rochester, MN). Mice were infected intracerebrally (2x104 PFU) with wild-type or modified TMEV viruses. Intracerebrally infected FVB RAG-/- mice were monitored daily and weighed every other day to determine morbidity, animals that appeared severely moribund or that had lost .30% of their body mass were sacrificed as required. All animals were housed in the Mayo Clinic Department of Comparative Medicine and cared for according to institutional and NIH guidelines for animals use and care. gy To generate infectious virus, we transfected 30 mg of vector DNA that contained the wild-type or modified virus sequences into 107 BHK cells using electroporation. Media was changed on day 1 and 4 after transfection and cells were monitored daily for signs of cytopathic effect. By day 7, the generation of productive virus infection was noted by the presence of mostly rounded cells floating in media with few cells remaining adherent. To verify the presence of processed viral protein products whole cell lysates and supernatant were collected separately and analyzed by Western blot to verify the presence of TMEV in either or both using a polyclonal antibody that recognizes virus capsid proteins [47]. Supernatants were titered by plaque assay using L929 cells prior to use in infection experiments. Mutant virus sequences derived from supernatants or mouse tissues were verified by amplification of virus RNA using RT-PCR with primers that flank the VP2121-130 coding region (Forward 59-CTTTCTCCCACATCCGCATT- CCTCTC and Reverse 59-GGTCCGGCTATCGTAGCGG- TAACCAG). The wild-type VP2121-130 (FHAGSLLVFM), S125A (FHA- GALLVFM), M130L (FHAGSLLVFL) and E749-57 (RAHY- NIVTF) peptides were manufactured to 95% purity by Elim Biopharm (Hayward, CA). The VP2121-130 peptide is the immunodominant MHC class I epitope derived from the VP2 capsid region of TMEV that binds to mouse H-2Db. S125A and M130L are variant peptides derived from the same VP2 region of mutant viruses. The E749-57 peptide is an irrelevant control H-2Db binding peptide derived from human papillomavirus E7 protein [44]. BHK-21 and L929 cell lines (American Type Culture Collec- tion, Manassas, VA) were maintained in DMEM (GIBCO Invitrogen, Grand Island, NY) containing 10% fetal bovine serum (GIBCO Invitrogen). MTT assays for cell death were performed on BHK-21 cells as described [45]. PerCP labeled anti-CD45 and PE-Cy7 or allophycocyanin (APC) labeled anti-mouse CD8 were purchased from BD Biosciences (San Diego, CA). Ethics statement This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocols were approved by the Institutional Animal Care and Use Committee of Mayo Clinic (#A43310). All mice were anesthetized with isoflurane prior to intracranial virus infection. The identification of these antigenic variants of the VP2121-130 sequence may have further implication for the design and use of attenuated live virus vaccine vectors for immunotherapy. We have April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org April 2014 \| Volume 9 \| Issue 4 \| e94332 9 Picornavirus Epitope Resistant to Immune Escape Peptide depletion of virus specific CD8+ T-cells To deplete antigen specific CD8+ T-cell responses we intravenously administered super physiologic doses of VP2 wild- type and mutant peptides prior to virus infection, a method previously described to ameliorate this response [19]. Briefly, one day prior to infection with wild-type Daniel’s strain of TMEV, we intravenously administered 0.1 mg of VP2121-130, S125A, M130L or E749-57 peptide three times with 4 hours rest between injections. On the following day mice were injected intracranially with 26104 PFU of wild-type TMEV. On day six after infection central Dr. Moses Rodriguez (Mayo Clinic, Rochester, MN). Homology mapping and virus stability prediction We used the homology mapping tool found on the Immune Epitope Database website (http://tools.immuneepitope.org/esm). This tool is designed to identify homologous linear and confor- mational epitopes from known 3 dimensional structures within the Protein Data Bank (http://www.rcsb.org). We used the PDB structure for the Daniel’s strain of TMEV (1TME) to search for homologous sequences to the immunodominant epitope region VP2121-130. Picornavirus Epitope Resistant to Immune Escape coding sequence for the Daniel’s strain of TMEV [46]. VP2 mutant TMEV vectors with modifications in the sequence encoding the VP2121-130 region were generated by using site- directed mutagenesis according to the manufacturer’s protocol (QuikChange II Site-Directed Mutagenesis Kit, Agilent, Santa Clara, CA). We introduced unique nucleotide sequences that encoded both silent and novel mutations within this region. All custom made oligonucleotides were manufactured by Integrated DNA Technologies (Carolville, IA). All virus encoding plasmids were verified by sequencing at the Mayo Clinic Advanced Genomics Technology Center. Mice, cell lines and reagents Mice, cell lines and reagents References Pavelko KD, Mendez-Fernandez Y, Bell MP, Hansen MJ, Johnson AJ, et al. (2012) Nonequivalence of classical MHC class I loci in ability to direct effective antiviral immunity. PLoS pathogens 8: e1002541. 25. McCright IJ, Tsunoda I, Libbey JE, Fujinami RS (2002) Mutation in loop I of VP1 of Theiler’s virus delays viral RNA release into cells and enhances antibody- mediated neutralization: a mechanism for the failure of persistence by the mutant virus. Journal of neurovirology 8: 100–110. 8. Brahic M, Bureau JF, Michiels T (2005) The genetics of the persistent infection and demyelinating disease caused by Theiler’s virus. Annual review of microbiology 59: 279–298. gy 9. Rodriguez M, David CS (1985) Demyelination induced by Theiler’s virus: influence of the H-2 haplotype. Journal of immunology 135: 2145–2148. 26. Lin X, Sato S, Patick AK, Pease LR, Roos RP, et al. (1998) Molecular characterization of a nondemyelinating variant of Daniel’s strain of Theiler’s virus isolated from a persistently infected glioma cell line. Journal of virology 72: 1262–1269. 10. Rodriguez M, Leibowitz J, David CS (1986) Susceptibility to Theiler’s virus- induced demyelination. Mapping of the gene within the H-2D region. The Journal of experimental medicine 163: 620–631. 27. Sato S, Zhang L, Kim J, Jakob J, Grant RA, et al. (1996) A neutralization site of DA strain of Theiler’s murine encephalomyelitis virus important for disease phenotype. Virology 226: 327–337. 11. Azoulay A, Brahic M, Bureau JF (1994) FVB mice transgenic for the H-2Db gene become resistant to persistent infection by Theiler’s virus. Journal of virology 68: 4049–4052. 28. Jnaoui K, Michiels T (1998) Adaptation of Theiler’s virus to L929 cells: mutations in the putative receptor binding site on the capsid map to neutralization sites and modulate viral persistence. Virology 244: 397–404. 12. Nicholson SM, Peterson JD, Miller SD, Wang K, Dal Canto MC, et al. (1994) BALB/c substrain differences in susceptibility to Theiler’s murine encephalo- myelitis virus-induced demyelinating disease. Journal of neuroimmunology 52: 19–24. 29. Grant RA, Filman DJ, Fujinami RS, Icenogle JP, Hogle JM (1992) Three- dimensional structure of Theiler virus. Proceedings of the National Academy of Sciences of the United States of America 89: 2061–2065. 13. Lyman MA, Lee HG, Kang BS, Kang HK, Kim BS (2002) Capsid-specific cytotoxic T lymphocytes recognize three distinct H-2D(b)-restricted regions of the BeAn strain of Theiler’s virus and exhibit different cytokine profiles. Journal of virology 76: 3125–3134. 30. References 1. Apanius V, Penn D, Slev PR, Ruff LR, Potts WK (1997) The nature of selection on the major histocompatibility complex. Critical reviews in immunology 17: 179–224. 1. Apanius V, Penn D, Slev PR, Ruff LR, Potts WK (1997) The nature of selection on the major histocompatibility complex. Critical reviews in immunology 17: 179–224. response against a single immunodominant viral peptide. European journal of immunology 33: 2501–2510. response against a single immunodominant viral peptide. European journal of immunology 33: 2501–2510. 20. Pavelko KD, Bell MP, Karyampudi L, Hansen MJ, Allen KS, et al. (2013) The epitope integration site for vaccine antigens determines virus control while maintaining efficacy in an engineered cancer vaccine. Molecular therapy: the journal of the American Society of Gene Therapy 21: 1087–1095. 2. Nikolich-Zugich J, Fremont DH, Miley MJ, Messaoudi I (2004) The role of mhc polymorphism in anti-microbial resistance. Microbes and infection/Institut Pasteur 6: 501–512. 3. Lienert K, Parham P (1996) Evolution of MHC class I genes in higher primates. Immunology and cell biology 74: 349–356. 21. Lewis MJ, Dagarag M, Khan B, Ali A, Yang OO (2012) Partial escape of HIV-1 from cytotoxic T lymphocytes during chronic infection. Journal of virology 86: 7459–7463. 4. Lambotte O, Boufassa F, Madec Y, Nguyen A, Goujard C, et al. (2005) HIV controllers: a homogeneous group of HIV-1-infected patients with spontaneous control of viral replication. Clinical infectious diseases: an official publication of the Infectious Diseases Society of America 41: 1053–1056. 22. Ruhl M, Chhatwal P, Strathmann H, Kuntzen T, Bankwitz D, et al. (2012) Escape from a dominant HLA-B*15-restricted CD8+ T cell response against hepatitis C virus requires compensatory mutations outside the epitope. Journal of virology 86: 991–1000. y 5. Bowen DG, Walker CM (2005) Adaptive immune responses in acute and chronic hepatitis C virus infection. Nature 436: 946-952. 23. Blinkova O, Kapoor A, Victoria J, Jones M, Wolfe N, et al. (2009) Cardioviruses are genetically diverse and cause common enteric infections in South Asian children. Journal of virology 83: 4631–4641. 6. Hill AV, Allsopp CE, Kwiatkowski D, Anstey NM, Twumasi P, et al. (1991) Common west African HLA antigens are associated with protection from severe malaria. Nature 352: 595–600. 24. Zurbriggen A, Hogle JM, Fujinami RS (1989) Alteration of amino acid 101 within capsid protein VP-1 changes the pathogenicity of Theiler’s murine encephalomyelitis virus. The Journal of experimental medicine 170: 2037–2049. 7. References Falk K, Rotzschke O, Deres K, Metzger J, Jung G, et al. (1991) Identification of naturally processed viral nonapeptides allows their quantification in infected cells and suggests an allele-specific T cell epitope forecast. The Journal of experimental medicine 174: 425–434. 14. Johnson AJ, Njenga MK, Hansen MJ, Kuhns ST, Chen L, et al. (1999) Prevalent class I-restricted T-cell response to the Theiler’s virus epitope Db:VP2121-130 in the absence of endogenous CD4 help, tumor necrosis factor alpha, gamma interferon, perforin, or costimulation through CD28. Journal of virology 73: 3702–3708. 31. Myoung J, Hou W, Kang B, Lyman MA, Kang JA, et al. (2007) The immunodominant CD8+ T cell epitope region of Theiler’s virus in resistant C57BL/6 mice is critical for anti-viral immune responses, viral persistence, and binding to the host cells. Virology 360: 159–171. 15. Lee WM, Monroe SS, Rueckert RR (1993) Role of maturation cleavage in infectivity of picornaviruses: activation of an infectosome. Journal of virology 67: 2110–2122. 32. Borson ND, Paul C, Lin X, Nevala WK, Strausbauch MA, et al. (1997) Brain- infiltrating cytolytic T lymphocytes specific for Theiler’s virus recognize H2Db molecules complexed with a viral VP2 peptide lacking a consensus anchor residue. Journal of virology 71: 5244–5250. 16. Bava KA, Gromiha MM, Uedaira H, Kitajima K, Sarai A (2004) ProTherm, version 4.0: thermodynamic database for proteins and mutants. Nucleic acids research 32: D120–121. J gy 33. Pritchard AE, Strom T, Lipton HL (1992) Nucleotide sequence identifies Vilyuisk virus as a divergent Theiler’s virus. Virology 191: 469–472. 17. Lobert PE, Escriou N, Ruelle J, Michiels T (1999) A coding RNA sequence acts as a replication signal in cardioviruses. Proceedings of the National Academy of Sciences of the United States of America 96: 11560–11565. 34. Lipton HL (2008) Human Vilyuisk encephalitis. Reviews in medical virology 18: 347–352. 35. Liang Z, Kumar AS, Jones MS, Knowles NJ, Lipton HL (2008) Phylogenetic analysis of the species Theilovirus: emerging murine and human pathogens. Journal of virology 82: 11545–11554. 18. Pavelko KD, Pease LR, David CS, Rodriguez M (2007) Genetic deletion of a single immunodominant T-cell response confers susceptibility to virus-induced demyelination. Brain pathology 17: 184–196. 36. Pavelko KD, Girtman MA, Mitsunaga Y, Mendez-Fernandez YV, Bell MP, et al. (2011) Theiler’s murine encephalomyelitis virus as a vaccine candidate for immunotherapy. PloS one 6: e20217. y p gy 19. We thank Kathleen S. Allen for expert technical assistance. We thank Kathleen S. Allen for expert technical assistance. Isolation CNS infiltrating lymphocyte Brain and spinal cord infiltrating lymphocytes from TMEV infected mice were recovered using previously described tech- niques and were analyzed by flow cytometry [51]. Absolute quantitation was calculated using CountBright Absolute counting beads (Life Technologies, Grand Isle, NY). Samples were analyzed on a BD LSR II flow cytometer (BD Biosciences, San Jose, CA) and analyzed using FloJo software (Ashland, OR). Single color stained splenocytes were used as compensation controls. Generation of TMEV VP2121-130 mutant viruses Generation of TMEV VP2121-130 mutant viruses Generation of TMEV VP2121-130 mutant viruses To generate mutant VP2 viruses we modified a cDNA vector described previously [20]. This vector contains the full length To generate mutant VP2 viruses we modified a cDNA vector described previously [20]. This vector contains the full length PLOS ONE \| www.plosone.org April 2014 \| Volume 9 \| Issue 4 \| e94332 10 Picornavirus Epitope Resistant to Immune Escape nervous system (CNS) tissues were harvested before isolation of tissue infiltrating lymphocytes. by Dunn’s method for non-parametric data. The Kaplan-Meier survival analysis was performed by the Log-Rank method with pairwise comparisons analyzed by the Holm-Sidak method. Significance was determined by p,0.05 for all tests. Statistical analysis was performed using SigmaPlot for Windows Version 11.0 software (Systat Software, San Jose, CA). Statistics All data were analyzed by One-way or Two-way ANOVA. Multiple comparisons between groups were performed using the Student-Newman-Keuls method for normally distributed data and Author Contributions Conceived and designed the experiments: KDP. Performed the experi- ments: MPB DNR KDP. Analyzed the data: MPB KDP. Contributed reagents/materials/analysis tools: MPB DNR AJJ KDP. Wrote the paper: KDP. Picornavirus Epitope Resistant to Immune Escape protects against a tumor induced by human papillomavirus type 16-transformed cells. European journal of immunology 23: 2242–2249. 37. Smith CL, Mirza F, Pasquetto V, Tscharke DC, Palmowski MJ, et al. (2005) Immunodominance of poxviral-specific CTL in a human trial of recombinant- modified vaccinia Ankara. Journal of immunology 175: 8431–8437. protects against a tumor induced by human papillomavirus type 16-transformed cells. European journal of immunology 23: 2242–2249. 45. Pavelko KD, Howe CL, Drescher KM, Gamez JD, Johnson AJ, et al. (2003) Interleukin-6 protects anterior horn neurons from lethal virus-induced injury. J Neurosci 23: 481–492. 38. Schirmbeck R, Reimann J, Kochanek S, Kreppel F (2008) The immunogenicity of adenovirus vectors limits the multispecificity of CD8 T-cell responses to vector-encoded transgenic antigens. Molecular therapy: the journal of the American Society of Gene Therapy 16: 1609–1616. J 46. Roos RP, Stein S, Ohara Y, Fu JL, Semler BL (1989) Infectious cDNA clones of the DA strain of Theiler’s murine encephalomyelitis virus. J Virol 63: 5492– 5496. y py 39. Yewdell JW (2006) Confronting complexity: real-world immunodominance in antiviral CD8+ T cell responses. Immunity 25: 533–543. 47. Njenga MK, Pavelko KD, Baisch J, Lin X, David C, et al. (1996) Theiler’s virus persistence and demyelination in major histocompatibility complex class II- deficient mice. Journal of virology 70: 1729–1737. 40. Hislop AD, Taylor GS, Sauce D, Rickinson AB (2007) Cellular responses to viral infection in humans: lessons from Epstein-Barr virus. Annual review of immunology 25: 587–617. 48. Rodriguez M, Leibowitz JL, Powell HC, Lampert PW (1983) Neonatal infection with the Daniels strain of Theiler’s murine encephalomyelitis virus. Laboratory investigation; a journal of technical methods and pathology 49: 672–679. 41. Obar JJ, Fuse S, Leung EK, Bellfy SC, Usherwood EJ (2006) Gammaherpes- virus persistence alters key CD8 T-cell memory characteristics and enhances antiviral protection. Journal of virology 80: 8303–8315. 49. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nature methods 9: 671–675. 42. Farrington LA, Smith TA, Grey F, Hill AB, Snyder CM (2013) Competition for antigen at the level of the APC is a major determinant of immunodominance during memory inflation in murine cytomegalovirus infection. Journal of immunology 190: 3410–3416. 50. Pavelko KD, Drescher KM, McGavern DB, David CS, Rodriguez M (2000) HLA-DQ polymorphism influences progression of demyelination and neurologic deficits in a viral model of multiple sclerosis. Molecular and cellular neurosciences 15: 495–509. gy 43. References Mendez-Fernandez YV, Johnson AJ, Rodriguez M, Pease LR (2003) Clearance of Theiler’s virus infection depends on the ability to generate a CD8+ T cell April 2014 \| Volume 9 \| Issue 4 \| e94332 April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org 11 Picornavirus Epitope Resistant to Immune Escape Valkenburg SA, Gras S, Guillonneau C, La Gruta NL, Thomas PG, et al. (2010) Protective efficacy of cross-reactive CD8+ T cells recognising mutant viral epitopes depends on peptide-MHC-I structural interactions and T cell activation threshold. PLoS pathogens 6: e1001039. 51. Deb C, Lafrance-Corey RG, Schmalstieg WF, Sauer BM, Wang H, et al. (2010) CD8+ T cells cause disability and axon loss in a mouse model of multiple sclerosis. PloS one 5: e12478. 44. Feltkamp MC, Smits HL, Vierboom MP, Minnaar RP, de Jongh BM, et al. (1993) Vaccination with cytotoxic T lymphocyte epitope-containing peptide PLOS ONE \| www.plosone.org April 2014 \| Volume 9 \| Issue 4 \| e94332 PLOS ONE \| www.plosone.org 12
https://openalex.org/W4240174600	https://peerj.com/articles/7504v0.3/submission	English	null	Peer Review #3 of "Comprehensive classification of the plant non-specific lipid transfer protein superfamily towards its sequence–structure–function analysis (v0.1)"	null	2,019	cc-by	25,991	Manuscript to be reviewed Manuscript to be reviewed Comprehensive classiﬁcation of the plant non-speciﬁc lipid transfer protein superfamily towards its Sequence – Structure – Function analysis PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 1 Comprehensive classif 2 lipid transfer protein su 3 Sequence – Structure – 4 5 Cécile Fleury1, Jérôme Gracy3, Marie-Fr 6 Dufayard2, Gilles Labesse3, Manuel Ruiz 7 8 1 INRA, UMR AGAP, F-34060 Montpel 9 2 CIRAD, UMR AGAP, F-34398, Montp 10 3 CBS CNRS Univ. Montpellier– INSER 11 12 Corresponding Author: 13 Frédéric de Lamotte1 14 Avenue Agropolis – TA A-108/03 – Mon 15 Email address: frederic.de-lamotte@inra 16 17 ABSTRACT 18 Background. Non-specific Lipid Transf 19 kingdom and constitute a superfamily of 20 sequences – and counting – have been ch 21 unclear. It has been clear for years that th 22 nutritional issues. Deciphering their func 23 from gene sequence to protein structure, 24 huge and growing number of new protein 25 meaningful knowledge from sequence-st 26 of new tools and approaches. 27 As nsLTPs show high evolutionary dive 28 superhelix structural fold, and as they are 29 development and defense, they are a stim 30 Methods. In this study we comprehensiv 31 including a phylogenetic analysis on can 32 structure modelling and functional annot 33 programs. Additionally, two integrative m 34 first was a new method for the detection 35 structure stabilization and ii) residues po 36 a structure-function classification based o 37 tree visualization interface. We also pres 38 Results. Following this new protocol, an 39 established and a new functional hypothe 1 Comprehensive classification of the plant non 2 lipid transfer protein superfamily towards its 3 Sequence – Structure – Function analysis 4 5 Cécile Fleury1, Jérôme Gracy3, Marie-Françoise Gautier1, Jean-Luc Pons3, Jean- 6 Dufayard2, Gilles Labesse3, Manuel Ruiz2, Frédéric de Lamotte1 7 8 1 INRA, UMR AGAP, F-34060 Montpellier, France 9 2 CIRAD, UMR AGAP, F-34398, Montpellier 10 3 CBS CNRS Univ. Montpellier– INSERM, Montpellier, F-34090, France 11 12 Corresponding Author: 13 Frédéric de Lamotte1 14 Avenue Agropolis – TA A-108/03 – Montpellier – F-34398 Cedex 5 – France 15 Email address: frederic.de-lamotte@inra.fr 16 17 ABSTRACT 18 Background. Non-specific Lipid Transfer Proteins (nsLTPs) are widely distribu 19 kingdom and constitute a superfamily of related proteins. Several hundreds of di 20 sequences – and counting – have been characterized so far, but their biological fu 21 unclear. It has been clear for years that they present a certain interest for agronom 22 nutritional issues. Comprehensive classiﬁcation of the plant non-speciﬁc lipid transfer protein superfamily towards its Sequence – Structure – Function analysis Cecile Fleury 1 , Jerome Gracy 2 , Marie-Francoise Gautier 1 , Jean-Luc Pons 2 , Jean-Francois Dufayard 3 , Gilles Labesse 2 , Manuel Ruiz 3 , Frederic de Lamotte Corresp. 1 1 UMR AGAP, INRA, Montpellier, France 2 CBS, CNRS Univ Montpellier INSERM, Montpellier, France 3 UMR AGAP, CIRAD, Montpellier, France 1 UMR AGAP, INRA, Montpellier, France 2 CBS, CNRS Univ Montpellier INSERM, Montpellier, France 3 UMR AGAP, CIRAD, Montpellier, France Corresponding Author: Frederic de Lamotte Email address: frederic.de-lamotte@inra.fr Corresponding Author: Frederic de Lamotte Email address: frederic.de-lamotte@inra.fr Corresponding Author: Frederic de Lamotte Email address: frederic.de-lamotte@inra.fr Background. Non-speciﬁc Lipid Transfer Proteins (nsLTPs) are widely distributed in the plant kingdom and constitute a superfamily of related proteins. Several hundreds of diﬀerent nsLTP sequences – and counting – have been characterized so far, but their biological functions remain unclear. It has been clear for years that they present a certain interest for agronomic and nutritional issues. Deciphering their functions means collecting and analyzing a variety of data from gene sequence to protein structure, from cellular localization to the physiological role. As a huge and growing number of new protein sequences are available nowadays, extracting meaningful knowledge from sequence-structure-function relationships calls for the development of new tools and approaches. As nsLTPs show high evolutionary divergence, but a conserved common right handed superhelix structural fold, and as they are involved in a large number of key roles in plant development and defense, they are a stimulating case study for validating such an approach. Methods. In this study we comprehensively investigated 797 nsLTP protein sequences, including a phylogenetic analysis on canonical protein sequences, three-dimensional (3D) structure modelling and functional annotation using several well-established bioinformatics programs. Additionally, two integrative methodologies using original tools were developed. The ﬁrst was a new method for the detection of i) conserved amino acid residues involved in structure stabilization and ii) residues potentially involved in ligand interaction. The second was a structure-function classiﬁcation based on the Evolutionary Trace Display method using a new tree visualization interface. We also present a new tool for visualizing phylogenetic trees. Results. Following this new protocol, an updated classiﬁcation of the nsLTP superfamily was established and a new functional hypothesis for key residues is suggested. Lastly, this work allows a better representation of the diversity of plant nsLTPs in terms of sequence, structure and function. 1 Comprehensive classification of the plant non-specific 2 lipid transfer protein superfamily towards its 3 Sequence – Structure – Function analysis 4 5 Cécile Fleury1, Jérôme Gracy3, Marie-Françoise Gautier1, Jean-Luc Pons3, Jean-François 6 Dufayard2, Gilles Labesse3, Manuel Ruiz2, Frédéric de Lamotte1 8 1 INRA, UMR AGAP, F-34060 Montpellier, France 9 2 CIRAD, UMR AGAP, F-34398, Montpellier 10 3 CBS CNRS Univ. Montpellier– INSERM, Montpellier, F-34090, France 11 11 12 Corresponding Author: 13 Frédéric de Lamotte1 14 Avenue Agropolis – TA A-108/03 – Montpellier – F-34398 Cedex 5 – France 15 Email address: frederic.de-lamotte@inra.fr 16 17 ABSTRACT 18 Background. Non-specific Lipid Transfer Proteins (nsLTPs) are widely distributed in the plant 19 kingdom and constitute a superfamily of related proteins. Several hundreds of different nsLTP 20 sequences – and counting – have been characterized so far, but their biological functions remain 21 unclear. It has been clear for years that they present a certain interest for agronomic and 22 nutritional issues. Deciphering their functions means collecting and analyzing a variety of data 23 from gene sequence to protein structure, from cellular localization to the physiological role. As a 24 huge and growing number of new protein sequences are available nowadays, extracting 25 meaningful knowledge from sequence-structure-function relationships calls for the development 26 of new tools and approaches. 27 As nsLTPs show high evolutionary divergence, but a conserved common right handed 28 superhelix structural fold, and as they are involved in a large number of key roles in plant 29 development and defense, they are a stimulating case study for validating such an approach. 30 Methods. In this study we comprehensively investigated 797 nsLTP protein sequences, 31 including a phylogenetic analysis on canonical protein sequences, three-dimensional (3D) 32 structure modelling and functional annotation using several well-established bioinformatics 33 programs. Additionally, two integrative methodologies using original tools were developed. The 34 first was a new method for the detection of i) conserved amino acid residues involved in 35 structure stabilization and ii) residues potentially involved in ligand interaction. The second was 36 a structure-function classification based on the Evolutionary Trace Display method using a new 37 tree visualization interface. We also present a new tool for visualizing phylogenetic trees. 38 Results. Following this new protocol, an updated classification of the nsLTP superfamily was 39 established and a new functional hypothesis for key residues is suggested. 13 Frédéric de Lamotte 14 Avenue Agropolis – TA A-108/03 – Montpellier – F-34398 Cedex 5 – France 18 Background. Non-specific Lipid Transfer Proteins (nsLTPs) are widely distributed in the plant 19 kingdom and constitute a superfamily of related proteins. Comprehensive classiﬁcation of the plant non-speciﬁc lipid transfer protein superfamily towards its Sequence – Structure – Function analysis Deciphering their functions means collecting and analyzing a v 23 from gene sequence to protein structure, from cellular localization to the physiol 24 huge and growing number of new protein sequences are available nowadays, ext 25 meaningful knowledge from sequence-structure-function relationships calls for t 26 of new tools and approaches. 27 As nsLTPs show high evolutionary divergence, but a conserved common right h 28 superhelix structural fold, and as they are involved in a large number of key role 29 development and defense, they are a stimulating case study for validating such a 30 Methods. In this study we comprehensively investigated 797 nsLTP protein sequ 31 including a phylogenetic analysis on canonical protein sequences, three-dimensi 32 structure modelling and functional annotation using several well-established bioi 33 programs. Additionally, two integrative methodologies using original tools were 34 first was a new method for the detection of i) conserved amino acid residues invo 35 structure stabilization and ii) residues potentially involved in ligand interaction. T 36 a structure-function classification based on the Evolutionary Trace Display meth 37 tree visualization interface. We also present a new tool for visualizing phylogene 38 Results. Following this new protocol, an updated classification of the nsLTP sup 39 established and a new functional hypothesis for key residues is suggested. 1 Comprehensive classification of the plant non-specific 2 lipid transfer protein superfamily towards its 3 Sequence – Structure – Function analysis 4 5 Cécile Fleury1, Jérôme Gracy3, Marie-Françoise Gautier1, Jean-Luc Pons3, Jean-François 6 Dufayard2, Gilles Labesse3, Manuel Ruiz2, Frédéric de Lamotte1 Several hundreds of different nsLTP 20 sequences – and counting – have been characterized so far, but their biological functions remain 21 unclear. It has been clear for years that they present a certain interest for agronomic and 22 nutritional issues. Deciphering their functions means collecting and analyzing a variety of data 23 from gene sequence to protein structure, from cellular localization to the physiological role. As a 24 huge and growing number of new protein sequences are available nowadays, extracting 25 meaningful knowledge from sequence-structure-function relationships calls for the development 26 of new tools and approaches. 38 Results. Following this new protocol, an updated classification of the nsLTP superfamily was 39 established and a new functional hypothesis for key residues is suggested. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 40 Lastly, this work allows a better representation of the diversity of plant nsLTPs in terms of 41 sequence, structure and function. 40 Lastly, this work allows a better representation of the diversity of plant nsLTPs in terms of 41 sequence, structure and function. 41 sequence, structure and function. 43 INTRODUCTION 44 Since the work of Kader (Kader et al., 1984; Kader, 1996), numerous proteins capable of 45 transferring lipids have been annotated as non-specific lipid transfer proteins (nsLTPs). Their 46 primary sequences are characterized by a conserved 8-Cysteine Motif (8CM) (C-Xn-C-Xn-CC- 47 Xn-CXC-Xn-C-Xn-C), which plays an important role in their structural scaffold (José-Estanyol 48 et al., 2004). Based on their molecular masses, plant nsLTPs were first separated into two types: 49 type I (9 kDa) and type II (7 kDa), which were distinct both in terms of primary sequence 50 identity and the disulfide bond pattern (Douliez et al., 2001). 51 Plant nsLTPs are ubiquitous proteins encoded by multigene families, as reported in different 52 phylogenetic studies. However, these studies involve a limited number of sequences and/or 53 species: fifteen nsLTPs identified in Arabidopsis (Arondel et al., 2000), restricted to Poaceae 54 (Jang et al., 2007), Solanaceae (Liu et al., 2010, D’Agostino et al., 2019), or Gossypium (Li et 55 al., 2016; Meng et al., 2018). Recently 189 nsLtp genes were identified in 3 Gossypium species 56 (Li et al., 2016) and 138 nsLtp genes in the single Gossypium hirsutum species (Meng et al., 57 2018) using traditional sequence homology approaches. As for Solanum lycopersicum, 51 Plant nsLTPs are ubiquitous proteins encoded by multigene families, as reported in different 52 phylogenetic studies. However, these studies involve a limited number of sequences and/or 53 species: fifteen nsLTPs identified in Arabidopsis (Arondel et al., 2000), restricted to Poaceae 54 (Jang et al., 2007), Solanaceae (Liu et al., 2010, D’Agostino et al., 2019), or Gossypium (Li et 55 al., 2016; Meng et al., 2018). Recently 189 nsLtp genes were identified in 3 Gossypium species 56 (Li et al., 2016) and 138 nsLtp genes in the single Gossypium hirsutum species (Meng et al., 57 2018) using traditional sequence homology approaches. As for Solanum lycopersicum, ) g q gy pp y p , 58 D’Agostino and collaborators identified 64 nsLtp gene sequences using an Hidden Markov 59 Model approach (D’Agostino et al., 2019). Around 200 nsLTPs have been identified in wheat, 60 rice and Arabidopsis genomes and classified into nine different types according to sequence 61 similarity (Boutrot et al., 2008). More extensive studies including ancestral plants indicate that 62 nsLTPs are also present in liverworts, mosses and ferns, but not present in algae (Edstam et al., 63 2011; Wang et al., 2012). Manuscript to be reviewed 90 From a functional point of view, plant nsLTPs are classified into different families depending on 91 the scope of interest and their properties (Liu et al. 2015). Plant nsLTPs belong to the Prolamin 92 superfamily (AF050), which includes the largest number of allergens (Radauer et al., 2008). 93 Indeed, several nsLTPs from fruits of the Rosaceae family, nuts or different vegetables are food 94 allergens, with fruit nsLTPs being mainly localized in the peel (Salcedo et al., 2007, D’Agostino 95 et al., 2019). 80 belong to the “Mainly alpha” class. They display the “Orthogonal Bundle” architecture and the 81 “Hydrophobic Seed Protein” topology. At this level, only one homologous superfamily called 82 “Plant lipid-transfer and hydrophobic proteins” can be found. The superfamily appears to contain 83 ten distinct protein sequences, lacking the A. thaliana nsLTP, but including the soybean 84 hydrophobic protein found in the SCOP database. Of the known nsLTP 3D structures, only 85 Boutrot’s type I, II and IV are represented. An interesting point to be noted is that two different 86 cysteine pairing patterns have been observed (which correspond to a single cysteine switch 87 between two disulfide bridges): C1-C6 and C5-C8 in type I structures; C1-C5 and C6-C8 in type 88 II and IV structures. However, C2-C7 and C3-C4 bridges are common to all known nsLTP 89 structures and the overall fold is conserved among the whole family g y 90 From a functional point of view, plant nsLTPs are classified into different families depending on 91 the scope of interest and their properties (Liu et al. 2015). Plant nsLTPs belong to the Prolamin 92 superfamily (AF050), which includes the largest number of allergens (Radauer et al., 2008). 93 Indeed, several nsLTPs from fruits of the Rosaceae family, nuts or different vegetables are food 94 allergens, with fruit nsLTPs being mainly localized in the peel (Salcedo et al., 2007, D’Agostino 95 et al., 2019). 90 From a functional point of view, plant nsLTPs are classified into different families depending on 91 the scope of interest and their properties (Liu et al. 2015). Plant nsLTPs belong to the Prolamin 92 superfamily (AF050), which includes the largest number of allergens (Radauer et al., 2008). Manuscript to be reviewed 93 Indeed, several nsLTPs from fruits of the Rosaceae family, nuts or different vegetables are food 94 allergens, with fruit nsLTPs being mainly localized in the peel (Salcedo et al., 2007, D’Agostino 95 et al., 2019). 96 Plant nsLTPs are members of the pathogenesis-related proteins and compose the PR14 family 97 (van Loon et al., 2006). Their role in plant defense mechanisms has been shown by the induction 98 of nsLtp gene expression following pathogen infections, overexpression in transgenic plants, or 99 their antimicrobial properties (Molina & García-Olmedo, 1993; Cammue et al., 1995; Li et al., 100 2003; Girault et al., 2008; Sun et al., 2008). A role in plant defense signaling pathways has also 101 been suggested for an Arabidopsis type IV nsLTP, which needs to form a complex with glycerol- 102 3-phosphate for its translocation and induction of systemic acquired resistance (Maldonado et al., 103 2002; Chanda et al., 2011). One wheat nsLTP competes with a fungal cryptogein receptor in 104 tobacco plasma membranes and, when the LTP is complexed with lipids, its interaction with the 105 membrane and its defense activity are enhanced (Buhot et al., 2001; Buhot et al., 2004). In 106 wheat, nsLtp genes display a complex expression pattern during the development of the seed 107 (Boutrot et al., 2005). NsLTPs may also be involved in plant defense mechanisms through their 108 participation in cuticle synthesis (Debono et al., 2009). This function is supported by their 109 extracellular localization (Thoma et al., 1993; Pyee et al., 1994), the expression of different 110 nsLtp genes in leaf epidermis (Sterk et al., 1991; Pyee & Kolattukudy, 1995; Clark & Bohnert, 111 1999), a positive correlation between nsLtp gene expression and cuticular wax deposition 112 (Cameron et al., 2006), and their ability to bind cutin monomers (i.e. hydroxylated fatty acids) 113 (Douliez et al., 2001). In addition, nsLtp gene transcripts are abundant or specifically present in 114 trichomes and one tobacco nsLTP seems to be required for lipid secretion from glandular 115 trichomes indicating that nsLTPs may play a role either in the secretion of essential oils or in 116 defense mechanism (Lange et al., 2000; Aziz et al., 2005; Choi et al., 2012). 43 INTRODUCTION However, the efforts made so far to classify the members of the nsLTP 64 superfamily were including proteins that do not satisfy the strict criteria of 8CM pattern (Edstam 65 et al., 2011; Wang et al., 2012). In comparison to previous studies, we computed the most 66 extensive phylogenetic analysis, sampling 797 nsLTP sequences from 123 different species. This 67 allows to enrich the phylogenic tree of many evolutionary events that would have been hidden 68 with more restrictive species choices. These events are essentially gene duplications and have a 69 major influence on the family evolution that could be correlated to 3D structure evolution. 70 From a structural point of view, the nsLTP family belongs to the all-alpha class in the SCOP 58 D’Agostino and collaborators identified 64 nsLtp gene sequences using an Hidden Markov 59 Model approach (D’Agostino et al., 2019). Around 200 nsLTPs have been identified in wheat, 60 rice and Arabidopsis genomes and classified into nine different types according to sequence 61 similarity (Boutrot et al., 2008). More extensive studies including ancestral plants indicate that 62 nsLTPs are also present in liverworts, mosses and ferns, but not present in algae (Edstam et al., 63 2011; Wang et al., 2012). However, the efforts made so far to classify the members of the nsLTP 64 superfamily were including proteins that do not satisfy the strict criteria of 8CM pattern (Edstam 65 et al., 2011; Wang et al., 2012). In comparison to previous studies, we computed the most 66 extensive phylogenetic analysis, sampling 797 nsLTP sequences from 123 different species. This 67 allows to enrich the phylogenic tree of many evolutionary events that would have been hidden 68 with more restrictive species choices. These events are essentially gene duplications and have a 69 major influence on the family evolution that could be correlated to 3D structure evolution. 70 From a structural point of view, the nsLTP family belongs to the all-alpha class in the SCOP PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 80 belong to the “Mainly alpha” class. They display the “Orthogonal Bundle” architecture and the 81 “Hydrophobic Seed Protein” topology. At this level, only one homologous superfamily called 82 “Plant lipid-transfer and hydrophobic proteins” can be found. The superfamily appears to contain 83 ten distinct protein sequences, lacking the A. thaliana nsLTP, but including the soybean 84 hydrophobic protein found in the SCOP database. Of the known nsLTP 3D structures, only 85 Boutrot’s type I, II and IV are represented. An interesting point to be noted is that two different 86 cysteine pairing patterns have been observed (which correspond to a single cysteine switch 87 between two disulfide bridges): C1-C6 and C5-C8 in type I structures; C1-C5 and C6-C8 in type 88 II and IV structures. However, C2-C7 and C3-C4 bridges are common to all known nsLTP 89 structures and the overall fold is conserved among the whole family. 90 From a functional point of view, plant nsLTPs are classified into different families depending on 91 the scope of interest and their properties (Liu et al. 2015). Plant nsLTPs belong to the Prolamin 92 superfamily (AF050), which includes the largest number of allergens (Radauer et al., 2008). 93 Indeed, several nsLTPs from fruits of the Rosaceae family, nuts or different vegetables are food 94 allergens, with fruit nsLTPs being mainly localized in the peel (Salcedo et al., 2007, D’Agostino 95 et al., 2019). 80 belong to the “Mainly alpha” class. They display the “Orthogonal Bundle” architecture and the 81 “Hydrophobic Seed Protein” topology. At this level, only one homologous superfamily called 82 “Plant lipid-transfer and hydrophobic proteins” can be found. The superfamily appears to contain 83 ten distinct protein sequences, lacking the A. thaliana nsLTP, but including the soybean 84 hydrophobic protein found in the SCOP database. Of the known nsLTP 3D structures, only 85 Boutrot’s type I, II and IV are represented. An interesting point to be noted is that two different 86 cysteine pairing patterns have been observed (which correspond to a single cysteine switch 87 between two disulfide bridges): C1-C6 and C5-C8 in type I structures; C1-C5 and C6-C8 in type 88 II and IV structures. However, C2-C7 and C3-C4 bridges are common to all known nsLTP 89 structures and the overall fold is conserved among the whole family. Manuscript to be reviewed y p y p ( , ) 127 Non-specific LTPs are possibly involved in a range of other biological processes, but their 128 physiological functions are not clearly understood. An analysis of gain of function or defective 129 plant mutants can address these issues (Maldonado et al., 2002; Chae et al., 2009). Site directed 130 mutagenesis (SDM) led to the identification of residues involved in their antifungal activity, lipid 131 binding and lipid transfer (Ge et al., 2003; Cheng et al., 2008; Sawano et al., 2008). However, 132 even if they remain extremely valuable, these approaches are time-consuming and have so far 133 been limited to a small number of proteins. They need to be computationally assisted, using 134 information emerging from big datasets. 135 Current bioinformatics programs such as GeneSilico Metaserver (Kurowski & Bujnicki, 2003) or 136 MESSA (Cong & Grishin, 2012) provide an overview of known information about protein 137 sequences, structures and functions. However, studying inner relationships into such complex 138 superfamilies of proteins as the nsLTP superfamily requires a knowledge visualization and 139 classification tool that still needs to be developed. 135 Current bioinformatics programs such as GeneSilico Metaserver (Kurowski & Bujnicki, 2003) or 136 MESSA (Cong & Grishin, 2012) provide an overview of known information about protein 137 sequences, structures and functions. However, studying inner relationships into such complex 138 superfamilies of proteins as the nsLTP superfamily requires a knowledge visualization and 139 classification tool that still needs to be developed. 140 This paper describes both the development of new tools together with the use of these tools to 141 improve our comprehension of the nsLTP superfamily. 140 This paper describes both the development of new tools together with the use of these tools to 141 improve our comprehension of the nsLTP superfamily. Manuscript to be reviewed Several nsLtp 117 genes are up or down-regulated by application of different abiotic stresses including low 118 temperature, drought, salinity and wounding (Wang et al., 2012; Treviño & O’Connell, 1998; 119 Gaudet et al., 2003; Maghuly et al., 2009). A cabbage nsLTP isolated from leaves stabilizes PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) 144 MATERIALS & METHODS 145 1/ NsLTP sequences Manuscript to be reviewed 162 identified by Boutrot (Boutrot et al., 2008), 797 non-redundant mature amino acid sequences 163 belonging to more than 120 plant species were kept for analysis. This dataset is available online 164 (doi:10.18167/DVN1/UNKLA6). 165 162 identified by Boutrot (Boutrot et al., 2008), 7 163 belonging to more than 120 plant species we 164 (doi:10.18167/DVN1/UNKLA6). 165 166 Sequence alignments and phylogenetic ana 167 In order to achieve the best alignment, the po 168 MAFFT (Katoh et al., 2002; Katoh & Toh, 2 169 respective parameters of 1.53 for gap openin 170 maximum iteration 16. 171 The two resulting multiple alignments were c 172 highlighted. To discriminate between the two 173 section), a restricted analysis was carried out 174 structure had previously been experimentally 175 were calculated, one by MUSCLE and one b 176 Labesse, 2004), each alignment was projecte 177 spatial distance of equivalent cysteine residu 178 these distances was selected as the best one. 179 Based on the best alignment, a phylogenetic 180 2010). Lastly, the tree was reconciled with th 181 (Dufayard et al., 2005). 182 183 2/ NsLTP three-dimensional structures 184 185 Three-dimensional structure modeling 186 For 10 out of the 797 nsLTP dataset, one or m 187 available and downloaded from the Protein D 188 structures were calculated for the other 787 p 189 perform homology modeling (Pons & Labess 190 each final structure model was evaluated usin 191 low quality (i.e. for which the cysteine scaffo 192 further analysis (see Table 1). 193 194 Structural classification 195 All the remaining good-quality theoretical str 196 structures composed the 3D structure pool of 197 by ViTo, this structural pool was used in all f 198 The structures were compared to each other i 199 similarity matching method of the MAMMO 200 calculated for each pair of structures, using th 201 in the lowest RMSD value. This superpositio 202 (Kabsch, 1976; Kabsch, 1978) implemented 162 identified by Boutrot (Boutrot et al., 2008), 797 non-redundant mature amino acid sequences 163 belonging to more than 120 plant species were kept for analysis. This dataset is available online 164 (doi:10.18167/DVN1/UNKLA6). Manuscript to be reviewed 165 166 Sequence alignments and phylogenetic analysis 167 In order to achieve the best alignment, the pool of 797 sequences was aligned using both the 168 MAFFT (Katoh et al., 2002; Katoh & Toh, 2010) and MUSCLE (Edgar, 2004) programs with 169 respective parameters of 1.53 for gap opening, 0.123 for gap extension and BLOSUM62 matrix; 170 maximum iteration 16. 171 The two resulting multiple alignments were compared and conflicts between the two were 172 highlighted. To discriminate between the two different cysteine patterns suggested (see Results 173 section), a restricted analysis was carried out using only the 10 nsLTPs for which at least one 174 structure had previously been experimentally determined. Two new 10-sequence alignments 175 were calculated, one by MUSCLE and one by MAFFT. Using the ViTo program (Catherinot & 176 Labesse, 2004), each alignment was projected on type I, II and IV nsLTP 3D structures, and the 177 spatial distance of equivalent cysteine residues was evaluated. The alignment that minimized 178 these distances was selected as the best one. 179 Based on the best alignment, a phylogenetic tree was calculated using PhyML (Guindon et al., 180 2010). Lastly, the tree was reconciled with the overall species tree using the Rap-Green program 181 (Dufayard et al., 2005). 182 183 2/ NsLTP three-dimensional structures 184 185 Three-dimensional structure modeling 186 For 10 out of the 797 nsLTP dataset, one or more experimentally determined 3D structures were 187 available and downloaded from the Protein Data Bank (http://www.rcsb.org/pdb). Theoretical 188 structures were calculated for the other 787 proteins using the @tome2 suite of programs to 189 perform homology modeling (Pons & Labesse, 2009) (http://atome.cbs.cnrs.fr). The quality of 190 each final structure model was evaluated using Qmean (Benkert et al., 2008). Structures with 191 low quality (i.e. for which the cysteine scaffold could not be fully modeled) were discarded from 192 further analysis (see Table 1). 193 194 Structural classification 195 All the remaining good-quality theoretical structures, together with the 10 experimental 196 structures composed the 3D structure pool of the study. Except for the cysteine pattern analysis 197 by ViTo, this structural pool was used in all further structural analysis. 198 The structures were compared to each other in a sequence-independent manner, using the 199 similarity matching method of the MAMMOTH program (Ortiz et al., 2002). Manuscript to be reviewed The RMSD was 200 calculated for each pair of structures, using the superposition between matched pairs that resulted 201 in the lowest RMSD value. This superposition was computed using the Kabsch rotation matrix 202 (Kabsch, 1976; Kabsch, 1978) implemented in the MaxCluster program (Herbert A. 147 Definition of the protein sequence set 148 A first pool of plant nsLTPs sequences was retrieved from the UniProtKB (Swiss-Prot + 149 TrEMBL) (http://www.uniprot.org), Phytozome (http://www.phytozome.net) and NCBI 150 databases (http://www.ncbi.nlm.nih.gov), using either Blast or keyword queries (“Plant lipid 151 transfer protein”, “viridiplantae lipid transfer protein”, “plant A9 protein”, “A9 like protein”, 152 “tapetum specific protein”, “tapetum specific”, “anther specific protein”, “A9 Fil1”). Original 153 data obtained on the Theobroma cacao genome were also investigated (Argout et al., 2011). 154 From this large pool of proteins, the plant nsLTP dataset was defined according to a new set of 155 criteria: (i) sequences from 60 to 150 residues long, including signal peptide; (ii) containing 156 strictly eight cysteine residues after removal of the signal peptide; (iii) cysteine residues 157 distributed in the 8CM pattern (C-Xn-C-Xn-CC-Xn-CXC-Xn-C-Xn-C). We excluded multi- 158 domain proteins, i.e. the hybrid proline-rich and hybrid glycine-rich proteins in which the signal 159 peptide is followed by a proline-rich or a glycine-rich domain of variable length (José-Estanyol 160 et al., 2004). For each sequence, the signal peptide was detected and removed using SignalP 3.0 161 (Bendtsen et al., 2004). In all, including the wheat, rice and Arabidopsis sequences previously PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 203 unpublished). We used the RMSD score matrix calculated by MaxCluster as input for the 204 FastME program (Desper & Gascuel, 2002) to calculate a structural distance tree. 203 unpublished). We used the RMSD score matrix calculated by MaxCluster as input for the 204 FastME program (Desper & Gascuel, 2002) to calculate a structural distance tree. 205 206 3/ NsLTP functional annotation 207 Extensive bibliographic work was carried out to collect and classify functional information 208 available in the literature about the nsLTPs of the dataset. Gene Ontology (GO), Plant Ontology 209 (PO) and Trait Ontology (TO) terms were collected from the Gramene Ontologies Database 210 (http://www.gramene.org/plant_ontology) and organized in a dedicated database, together with 211 the bibliographic references when available. The database was also enriched with additional 212 information, such as methods used for gene expression studies (northern, RT-PCR or microarray 213 data, in situ hybridization), protein purification, in vitro or in planta antifungal and antibacterial 214 activity, lipid binding or transport (fluorescence binding assay or in vitro lipid transfer). 215 Information about tissues and organs used in cDNA libraries was collected from the NCBI 216 databases (http://www.ncbi.nlm.nih.gov). A dataset with all the annotations is available online 217 (http://dx.doi.org/10.18167/DVN1/1O5UAK. 218 219 4/Integrative method 1: sequence -> structure -> function 220 This method seeks to identify common ligand binding properties in nsLTPs clustered by 221 sequence similarity. 222 223 Sequence consensus for each nsLTP type 224 797 nsLTP sequences were clustered by type on the basis of regular expressions derived from the 225 consensus motifs described in (Boutrot et al., 2008). Each type subfamily was then aligned 226 individually and the resulting sequence profiles were globally aligned using MUSCLE. For each 227 type subfamily, the most frequent amino acids were selected at each alignment position to build 228 the consensus sequence. A consensus amino acid was replaced by a gap if more than half of the 229 sequences were aligned with a deletion at the considered position. 230 231 NsLTP sequence-structure analysis using Frequently Aligned Symbol Tree (FAST) 232 An original tool was designed to highlight conserved amino acid positions specific to each 233 nsLTP phylogenetic type, and which might be decisive for their function. The algorithm relied 234 on a statistical analysis of each alignment row, after the sequences had been clustered according 235 to their phylogenetic distances. 166 Sequence alignments and phylogenetic analysis 177 spatial distance of equivalent cysteine residues was evaluated. The alignment that minimized 178 these distances was selected as the best one. 179 Based on the best alignment, a phylogenetic tree was calculated using PhyML (Guindon et al., 180 2010). Lastly, the tree was reconciled with the overall species tree using the Rap-Green program 181 (Dufayard et al., 2005). 182 183 2/ NsLTP three-dimensional structures 184 179 Based on the best alignment, a phylogenetic tree was calculated using PhyML (Guindon et al., 180 2010). Lastly, the tree was reconciled with the overall species tree using the Rap-Green program 181 (Dufayard et al., 2005). 179 Based on the best alignment, a phylogenetic tree was calculated using PhyML (Guindon et al., 180 2010). Lastly, the tree was reconciled with the overall species tree using the Rap-Green program 181 (Dufayard et al., 2005). PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 236 For each type subfamily, the most frequent amino acids were selected at each alignment position 237 to build the consensus sequence. A consensus amino acid was replaced by a gap if more than half 238 of the sequences were aligned with a deletion at the considered position. We then calculated the 239 amino acid conservations and specificities over each column of the multiple sequence alignment 240 to delineate the functionally important residues in each nsLTP subfamily. This statistical analysis 241 is explained in Figure 3. Manuscript to be reviewed 242 In order to visualize the conserved and divergent regions of the sequences, different color ranges 243 were assigned to the nsLTP phylogenetic subfamilies. Conserved amino acid positions along the 244 whole alignment (CCP: Conserved Core Positions) are represented in grey/black, while 245 specifically conserved positions among proteins of the same subfamily (SDP: Specificity 246 Determining Positions) are represented in saturated colors corresponding to the family ones. The 247 tool enabled scrolling down of the alignment to easily identify both types of conserved positions 248 and two distant parts of the alignment could be displayed together to compare distant 249 phylogenetic subfamilies. 247 tool enabled scrolling down of the alignment to easily identify both types of conserved positions 248 and two distant parts of the alignment could be displayed together to compare distant 249 phylogenetic subfamilies. 250 Contacts with ligands, solvent accessibility and other parameters could also be displayed above 251 the alignment. Using the Jmol interface, conserved amino acid residues could be projected on 252 nsLTP representative 3D structures, so that the potential role of each position could be 253 interpreted geometrically. 254 255 5/ Integrative method 2: function -> structure -> sequence 256 Structural Trace Display is a method, based on Evolutionary Trace Display (ETD, Erdin et 257 al.,2010), that seeks to identify common structural (1D, 3D) properties in nsLTPs sharing similar 258 functions. 259 260 Clustering of the structure tree 261 As in a phylogenetic tree, nsLTPs in the structure tree were clustered according to their 262 similarity. In the case of this particular tree, the similarity between nsLTPs was measured by a 263 spatial distance in angströms (see paragraph 2/ NsLTP three-dimensional structures / Structural 264 classification). Decreasing distance cutoffs ranged from 11.5 Å (one cluster containing all nsLTP 265 structures) to 0.5 Å. Each cutoff application caused a division of the tree into one or more sub- 266 trees that contained leaves (i.e. nsLTP structures) whose structural proximity altogether 267 (represented by the pairwise RMSDs) was up to the value of the applied cutoff. 268 269 InTreeGreat: an integrative tree visualization tool 270 We developed an integrative tree visualization tool called InTreeGreat in order to display the 271 whole or some parts of either sequence or structure distance trees. 272 InTreeGreat was implemented using PHP and Javascript, in order to generate and manipulate an 273 SVG graphical object. Manuscript to be reviewed 274 The main objective of this tool is to graphically highlight correlations between 3D structures, 275 evolution, functional annotations or any available heterogeneous data. In the context of this 276 study, the interface was able to retrieve information from the nsLTP database to annotate the 277 tree. 278 InTreeGreat includes functionalities such as tree coloration, fading, and collapsing. 279 Heterogeneous data related to sequences (e.g. annotations, nsLTP classification) can be 280 displayed in colored boxes, aligned to the tree. 281 282 Cluster Selection 250 Contacts with ligands, solvent accessibility and other parameters could also be displayed above 251 the alignment. Using the Jmol interface, conserved amino acid residues could be projected on 252 nsLTP representative 3D structures, so that the potential role of each position could be 253 interpreted geometrically. sequence-structure analysis using Frequently Aligned Symbol Tree (FAST) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) 255 5/ Integrative method 2: function -> structure -> sequence 256 Structural Trace Display is a method, based on Evolutionary Trace Display (ETD, Erdin et 257 al.,2010), that seeks to identify common structural (1D, 3D) properties in nsLTPs sharing similar 258 functions. Manuscript to be reviewed 283 Using InTreeGreat to investigate our annotated structure tree, we looked for clusters of nsLTPs 284 sharing the same kind of functional annotations. We focused our attention on one interesting 285 functional role highlighted in several nsLTPs: the implication in plant defense mechanisms 286 against pathogens (bacteria and/or fungus). In order to highlight structure-function relationships 287 we studied three groups of nsLTPs (see Results section for details): (i) the “defense cluster” (43 288 proteins, distance cutoff = 1.5 Å); (ii) the cluster containing all type I fold proteins (402 protein 289 distance cutoff = 3 Å); (iii) a group manually composed of all type I fold nsLTPs for which a 290 functional role in defense and/or resistance against pathogens had been reported in the literature 291 (28 proteins). 292 Within each of these 3 clusters, the protein structure showing the shortest RMS distance from al 293 the others was selected as the representative structure of the cluster for the structural trace 294 calculation. 295 296 Structural Trace calculation 297 A structure-based sequence alignment was carried out on the nsLTP structures by Mustang 298 software (Konagurthu et al., 2006). 299 For each previously selected structural cluster, the corresponding set of protein sequences was 300 extracted from the multiple structural alignment of the nsLTPs. The Evolutionary Trace (ET) 301 method (Lichtarge et al., 1996) was applied: the partial multiple sequence alignment was 302 submitted as input for the ETC program (locally installed, 303 http://mammoth.bcm.tmc.edu/ETserver.html) together with the representative structure of the 304 cluster (selected as described in the previous paragraph). 305 The “evolutionary” traces based on the structural alignments corresponding to the three nsLTP 306 clusters were then compared to each other. To that end, the 30% top-ranked residues of the 307 defense cluster trace were considered as constitutive of the reference trace (i.e. 27 most 308 conserved amino acid residues) and their ranking and scores in the two other traces were 309 analyzed. The results were compiled in a table and graphically visualized using PyMOL 310 (http://www.pymol.org/). 311 312 313 RESULTS 314 1/ NsLTP sequences analysis 315 316 NsLTP dataset 317 Over the last four decades numerous proteins, whose ability to transfer lipids has not always 318 been demonstrated, have been annotated as nsLTPs on the basis of sequence similarity. 260 Clustering of the structure tree 261 As in a phylogenetic tree, nsLTPs in the structure tree were clustered according to their 262 similarity. In the case of this particular tree, the similarity between nsLTPs was measured by a 263 spatial distance in angströms (see paragraph 2/ NsLTP three-dimensional structures / Structural 264 classification). Decreasing distance cutoffs ranged from 11.5 Å (one cluster containing all nsLTP 265 structures) to 0.5 Å. Each cutoff application caused a division of the tree into one or more sub- 266 trees that contained leaves (i.e. nsLTP structures) whose structural proximity altogether 267 (represented by the pairwise RMSDs) was up to the value of the applied cutoff. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed In order 319 to understand more clearly the functional characteristics and the inner variability of this family, 320 we focused the study on the monodomain proteins, which present the strict and only nsLTP 321 domain, i.e. the eight-cysteine residues arranged in four disulfide bridges. In total, including the 322 wheat, rice and Arabidopsis sequences previously identified (Boutrot et al., 2008), together with 303 http://mammoth.bcm.tmc.edu/ETserver.html) together with the representative structure of the 304 cluster (selected as described in the previous paragraph). PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed For these reasons we think that the sequence alignment calculated 351 MAFFT is more relevant. 352 This compound approach allowed us to sort the 797 sequences unambiguously into two m 353 families. 354 355 NsLTP sequence classification 356 Our dataset was mainly composed of nsLTPs from angiosperm species (19 monocotyledon 357 species and 83 eudicotyledonous species) plus five gymnosperm species (35 sequences), o 358 lycophyte species (34 sequences) and two bryophyte species (17 sequences). The monocot 359 i l t d b P l LTP (256 t f 270 ) h 323 sequences from the UniProt (Swiss-Prot/TrEMBL), NCBI and Phytozome databases, 797 non- 324 redundant mature nsLTP sequences belonging to more than 120 plant species were kept for 325 analysis. This first step allowed the selection of a relevant set of proteins covering variability in 326 the nsLTP family. The number of sequences (797) was also large enough to challenge any 327 analysis method we used during this study. 328 329 Sequence alignment and Cysteine pattern 330 The alignment of all non-redundant protein sequences for which the 3D structure was 331 experimentally determined (10 sequences) was carried out twice, using the MUSCLE program 332 on the one hand, and the MAFFT program on the other hand. The resulting alignments obtained 333 with standard settings are shown on Figures 1A and 1B. 334 In both cases, cysteine residues of the 8CM aligned quite well among the three represented types 335 of nsLTPs (types I, II and IV), except for the Cys5-X-Cys6 (CXC) pattern region (where X 336 stands for any amino acid residue). MUSCLE did align type I Cys5 with types II and IV Cys5’, 337 as well as type I Cys6 with types II and IV Cys6’ (Figure 1A), just as previous studies typically 338 showed (Liu et al., 2010; Siverstein et al., 2007). However, in the alignment carried out by 339 MAFFT (Figure 1B), Cys5 of type I nsLTPs was equivalent to Cys6’ of type II and IV nsLTPs, 340 and not to the corresponding Cys5’. 341 To determine which of these two alignments better aligned nsLTP type I and II CXC patterns, 342 we used ViTO program to compare the effects on the 3D structures of both sequence alignments. Manuscript to be reviewed 343 The 3D projection of MUSCLE sequence alignment (Figure 1C) showed that the attempt to 344 spatially superimpose Cys5 and Cys6 of type I nsLTPs with Cys5’ and Cys6’ of type II nsLTPs 345 respectively, was very expensive in term of Cys C-alpha RMSD value (7.32 Ǻ). 346 On the contrary, according to the 3D projection of MAFFT sequence alignment (Figure 1D), 347 type I Cys5 could be well superimposed with type II Cys6’, with a Cys C-alpha RMSD value 348 dropping to 2.15 Å. Furthermore, type I hydrophylic X residue was exposed to the solvent, 349 whereas type II apolar X residue was orientated toward the core of the protein, increasing the 350 stability of the proteins. For these reasons we think that the sequence alignment calculated by 351 MAFFT is more relevant. 352 This compound approach allowed us to sort the 797 sequences unambiguously into two main 353 families. 354 355 NsLTP sequence classification 356 Our dataset was mainly composed of nsLTPs from angiosperm species (19 monocotyledonous 357 species and 83 eudicotyledonous species) plus five gymnosperm species (35 sequences) one 323 sequences from the UniProt (Swiss-Prot/TrEMBL), NCBI and Phytozome databases, 797 non- 324 redundant mature nsLTP sequences belonging to more than 120 plant species were kept for 325 analysis. This first step allowed the selection of a relevant set of proteins covering variability in 326 the nsLTP family. The number of sequences (797) was also large enough to challenge any 327 analysis method we used during this study. 328 323 sequences from the UniProt (Swiss-Prot/TrEMBL), NCBI and Phytozome databases, 797 non- 324 redundant mature nsLTP sequences belonging to more than 120 plant species were kept for 325 analysis. This first step allowed the selection of a relevant set of proteins covering variability in 326 the nsLTP family. The number of sequences (797) was also large enough to challenge any 327 analysis method we used during this study. Manuscript to be reviewed 323 sequences from the UniProt (Swiss-Prot/TrEMBL), NCBI and Phytozome databases, 797 324 redundant mature nsLTP sequences belonging to more than 120 plant species were kept fo 325 analysis. This first step allowed the selection of a relevant set of proteins covering variabil 326 the nsLTP family. The number of sequences (797) was also large enough to challenge any 327 analysis method we used during this study. 328 329 Sequence alignment and Cysteine pattern 330 The alignment of all non-redundant protein sequences for which the 3D structure was 331 experimentally determined (10 sequences) was carried out twice, using the MUSCLE prog 332 on the one hand, and the MAFFT program on the other hand. The resulting alignments obt 333 with standard settings are shown on Figures 1A and 1B. 334 In both cases, cysteine residues of the 8CM aligned quite well among the three represented 335 of nsLTPs (types I, II and IV), except for the Cys5-X-Cys6 (CXC) pattern region (where X 336 stands for any amino acid residue). MUSCLE did align type I Cys5 with types II and IV C 337 as well as type I Cys6 with types II and IV Cys6’ (Figure 1A), just as previous studies typi 338 showed (Liu et al., 2010; Siverstein et al., 2007). However, in the alignment carried out by 339 MAFFT (Figure 1B), Cys5 of type I nsLTPs was equivalent to Cys6’ of type II and IV nsL 340 and not to the corresponding Cys5’. 341 To determine which of these two alignments better aligned nsLTP type I and II CXC patte 342 we used ViTO program to compare the effects on the 3D structures of both sequence align 343 The 3D projection of MUSCLE sequence alignment (Figure 1C) showed that the attempt t 344 spatially superimpose Cys5 and Cys6 of type I nsLTPs with Cys5’ and Cys6’ of type II ns 345 respectively, was very expensive in term of Cys C-alpha RMSD value (7.32 Ǻ). 346 On the contrary, according to the 3D projection of MAFFT sequence alignment (Figure 1D 347 type I Cys5 could be well superimposed with type II Cys6’, with a Cys C-alpha RMSD va 348 dropping to 2.15 Å. Furthermore, type I hydrophylic X residue was exposed to the solvent 349 whereas type II apolar X residue was orientated toward the core of the protein, increasing 350 stability of the proteins. 329 Sequence alignment and Cysteine pattern 330 The alignment of all non-redundant protein sequences for which the 3D structure was 331 experimentally determined (10 sequences) was carried out twice, using the MUSCLE program 332 on the one hand, and the MAFFT program on the other hand. The resulting alignments obtained 333 with standard settings are shown on Figures 1A and 1B. 334 In both cases, cysteine residues of the 8CM aligned quite well among the three represented types 335 of nsLTPs (types I, II and IV), except for the Cys5-X-Cys6 (CXC) pattern region (where X 336 stands for any amino acid residue). MUSCLE did align type I Cys5 with types II and IV Cys5’, 337 as well as type I Cys6 with types II and IV Cys6’ (Figure 1A), just as previous studies typically 338 showed (Liu et al., 2010; Siverstein et al., 2007). However, in the alignment carried out by 339 MAFFT (Figure 1B), Cys5 of type I nsLTPs was equivalent to Cys6’ of type II and IV nsLTPs, 340 and not to the corresponding Cys5’. 337 as well as type I Cys6 with types II and IV Cys6’ (Figure 1A), just as previous studies typically 338 showed (Liu et al., 2010; Siverstein et al., 2007). However, in the alignment carried out by 339 MAFFT (Figure 1B), Cys5 of type I nsLTPs was equivalent to Cys6’ of type II and IV nsLTPs, 340 and not to the corresponding Cys5’. 339 MAFFT (Figure 1B), Cys5 of type I nsLTPs was equivalent to Cys6’ of type II and IV nsLTPs, 340 and not to the corresponding Cys5’. 341 To determine which of these two alignments better aligned nsLTP type I and II CXC patterns, 342 we used ViTO program to compare the effects on the 3D structures of both sequence alignments. 343 The 3D projection of MUSCLE sequence alignment (Figure 1C) showed that the attempt to 344 spatially superimpose Cys5 and Cys6 of type I nsLTPs with Cys5’ and Cys6’ of type II nsLTPs 345 respectively, was very expensive in term of Cys C-alpha RMSD value (7.32 Ǻ). 352 This compound approach allowed us to sort the 797 sequences unambiguously into two main 353 families. Manuscript to be reviewed 363 defined on A. thaliana, T. aestivum and O. sativa nsLTP sequences, in nine types (Boutrot et al., 364 2008). The main differences were the identification of a new group (named type XI), including 365 23 sequences, and that Boutrot’s type VII nsLTPs disappeared from our dataset. Indeed, the 366 latter did not satisfy the 8CM criteria as they have only seven cysteine residues in their 367 sequences. For the same reason, Wang’s A, B, C and D types (Wang et al., 2012) were not 368 represented in our classification. 369 370 Type I nsLTPs formed a well-supported monophyletic group (branch support of 0.879) and 371 predominated over the other types, as they accounted for more than half of our dataset (417 out 372 of 797 sequences). This was also observed by Wang (Wang et al., 2012) with a set of 595 373 nsLTPs. Conversely, in Solanaceae, the most abundant nsLTPs belong to a type referred to as 374 type X by Wang (70 out of 135 sequences) and which seems specific to that plant family (Liu et 375 al., 2010) but was not present in our dataset. To avoid any confusion, we did not used type X 376 denomination in this work. Type II nsLTPs were the second most abundant type (126 sequences) 377 followed by type V (70 sequences) and type VI (60 sequences). Type IX (12 sequences) was 378 mainly composed of Physcomitrella patens nsLTPs and type XI (23 sequences) was mainly 379 composed of nsLTPs from eudicot species. Twelve nsLTPs were not included in any of the 380 identified types: these were mainly P. patens (6 sequences) and S. moellendorfii (4 sequences) 381 proteins (Figure 2). 382 383 Type XI were grouped in a cluster of 23 sequences in the phylogenetic tree, fairly well supported 384 by a branch of 0.879 aLRT SH-like score. Type XI appeared between type I and the other types, 385 but even though type XI and I were grouped together in the tree, it remained unclear which of the 386 3 groups (type I, type XI, and other types) diverged first. 387 388 All nsLTP types were represented in eudicots while types IX, X (in Wang’s nomenclature) and 389 XI were not identified in monocot species. Within the lycophyte and bryophyte species, no type 390 II, III, IV nor VIII nsLTPs were identified. 355 NsLTP sequence classification 356 Our dataset was mainly composed of nsLTPs from angiosperm species (19 monocotyledonous 357 species and 83 eudicotyledonous species) plus five gymnosperm species (35 sequences), one 358 lycophyte species (34 sequences) and two bryophyte species (17 sequences). The monocot 359 sequences were mainly represented by Poales nsLTPs (256 out of 270 sequences) whereas Rosid 360 nsLTPs were the most abundant (364 out of 436 sequences) within eudicots. 361 The phylogenetic analysis showed that the pool of proteins clustered into nine different types, all 362 highly supported (branch support >0.84). This result mostly confirmed Boutrot’s classification, 361 The phylogenetic analysis showed that the pool of proteins clustered into nine different types, all 362 highly supported (branch support >0.84). This result mostly confirmed Boutrot’s classification, PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 402 According to Yi and coworkers (Yi et al., 2009), Allium nsLTPs may constitute a novel type of 403 nsLTPs harboring a C-terminal pro-peptide localized in endomembrane compartments. In the 404 prolamin superfamily tree of Radauer and Breiteneder (Radauer & Breiteneder, 2007), the Allium 405 cepa nsLTP (192_ALLCE) is closed but not included in the type I nsLTPs. In our phylogenetic 406 tree, the three nsLTPs from Allium species were classified as type I. The 501_MEDTR medicago 407 nsLTP was suggested to belong to a new nsLTP subfamily involved in lipid signaling (Pii et al., 408 2010) like Arabidopsis DIR1 (151_typeIV_ARATH). In our phylogenetic tree, both proteins 409 were identified as type IV nsLTPs. 410 411 The Theobroma cacao genome contains at least 46 nsLtp genes distributed across the ten 412 chromosomes. Several T. cacao nsLtp genes are organized in clusters, as observed in the rice, 413 Arabidopsis and sorghum genomes (Boutrot et al., 2008; Wang et al., 2012). Apart from nine 414 sequences that were classified in the new type XI, all other T. cacao nsLTPs were classified 415 within the previously identified types and belonged mainly to type I (14 sequences), type VI (7 416 sequences) and type V (6 sequences). 417 418 It is worth noting that all the nsLTPs identified as allergens (IgE binding) were type I, except one 419 type II nsLTP (545_BRACM). The 501_MEDTR nsLTP was also suggested to play a role in the 420 root nodulation process (Pii et al., 2009; Pii et al., 2013). Lipid signaling (lyso- 421 phosphatidylcholine) has been reported to be involved in symbiosis (Bucher et al., 2009). 422 This analysis was the most extensive so far and confirmed most of Boutrot’s classification, but 423 complements it due to a larger dataset and a more detailed phylogeny analysis. 424 425 2/ NsLTP structure analysis 426 427 NsLTP structure modeling 428 Given the nsLTP fold conservation and the quality of the available experimental structures, 429 reliable models could be obtained for all nsLTPs using the comparative modeling method, 430 although the sequence identity observed among all nsLTP sequences was only in the range of 431 25%. 432 Models deduced by fold-recognition using the @TOME-2 server displayed overall good quality, 433 as shown in Table 1 summarizing the Qmean scores. Manuscript to be reviewed In the same way, no type III, VIII, IX or XI were 391 identified within gymnosperm species. Ten out of the 16 moss P. patens nsLTPs were type IX, 392 the other 6 remained un-typed, and the only liverwort Marchantia polymorpha nsLTP was a type 393 VI. The 34 S. moellendorfii sequences were mainly types V and VI (15 and 7, respectively) and 394 seven nsLTPs belonged to the new type XI. The P. patens and S. moellendorfii nsLTPs formed 395 independent branches or were located at the same branch as type V in Wang’s phylogenetic tree 396 (Wang et al., 2012) and were included in type D in Edstam’s classification (Edstam et al., 2011). 397 However, Edstam’s type D included rice and Arabidopsis type IV, V and VI nsLTPs. Edstam’s 398 type G nsLTPs, which corresponded to GPI-anchored LTPs and types J and K, which did not fit 399 our molecular mass criteria or contain more than one 8CM motif were not included in our 400 dataset. 401 Manuscript to be review 363 defined on A. thaliana, T. aestivum and O. sativa nsLTP sequences, in nine types (Boutrot et al., 364 2008). The main differences were the identification of a new group (named type XI), including 365 23 sequences, and that Boutrot’s type VII nsLTPs disappeared from our dataset. Indeed, the 366 latter did not satisfy the 8CM criteria as they have only seven cysteine residues in their 367 sequences. For the same reason, Wang’s A, B, C and D types (Wang et al., 2012) were not 368 represented in our classification. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 442 NsLTP sequence – structure relationships 443 In order to challenge the structure – function relationship analysis on such a big set, we decided 444 to develop a new tool called FAST, which builds consensus sequences for each family, and 445 highlights the sequence conservation and specificities on the alignment and the associated 3D 446 structures. 447 Figure 3 shows the consensus sequence alignment for all nsLTP types. The pool of 797 448 sequences was clustered by type on the basis of regular expressions derived from the consensus 449 motifs described by Boutrot and coworkers (Boutrot et al., 2008). Each type subfamily was then 450 aligned individually and the resulting sequence profiles were globally aligned using MUSCLE. 451 452 Many residues specifically conserved in type I nsLTP1 corresponded to important folding 453 differences between type I nsLTPs on the one side and all other LTP types on the other side. In 454 the following sections, we list type I nsLTP-specific residues whose differential conservation 455 was supported by structural or experimental data. 456 457 First, Gly37, which was specifically conserved in type I nsLTPs, allowed very tight contact of 458 helix 1 and helix 2, which were connected by the disulfide bridge Cys17-Cys34. The closest 459 backbone distance between position 13 of helix 1 and position 37 of helix 2 was 3.34 Å in a type 460 I nsLTP structure (PDB code 1mid) while it was 6.45 Å in a type II nsLTP structure (PDB code 461 1tuk). These increased helix distances closed the ligand tunnel, which was opened in type I 462 nsLTPs between helix 1 and helix 3, and created two distinct cavities separated by a septum in 463 type II nsLTPs (Hoh et al., 2005). Larger distances between helix 1 and helix 2 were predicted in 464 all nsLTP sequences where Gly37 was mutated into larger residues (i.e. all types but I and XI) 465 and should cause major rearrangement of the ligand cavity entrance on this side of the proteins. 466 Arginine and lysine residues at position 51 and bulky hydrophobic residues at positions 87 and 467 89 were two other conserved specificities among type I nsLTPs. The side chains at position 51 468 had type I-specific polar interactions with the ligand at the cavity entrance near the C-terminal 469 loop, which were not found in other nsLTP types, as detailed later in Figure 4. Manuscript to be reviewed For 96% of the models, Qmean scores were 434 above 0.4, and 57% of the models obtained scores ranging from 0.5 to 0.9., corresponding to 435 scores for high-resolution proteins. 436 For 121 theoretical structures, the polypeptide chain could not be fully built and the resulting 437 models were lacking at least one of the 8 cysteine residues. Such models were discarded and 438 l th l t l f 677 t t k t f f th l i 402 According to Yi and coworkers (Yi et al., 2009), Allium nsLTPs may constitute a novel type of 403 nsLTPs harboring a C-terminal pro-peptide localized in endomembrane compartments. In the 404 prolamin superfamily tree of Radauer and Breiteneder (Radauer & Breiteneder, 2007), the Allium 405 cepa nsLTP (192_ALLCE) is closed but not included in the type I nsLTPs. In our phylogenetic 406 tree, the three nsLTPs from Allium species were classified as type I. The 501_MEDTR medicago 407 nsLTP was suggested to belong to a new nsLTP subfamily involved in lipid signaling (Pii et al., 408 2010) like Arabidopsis DIR1 (151_typeIV_ARATH). In our phylogenetic tree, both proteins 409 were identified as type IV nsLTPs. 410 holine) has been reported to be involved in symbiosis (Bucher et al., 2009). 422 This analysis was the most extensive so far and confirmed most of Boutrot’s classification, but 423 complements it due to a larger dataset and a more detailed phylogeny analysis. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Marginally, the most frequent 501 polar contacts involved the side chains of conserved arginines at position 46 of the type I nsLTP 502 alignment, lysines at position 54, aspartic acids at position 90, and various polar atoms of 503 histidines, lysines and asparagines at position 37. It should be stressed that none of these polar 504 interactions were shared by more than 31% of the protein-ligand complexes (fewer than 6/19 505 PDB structures) although the least similar protein pair from the 19 structure set shared 67% 506 sequence identities. This low level of polar contact conservation in homologous proteins with 507 very similar sequences clearly indicated that no specific polar interactions anchored the protein- 508 ligand complexes in particular conformations. From this statistical analysis of protein-ligand 509 polar contacts that did not exhibit a preferential cavity region for the interaction with the ligand 510 polar heads, it could not be concluded that there was a preferred ligand orientation in the type I 511 nsLTP tunnel. This observation was supported by recent protein-docking simulations and protein 512 binding evaluations, which also concluded on a lack of preferred orientations of the ligand in the 513 cavities of type I nsLTPs, and clear dominance of hydrophobic interactions in the protein-ligand 514 interface (Pacios et al., 2012). 515 16 Lastly, positions 82 to 94, which corresponded to the C-terminal loop, included some more 17 residues specifically conserved in nsLTPs. This loop was much longer in type I nsLTPs than in 18 other types, and had a major impact on the orientation of the ligand in the cavity, as shown in 19 Figure 4. Manuscript to be reviewed 482 in type I nsLTPs, while larger hydrophobic residues almost always occupied this buried position 483 in other nsLTP types. 484 485 The mutation to alanine of the residue at position 63 (residue 45 in Cheng's article) was also 486 shown experimentally to be destabilizing in type II nsLTPs (Cheng et al., 2008). This position 487 was occupied by large hydrophobic residues in all nsLTPs but types I and V, where alanine 488 residues were frequent, and type III, where it corresponded to a deletion of 12 consecutive 489 residues. 490 Other residues specifically conserved in type I nsLTPs were helix N-capping Thr6 and Thr47, 491 whose side chains formed stabilizing hydrogen bonds with the protein backbone, and Tyr20, 492 which was the center of a conserved hydrophobic cluster with Pro30 and Leu/Ile79. The 493 interaction of Tyr20 with Pro30 was experimentally confirmed by the large up field shift of 494 Pro30 (Hα, Hδ) protons (Poznanski et al., 1999). This conserved cluster was stabilizing the 495 interface between helices 1 and 4, but did not participate in the ligand cavity. This particular 496 helix interface was also observed in nsLTP types III, VI, VIII and XI. 497 498 We then analyzed the atomic interactions observed between type I nsLTPs and their associated 499 ligands in 19 PDB structures (supplementary data). Most contacts involved hydrophobic side 500 chains of the type I nsLTP proteins and carbons of the ligands. Marginally, the most frequent 501 polar contacts involved the side chains of conserved arginines at position 46 of the type I nsLTP 502 alignment, lysines at position 54, aspartic acids at position 90, and various polar atoms of 503 histidines, lysines and asparagines at position 37. It should be stressed that none of these polar 504 interactions were shared by more than 31% of the protein-ligand complexes (fewer than 6/19 505 PDB structures) although the least similar protein pair from the 19 structure set shared 67% 506 sequence identities. This low level of polar contact conservation in homologous proteins with 507 very similar sequences clearly indicated that no specific polar interactions anchored the protein- 508 ligand complexes in particular conformations. Manuscript to be reviewed 470 471 In addition, in type I nsLTPs, the 5th and 6th cysteine residues belonged to helix 3 and were 472 bridged with the first and 8th cysteines, respectively. These two-disulfide bridges tightened both 473 sequence termini to the protein core. Conversely, in types II and IV nsLTPs, the 5th and 6th 474 cysteines showed permuted bridging partners (to 8th and 1st cysteines, respectively). The 475 intermediate residue connecting the 5th and 6th cysteines was exposed to solvent in type I 476 nsLTPs, while it was replaced by a bulky hydrophobic residue interacting with the ligand in the 477 type II and IV nsLTP core at position 54 of the alignment. It was shown by site-directed 478 mutagenesis that the replacement of this intermediate residue by an alanine residue perturbed 442 NsLTP sequence – structure relationships 471 In addition, in type I nsLTPs, the 5th and 6th cysteine residues belonged to helix 3 and were 472 bridged with the first and 8th cysteines, respectively. These two-disulfide bridges tightened both 473 sequence termini to the protein core. Conversely, in types II and IV nsLTPs, the 5th and 6th 474 cysteines showed permuted bridging partners (to 8th and 1st cysteines, respectively). The 475 intermediate residue connecting the 5th and 6th cysteines was exposed to solvent in type I 476 nsLTPs, while it was replaced by a bulky hydrophobic residue interacting with the ligand in the 477 type II and IV nsLTP core at position 54 of the alignment. It was shown by site-directed 478 mutagenesis that the replacement of this intermediate residue by an alanine residue perturbed 479 folding, ligand binding and lipid transfer activity in type II nsLTPs (Cheng et al., 2008). The 480 position 54 in our alignment corresponds to residue 36 in Cheng's article. In the light of these 481 experiments, it is therefore interesting to note that alanine residues were frequent at position 54 PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed From this statistical analysis of protein-ligand 509 polar contacts that did not exhibit a preferential cavity region for the interaction with the ligand 510 polar heads, it could not be concluded that there was a preferred ligand orientation in the type I 511 nsLTP tunnel. This observation was supported by recent protein-docking simulations and protein 512 binding evaluations, which also concluded on a lack of preferred orientations of the ligand in the 513 cavities of type I nsLTPs, and clear dominance of hydrophobic interactions in the protein-ligand 514 interface (Pacios et al., 2012). 515 516 Lastly, positions 82 to 94, which corresponded to the C-terminal loop, included some more 517 residues specifically conserved in nsLTPs. This loop was much longer in type I nsLTPs than in 482 in type I nsLTPs, while larger hydrophobic residues almost always occupied this buried position 483 in other nsLTP types. 484 485 The mutation to alanine of the residue at position 63 (residue 45 in Cheng's article) was also 486 shown experimentally to be destabilizing in type II nsLTPs (Cheng et al., 2008). This position 487 was occupied by large hydrophobic residues in all nsLTPs but types I and V, where alanine 488 residues were frequent, and type III, where it corresponded to a deletion of 12 consecutive 489 residues. 490 Other residues specifically conserved in type I nsLTPs were helix N-capping Thr6 and Thr47, 491 whose side chains formed stabilizing hydrogen bonds with the protein backbone, and Tyr20, 492 which was the center of a conserved hydrophobic cluster with Pro30 and Leu/Ile79. The 493 interaction of Tyr20 with Pro30 was experimentally confirmed by the large up field shift of 494 Pro30 (Hα, Hδ) protons (Poznanski et al., 1999). This conserved cluster was stabilizing the 495 interface between helices 1 and 4, but did not participate in the ligand cavity. This particular 496 helix interface was also observed in nsLTP types III, VI, VIII and XI. 497 482 in type I nsLTPs, while larger hydrophobic residues almost always occupied this buried position 483 in other nsLTP types. 484 498 We then analyzed the atomic interactions observed between type I nsLTPs and their associated 499 ligands in 19 PDB structures (supplementary data). Most contacts involved hydrophobic side 500 chains of the type I nsLTP proteins and carbons of the ligands. Manuscript to be reviewed Lastly, the capping 545 hydrophic residues at positions 54 and 51 of types II and IV nsLTPs were also observed in all the 546 other nsLTP types. These conserved differences between type I and other types of nsLTP 547 sequences indicated with high confidence that the global fold of type I LTP differed from the 548 fold of the other nsLTP types and that the ligand cavity entries in type I nsLTPs were uniquely 549 located. 550 The fold of type I nsLTPs will be hereafter referred to as “Type-1 fold” and the alternative fold 551 of Types II to XI will be referred to as “Type-2 fold”. (in other words: Roman numerals I to XI 552 correspond to phylogeny analysis while Arabic numerals 1 or 2 refer to structural analysis) 553 554 The preceding analysis of the evolutive conservations specific to type I nsLTPs revealed many 555 residues whose role could be explained by local structural differences with the available types II 556 and IV nsLTP structures. This comparative structure analysis confirmed the clear separation 557 between type I and all the other nsLTP types initially observed in the phylogenetic tree inferred 558 from a multiple sequence alignment of the 797 available proteins. The key residues were usually 559 present in type I nsLTPs only and suggested that many structural differences observed when 560 comparing type I versus types II and IV nsLTPs should also be observed versus other nsLTP 522 The potential impact of variability within the nsLTP family on the tree dimensional structure of 523 the proteins was further investigated. As shown in Figure 4, the ligand cavity opening near the C- 524 Terminal loop was very different when we compared the nsLTP structures of type I versus those 525 of types II and IV. The C-terminal loops connected the 4 helices to the 3 helices through the 526 disulfide bridge between cysteine residues localized at alignment positions 95 and 55. Both 527 helices 2 and 3 and the C-terminal loop were longer in type I than in types II and IV nsLTPs. In 528 the type I nsLTPs, these elongations created a ligand cavity entrance along an axis perpendicular 529 to the figure plane, while in types II and IV nsLTPs, the entrance was approximately parallel 530 with the figure plane. Manuscript to be reviewed Consequently, ligands would access the cavities on opposite sides of the C- 531 terminal loop in type I versus types II and IV nsLTPs. Helix 2 and 3 were extended by an extra 532 turn in type I nsLTPs comparatively to the structures of the other types. Moreover, the small 533 space left in between helices 2 and 3 and the C-term loop was capped in types II and IV by bulky 534 hydrophobic residues (Phe54 in 1tuk and Phe51 in 2rkn), while that position was occupied by a 535 positively charged lysine or arginine in type I nsLTPs (red colored Arg51 in 1mid), whose side 536 chain formed a hydrogen bond with the polar tail of the ligand. 537 538 The structural differences observed between type I nsLTPs versus types II and IV can be 539 generalized to other nsLTP types by looking at the alignment of consensus sequences in Figure 540 3. First, the extension of helices 2 and 3 in type I nsLTPs corresponded to a 6- to 8-residue 541 insertion in the consensus sequence alignment, which differentiated type I from every other type 542 of nsLTPs. Secondly, the C-terminal loop connecting the last two cysteine residues was, on 543 average, 13 residues long in type I nsLTPs, while this loop was shortened to 6, 6, 7, 12, 9, 8, 6 544 and 9 residues long in types II, III, IV, V, VI, VIII, IX and XI, respectively. Lastly, the capping 545 hydrophic residues at positions 54 and 51 of types II and IV nsLTPs were also observed in all the 546 other nsLTP types. These conserved differences between type I and other types of nsLTP 547 sequences indicated with high confidence that the global fold of type I LTP differed from the 548 fold of the other nsLTP types and that the ligand cavity entries in type I nsLTPs were uniquely 549 located. 550 The fold of type I nsLTPs will be hereafter referred to as “Type-1 fold” and the alternative fold 551 of Types II to XI will be referred to as “Type-2 fold”. Manuscript to be reviewed 522 The potential impact of variability within the nsLTP family on the tree dimensional structure of 523 the proteins was further investigated. As shown in Figure 4, the ligand cavity opening near the C- 524 Terminal loop was very different when we compared the nsLTP structures of type I versus those 525 of types II and IV. The C-terminal loops connected the 4 helices to the 3 helices through the 526 disulfide bridge between cysteine residues localized at alignment positions 95 and 55. Both 527 helices 2 and 3 and the C-terminal loop were longer in type I than in types II and IV nsLTPs. In 528 the type I nsLTPs, these elongations created a ligand cavity entrance along an axis perpendicular 529 to the figure plane, while in types II and IV nsLTPs, the entrance was approximately parallel 530 with the figure plane. Consequently, ligands would access the cavities on opposite sides of the C- 531 terminal loop in type I versus types II and IV nsLTPs. Helix 2 and 3 were extended by an extra 532 turn in type I nsLTPs comparatively to the structures of the other types. Moreover, the small 533 space left in between helices 2 and 3 and the C-term loop was capped in types II and IV by bulky 534 hydrophobic residues (Phe54 in 1tuk and Phe51 in 2rkn), while that position was occupied by a 535 positively charged lysine or arginine in type I nsLTPs (red colored Arg51 in 1mid), whose side 536 chain formed a hydrogen bond with the polar tail of the ligand. 537 538 The structural differences observed between type I nsLTPs versus types II and IV can be 539 generalized to other nsLTP types by looking at the alignment of consensus sequences in Figure 540 3. First, the extension of helices 2 and 3 in type I nsLTPs corresponded to a 6- to 8-residue 541 insertion in the consensus sequence alignment, which differentiated type I from every other type 542 of nsLTPs. Secondly, the C-terminal loop connecting the last two cysteine residues was, on 543 average, 13 residues long in type I nsLTPs, while this loop was shortened to 6, 6, 7, 12, 9, 8, 6 544 and 9 residues long in types II, III, IV, V, VI, VIII, IX and XI, respectively. 562 guide the choice of templates when nsLTP types with unknown structures are modeled by 563 homology. 563 homology. 564 565 Structure classification 566 In order to correlate the evolution of protein sequences and the impact on the corresponding 567 structures, we produced a circular tree according to structural distances (Fig. 5). Whereas type I 568 remained together in this second classification, other phylogenetic types were relatively scattered 569 in the tree. A majority of type II nsLTPs remained together in this tree, as was also the case for 570 type IV and type III, but no clear and reliable segregation between all non-type I nsLTPs could 571 be made. Looking at the 3D structures allowed us to confirm the hypothesis that only two major 572 structural types could be distinguished. They will be hereafter referred to as “Type-1 fold” and 573 “Type-2 fold”. 574 Several studies also showed that type I and type II nsLTPs differed through the characteristics of 575 the residue standing between Cys5 and Cys6, being respectively hydrophilic in type I and apolar 576 in type II proteins (Douliez et al., 2001; Marion et al., 2004). Based on the multiple sequence 577 alignment of the 797 nsLTPs and observation of the nature of the central residue in the CXC 578 pattern, together with the observations made in the preceding sequence-structure analysis, we 579 suggest that types III, IV, V, VI, VIII, IX and XI nsLTP C5 and C6 residues will adopt the same 580 spatial conformation as type II proteins, i.e. the “Type-2 fold”. 581 582 NsLTP structure-function relationship 583 Dealing with big datasets can be cumbersome and requires a very efficient interface. To address 584 this challenge, we developed InTreeGreat, a Javascript/PHP interface, compatible with every 585 standard web navigator. It is able to display and explore any tree and to deal with branch and leaf 586 coloring, branch lengths, branch support (or any other branch labels), and can aggregate 587 heterogeneous data (annotations, expression profiles, etc.). Figure 6 shows how InTreeGreat can 588 be used to display phylogenic trees together with various types of annotations. 589 590 Among the annotated nsLTPs (433 out of 797), we focused on those that had been reported for 591 their role in plant defense and/or resistance against pathogens (bacteria and/or fungi). To 592 simplify, we shall hereafter refer to them as “defense nsLTPs” in the present discussion. Manuscript to be reviewed 562 guide the choice of templates when nsLTP types with unknown 563 homology. 564 565 Structure classification 566 In order to correlate the evolution of protein sequences and the 567 structures, we produced a circular tree according to structural d 568 remained together in this second classification, other phylogene 569 in the tree. A majority of type II nsLTPs remained together in t 570 type IV and type III, but no clear and reliable segregation betw 571 be made. Looking at the 3D structures allowed us to confirm th 572 structural types could be distinguished. They will be hereafter r 573 “Type-2 fold”. 574 Several studies also showed that type I and type II nsLTPs diff 575 the residue standing between Cys5 and Cys6, being respectivel 576 in type II proteins (Douliez et al., 2001; Marion et al., 2004). B 577 alignment of the 797 nsLTPs and observation of the nature of t 578 pattern, together with the observations made in the preceding s 579 suggest that types III, IV, V, VI, VIII, IX and XI nsLTP C5 and 580 spatial conformation as type II proteins, i.e. the “Type-2 fold”. 581 582 NsLTP structure-function relationship 583 Dealing with big datasets can be cumbersome and requires a ve 584 this challenge, we developed InTreeGreat, a Javascript/PHP int 585 standard web navigator. It is able to display and explore any tre 586 coloring, branch lengths, branch support (or any other branch l 587 heterogeneous data (annotations, expression profiles, etc.). Figu 588 be used to display phylogenic trees together with various types 589 590 Among the annotated nsLTPs (433 out of 797), we focused on 591 their role in plant defense and/or resistance against pathogens ( 592 simplify, we shall hereafter refer to them as “defense nsLTPs” 593 investigating structural similarities between the 31 identified de 594 dataset, we attempted to identify key amino acid residues that w 595 SDM experiments as they may bestow their functional properti 596 Looking at the distribution of the defense nsLTPs in our structu 597 observed that they were predominantly found in the type I part 598 only 3 defense nsLTPs with a type II (85, 151, 501 - UniProtKB 599 transfer protein 2G, Q8W453: Putative lipid-transfer protein D 600 protein). We therefore preferred to focus on the Type-1 fold ns 601 trace(s) inside this important subfamily of nsLTPs. Manuscript to be reviewed (in other words: Roman numerals I to XI 552 correspond to phylogeny analysis while Arabic numerals 1 or 2 refer to structural analysis) 553 550 The fold of type I nsLTPs will be hereafter referred to as “Type-1 fold” and the alternative fold 551 of Types II to XI will be referred to as “Type-2 fold”. (in other words: Roman numerals I to XI 552 correspond to phylogeny analysis while Arabic numerals 1 or 2 refer to structural analysis) 554 The preceding analysis of the evolutive conservations specific to type I nsLTPs revealed many 555 residues whose role could be explained by local structural differences with the available types II 556 and IV nsLTP structures. This comparative structure analysis confirmed the clear separation 557 between type I and all the other nsLTP types initially observed in the phylogenetic tree inferred 558 from a multiple sequence alignment of the 797 available proteins. The key residues were usually 559 present in type I nsLTPs only and suggested that many structural differences observed when 560 comparing type I versus types II and IV nsLTPs should also be observed versus other nsLTP 561 types, particularly regarding ligand orientation and cavity entrances. This observation should PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 562 guide the choice of templates when nsLTP types with unknown structures are modeled by 563 homology. 564 565 Structure classification 566 In order to correlate the evolution of protein sequences and the impact on the corresponding 567 structures, we produced a circular tree according to structural distances (Fig. 5). Whereas type I 568 remained together in this second classification, other phylogenetic types were relatively scattered 569 in the tree. A majority of type II nsLTPs remained together in this tree, as was also the case for 570 type IV and type III, but no clear and reliable segregation between all non-type I nsLTPs could 571 be made. Looking at the 3D structures allowed us to confirm the hypothesis that only two major 572 structural types could be distinguished. They will be hereafter referred to as “Type-1 fold” and 573 “Type-2 fold”. 574 Several studies also showed that type I and type II nsLTPs differed through the characteristics of 575 the residue standing between Cys5 and Cys6, being respectively hydrophilic in type I and apolar 576 in type II proteins (Douliez et al., 2001; Marion et al., 2004). Based on the multiple sequence 577 alignment of the 797 nsLTPs and observation of the nature of the central residue in the CXC 578 pattern, together with the observations made in the preceding sequence-structure analysis, we 579 suggest that types III, IV, V, VI, VIII, IX and XI nsLTP C5 and C6 residues will adopt the same 580 spatial conformation as type II proteins, i.e. the “Type-2 fold”. 581 562 guide the choice of templates when nsLTP types with unknown structures are modeled by 563 homology. By 593 investigating structural similarities between the 31 identified defense nsLTPs in our annotated 594 dataset, we attempted to identify key amino acid residues that would be good candidates for 595 SDM experiments as they may bestow their functional properties on these proteins. 596 Looking at the distribution of the defense nsLTPs in our structural classification (Figure 6) we 597 observed that they were predominantly found in the type I part of the tree (28 proteins), with 598 only 3 defense nsLTPs with a type II (85, 151, 501 - UniProtKB - P82900: Non-specific lipid- 599 transfer protein 2G, Q8W453: Putative lipid-transfer protein DIR1, O24101: Lipid transfer 600 protein). We therefore preferred to focus on the Type-1 fold nsLTPs and study the structural 601 trace(s) inside this important subfamily of nsLTPs. 582 NsLTP structure-function relationship 582 NsLTP structure-function relationship 583 Dealing with big datasets can be cumbersome and requires a very efficient interface. To address 584 this challenge, we developed InTreeGreat, a Javascript/PHP interface, compatible with every 585 standard web navigator. It is able to display and explore any tree and to deal with branch and leaf 586 coloring, branch lengths, branch support (or any other branch labels), and can aggregate 587 heterogeneous data (annotations, expression profiles, etc.). Figure 6 shows how InTreeGreat can 588 be used to display phylogenic trees together with various types of annotations. 589 590 Among the annotated nsLTPs (433 out of 797), we focused on those that had been reported for 591 their role in plant defense and/or resistance against pathogens (bacteria and/or fungi). To 592 simplify, we shall hereafter refer to them as “defense nsLTPs” in the present discussion. By 593 investigating structural similarities between the 31 identified defense nsLTPs in our annotated 594 dataset, we attempted to identify key amino acid residues that would be good candidates for 595 SDM experiments as they may bestow their functional properties on these proteins. 596 Looking at the distribution of the defense nsLTPs in our structural classification (Figure 6) we 597 observed that they were predominantly found in the type I part of the tree (28 proteins), with 598 only 3 defense nsLTPs with a type II (85, 151, 501 - UniProtKB - P82900: Non-specific lipid- 599 transfer protein 2G, Q8W453: Putative lipid-transfer protein DIR1, O24101: Lipid transfer 600 protein). We therefore preferred to focus on the Type-1 fold nsLTPs and study the structural 601 trace(s) inside this important subfamily of nsLTPs. p 583 Dealing with big datasets can be cumbersome and requires a very efficient interface. To address 584 this challenge, we developed InTreeGreat, a Javascript/PHP interface, compatible with every 585 standard web navigator. It is able to display and explore any tree and to deal with branch and leaf 586 coloring, branch lengths, branch support (or any other branch labels), and can aggregate 587 heterogeneous data (annotations, expression profiles, etc.). Figure 6 shows how InTreeGreat can 588 be used to display phylogenic trees together with various types of annotations. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed In the trace obtained for the group composed by all the other 625 Type-1 fold defense nsLTPs, both residues Asp and Ile were among the 4 best ranked residues 626 after the 8 Cys residues and also showed good coverage and rvEt scores (Table 2). 627 628 Three other residues located at positions 137, 154 and 266 of the structural alignment were 629 differently conserved in the three clusters. Interestingly, these three positions showed good 630 conservation ranking, but the variability of the three corresponding residues was notably higher 631 in the Type-1 fold cluster. Indeed, in the defense cluster trace, position 137 was occupied either 632 by a valine or by an alanine residue (Val7 in protein 525) and position 154 was occupied either 633 by a leucine or by a valine residue (Leu11 in protein 525). Thus, both positions were occupied by 634 hydrophobic residues in defense proteins, which was not always the case in Type-1 fold proteins 635 (Table 2). In the same way, position 266 was occupied either by an arginine or a lysine residue 636 (both positively charged residues) (Lys46 in protein 525) in defense proteins, but allowed greater 637 variability in terms of physicochemical properties in the other proteins harboring a Type-1 fold. 638 The fact that these three positions of the structural alignment belonged to the top 30% most 602 The cluster containing all Type-1 fold defense nsLTPs corresponded to the whole type I part of 603 the tree (402 members). The corresponding structural trace was calculated, but it could not be 604 linked to the defense function, as the proportion of annotated nsLTPs with a defense function 605 was too low (28 out of 402, i.e. 7%). 606 In order to obtain a meaningful trace of the potential defense function, we needed to select a 607 cluster with a higher proportion of annotated defense nsLTPs. The best cluster we could find was 608 a relatively small cluster (43 members) of proteins with a structural distance no greater than 1.5Å 609 (i.e. 1.5 cut off), which contained 33% of the defense nsLTPs (i.e. 10 out of 31 proteins). This 610 cluster will be referred to as “defense cluster” in the further discussion. 611 611 612 The structural trace of the defense cluster showed several differences in comparison with the 613 structural trace of the Type-1 fold cluster (Table 2). Manuscript to be reviewed Apart from the 8 Cys residues that were 614 common to all nsLTPs, the 30% top ranked (i.e. 27 residues) most conserved residues were not 615 the same, or did not come in the same order in both traces. According to the defense cluster 616 trace, residue Asp at position 259 of the alignment (Asp45 in protein 525) was as strongly 617 conserved as the 8 Cys residues. Residue Ile at position 402 (Ile80 in protein 525) was among the 618 4 best ranked residues after the 8 Cys residues and obtained a very low coverage, variability and 619 rvET score. In terms of the ranking of these two (amino acid) residues in the Type-1 fold nsLTP 620 trace, they appeared to occur much later in the ranking (20th and 21st rank, respectively) with 621 much higher rvET scores and large variability in terms of the number and physico-chemical 622 properties of the residues (Table 2). It can be suggested that these two residues were not critical 623 for maintaining structure integrity, but could bestow functional specificity on the proteins 624 classified in the defense cluster. In the trace obtained for the group composed by all the other 625 Type-1 fold defense nsLTPs, both residues Asp and Ile were among the 4 best ranked residues 626 after the 8 Cys residues and also showed good coverage and rvEt scores (Table 2). 627 628 Three other residues located at positions 137, 154 and 266 of the structural alignment were 629 differently conserved in the three clusters. Interestingly, these three positions showed good 630 conservation ranking, but the variability of the three corresponding residues was notably higher 631 in the Type-1 fold cluster. Indeed, in the defense cluster trace, position 137 was occupied either 632 by a valine or by an alanine residue (Val7 in protein 525) and position 154 was occupied either 633 by a leucine or by a valine residue (Leu11 in protein 525). Thus, both positions were occupied by 634 hydrophobic residues in defense proteins, which was not always the case in Type-1 fold proteins 635 (Table 2). In the same way, position 266 was occupied either by an arginine or a lysine residue 636 (both positively charged residues) (Lys46 in protein 525) in defense proteins, but allowed greater 637 variability in terms of physicochemical properties in the other proteins harboring a Type-1 fold. Manuscript to be reviewed Manuscript to be reviewed 602 The cluster containing all Type-1 fold defense nsLTPs corresponded to the whole type I part of 603 the tree (402 members). The corresponding structural trace was calculated, but it could not be 604 linked to the defense function, as the proportion of annotated nsLTPs with a defense function 605 was too low (28 out of 402, i.e. 7%). 606 In order to obtain a meaningful trace of the potential defense function, we needed to select a 607 cluster with a higher proportion of annotated defense nsLTPs. The best cluster we could find was 608 a relatively small cluster (43 members) of proteins with a structural distance no greater than 1.5Å 609 (i.e. 1.5 cut off), which contained 33% of the defense nsLTPs (i.e. 10 out of 31 proteins). This 610 cluster will be referred to as “defense cluster” in the further discussion. 611 612 The structural trace of the defense cluster showed several differences in comparison with the 613 structural trace of the Type-1 fold cluster (Table 2). Apart from the 8 Cys residues that were 614 common to all nsLTPs, the 30% top ranked (i.e. 27 residues) most conserved residues were not 615 the same, or did not come in the same order in both traces. According to the defense cluster 616 trace, residue Asp at position 259 of the alignment (Asp45 in protein 525) was as strongly 617 conserved as the 8 Cys residues. Residue Ile at position 402 (Ile80 in protein 525) was among the 618 4 best ranked residues after the 8 Cys residues and obtained a very low coverage, variability and 619 rvET score. In terms of the ranking of these two (amino acid) residues in the Type-1 fold nsLTP 620 trace, they appeared to occur much later in the ranking (20th and 21st rank, respectively) with 621 much higher rvET scores and large variability in terms of the number and physico-chemical 622 properties of the residues (Table 2). It can be suggested that these two residues were not critical 623 for maintaining structure integrity, but could bestow functional specificity on the proteins 624 classified in the defense cluster. Manuscript to be reviewed 644 645 Figure 7 shows the five residues highlighted in Table 2 in the 3D structural context of the 646 representative protein of the defense cluster (protein 525). In this protein, conserved residues 647 Asp and Ile were located at positions 45 and 80, respectively. The two small hydrophobic 648 residues were Val7 and Leu11 and the positively charged residue was Lys46. All five key 649 residues were located around the ligand cavity (Figure 7), which allowed either guidance or 650 direct contact with the lipid. This observation was consistent with the suggested hypothesis. 651 652 NsLTP sequence-structure analyses using either FAST or STD revealed some key residues or 653 key positions (in type I: Gly37, Arg/Lys51, bulky hydrophobic residues 87 and 89, Ala54, Thr6, 654 Thr47, Tyr20, Pro30, Leu/Ile79, longer C-terminal loop; large hydrophobic residue 63 in types 655 II, III, IV, VI, VIII, IX nsLTPs). The structural trace analysis highlighted other amino acid 656 residues (in type I defense/resistance nsLTPs: Asp45, Ile80, Val/Ala7, Leu/Val11, Arg/Lys46). It 657 is important to note that these two complementary analyses by FAST and STD were not meant to 658 lead to the same kind of conclusions. Indeed, using sequence information projected on the 3D 659 structure, the first method revealed nsLTP-type-specific amino acid residues that could be 660 involved in structure stabilization and/or ligand interaction, given their structural context. The 661 second method however considered a set of functionally close nsLTPs sharing a very similar 662 structure and highlighted over-representatively conserved amino acid residues that might thus 663 bestow functional specificity on these proteins. These two approaches took inverse directions in 664 the path sequence – structure – function. The “sequence-to-function” method would lead to 665 more precise conclusions if more data about the inner structural mechanisms of lipid binding 666 were available (only a few structures of nsLTP-lipid complexes have been experimentally 667 determined so far). The “function-to-sequence” method would give us a better overview of the 668 range of nsLTP activities if the functional data were not so rare and heterogeneous. Manuscript to be reviewed 638 The fact that these three positions of the structural alignment belonged to the top 30% most 639 conserved among all Type-1 fold nsLTPs suggested their importance in these proteins. However, 640 because the variability at these three positions was very small among defense nsLTPs and 641 because the physico-chemical property was strongly conserved, we suspected that residues 642 located at positions 137, 154 and 266 of the structural alignment might bestow functional 643 specificity, at least in the case of defense/resistance proteins. 612 The structural trace of the defense cluster showed several differences in comparison with the 613 structural trace of the Type-1 fold cluster (Table 2). Apart from the 8 Cys residues that were 614 common to all nsLTPs, the 30% top ranked (i.e. 27 residues) most conserved residues were not 615 the same, or did not come in the same order in both traces. According to the defense cluster 616 trace, residue Asp at position 259 of the alignment (Asp45 in protein 525) was as strongly 617 conserved as the 8 Cys residues. Residue Ile at position 402 (Ile80 in protein 525) was among the 618 4 best ranked residues after the 8 Cys residues and obtained a very low coverage, variability and 619 rvET score. In terms of the ranking of these two (amino acid) residues in the Type-1 fold nsLTP 620 trace, they appeared to occur much later in the ranking (20th and 21st rank, respectively) with 621 much higher rvET scores and large variability in terms of the number and physico-chemical 622 properties of the residues (Table 2). It can be suggested that these two residues were not critical 623 for maintaining structure integrity, but could bestow functional specificity on the proteins 624 classified in the defense cluster. In the trace obtained for the group composed by all the other 625 Type-1 fold defense nsLTPs, both residues Asp and Ile were among the 4 best ranked residues 626 after the 8 Cys residues and also showed good coverage and rvEt scores (Table 2). 627 PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 669 670 However, we assumed that this combination of approaches i) allowed structure-sequence 671 analysis for large multigene families, ii) could reveal structural patterns related to functions that 672 were not revealed so far, as alignments would have been limited to primary sequences only and 673 iii) allowed a comparison of groups composed of proteins with an evolutionary connection with 674 groups displaying structural similarity. 675 676 677 DISCUSSION 678 679 A) We combined two powerful alignment algorithms (MAFFT and MUSCLE) together with a 680 3D projection of the impact of alignment on the structure of proteins (VITO). Real-time 681 monitoring of the impact of gap positions and lengths on the resulting 3D model offered the 682 possibility of discriminating between various alignment possibilities. This allowed us to provide 683 definitive insight into the old debate about the CXC pattern and its implication for the structure PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed This was understandable, because phylogeny compares the different 713 proteins with a much larger number of parameters (site-to-site mutation, classification of sites by 714 mutation rate, use of refined distance matrix, etc.) while the structure tree only uses the RMSD 715 of the structures taken 2 by 2. While this innovative information was very interesting, it could 716 potentially be improved if we had templates from each sub-family for the generation of 717 molecular models (experimental structures are available for Type I, II and IV). Indeed, at this 718 level of analysis, it is conceivable that models obtained from experimental structures for the 719 other types (III, V, VI, VII, VIII, IX and XI) would provide improved models allowing the 720 detection of other key residues. 721 722 684 of LTPs (Douliez et al. 2000). The resulting alignment allowed us to classify unambiguously all 685 797 sequences in main two nsLTP families. 686 687 B) The phylogenetic analysis was the most extensive to date, including 797 nsLTP sequences. 688 This was a much more complete description than the previous one (195 sequences, Wang et al., 689 2012). 690 This phylogenetic analysis was conducted from a clearly defined dataset: sequences were 691 selected using unambiguous parameters optimizing the quality of the output tree, also 692 considering our 3D structural integration objective. Although GPI-Anchored LTP could have 693 been included in this study, their incomplete homology with other LTPs and the lack of any 694 experimental 3D structure, convinced us not to include them. Thanks to this choice, alignment 695 quality was preserved, and a better-quality 3D structural model are used. This analysis allowed 696 us to classify unambiguously all 797 sequences in the main two nsLTP families, complementing 697 and reinforcing the former classification by Boutrot (Boutrot et al. 2008). 698 699 C) The production of more than 600 3D structural models and the collection of numerous 700 functional annotations enabled progress to be made in the study of structure-function 701 relationships of nsLTPs. The re-use of the ETD method in a close and adapted form (STD) led to 702 the identification of amino acids involved in the functional specialization of some nsLTPs. 703 STD allowed us to highlight amino acids specific to certain functions. One of the limiting points 704 of this analysis remained the publication bias. Manuscript to be reviewed 684 of LTPs (Douliez et al. 2000). The resulting alignment allowed us to classify unambiguously all 685 797 sequences in main two nsLTP families. 686 687 B) The phylogenetic analysis was the most extensive to date, including 797 nsLTP sequences. 688 This was a much more complete description than the previous one (195 sequences, Wang et al., 689 2012). 690 This phylogenetic analysis was conducted from a clearly defined dataset: sequences were 691 selected using unambiguous parameters optimizing the quality of the output tree, also 692 considering our 3D structural integration objective. Although GPI-Anchored LTP could have 693 been included in this study, their incomplete homology with other LTPs and the lack of any 694 experimental 3D structure, convinced us not to include them. Thanks to this choice, alignment 695 quality was preserved, and a better-quality 3D structural model are used. This analysis allowed 696 us to classify unambiguously all 797 sequences in the main two nsLTP families, complementing 697 and reinforcing the former classification by Boutrot (Boutrot et al. 2008). 698 699 C) The production of more than 600 3D structural models and the collection of numerous 700 functional annotations enabled progress to be made in the study of structure-function 701 relationships of nsLTPs. The re-use of the ETD method in a close and adapted form (STD) led to 702 the identification of amino acids involved in the functional specialization of some nsLTPs. 703 STD allowed us to highlight amino acids specific to certain functions. One of the limiting points 704 of this analysis remained the publication bias. Indeed, the annotations were not evenly 705 distributed among available sequences, nor was it possible to distinguish between an unsearched 706 function and a function not found. It seemed difficult to propose a solution to circumvent this 707 bias (Douliez et al. 2000). 708 709 D) The structure tree clearly showed that all Type I ns-LTPs adopted the same folding (Type-1 710 fold), while all the other proteins adopted the second fold (Type-2 fold). This approach seemed 711 very interesting but did not offer the same level of detail and the same analytical power as the 712 phylogenetic approach. Manuscript to be reviewed 758 To researchers who may not grasp the importance of the protein structure-function relationships 759 we would like to insist on three main contributions of the methods presented in this work: 760 - The structural classification agrees with the sequence classification by phylogenetic types. 761 - The sequence-structure analysis highlights key-residues explaining the specificities of the 762 different folds. 763 - The structure-function analysis based on the evolutionary trace of the aligned sequences can 764 show the functional signature of a subfamily of proteins. 725 Plant non-specific Lipid Transfer Proteins constitute a complex family of proteins whose 726 biological functions are far from well understood. However, it has become clear for years that 727 they are of increasing interest for agronomical and nutritional issues. 728 Experimental approaches are irreplaceable for accessing their inner functional mechanisms. 729 However, such methods are expensive and time-consuming. Furthermore, they produce a large 730 amount of heterogeneous data. For all these reasons, resorting to bioinformatics methods has 731 long become necessary to organize and analyze existing data, and/or model and hypothesize new 732 data. 733 This paper presented a new methodology based on the combination of either classical or original 734 bioinformatics approaches, using various computational tools to extract information and suggest 735 new hypotheses from a large pool of experimental data about the plant nsLTP superfamily of 736 proteins. 737 In this paper, we: 738 1) Suggested a new definition of the nsLTP superfamily, with a set of criteria based on sequence 739 length, sequence composition (e.g. Cys involved in SS bonds) and structure (monodomain). 740 2) Confirmed and enriched Boutrot’s phylogenetic classification of plant nsLTP sequences. 741 3) Demonstrated the need for a small shift in the CXC alignment that reflected the existence of 742 two main distinct nsLTP folds. 743 4) Calculated 666 good quality theoretical three-dimensional structures of nsLTPs. 744 5) Developed an original alignment tool to detect conserved and specific positions among the 745 different phylogenetic types of nsLTPs. 746 6) Used the latter tool to reveal some key residues. 747 7) Suggested a new structure-based classification of the 676 nsLTP structures now available (10 748 experimental + 666 theoretical), which that allow clustering by structural similarity. 749 8) Annotated all available information about the function. 750 9) Developed an original interface allowing quick visualization of several types of annotations 751 on any phylogenetic tree. Manuscript to be reviewed 725 Plant non-specific Lipid Transfer Proteins constitute a complex family of proteins whose 726 biological functions are far from well understood. However, it has become clear for years that 727 they are of increasing interest for agronomical and nutritional issues. 728 Experimental approaches are irreplaceable for accessing their inner functional mechanisms. 729 However, such methods are expensive and time-consuming. Furthermore, they produce a large 730 amount of heterogeneous data. For all these reasons, resorting to bioinformatics methods has 731 long become necessary to organize and analyze existing data, and/or model and hypothesize new 732 data. 733 This paper presented a new methodology based on the combination of either classical or original 734 bioinformatics approaches, using various computational tools to extract information and suggest 735 new hypotheses from a large pool of experimental data about the plant nsLTP superfamily of 736 proteins. 737 In this paper, we: 738 1) Suggested a new definition of the nsLTP superfamily, with a set of criteria based on sequence 739 length, sequence composition (e.g. Cys involved in SS bonds) and structure (monodomain). 740 2) Confirmed and enriched Boutrot’s phylogenetic classification of plant nsLTP sequences. 741 3) Demonstrated the need for a small shift in the CXC alignment that reflected the existence of 742 two main distinct nsLTP folds. 743 4) Calculated 666 good quality theoretical three-dimensional structures of nsLTPs. 744 5) Developed an original alignment tool to detect conserved and specific positions among the 745 different phylogenetic types of nsLTPs. 746 6) Used the latter tool to reveal some key residues. 747 7) Suggested a new structure-based classification of the 676 nsLTP structures now available (10 748 experimental + 666 theoretical), which that allow clustering by structural similarity. 749 8) Annotated all available information about the function. 750 9) Developed an original interface allowing quick visualization of several types of annotations 751 on any phylogenetic tree. 752 10) Revealed, using structural trace analysis, potential specific amino acid residues involved in 753 plant defense and/or resistance against pathogens 754 755 Our work was made more difficult by the problems of annotation bias for which we did not 756 expect a practical solution. However, it seemed that some of our results could be improved if we 757 had additional experimental structures for all types of nsLTPs. Manuscript to be reviewed Indeed, the annotations were not evenly 705 distributed among available sequences, nor was it possible to distinguish between an unsearched 706 function and a function not found. It seemed difficult to propose a solution to circumvent this 707 bias (Douliez et al. 2000). 708 709 D) The structure tree clearly showed that all Type I ns-LTPs adopted the same folding (Type-1 710 fold), while all the other proteins adopted the second fold (Type-2 fold). This approach seemed 711 very interesting but did not offer the same level of detail and the same analytical power as the 712 phylogenetic approach. This was understandable, because phylogeny compares the different 713 proteins with a much larger number of parameters (site-to-site mutation, classification of sites by 714 mutation rate, use of refined distance matrix, etc.) while the structure tree only uses the RMSD 715 of the structures taken 2 by 2. While this innovative information was very interesting, it could 716 potentially be improved if we had templates from each sub-family for the generation of 717 molecular models (experimental structures are available for Type I, II and IV). Indeed, at this 718 level of analysis, it is conceivable that models obtained from experimental structures for the 719 other types (III, V, VI, VII, VIII, IX and XI) would provide improved models allowing the 720 detection of other key residues. 721 722 684 of LTPs (Douliez et al. 2000). The resulting alignment allowed us to classify unambiguously all 685 797 sequences in main two nsLTP families. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Ma, Z., Zhang, Y., 791 Brown S Bourge M Golser W Song X Clement D Rivallan R Tahi M Akaza J M 789 S.C., Carlson, J.E., Sallet, E., Schiex, T., Dievart, A., Kramer, M., Gelley, L., Shi, Z. Bérard, A., , , , , , , , , , , , y, , , , , 790 Viot, C., Boccara, M., Risterucci, A-M. Guignon, V., Sabau, X., Axtell, M.J. Ma, Z., Zhang, Y., 791 Brown S Bourge M Golser W Song X Clement D Rivallan R Tahi M Akaza J M 790 Viot, C., Boccara, M., Risterucci, A-M. Guignon, V., Sabau, X., Axtell, M.J. Ma, Z., Zhang, Y., 91 Brown, S., Bourge, M., Golser, W., Song, X., Clement, D., Rivallan, R., Tahi 792 Pitollat, B., Gramacho, K., D’Hont, A., Brunel, D., Infante, D., Kebe, I., Costet, P., Wing, R., 793 McCombie, W.R., Guiderdoni, E., Quetier, F., Panaud, O., Wincker, P., Bocs, S., and Lanaud, C. 794 (2011). The genome of Theobroma cacao. Nat. Genet. 43, 101–108. 794 (2011). The genome of Theobroma cacao. Nat. Genet. 43, 101–108. 795 Arondel, V., Vergnolle, C., Cantrel, C., and Kader, J. (2000). Lipid transfer proteins are encoded 796 by a small multigene family in Arabidopsis thaliana. Plant Sci. Int. J. Exp. Plant Biol. 157, 1–12. 797 Aziz, N., Paiva, N.L., May, G.D., and Dixon, R.A. (2005). Transcriptome analysis of alfalfa 798 glandular trichomes. Planta 221, 28–38. 795 Arondel, V., Vergnolle, C., Cantrel, C., and Kader, J. (2000). Lipid transfer proteins are encoded 796 by a small multigene family in Arabidopsis thaliana. Plant Sci. Int. J. Exp. Plant Biol. 157, 1–12. 796 by a small multigene family in Arabidopsis thaliana. Plant Sci. Int. J. Exp. Plant Biol. 157, 1 12. 797 Aziz, N., Paiva, N.L., May, G.D., and Dixon, R.A. (2005). Transcriptome analysis of alfalfa 798 glandular trichomes. Planta 221, 28–38. 799 Bendtsen, J.D., Nielsen, H., von Heijne, G., and Brunak, S. (2004). Improved prediction of 800 signal peptides: SignalP 3.0. J. Mol. Biol. 340, 783–795. 801 Benkert, P., Tosatto, S.C.E., and Schomburg, D. (2008). QMEAN: A comprehensive scoring 802 function for model quality assessment. Proteins 71, 261–277. 801 Benkert, P., Tosatto, S.C.E., and Schomburg, D. (2008). QMEAN: A comprehensive scoring 802 function for model quality assessment. Proteins 71, 261–277. 803 Benkert, P., Biasini, M., and Schwede, T. (2011). Toward the estimation of the absolute quality 804 of individual protein structure models. Bioinforma. Oxf. Engl. Manuscript to be reviewed 765 Furthermore, the structural dichotomy between type I nsLTPs and all the others may go 766 unnoticed by anyone who would focus on one sequence and who would not conduct a combined 767 analysis on a large set of sequences. We encourage researchers studying nsLTPs to use this 768 approach combining sequence alignment, phylogeny and structural biochemistry. This would 769 enhance the power of the analysis by drawing a connection between primary sequence, three- 770 dimensional structure and function. 771 More broadly, we consider that this type of combined approach should be favored for any study 772 involving a multigenic family. 777 This work was supported by the CIRAD - UMR AGAP HPC Data Centre of the South Green 778 Bioinformatics platform. 777 This work was supported by the CIRAD - UMR AGAP HPC Data Centre of the South Green 778 Bioinformatics platform 777 This work was supported by the CIRAD - UMR AGAP HPC Data Centre 779 The authors are thankful to Dr. Franck Molina for his key role at the beginning of this project 780 and all the fruitful and friendly discussions. 779 The authors are thankful to Dr. Franck Molina for his key role at the beginning of this project 780 and all the fruitful and friendly discussions. 781 We are thankful to Peter Biggins for the careful and critical review of this manuscript. 782 781 We are thankful to Peter Biggins for the careful and critical review of this manuscript. 782 781 We are thankful to Peter Biggins for the careful and critical review of this manuscript. 782 781 We are thankful to Peter Biggins for the careful and critical review of this m 782 786 Argout, X., Salse, J., Aury, J-M., Guiltinan, M.J.,Droc, G., Gouzy, J., Alleg 787 Legavre, T., Maximova, S.N., Abrouk, M., Murat, F., Fouet, O., Poulain, J., Ruiz, M., Roguet, 788 Y., Rodier-Goud, M., Barbosa-Neto, J.F., Sabot, F., Kudrna, D., Ammiraju 788 ., od e Goud, ., a bosa Ne o, J. ., Sabo , ., ud a, ., aju, J.S.S., Sc us e , 789 S.C., Carlson, J.E., Sallet, E., Schiex, T., Dievart, A., Kramer, M., Gelley, L., Shi, Z. Bérard, A., 790 Viot, C., Boccara, M., Risterucci, A-M. Guignon, V., Sabau, X., Axtell, M.J. Manuscript to be reviewed 752 10) Revealed, using structural trace analysis, potential specific amino acid residues involved in 753 plant defense and/or resistance against pathogens 754 755 Our work was made more difficult by the problems of annotation bias for which we did not 756 expect a practical solution. However, it seemed that some of our results could be improved if we 757 had additional experimental structures for all types of nsLTPs. 758 To researchers who may not grasp the importance of the protein structure-function relationships 759 we would like to insist on three main contributions of the methods presented in this work: 760 - The structural classification agrees with the sequence classification by phylogenetic types. 761 - The sequence-structure analysis highlights key-residues explaining the specificities of the 762 different folds. 763 - The structure-function analysis based on the evolutionary trace of the aligned sequences can 764 show the functional signature of a subfamily of proteins. 725 Plant non-specific Lipid Transfer Proteins constitute a complex family of proteins whose 726 biological functions are far from well understood. However, it has become clear for years that 727 they are of increasing interest for agronomical and nutritional issues. 4) Calculated 666 good quality theoretical three-dimensional structures of nsLTPs. 744 5) Developed an original alignment tool to detect conserved and specific positions among the 745 different phylogenetic types of nsLTPs. 747 7) Suggested a new structure-based classification of the 676 nsLTP structures now available (10 748 experimental + 666 theoretical), which that allow clustering by structural similarity. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 805 Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, H., Shindyalov, I.N., 806 and Bourne, P.E. (2000). The Protein Data Bank. Nucleic Acids Res. 28, 235–242. 805 Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, 806 and Bourne, P.E. (2000). The Protein Data Bank. Nucleic Acids Res. 28, 235–242. 807 Boutrot, F., Guirao, A., Alary, R., Joudrier, P., and Gautier, M.-F. (2005). Wheat non-specific 808 lipid transfer protein genes display a complex pattern of expression in developing seeds. 809 Biochim. Biophys. Acta 1730, 114–125. 810 Boutrot, F., Chantret, N., and Gautier, M.-F. (2008). Genome-wide analysis of the rice and 811 Arabidopsis non-specific lipid transfer protein (nsLtp) gene families and identification of wheat 812 nsLtp genes by EST data mining. BMC Genomics 9, 86. 810 Boutrot, F., Chantret, N., and Gautier, M.-F. (2008). Genome-wide analysis of the rice and 811 Arabidopsis non-specific lipid transfer protein (nsLtp) gene families and identification of wheat 812 nsLtp genes by EST data mining. BMC Genomics 9, 86. 812 nsLtp genes by EST data mining. BMC Genomics 9, 86 813 Bucher, M., Wegmüller, S., and Drissner, D. (2009). Chasing the structures of small molecules 814 in arbuscular mycorrhizal signaling. Curr. Opin. Plant Biol. 12, 500–507. 815 Buhot, N., Douliez, J.P., Jacquemard, A., Marion, D., Tran, V., Maume, B. 816 Ponchet, M., Mikès, V., Kader, J.C., and Blein J-P. (2001). A lipid transfer protein binds to a 817 receptor involved in the control of plant defence responses. FEBS Lett. 509, 27–30. 818 Buhot, N., Gomès, E., Milat, M.-L., Ponchet, M., Marion, D., Lequeu, J., Delrot, S., Coutos- 819 Thévenot, P., and Blein, J.-P. (2004). Modulation of the biological activity of a tobacco LTP1 by 820 lipid complexation. Mol. Biol. Cell 15, 5047–5052. 819 Thévenot, P., and Blein, J.-P. (2004). Modulation of the biological activity of a tobacco LTP1 by 820 lipid complexation. Mol. Biol. Cell 15, 5047–5052. 821 Cameron, K.D., Teece, M.A., and Smart, L.B. (2006). Increased accumulation of cuticular wax 822 and expression of lipid transfer protein in response to periodic drying events in leaves of tree 823 tobacco. Plant Physiol. 140, 176–183. 824 Cammue, B.P., Thevissen, K., Hendriks, M., Eggermont, K., Goderis, I.J., 824 Cammue, B.P., Thevissen, K., Hendriks, M., Eggermont, K., Goderis, I.J., Proost, P., Van 825 Damme, J., Osborn, R.W., Guerbette, F., and Kader, J.C. (1995). Manuscript to be reviewed 27, 343–350. 803 Benkert, P., Biasini, M., and Schwede, T. (2011). Toward the estimation of the absolute quality 804 of individual protein structure models. Bioinforma. Oxf. Engl. 27, 343–350. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Mini Review: Structure, 857 Biological and Technological Functions of Lipid Transfer Proteins and Indolines, the Major 858 Lipid Binding Proteins from Cereal Kernels. Journal of Cereal Science, 32(1), 1–20. 858 Lipid Binding Proteins from Cereal Kernels. Journal of Cereal Science, 32(1), 1–20. 859 Douliez, J.P., Pato, C., Rabesona, H., Mollé, D., and Marion, D. (2001b). Disulfide bond 860 assignment, lipid transfer activity and secondary structure of a 7-kDa plant lipid transfer protein, 861 LTP2. Eur. J. Biochem. 268, 1400–1403. 862 Douliez, J.P., Jégou, S., Pato, C., Mollé, D., Tran, V., and Marion, D. (2001a). Binding of two 863 mono-acylated lipid monomers by the barley lipid transfer protein, LTP1, as viewed by 864 fluorescence, isothermal titration calorimetry and molecular modelling. Eur. J. Biochem. 268, 865 384–388. 866 Dufayard, J.-F., Duret, L., Penel, S., Gouy, M., Rechenmann, F., and Perrière, G. (2005). Tree 867 pattern matching in phylogenetic trees: automatic search for orthologs or paralogs in homologous 868 gene sequence databases. Bioinforma. Oxf. Engl. 21, 2596–2603. 869 Edgar, R.C. (2004). MUSCLE: a multiple sequence alignment method with reduced time and 870 space complexity. BMC Bioinformatics 5, 113. 871 Edstam, M.M., Viitanen, L., Salminen, T.A., and Edqvist, J. (2011). Evolutionary history of the 872 ifi li id t f t i M l Pl t 4 947 964 871 Edstam, M.M., Viitanen, L., Salminen, T.A., and Edqvist, J. (2011). Evolutionary history of the 872 non-specific lipid transfer proteins. Mol. Plant 4, 947–964. 872 non-specific lipid transfer proteins. Mol. Plant 4, 947–9 873 Erdin, S., Ward, R.M., Venner, E., and Lichtarge, O. (2010). Evolutionary trace annotation of 874 protein function in the structural proteome. J. Mol. Biol. 396, 1451–1473. 875 Gao, M., Yang, F., Zhang, L., Su, Z., and Huang, Y. (2017). Exploring the sequence-structure- 876 function relationship for the intrinsically disordered βγ-crystallin Hahellin. J. Biomol. Struct. 877 Dyn. 1–11. 878 Ge, X., Chen, J., Sun, C., Cao, K. (2003). Preliminary study on the structural basis of the 879 tif l ti it f i e li id t fe tei P tei E 6 387 390 878 Ge, X., Chen, J., Sun, C., Cao, K. (2003). Preliminary study on the structural basis of the 879 antifungal activity of a rice lipid transfer protein. Protein Eng. 6, 387–390. 880 Girault, T., François, J., Rogniaux, H., Pascal, S., Delrot, S., Coutos-Thévenot, P., and Gomès, E. 881 (2008). Manuscript to be reviewed 844 Clark, A.M., and Bohnert, H.J. (1999). Cell-specific expression of genes of the lipid transfer 845 protein family from Arabidopsis thaliana. Plant Cell Physiol. 40, 69–76. 44 Clark, A.M., and Bohnert, H.J. (1999). Cell-specific expression of genes of th 845 protein family from Arabidopsis thaliana. Plant Cell Physiol. 40, 69–76. 846 Cong, Q., and Grishin, N.V. (2012). MESSA: MEta-Server for protein Sequence Analysis. BMC 847 Biol. 10, 82. 848 D’Agostino, N. ,Buonanno, M., Ayoub, J., Barone, A., Monti, S. M., Rigano, Maria M. (2019). 849 Identification of non-specific Lipid Transfer Protein gene family members in Solanum 848 D’Agostino, N. ,Buonanno, M., Ayoub, J., Barone, A., Monti, S. M., Rigano, Maria M. (2019). 849 Identification of non-specific Lipid Transfer Protein gene family members in Solanum 850 lycopersicum and insights into the features of Sola l 3 protein. Sci. Reports. 9. 1607.Debono, A., 851 Yeats, T.H., Rose, J.K.C., Bird, D., Jetter, R., Kunst, L., and Samuels, L. (2009). Arabidopsis 852 LTPG is a glycosylphosphatidylinositol-anchored lipid transfer protein required for export of 853 lipids to the plant surface. Plant Cell 21, 1230–1238. 849 Identification of non-specific Lipid Transfer Protein gene family members in Solanum 850 lycopersicum and insights into the features of Sola l 3 protein. Sci. Reports. 9. 1607.Debono, A., 851 Yeats, T.H., Rose, J.K.C., Bird, D., Jetter, R., Kunst, L., and Samuels, L. (2009). Arabidopsis 851 Yeats, T.H., Rose, J.K.C., Bird, D., Jetter, R., Kunst, L., and Samuels, L. (2009). Arabidopsis 852 LTPG is a glycosylphosphatidylinositol-anchored lipid transfer protein required for export of 853 lipids to the plant surface Plant Cell 21 1230–1238 853 lipids to the plant surface. Plant Cell 21, 1230–1238. 854 Desper, R., and Gascuel, O. (2002). Fast and accurate phylogeny reconstruction algorithms based 855 on the minimum-evolution principle. J. Comput. Biol. J. Comput. Mol. Cell Biol. 9, 687–705. 854 Desper, R., and Gascuel, O. (2002). Fast and accurate phylogeny reconstru 855 on the minimum-evolution principle. J. Comput. Biol. J. Comput. Mol. Cell Biol. 9, 687–705. 856 Douliez, J.-P., Michon, T., Elmorjani, K., & Marion, D. (2000). Mini Review: Structure, 857 Biological and Technological Functions of Lipid Transfer Proteins and Indolines, the Major 858 Lipid Binding Proteins from Cereal Kernels. Journal of Cereal Science, 32(1), 1–20. 856 Douliez, J.-P., Michon, T., Elmorjani, K., & Marion, D. (2000). Manuscript to be reviewed A potent antimicrobial protein 826 from onion seeds showing sequence homology to plant lipid transfer proteins. Plant Physiol. 109, 827 445–455. 828 Catherinot, V., and Labesse, G. (2004). ViTO: tool for refinement of protein sequence-structure 829 alignments. Bioinforma. Oxf. Engl. 20, 3694–3696. 828 Catherinot, V., and Labesse, G. (2004). ViTO: tool for refinement of protein sequence-structure 829 alignments. Bioinforma. Oxf. Engl. 20, 3694–3696. 830 Chae, K., and Lord, E.M. (2011). Pollen tube growth and guidance: roles of small, secreted 831 proteins. Ann. Bot. 108, 627–636. 832 Chae, K., Kieslich, C.A., Morikis, D., Kim, S.-C., and Lord, E.M. (2009). A gain-of-function 833 mutation of Arabidopsis lipid transfer protein 5 disturbs pollen tube tip growth and fertilization. 834 Plant Cell 21, 3902–3914. 835 B. Chanda, Ye Xia, M. Kumar Mandal, K. Yu, K‐T. Sekine, Q. Gao, D. Selote, Y. Hu, A. 836 Stromberg, D. Navarre, A. Kachroo and P. Kachroo. (2011). Glycerol-3-phosphate is a critical 837 mobile inducer of systemic immunity in plants. Nat. Genet. 43, 421–427. 838 Cheng, C.-S., Chen, M.-N., Lai, Y.-T., Chen, T., Lin, K.-F., Liu, Y.-J., and Lyu, P.-C. (2008). 839 Mutagenesis study of rice nonspecific lipid transfer protein 2 reveals residues that contribute to 840 structure and ligand binding. Proteins 70, 695–706. 841 Choi, Y.E., Lim, S., Kim, H.-J., Han, J.Y., Lee, M.-H., Yang, Y., Kim, J.-A., and Kim, Y.-S. 842 (2012). Tobacco NtLTP1, a glandular-specific lipid transfer protein, is required for lipid 841 Choi, Y.E., Lim, S., Kim, H.-J., Han, J.Y., Lee, M.-H., Yang, Y., Kim, J.-A., and Kim, Y.-S. 842 (2012) T b N LTP1 l d l ifi li id f i i i d f li id 841 Choi, Y.E., Lim, S., Kim, H.-J., Han, J.Y., Lee, M.-H., Yang, Y., Kim, J.-A., and Kim, Y.-S. 842 (2012). Tobacco NtLTP1, a glandular-specific lipid transfer protein, is required for lipid 842 (2012). Tobacco NtLTP1, a glandular-specific lipid transfer protein, is required for lipid 843 secretion from glandular trichomes. Plant J. Cell Mol. Biol. 70, 480–491. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 884 Guindon, S., Dufayard, J.-F., Lefort, V., Anisimova, M., Hordijk, W., and G 884 Guindon, S., Dufayard, J.-F., Lefort, V., Anisimova, M., Hordijk, W., and Gascuel, O. (2010). 885 New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the 886 performance of PhyML 3.0. Syst. Biol. 59, 307–321. 885 New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the 886 performance of PhyML 3.0. Syst. Biol. 59, 307–321. 887 Henikoff, S., and Henikoff, J.G. (1994). Position-based sequence weights. J. Mol. Biol. 243, 888 574–578. 889 Herbert A, unpublished. Max Cluster. Available 890 at http://www.sbg.bio.ic.ac.uk/maxcluster (accessed 2019-08/08). 890 at http://www.sbg.bio.ic.ac.uk/maxcluster (accessed 2019-08/08). 891 Hoh, F., Pons, J.-L., Gautier, M.-F., de Lamotte, F., and Dumas, C. (2005). 891 Hoh, F., Pons, J.-L., Gautier, M.-F., de Lamotte, F., and Dumas, C. (2005). Structure of a 892 liganded Type-2 non-specific lipid-transfer protein from wheat and the molecular basis of lipid 893 binding. Acta Crystallogr. D Biol. Crystallogr. 61, 397–406. 892 liganded Type-2 non-specific lipid-transfer protein from wheat and the molecular basis of lipid 892 liganded Type-2 non-specific lipid-transfer protein from wheat and the molecular basis of lipid 893 binding. Acta Crystallogr. D Biol. Crystallogr. 61, 397–406. 893 binding. Acta Crystallogr. D Biol. Crystallogr. 61, 397–406. 893 binding. Acta Crystallogr. D Biol. Crystallogr. 61, 397–406. 894 Huson, D.H., and Scornavacca, C. (2012). Dendroscope 3: an interactive to 894 Huson, D.H., and Scornavacca, C. (2012). Dendroscope 3: an interactive tool for rooted 895 phylogenetic trees and networks. Syst. Biol. 61, 1061–1067. 895 phylogenetic trees and networks. Syst. Biol. 61, 1061–1067. 895 phylogenetic trees and networks. Syst. Biol. 61, 1061–1067. 896 Jang, C.S., Jung, J.H., Yim, W.C., Lee, B.-M., Seo, Y.W., and Kim, W. (2007). Divergence of 897 genes encoding non-specific lipid transfer proteins in the poaceae family. Mol. Cells 24, 215– 898 223. 899 José-Estanyol, M., Gomis-Rüth, F.X., and Puigdomènech, P. (2004). The eight-cysteine motif, a 900 versatile structure in plant proteins. Plant Physiol. Biochem. PPB 42, 355–365. 901 Kader, J.-C. (1996). LIPID-TRANSFER PROTEINS IN PLANTS. Annu. Rev. Plant Physiol. 902 Plant Mol. Biol. 47, 627–654. 901 Kader, J. C. (1996). LIPID TRANSFER PROTEINS IN PLANTS. Annu. Rev. Plant Physiol. 902 Plant Mol. Biol. 47, 627–654. 903 Kader, J.C., Julienne, M., and Vergnolle, C. (1984). Purification and characterization of a 904 spinach-leaf protein capable of transferring phospholipids from liposomes to mitochondria or 905 chloroplasts. Eur. J. Biochem. 139, 411–416. 906 Katoh, K., and Toh, H. (2010). Manuscript to be reviewed Exogenous application of a lipid transfer protein-jasmonic acid complex induces 880 Girault, T., François, J., Rogniaux, H., Pascal, S., Delrot, S., Coutos-Thévenot, P., and Gomès, E ( ) g pp p p j p 882 protection of grapevine towards infection by Botrytis cinerea. Plant Physiol. Biochem. PPB 46, 883 140–149. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 25 Liu, W., Huang, D., Liu, K., Hu, S., Yu, J., Gao, G., and Song, S. (2010). Disc 926 identification and comparative analysis of non-specific lipid transfer protein (nsLtp) family in 926 identification and comparative analysis of non-specific lipid transfer protein (nsLtp) family in 927 Solanaceae. Genomics Proteomics Bioinformatics 8, 229–237. 27 Solanaceae. Genomics Proteomics Bioinformatics 8, 229–237. 928 van Loon, L.C., Rep, M., and Pieterse, C.M.J. (2006). Significance of inducible defense-related 929 proteins in infected plants. Annu. Rev. Phytopathol. 44, 135–162. , , p, , , ( ) g 929 proteins in infected plants. Annu. Rev. Phytopathol. 44, 135–162. 930 Maghuly, F., Borroto-Fernandez, E.G., Khan, M.A., Herndl, A., Marzban, G., and Laimer, M. 931 (2009). Expression of calmodulin and lipid transfer protein genes in Prunus incisa x serrula 932 under different stress conditions. Tree Physiol. 29, 437–444. 930 Maghuly, F., Borroto-Fernandez, E.G., Khan, M.A., Herndl, A., Marzban, G., and Laimer, M. 931 (2009). Expression of calmodulin and lipid transfer protein genes in Prunus incisa x serrula 932 d diff t t diti T Ph i l 29 437 444 933 Maldonado, A.M., Doerner, P., Dixon, R.A., Lamb, C.J., and Cameron, R.K. (2002). A putative 934 lipid transfer protein involved in systemic resistance signalling in Arabidopsis. Nature 419, 399– 935 403. 936 Meng, C. Yan, Y., Liu, Z., Chen, L., Zhang, Y., Li, X., Wu, L., Zhang, G., Wang, X., Ma, Z. 937 (2018). Systematic Analysis of Cotton Non-specific Lipid Transfer Protein Family Revealed a 936 Meng, C. Yan, Y., Liu, Z., Chen, L., Zhang, Y., Li, X., Wu, L., Zhang, G., Wang, X., Ma, Z. 936 Meng, C. Yan, Y., Liu, Z., Chen, L., Zhang, Y., Li, X., Wu, L., Zhang, G., Wang, X., Ma, Z. 937 (2018). Systematic Analysis of Cotton Non-specific Lipid Transfer Protein Family Revealed a 938 Special Group That Is Involved in Fiber Elongation. Front. in Plant Sci. 9. 1285. 937 (2018). Systematic Analysis of Cotton Non-specific Lipid Transfer Protein ( ) y y p p y 938 Special Group That Is Involved in Fiber Elongation. Front. in Plant Sci. 9. 1285. 939 Molina, A., and García-Olmedo, F. (1993). Developmental and pathogen-induced expression of 940 three barley genes encoding lipid transfer proteins. Plant J. Cell Mol. Biol. 4, 983–991. 941 Mollet, J.C., Park, S.Y., Nothnagel, E.A., and Lord, E.M. (2000). Manuscript to be reviewed A lily stylar pectin is necessary 942 for pollen t be adhesion to an in itro st lar matri Plant Cell 12 1737 1750 941 Mollet, J.C., Park, S.Y., Nothnagel, E.A., and Lord, E.M. (2000). A lily stylar pectin is necessary 942 for pollen tube adhesion to an in vitro stylar matrix. Plant Cell 12, 1737–1750. 942 for pollen tube adhesion to an in vitro stylar matrix. Plant Cell 12, 1737–17 943 Murzin, A.G., Brenner, S.E., Hubbard, T., and Chothia, C. (1995). SCOP: a structural 944 classification of proteins database for the investigation of sequences and structures. J. Mol. Biol. 945 247, 536–540. 946 Nieuwland, J., Feron, R., Huisman, B.A.H., Fasolino, A., Hilbers, C.W., Derksen, J., and 947 Mariani, C. (2005). Lipid transfer proteins enhance cell wall extension in tobacco. Plant Cell 17, 948 2009–2019. 949 Orengo, C.A., Michie, A.D., Jones, S., Jones, D.T., Swindells, M.B., and Thornton, J.M. (1997). 950 CATH--a hierarchic classification of protein domain structures. Struct. Lond. Engl. 1993 5, 951 1093–1108. 952 Ortiz, A.R., Strauss, C.E.M., and Olmea, O. (2002). MAMMOTH (matching molecular models 953 obtained from theory): an automated method for model comparison. Protein Sci. Publ. Protein 954 Soc. 11, 2606–2621. chinnikov, S., Park, H., Varghese, N., Huang, P.-S., Pavlopoulos, G.A., Kim, D.E., 956 Kamisetty, H., Kyrpides, N.C., and Baker, D. (2017). Protein structure determination using 957 metagenome sequence data. Science 355, 294–298. 958 Liu F, Zhang X, Lu C, Zeng X, Li Y, Fu D, Wu G. Non-specific lipid transfer proteins in plants: 959 presenting new advances and an integrated functional analysis. J Exp. Bot. (2015) 960 Pacios, L.F., Gómez-Casado, C., Tordesillas, L., Palacín, A., Sánchez-Monge, R., and Díaz- 960 Pacios, L.F., Gómez-Casado, C., Tordesillas, L., Palacín, A., Sánchez-Monge, R., and Díaz- 961 Perales, A. (2012). Computational study of ligand binding in lipid transfer proteins: Structures, 962 interfaces, and free energies of protein-lipid complexes. J. Comput. Chem. 33, 1831–1844. , , , , , , , , g , , 961 Perales, A. (2012). Computational study of ligand binding in lipid transfer proteins: Structures, 962 i f d f i f i li id l J C Ch 33 1831 1844 963 Park, S.Y., Jauh, G.Y., Mollet, J.C., Eckard, K.J., Nothnagel, E.A., Walling, L.L., and Lord, 964 E.M. (2000). Manuscript to be reviewed Parallelization of the MAFFT multiple sequence alignment 907 program. Bioinforma. Oxf. Engl. 26, 1899–1900. 907 program. Bioinforma. Oxf. Engl. 26, 1899–1900. 908 Katoh, K., Misawa, K., Kuma, K., and Miyata, T. (2002). MAFFT: a novel method for rapid 909 multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 30, 3059–3066. 908 Katoh, K., Misawa, K., Kuma, K., and Miyata, T. (2002). MAFFT: a novel method for rapid 909 lti l li t b d f t F i t f N l i A id R 30 3059 3066 908 Katoh, K., Misawa, K., Kuma, K., and Miyata, T. (2002). MAFFT: a novel method for rapid 909 multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 30, 3059–3066. 910 Konagurthu, A.S., Whisstock, J.C., Stuckey, P.J., and Lesk, A.M. (2006). 910 Konagurthu, A.S., Whisstock, J.C., Stuckey, P.J., and Lesk, A.M. (2006). 911 : a multiple structural alignment algorithm. Proteins 64, 559–574. 912 Kurowski, M.A., and Bujnicki, J.M. (2003). GeneSilico protein structure prediction meta-server. 913 Nucleic Acids Res. 31, 3305–3307. 914 Lange, B.M., Wildung, M.R., Stauber, E.J., Sanchez, C., Pouchnik, D., and Croteau, R. (2000). 915 Probing essential oil biosynthesis and secretion by functional evaluation of expressed sequence 916 tags from mint glandular trichomes. Proc. Natl. Acad. Sci. U. S. A. 97, 2934–2939. 917 Li, F., Fan, K., Ma, F., Yue, E., Bibi, N., Wang, M., Shen, H., Hasan, Md., Wang, X. (2016). 918 Genomic Identification and Comparative Expansion Analysis of the Non-Specific Lipid Transfer 919 Protein Gene Family in Gossypium. Sci Reports. 6. 1-15. 920 Li, X., Gasic, K., Cammue, B., Broekaert, W., and Korban, S.S. (2003). Transgenic rose lines 921 harboring an antimicrobial protein gene, Ace-AMP1, demonstrate enhanced resistance to 922 powdery mildew ( Sphaerotheca pannosa). Planta 218, 226–232. 923 Lichtarge, O., Bourne, H.R., and Cohen, F.E. (1996). An evolutionary trace method defines 924 binding surfaces common to protein families. J. Mol. Biol. 257, 342–358. g , , , , , ( ) y 924 binding surfaces common to protein families. J. Mol. Biol. 257, 342–358. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 66 Pii, Y., Astegno, A., Peroni, E., Zaccardelli, M., Pandolfini, T., and Crimi, M. 966 Pii, Y., Astegno, A., Peroni, E., Zaccardelli, M., Pandolfini, T., and Crimi, M. (2009). The 967 Medicago truncatula N5 gene encoding a root-specific lipid transfer protein is required for the 968 symbiotic interaction with Sinorhizobium meliloti. Mol. Plant-Microbe Interact. MPMI 22, 969 1577–1587. 970 Pii, Y., Molesini, B., and Pandolfini, T. (2013). The involvement of Medicago truncatula non- 971 specific lipid transfer protein N5 in the control of rhizobial infection. Plant Signal. Behav. 8, 972 e24836. 973 Pons, J.-L., and Labesse, G. (2009). @TOME-2: a new pipeline for comparative modeling of 974 protein-ligand complexes. Nucleic Acids Res. 37, W485-491. 975 Poznanski, J., Sodano, P., Suh, S.W., Lee, J.Y., Ptak, M., and Vovelle, F. (1999). Solution 976 structure of a lipid transfer protein extracted from rice seeds. Comparison with homologous 977 proteins. Eur. J. Biochem. 259, 692–708. 978 Pyee, J., and Kolattukudy, P.E. (1995). The gene for the major cuticular wax-associated protein 979 and three homologous genes from broccoli (Brassica oleracea) and their expression patterns. 980 Plant J. Cell Mol. Biol. 7, 49–59. 981 Pyee, J., Yu, H., and Kolattukudy, P.E. (1994). Identification of a lipid transfer protein as the 982 major protein in the surface wax of broccoli (Brassica oleracea) leaves. Arch. Biochem. Biophys. 983 311, 460–468. 984 Qin, X., Liu, Y., Mao, S., Li, T., Wu, H., Chu, C., and Wang, Y. (2011). Genetic transformation 985 of lipid transfer protein encoding gene in Phalaenopsis amabilis to enhance cold resistance. 986 Euphytica 177, 33–43. 987 Radauer, C., and Breiteneder, H. (2007). Evolutionary biology of plant food allergens. J. Allergy 988 Clin. Immunol. 120, 518–525. 989 Radauer, C., Bublin, M., Wagner, S., Mari, A., and Breiteneder, H. (2008). Allergens are 990 distributed into few protein families and possess a restricted number of biochemical functions. J. 991 Allergy Clin. Immunol. 121, 847-852.e7. 992 G. J. Rocklin, T. M. Chidyausiku, I. Goreshnik, A. Ford, S. Houliston, A. Lemak, L. Carter, R. 993 Ravichandran, V. K. Mulligan, A. Chevalier, C. H. Arrowsmith, and D. Baker (2017). Global 994 analysis of protein folding using massively parallel design, synthesis, and testing. Science 357, 995 168–175. 996 Salcedo, G., Sánchez-Monge, R., Barber, D., and Díaz-Perales, A. (2007). Plant non-specific 997 lipid transfer proteins: an interface between plant defence and human allergy. Biochim. Biophys. 998 Acta 1771, 781–791. Manuscript to be reviewed A lipid transfer-like protein is necessary for lily pollen tube adhesion to an in vitro 963 Park, S.Y., Jauh, G.Y., Mollet, J.C., Eckard, K.J., Nothnagel, E.A., Walling, L.L., and Lord, 964 E.M. (2000). A lipid transfer-like protein is necessary for lily pollen tube ad 964 E.M. (2000). A lipid transfer-like protein is necessary for lily pollen tube adhesion to an in vitro 965 stylar matrix. Plant Cell 12, 151–164. 964 E.M. (2000). A lipid transfer like protein is necessary for lily pollen tube adhesion to an in vitro 965 stylar matrix. Plant Cell 12, 151–164. 965 stylar matrix. Plant Cell 12, 151–164. 965 stylar matrix. Plant Cell 12, 151–164. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed 999 Sawano, Y., Hatano, K., Miyakawa, T., Komagata, H., Miyauchi, Y., Yamazaki, H., and 1000 Tanokura, M. (2008). Proteinase inhibitor from ginkgo seeds is a member of the plant 1001 nonspecific lipid transfer protein gene family. Plant Physiol. 146, 1909–1919. 1002 Silverstein, K.A.T., Moskal, W.A., Wu, H.C., Underwood, B.A., Graham, M.A., Town, C.D., 1003 and VandenBosch, K.A. (2007). Small cysteine-rich peptides resembling antimicrobial peptides 1004 have been under predicted in plants Plant J Cell Mol Biol 51 262 280 S , , , W , W , C , U , , G , , , C , 1003 and VandenBosch, K.A. (2007). Small cysteine-rich peptides resembling antimicrobial peptides 1004 have been under-predicted in plants. Plant J. Cell Mol. Biol. 51, 262–280. 1004 have been under-predicted in plants. Plant J. Cell Mol. Biol. 51, 262–280. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) 1024 (nsLTPDB). BMC Genomics 13 Suppl 1, S9. 1024 (nsLTPDB). BMC Genomics 13 Suppl 1, S9. 1025 Manuscript to be reviewed Manuscript to be reviewed 1005 Sror, H.A.M., Tischendorf, G., Sieg, F., Schmitt, J.M., and Hincha, D.K. (2003). Cryoprotectin 1006 protects thylakoids during a freeze-thaw cycle by a mechanism involving stable membrane 1007 binding. Cryobiology 47, 191–203. 1008 Sterk, P., Booij, H., Schellekens, G.A., Van Kammen, A., and De Vries, S.C. (1991). Cell- 1009 specific expression of the carrot EP2 lipid transfer protein gene. Plant Cell 3, 907–921. 1010 Sun, J.-Y., Gaudet, D.A., Lu, Z.-X., Frick, M., Puchalski, B., and Laroche, A. (2008). 1011 Characterization and antifungal properties of wheat nonspecific lipid transfer proteins. Mol. 1012 Plant-Microbe Interact. MPMI 21, 346–360. 1010 Sun, J.-Y., Gaudet, D.A., Lu, Z.-X., Frick, M., Puchalski, B., and Laroche, A. (2008). 1011 Characterization and antifungal properties of wheat nonspecific lipid transfer proteins. Mol. 1012 Plant-Microbe Interact. MPMI 21, 346–360. 1013 Thoma, S., Kaneko, Y., and Somerville, C. (1993). A non-specific lipid transfer protein from 1014 Arabidopsis is a cell wall protein. Plant J. Cell Mol. Biol. 3, 427–436. 1013 Thoma, S., Kaneko, Y., and Somerville, C. (1993). A non-specific lipid transfer protein from 1014 Arabidopsis is a cell wall protein. Plant J. Cell Mol. Biol. 3, 427–436. 1015 Trevino, M.B., and OConnell, M.A. (1998). Three drought-responsive members of the 1016 nonspecific lipid-transfer protein gene family in Lycopersicon pennellii show different 1017 developmental patterns of expression. Plant Physiol. 116, 1461–1468. 1018 Wang, H.W., Hwang, S.-G., Karuppanapandian, T., Liu, A., Kim, W., and Jang, C.S. (2012a). 1019 Insight into the molecular evolution of non-specific lipid transfer proteins via comparative 1020 analysis between rice and sorghum. DNA Res. Int. J. Rapid Publ. Rep. Genes Genomes 19, 179– 1021 194. 1019 Insight into the molecular evolution of non-specific lipid transfer proteins via comparative 1020 analysis between rice and sorghum. DNA Res. Int. J. Rapid Publ. Rep. Genes Genomes 19, 179– 1021 194. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Figure 1 Eﬀect of alternate cysteine residue alignments on the superposition of type I and II nsLTP experimentally determined structures. (A) Common alignment of Cys5 (type I), Cys5’ (types II and IV) (green) and Cys6 (type I), Cys6’ (types II and IV) (magenta) of nsLTP sequences generated by MUSCLE. Only nsLTPs (PDB IDs) with known experimental structures were considered. (C) 3D projection of this alignment leads to an RMSD of 7.32 Å between type I (blue backbone) Cys6 and type II (pink backbone) Cys6’, colorized as in (A). (B) Type I, II and IV nsLTP alignment generated by the MAFFT program, suggesting that type I Cys5 (dark green) corresponds to type II Cys6’ (light green) rather than type II Cys5’. (D) 3D projection of this alignment leads to an RMSD of 2.15 Å between type I Cys5 and type II Cys6’, colorized as in (B). Note that type IV nsLTPs are structurally close to type II nsLTPs. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Figure 2 NsLTP sequence classiﬁcation. Manuscript to be reviewed Manuscript to be reviewed NsLTP sequence classiﬁcation. Dendrogram built on MAFFT alignment of the 797 nsLTP sequences, using Dendroscope program (Huson and Scornavacca 2012). The diﬀerent nsLTP types are displayed using various colors and the number of sequences in each type is speciﬁed in parenthesis. Branch support values of each group are indicated on the corresponding nodes. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Consensus sequence alignment for all nsLTP types. The indicated amino acids are the most frequent for each type of nsLTP and vertical arrows indicate residues analyzed in detail in the text below. The residues are colored as follows: 1. The sequences were sorted according to the FastME phylogenetic tree order. 2. A color was assigned to each sorted sequence according to a rainbow color gradient order. 3. Let i be a position of the alignment, a(i) be an amino acid at position i and A(a(i),i) be the amino acid cluster to which a(i) belongs and which has the lowest Fisher’s test probability FP(A(a(i),i)) relatively to any tree cluster. The color of each aligned amino acid a(i) is coded using the Hue-Saturation-Value color space: the color hue of a(i) corresponds to the hue of the median sequence containing an amino acid of the cluster A(a(i),i) at position i. The color saturation of a(i) is proportional to the speciﬁcity score -log(FP(A(a(i),i))). The color value or intensity of a(i) is reversely proportional to the conservation score of the column i. The more conserved an amino acid cluster is, the darker its color will be, and the more speciﬁc to a phylogenetic group an amino acid cluster is, the more saturated its color will be. Consequently, the globally conserved amino acid clusters are highlighted as dark grey or black columns in the sequence alignment while the amino acids cluster speciﬁc to a subgroup of related sequences are highlighted by aligned amino acids with same saturated colors. Furthermore, the amino acid speciﬁcally conserved in a given protein can then easily be detected by looking at the residues with colors similar to the sequence name color. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Figure 4 Cartoon representation of the crystallographic structures 1mid (type I), 1tuk (type II) and 2rkn (type IV). The residues are numbered and colored as in the multiple sequence alignment of Figure 3. The ligands are represented as ball and sticks (carbon in white, oxygen in red). Some determining amino acid side chains are also displayed. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Figure 6 InTreeGreat view of the structure tree. Manuscript to be reviewed NsLTP structure classiﬁcation. Dendrogram built on Mustang structure-based sequence alignment of the 727 nsLTPs for which a reliable 3D model has been calculated. The two main fold types are displayed in red (type 1 fold) and black (type 2 fold). In order to study their distribution in term of structural families, nsLTP structures are colored according to the previously determined phylogenetic type they belong to (same colors as used in ﬁgure 2). Phylogenetic type I nsLTPs display the type 1 fold and all other nsLTPs follow the type 2 fold. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Manuscript to be reviewed PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) InTreeGreat view of the structure tree. The left pane shows the phylogenic tree of the nsLTP structures colored according to type and the right pane represents a close-up of the Type I (colored in red) part of the tree. For clarity, some sub tree parts for which no annotation was available have been collapsed. They appear as grey triangle and the number of structures they contain is indicated. NsLTPs for which a functional annotation is available are highlighted with a grey box in the left column. On the right side of the tree several columns appear that correspond to annotations (PO, GO), number of leaves in a collapsed sub-tree together with colored boxes. The ﬁrst column of boxes shows alternative colors to enhance the clusters, the other ones correspond to each keyword selected among the annotations of the database (here: “defense” or “resistance”). Keywords “defense” or “resistance” used in functional annotation are highlighted with a colored box (blue and red respectively). The “defense cluster” (see text) has been enlarged (black border) for a better view. Manuscript to be reviewed PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Manuscript to be reviewed Manuscript to be reviewed Table 1(on next page) Figure 7 Conserved amino acid residues among the so-called defense cluster, displayed on the 3D structure of nsLTP 525, (“LTP”, UniProtKB - Q1KMV1). The more the residue is conserved in the 3D alignment, the redder its color appears, then orange, yellow and green. Residues with no signiﬁcative conservation appears in white on the ﬁgure. Residues highlighted in table 2 and which potential functional implication is discussed (see text) are labeled on the ﬁgure. PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Qmean scores obtained by the 797 theoretical models of nsLTPs of this study. Qmean scores obtained by the 797 theoretical models of nsLTPs of this study. Models obtained by @tome2 present an overall good quality as shown in Table 1 that summarizes the Qmean scores. For 95,85% of the models, Qmean scores are above 0.4 and 57% of the models obtained scores ranging from 0.5 to 0.9, which correspond to scores for high-resolution proteins. It is known that disordered protein regions are very ﬂexible regions. While submitted to automatic evaluation, these ﬂexible regions will be considered as regions of bad quality modeling, leading to lower Qmean scores (Benkert, Tosatto et al. 2008; Models obtained by @tome2 present an overall good quality as shown in Table 1 that summarizes the Qmean scores. For 95,85% of the models, Qmean scores are above 0.4 and 57% of the models obtained scores ranging from 0.5 to 0.9, which correspond to scores for high-resolution proteins. It is known that disordered protein regions are very ﬂexible regions. While submitted to automatic evaluation, these ﬂexible regions will be considered as regions of bad quality modeling, leading to lower Qmean scores (Benkert, Tosatto et al. 2008; Benkert, Biasini et al. 2011). Small proteins tend to have lower scores than larger proteins, because of the lower proportion of secondary structures compared to random coils. However, the set of theoretical models calculated by @tome2 obtained overall good Qmean scores.NB: for 121 theoretical structures, the polypeptide chain could not be fully built and the resulting models were lacking at least one of the 8 cysteine residues. Such models were discarded and a new pool of 677 structures was retained for further analysis.The models are available at: http://atome.cbs.cnrs.fr/AT2B/SERVER/LTP.html Manuscript to be reviewed Table 2(on next page) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Qmean score (Q) Nb. of models Dataset proportion Q < 0.2 0 0% 0.2 < Q < 0.3 1 0.1% 0.3 < Q < 0.4 30 4% 0.4 < Q < 0.5 173 22% 0.5 < Q < 0.6 309 39% 0.6 < Q < 0.7 221 28% 0.7 < Q < 0.8 43 5% 0.8 < Q < 0.9 9 1% 0.9 < Q 1 0.1% PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Compared analysis of Evolutionary Trace of three groups of nsLTPs. Compared analysis of Evolutionary Trace of three groups of nsLTPs. Compared analysis of Evolutionary Trace of three groups of nsLTPs: (A) the defense cluster (43 proteins), (B) the cluster containing all type 1 fold nsLTPs (402 proteins) and (C) a group composed by all type 1 fold defense/resistance nsLTPs, including those which do not belong to the defense cluster (28 proteins). This table lists the 30% top-ranked (= most conserved) residues identiﬁed in the defense cluster trace and shows, by comparison, the ranking of these same residues in the other two traces, together with their coverage, variability, and rvET score. Residue positions in the reference proteins and in the structure-based sequence alignment are also indicated. Alignment position is the same in all three groups because all three alignments used to perform the traces are extracted from the general multiple alignments of all 797 nsLTPs of the study. Five residues are highlighted for they are diﬀerently conserved in the three clusters of proteins (see text). PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed A Defense cluster (ref. prot. = 525) Rank Residue Number Alignment Position Residue Coverage Variability rvET score 1 4 93 C 0.10000 C 1.00 1 14 159 C 0.10000 C 1.00 1 29 228 C 0.10000 C 1.00 1 30 229 C 0.10000 C 1.00 1 45 259 D 0.10000 D 1.00 1 50 275 C 0.10000 C 1.00 1 52 277 C 0.10000 C 1.00 1 72 372 C 0.10000 C 1.00 1 86 432 C 0.10000 C 1.00 10 7 137 V 0.13333 AV 1.11 11 32 231 G 0.13333 SG 1.11 12 80 402 I 0.13333 VI 1.11 13 69 367 P 0.14444 PA 1.17 14 36 236 L 0.15556 LV 1.28 15 17 165 Y 0.16667 FY 1.59 16 74 374 V 0.17778 LVIA 1.75 17 11 154 L 0.18889 LV 1.83 18 54 289 K 0.20000 VKQ 1.93 19 65 360 A 0.21111 TALV 2.01 20 40 247 A 0.22222 TAV 2.13 21 1 63 A 0.23333 .AD 2.15 22 33 232 V 0.24444 AVI 2.29 23 68 364 I 0.25556 LI 2.50 24 43 256 T 0.26667 TPMAS 2.61 25 61 344 N 0.27778 KNSV 2.65 26 47 268 Q 0.28889 RQK 2.71 27 46 266 K 0.30000 RK 2.75 B Fold 1 nsLTPs (ref. prot. Manuscript to be reviewed = 525) Rank Residue Number Alignment Position Residue Coverage Variability rvET score 1 4 93 C 0.11111 C 1.00 1 14 159 C 0.11111 C 1.00 1 29 228 C 0.11111 C 1.00 1 30 229 C 0.11111 C 1.00 1 50 275 C 0.11111 C 1.00 1 52 277 C 0.11111 C 1.00 1 72 372 C 0.11111 C 1.00 1 86 432 C 0.11111 C 1.00 1 7 137 V 0.11111 V 1.00 1 69 367 P 0.11111 P 1.00 11 45 259 D 0.13333 DL 1.15 12 80 402 I 0.13333 IW 1.15 13 74 374 V 0.15556 VIN 1.39 16 17 165 Y 0.18889 YF 1.67 17 36 236 L 0.18889 LI 1.67 18 32 231 G 0.20000 GAV 1.76 20 54 289 K 0.22222 KVQ 1.93 22 65 360 A 0.24444 AVF 2.04 23 40 247 A 0.25556 ATVS 2.05 25 33 232 V 0.27778 VALI 2.59 27 61 344 N 0.30000 NVDR 2.88 30 46 266 K 0.33333 KR 3.25 31 11 154 L 0.34444 LIVM 3.26 36 43 256 T 0.40000 TPQRS 3.72 38 68 364 I 0.42222 ILV 3.95 44 47 268 Q 0.48889 QRK 4.61 45 1 63 A 0.50000 A.QV 4.63 13 36 236 L 0.13978 LVIM 13.58 14 77 374 V 0.15054 VLTAINP 13.66 15 17 165 Y 0.16129 YFAH 13.82 16 40 247 A 0.17204 ATSVIRPL 14.49 17 68 360 A 0.18280 ATVLFIM 14.52 18 71 364 I 0.19355 LIVTAPFM 14.53 19 54 289 K 0.20430 KVQIERLMHTS 15.40 20 45 259 D 0.21505 DAENITLRG.K 15.74 21 83 402 I 0.22581 IVFPLTAKW 15.92 29 33 232 V 0.31183 VAILSM 21.38 32 47 268 R 0.34409 KQRVEMIYSH 24.45 34 11 154 I 0.36559 VLMIFATP 25.38 42 64 344 N 0.45161 NGKQDASTLERVFYI 54.16 56 43 256 T 0.60215 TAPGRSQKDHVMI.LFY 38.13 61 1 63 A 0.65591 .AHETDVPSGFQL 39.96 C Defense nsLTPs outside cluster (ref. prot. Compared analysis of Evolutionary Trace of three groups of nsLTPs. = 437) Rank Residue Number Alignment Position Residue Coverage Variability rvET score 1 14 159 C 0.05376 C 1.00 1 29 228 C 0.05376 C 1.00 1 30 229 C 0.05376 C 1.00 1 50 275 C 0.05376 C 1.00 1 52 277 C 0.05376 C 1.00 6 75 372 C 0.06452 CR 1.75 7 4 93 C 0.07527 CA 3.00 8 89 432 C 0.08602 CDN 4.36 9 72 367 P 0.09677 PASLQG 7.27 10 46 266 R 0.10753 RKTAPIQD 11.55 11 7 137 V 0.11828 VALISGT 11.81 12 32 231 G 0.12903 GSAEQVHR 13.26 A A Defense cluster (ref. prot. = 525) B PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019) Manuscript to be reviewed Manuscript to be reviewed 13 36 236 L 0.13978 LVIM 13.58 14 77 374 V 0.15054 VLTAINP 13.66 15 17 165 Y 0.16129 YFAH 13.82 16 40 247 A 0.17204 ATSVIRPL 14.49 17 68 360 A 0.18280 ATVLFIM 14.52 18 71 364 I 0.19355 LIVTAPFM 14.53 19 54 289 K 0.20430 KVQIERLMHTS 15.40 20 45 259 D 0.21505 DAENITLRG.K 15.74 21 83 402 I 0.22581 IVFPLTAKW 15.92 29 33 232 V 0.31183 VAILSM 21.38 32 47 268 R 0.34409 KQRVEMIYSH 24.45 34 11 154 I 0.36559 VLMIFATP 25.38 42 64 344 N 0.45161 NGKQDASTLERVFYI 54.16 56 43 256 T 0.60215 TAPGRSQKDHVMI.LFY 38.13 61 1 63 A 0.65591 .AHETDVPSGFQL 39.96 C Defense nsLTPs outside cluster (ref. prot. Manuscript to be reviewed = 525) Rank Residue Number Alignment Position Residue Coverage Variability rvET score 1 4 93 C 0.11111 C 1.00 1 14 159 C 0.11111 C 1.00 1 29 228 C 0.11111 C 1.00 1 30 229 C 0.11111 C 1.00 1 50 275 C 0.11111 C 1.00 1 52 277 C 0.11111 C 1.00 1 72 372 C 0.11111 C 1.00 1 86 432 C 0.11111 C 1.00 1 7 137 V 0.11111 V 1.00 1 69 367 P 0.11111 P 1.00 11 45 259 D 0.13333 DL 1.15 12 80 402 I 0.13333 IW 1.15 13 74 374 V 0.15556 VIN 1.39 16 17 165 Y 0.18889 YF 1.67 17 36 236 L 0.18889 LI 1.67 18 32 231 G 0.20000 GAV 1.76 20 54 289 K 0.22222 KVQ 1.93 22 65 360 A 0.24444 AVF 2.04 23 40 247 A 0.25556 ATVS 2.05 25 33 232 V 0.27778 VALI 2.59 27 61 344 N 0.30000 NVDR 2.88 30 46 266 K 0.33333 KR 3.25 31 11 154 L 0.34444 LIVM 3.26 36 43 256 T 0.40000 TPQRS 3.72 38 68 364 I 0.42222 ILV 3.95 44 47 268 Q 0.48889 QRK 4.61 45 1 63 A 0.50000 A.QV 4.63 C 1 PeerJ reviewing PDF \| (2019:04:36959:2:1:ACCEPTED 8 Jul 2019)
https://openalex.org/W3170050807	https://www.nature.com/articles/s42003-021-02267-y.pdf	English	null	Kinetic mechanism of Na+-coupled aspartate transport catalyzed by GltTk	Communications biology	2,021	cc-by	12,876	1 Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Groningen, The Netherlands. 2 Institute of Medical Microbiology, University of Zurich, Zurich, Switzerland. 3 Moscow Institute of Physics and Technology, Dolgoprudny, Russia. 4Present address: ZoBio BV, Leiden, The Netherlands. 5Present address: Institute of Medical Microbiology, University of Zurich, Zurich, Switzerland. ✉email: d.j.slotboom@rug.nl COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio Kinetic mechanism of Na+-coupled aspartate transport catalyzed by GltTk Gianluca Trinco1, Valentina Arkhipova1,4, Alisa A. Garaeva1,5, Cedric A. J. Hutter 2, Markus A. Seeger 2, Albert Guskov 1,3 & Dirk J. Slotboom 1✉ It is well-established that the secondary active transporters GltTk and GltPh catalyze coupled uptake of aspartate and three sodium ions, but insight in the kinetic mechanism of transport is fragmentary. Here, we systematically measured aspartate uptake rates in proteoliposomes containing puriﬁed GltTk, and derived the rate equation for a mechanism in which two sodium ions bind before and another after aspartate. Re-analysis of existing data on GltPh using this equation allowed for determination of the turnover number (0.14 s−1), without the need for error-prone protein quantiﬁcation. To overcome the complication that puriﬁed transporters may adopt right-side-out or inside-out membrane orientations upon reconstitution, thereby confounding the kinetic analysis, we employed a rapid method using synthetic nanobodies to inactivate one population. Oppositely oriented GltTk proteins showed the same transport kinetics, consistent with the use of an identical gating element on both sides of the mem- brane. Our work underlines the value of bona ﬁde transport experiments to reveal mechanistic features of Na+-aspartate symport that cannot be observed in detergent solu- tion. Combined with previous pre-equilibrium binding studies, a full kinetic mechanism of structurally characterized aspartate transporters of the SLC1A family is now emerging. 1 Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Groningen, The Netherlands. 2 Institute of Medical Microbiology, University of Zurich, Zurich, Switzerland. 3 Moscow Institute of Physics and Technology, Dolgoprudny, Russia. 4Present address: ZoBio BV, Leiden, The Netherlands. 5Present address: Institute of Medical Microbiology, University of Zurich, Zurich, Switzerland. ✉email: d.j.slotboom@rug.nl 1 ARTICLE COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y E xcitatory amino acid transporters (EAATs) of the solute carrier family 1A (SLC1A) take up the neurotransmitter L-glutamate from the synaptic environment, which is necessary to keep the extracellular concentration low and prevent neurotoxicity1,2. EAATs couple uptake of one amino acid substrate molecule to the co- transport of three sodium ions and one proton and counter-transport of one potassium ion3–6. Thus, glutamate gradients of a million-fold across the membrane under resting conditions can be sustained. The closely related archaeal transporters GltPh from Pyrococcus horikoshii and GltTk from Thermococcus kodakarensis of the SLC1A family (78% sequence identity to each other, ~36% sequence identity to EAATs) take up aspartate rather than glutamate in symport with three sodium ions and are not coupled to potassium or proton transport7–11. Results T d To study the kinetic mechanism of Na+-coupled aspartate transport by GltTk, we used a classical enzymology method, in which we measured the initial uptake rates of radiolabelled L-aspartate into proteoliposomes reconstituted with puriﬁed GltTk, as a function of the external concentrations of L-aspartate and Na+. To ensure initial rate conditions, aspartate and Na+ were absent from the lumen of the liposomes at the onset of the experiment, and the rate was determined from the linear part of the uptake experiment (Supplementary Fig. 1). In the ﬁrst set of experiments (presented in Figs. 1–3 and Tables 1–3), we measured the combined transport activity of proteins with right-side-out and inside-out membrane orientation in the liposomes, as we did not inactivate either of the two populations. These experiments can be compared with binding experiments performed in detergent-solution where the sidedness is absent. In follow-up experiments (presented in Fig. 4 and Table 4), we silenced one of the two populations, which allows for compar- ison with experiments in which the transporters had been ﬁxed in a single orientation by crosslinking35,36. Transport assays in proteoliposomes have revealed that both GltPh and GltTk catalyze electrogenic transport with a strict stoi- chiometry of three co-transported Na+ ions per aspartate14,32. Data from studies on equilibrium binding and pre-equilibrium kinetics of binding with the solubilized proteins in detergent solution have shown that co-transported ions and aspartate bind in a highly cooperative way, which is crucial to ensure thermo- dynamic coupling10,15,33–37. These experiments have indicated that most likely two sodium ions bind ﬁrst, then aspartate, and ﬁnally the third sodium ion. The binding of the last Na+ leads to gate closure of HP2, which in turn is a prerequisite for elevator- like conformational changes that translocate the bound cargo across the membrane. Structures of GltPh and GltTk have provided a qualitative explanation for the observed binding order. Two of the sodium binding sites (named Na1 and Na3) are buried deep in the proteins15. A substantial conformation rearrangement in the apo-protein (most pronounced in the conserved unwound region of TM7) is required to create the geometry needed for sodium binding, which makes this step slow. The conformational rear- rangement, which is stabilized by the binding of the two sodium ions, also affects residues involved in aspartate binding. While the apo-state does not have a measurable afﬁnity for aspartate, sodium binding creates a high-afﬁnity site for the amino acid substrate. Kinetic mechanism of Na+-coupled aspartate transport catalyzed by GltTk These prokaryotic homologs of the neurotransmitter transporters have been instrumental in delineating shared structural features of this transporter family7–9,12–22. E for the mechanistic characterization of enzymes, leading to insight into the order of binding of substrates38. This method has not been used much on puriﬁed membrane transporters, in part because it is often impossible to control the orientation of the reconstituted transporters in proteoliposomes, leading to mixed populations, thus complicating kinetic analysis. To overcome this latter problem, we isolated an inhibitory synthetic nanobody39, which we used to inactivate membrane transporters oriented in one of the two possible orientations. This method is rapid, generally applicable, and does not depend on mutagenesis and chemical modiﬁcations. The work presented here, combined with the results of previous pre-equilibrium binding studies36, allows for the determination of accurate turnover numbers, which we illustrate by analyzing available data for aspartate transport by GltPh. SLC1A family proteins are homotrimers, with independently operating protomers19,21,23–29, each organized in two domains. A rigid scaffold domain mediates all the contacts with the neigh- boring protomers, and a peripheral transport domain binds the amino acid substrate and cations13,16,30,31. The transport domains are mobile and move through the lipid bilayer (alike an elevator) when translocating the amino acid substrate and co- transported ions across the membrane13. During movement of the transport domain, the substrate-binding site is occluded from the solvent and shielded by the tips of two pseudo-symmetrical helical hairpins (HP1 and HP2). The latter hairpin is a gating element that can hinge between a closed position (taken during elevator movements) and an open position (allowing loading or release of the substrate and co-transported ions). The extent of the elevator-like movement of the transport domain is so large (~20 Å in GltTk) that HP2 acts as a gating element both on the extracellular and the intracellular side of the membrane. COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio Results T d The last sodium ion binds to a site in direct contact with the HP2 gate (Na2), and locks the gate in the closed position, with aspartate and the three sodium ions occluded from the environment. g y g We managed to determine accurate transport rates using external Na+ concentrations in the range between 5 and 300 mM and aspartate concentrations between 50 nM and 100 µM (Table 1). The upper and lower boundaries of the concentration ranges were set by practical considerations. Aspartate concentrations higher than 100 µM required large dilution of the radiolabelled amino acid with unlabeled aspartate, which caused poor signal-to-noise levels in the uptake experiments. Na+ concentrations higher than 300 mM could not be used, because the preparation of proteoliposomes in buffer containing high salt concentrations prevented the formation of a ﬁrm pellet when centrifuged, therefore making it impossible to reach the necessary protein concentration for the experiments. In the low concentration regime, conditions in which 1 mM Na+ was used in combination with aspartate concentrations lower than 1 µM resulted in poor signal-to-noise ratios. Despite these limitations, the range of concentrations was sufﬁcient to provide insight into the kinetic mechanism. The results of the uptake experiments in liposomes with mixed protein orientations are summarized in Table 1, where each row contains the initial rates of transport (v0) at a ﬁxed sodium concentration, but with increasing aspartate concentrations. When analyzing v0 as a function of the aspartate concentration row-by-row, we found that rectangular hyperbolic functions ﬁtted the data well (Fig. 1a), which allowed for the determination of the apparent maximal rates of transport (vAsp max (app)) and apparent Michaelis-Menten constants (KAsp M (app)) (Table 2). The super- script “Asp” indicates that the aspartate was varied, while the sodium concentration was kept constant (hence “apparent”). Despite the enormous amount of data on structure and binding mechanism in detergent solution, insight into the kinetic mechan- ism under translocating conditions is fragmented and incomplete. Here, we set out to determine the kinetic mechanism of aspartate transport by using puriﬁed GltTk, reconstituted in proteoliposomes. Results T d We measured initial rates of transport at a wide range of substrate and co-ion concentrations, a method that has been used extensively Each column in Table 1 contains the measured initial rates of aspartate transport at increasing Na+ concentrations while 2 COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y ARTICLE 0 50 100 0 20 40 60 [Asp] (µM) v0 (mol Asp (mol GltTk)-1 min-1) 300 mM Na+ 200 mM Na+ 100 mM Na+ 50 mM Na+ 25 mM Na+ 10 mM Na+ 5 mM Na+ 0 100 200 300 0 20 40 60 [Na+] (mM) v0 (mol Asp (mol GltTk)-1 min-1) 0.05 µM Asp 0.1 µM Asp 0.5 µM Asp 1 µM Asp 5 µM Asp 10 µM Asp 50 µM Asp 100 µM Asp a b Fig. 1 L-Asp transport rates catalyzed by puriﬁed and reconstituted GltTk as a function of the concentrations of Na+ and L-aspartate. The rates represent the combined contributions of right-side-out and inside-out oriented proteins. a Aspartate-dependent measurements at different ﬁxed Na+ concentrations. The lines represent ﬁts of the Michaelis-Menten equation to the data for uptake at Na+ concentrations of 5 mM (red), 10 mM (orange), 25 mM (yellow), 50 mM (green), 100 mM (cyan), 200 mM (blue), 300 mM (purple). b Sodium-dependence of transport at ﬁxed L-Asp concentrations. The lines represent ﬁts of the Hill equation to the data for uptake at 0.05 µM (red), 0.1 µM (orange), 0.5 µM (yellow), 1 µM (green), 5 µM (cyan), 10 µM (light blue), 50 µM (blue), 100 µM (purple). Each uptake rate represents the average of three independent biological replicates, each constituted by two technical replicates, and the standard error of the mean is shown. maintaining a ﬁxed aspartate concentration. Analysis of these rates as a function of the Na+ concentration revealed sigmoidal depen- dencies (Fig. 1b). The Hill equation was used to ﬁt the data, yielding values for vNa max (app) and KNa M (app) and the apparent Hill coefﬁ- cient nHill (app) (Table 3). In this case, the superscript “Na” indi- cates sodium-dependent measurements, and “apparent” indicates that the measurements were done at a ﬁxed L-aspartate concentration. Results T d 0 50 100 0 20 40 60 [Asp] (µM) v0 (mol Asp (mol GltTk)-1 min-1) 300 mM Na+ 200 mM Na+ 100 mM Na+ 50 mM Na+ 25 mM Na+ 10 mM Na+ 5 mM Na+ a For interpretation of the uptake data in the framework of a kinetic model of transport, the apparent afﬁnities for L-aspartate (KAsp M (app)) and the apparent maximal rates (vAsp max (app) and vNa max (app)) are the most informative parameters. KNa M (app) and nHill (app) contain less useful information for discrimination between different mechanisms (as discussed in references 38,40), and therefore these values were not used further here. 0 100 200 300 0 20 40 60 [Na+] (mM) v0 (mol Asp (mol GltTk)-1 min-1) 0.05 µM Asp 0.1 µM Asp 0.5 µM Asp 1 µM Asp 5 µM Asp 10 µM Asp 50 µM Asp 100 µM Asp b KM analysis reveals a complex mechanism. Information on the kinetic mechanism is contained both in the dependence of the apparent afﬁnity for aspartate KAsp M (app) on the Na+ concentration and in potential differences between KAsp M (app) and the equilibrium constant for L-aspartate binding (KAsp D (app)). Table 2 shows that KAsp M (app) was strongly dependent on the Na+ concentration at concentrations below 50 mM Na+. At Na+ concentrations above 100 mM, KAsp M (app) became independent of the Na+ concentration and leveled off to 0.7 μM (Table 2), a value that is important for mechanistic interpretation (discussed in detail below). A [Na+] (mM) Fig. 1 L-Asp transport rates catalyzed by puriﬁed and reconstituted GltTk as a function of the concentrations of Na+ and L-aspartate. The rates Fig. 1 L-Asp transport rates catalyzed by puriﬁed and reconstituted GltTk as a function of the concentrations of Na+ and L aspartate The rates Fig. 1 L-Asp transport rates catalyzed by puriﬁed and reconstituted GltTk as a function of the concentrations of Na+ and L-aspartate. The rates represent the combined contributions of right-side-out and inside-out oriented proteins. a Aspartate-dependent measurements at different ﬁxed Na+ concentrations. The lines represent ﬁts of the Michaelis-Menten equation to the data for uptake at Na+ concentrations of 5 mM (red), 10 mM (orange), 25 mM (yellow), 50 mM (green), 100 mM (cyan), 200 mM (blue), 300 mM (purple). b Sodium-dependence of transport at ﬁxed L-Asp concentrations. Results T d The lines represent ﬁts of the Michaelis-Menten equation to the data for uptake at Na+ concentrations of 5 mM (red), 10 mM (orange), 25 mM (yellow), 50 mM (green), 100 mM (cyan), 200 mM (blue), 300 mM (purple). b Sodium-dependence of transport at ﬁxed L-Asp concentrations. The lines represent ﬁts of the Hill equation to the data for uptake at 0.05 µM (red), 0.1 µM (orange), 0.5 µM (yellow), 1 µM (green), 5 µM (cyan), 10 µM (light blue), 50 µM (blue), 100 µM (purple). Each uptake rate represents the average of three independent biological replicates, each constituted by two technical replicates, and the standard error of the mean is shown. 0.01 0.1 1 10-7 10-6 10-5 10-4 [Na+] (M) KM or KD (M) KD KM Fig. 2 Dependence of the apparent afﬁnities for L-Asp (KASP M (app)) on the sodium ion concentration (black symbols). Dashed lines represent linear ﬁts to data points in the low and high sodium ion concentration regimes. Each value represents the average of three independent biological replicates, and two technical replicates. The standard error of the mean is shown. For comparison, the dissociation constants KAsp D (app) determined previously are also plotted (orange)15. The data represent the combined contributions of right-side-out and inside-out oriented proteins. 0.01 0.1 1 10-7 10-6 10-5 10-4 [Na+] (M) KM or KD (M) KD KM Mechanistic interpretation of transport data using the steady- state approximation. Because the rapid equilibrium approx- imation was found invalid, we turned to analysis based on the steady-state assumption. While the Michaelis–Menten or Hill equations can always describe the substrate dependencies of the uptake rates when the rapid equilibrium approximation is valid, it is possible to ﬁnd more complex relations in the steady-state analysis, depending on the details of the kinetic mechanism38. For instance, for some kinetic mechanisms, vmax values may be (local) Fig. 2 Dependence of the apparent afﬁnities for L-Asp (KASP M (app)) on Fig. 2 Dependence of the apparent afﬁnities for L-Asp (KASP M (app)) on the sodium ion concentration (black symbols). Dashed lines represent linear ﬁts to data points in the low and high sodium ion concentration regimes. Each value represents the average of three independent biological replicates, and two technical replicates. The standard error of the mean is shown. For comparison, the dissociation constants KAsp D (app) determined previously are also plotted (orange)15. Results T d The lines represent ﬁts of the Hill equation to the data for uptake at 0.05 µM (red), 0.1 µM (orange), 0.5 µM (yellow), 1 µM (green), 5 µM (cyan), 10 µM (light blue), 50 µM (blue), 100 µM (purple). Each uptake rate represents the average of three independent biological replicates, each constituted by two technical replicates, and the standard error of the mean is shown. p Comparison of the KAsp M (app) values with the equilibrium constants for L-aspartate binding (KAsp D (app)), determined previously by isothermal titration calorimetry (ITC) in detergent solution15, revealed almost identical values at a Na+ concentration of 100 mM, but large differences at higher and lower concentrations of Na+ (Fig. 2). While KAsp M (app) was an order of magnitude higher than the equilibrium constant for binding at a sodium concentra- tion of 300 mM (7.0 × 102 nM versus 75 nM), at concentrations below ~75 mM, KAsp M (app) was up to an order of magnitude lower than the KAsp D (app). Such discrepancies are indicative of a complex kinetic mechanism that cannot be interpreted in the conceptual framework of the rapid equilibrium approximation, which is based on the assumption that the transport step (described by turnover number kcat) is much slower than the establishment of the binding equilibrium of sodium ions and aspartate, described by the equilibrium constants KD. The rapid equilibrium assumption was previously also dismissed for aspartate transport by GltPh, based on a more limited comparison of KM and KD10, and on pre- equilibrium binding experiments in detergent solution36, but the data presented here, based on the comparison of transport and binding experiments over a broad range of sodium concentrations, revealed a variable ratio between KAsp M (app) and KAsp D (app) depending on the Na+ concentration, which is indicative of kinetic complexity. It is noteworthy that the rapid equilibrium assumption might hold at very low Na+ concentration36, but as discussed above, the sensitivity of the radiolabel-based transport assays is not high enough to measure aspartate uptake in such conditions. as a function of the concentrations of Na+ and L-aspartate. The rates represent the combined contributions of right-side-out and inside-out oriented proteins. a Aspartate-dependent measurements at different ﬁxed Na+ concentrations. Results T d Each uptake rate represents the average of three independent biological replicates and independent reconstitutions, each with two technical replicates. The standard error of the mean is indicated. Table 1 Initial rates of L-aspartate uptake by GltTk reconstituted in proteoliposomes. In the ﬁrst column, the concentrations in parentheses indicate the amount of choline chloride used in the external reaction buffer to balance the osmotic and ionic strength. Each uptake rate represents the average of three independent biological replicates and independent reconstitutions, each with two technical replicates. The standard error of the mean is indicated. Results T d [L-Asp]→[Na+↓] 0.05 µM 0.1 µM 0.5 µM 1 µM 5 µM 10 µM 50 µM 100 µM 1 mM (299 mM) Not determined Not determined Not determined 0.03 ± 2.6 10−2 min−1 0.51 ± 0.3 min−1 0.25 ± 0.3 min−1 2.85 ± 1.2 min−1 10.93 ± 4.4 min−1 5 mM (295 mM) 0.05 ± 1.6 10−2 min−1 0.06 ± 2.1 10−2 min−1 0.59 ± 0.1 min−1 0.96 ± 0.2 min−1 3.62 ± 0.7 min−1 6.48 ± 0.9 min−1 16.59 ± 2.8 min−1 21.62 ± 3.2 min−1 10 mM (290 mM) 0.21 ± 5.5 10−2 min−1 0.28 ± 0.1 min−1 2.16 ± 0.4 min−1 3.75 ± 0.6 min−1 8.19 ± 1.2 min−1 12.60 ± 1.9 min−1 25.08 ± 5.0 min−1 30.51 ± 4.1 min−1 25 mM (275 mM) 0.95 ± 0.2 min−1 1.46 ± 0.5 min−1 8.53 ± 1.2 min−1 11.56 ± 1.7 min−1 19.09 ± 2.0 min−1 26.39 ± 3.0 min−1 40.58 ± 6.4 min−1 43.59 ± 6.9 min−1 50 mM (250 mM) 1.88 ± 0.4 min−1 3.34 ± 1.0 min−1 16.79 ± 2.6 min−1 21.26 ± 2.3 min−1 31.28 ± 3.5 min−1 39.30 ± 4.9 min−1 43.69 ± 8.3 min−1 50.21 ± 4.4 min−1 100 mM (200 mM) 2.75 ± 0.5 min−1 4.79 ± 1.5 min−1 23.46 ± 2.5 min−1 30.47 ± 3.3 min−1 38.90 ± 3.5 min−1 52.16 ± 7.7 min−1 53.52 ± 7.9 min−1 68.13 ± 3.5 min−1 200 mM (100 mM) 3.06 ± 0.6 min−1 4.83 ± 1.5 min−1 31.27 ± 6.4 min−1 32.55 ± 3.8 min−1 37.29 ± 4.0 min−1 49.80 ± 9.0 min−1 51.90 ± 6.7 min−1 65.40 ± 7.5 min−1 300 mM (0 mM) 2.93 ± 0.7 min−1 4.96 ± 1.6 min−1 27.72 ± 5.3 min−1 28.95 ± 4.0 min−1 40.12 ± 5.4 min−1 46.98 ± 7.9 min−1 52.69 ± 8.5 min−1 61.77 ± 2.9 min−1 initial rates are described well by rectangular hyperbola and sig- moidal curves (Figs. 1 and 2), suggesting that such random steps do not play a signiﬁcant role, at least not in the concentration regime that we used. This notion is consistent with the kinetic binding model derived for detergent-solubilized GltPh, where the binding of sodium ions and aspartate to GltPh was found to be ordered in the concentration range between 5 and 300 mM Na+, with two sodium ions binding before and one after aspartate36. Results T d The data represent the combined contributions of right-side-out and inside-out oriented proteins. MMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio 3 COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y ARTICLE 0 100 200 300 0 20 40 60 [Na+] (mM) Vmax Asp (app) (mol Asp (mol GltTk)-1 min-1) 0 50 100 0 20 40 60 80 [Asp] (µM) Vmax Na (app) (mol Asp (mol GltTk)-1 min-1) kcat/k3 = 0.70 +/- 0.3 µM a b Fig. 3 Dependence of the maximal rates L-Asp transport rate on the concentrations of Na+ and L-aspartate. a Dependence of the maximal rates of transport vNa max (app) from Fig. 1a on the concentration of L-Asp. b Dependence of the maximal rates of transport vAsp max (app) from Fig. 1b on the concentration of sodium ions. Solid and dashed lines represent ﬁts of rectangular hyperbolic functions to the data and the 95% conﬁdence intervals, respectively. In panel A the ﬁtted value of kcat/k3 (Eq. (13)) is indicated. Each value represents the average of three independent biological replicates, and two technical replicates. The standard error of the mean is shown. The data represent the combined contributions of right-side-out and inside-out oriented proteins. 0 50 100 0 20 40 60 80 [Asp] (µM) Vmax Na (app) (mol Asp (mol GltTk)-1 min-1) kcat/k3 = 0.70 +/- 0.3 µM a 0 100 200 300 0 20 40 60 [Na+] (mM) Vmax Asp (app) (mol Asp (mol GltTk)-1 min-1) b Fig. 3 Dependence of the maximal rates L-Asp transport rate on the concentrations of Na+ and L-aspartate. a Dependence of the maximal rates of transport vNa max (app) from Fig. 1a on the concentration of L-Asp. b Dependence of the maximal rates of transport vAsp max (app) from Fig. 1b on the concentration of sodium ions. Solid and dashed lines represent ﬁts of rectangular hyperbolic functions to the data and the 95% conﬁdence intervals, respectively. In panel A the ﬁtted value of kcat/k3 (Eq. (13)) is indicated. Each value represents the average of three independent biological replicates, and two technical replicates. The standard error of the mean is shown. The data represent the combined contributions of right-side-out and inside-out oriented proteins. Table 1 Initial rates of L-aspartate uptake by GltTk reconstituted in proteoliposomes. In the ﬁrst column, the concentrations in parentheses indicate the amount of choline chloride used in the external reaction buffer to balance the osmotic and ionic strength. Results T d This ordered binding mechanism is also consistent with a recent study that revealed a conformational selection step leading to the binding of the ﬁrst sodium ion34. Therefore, we chose to analyze the transport data presented here with the following kinetic model: Table 2 vAsp max (app) and KAsp M (app) values for aspartate dependent uptakes obtained at constant [Na+]. [Na+] vASP max (app) (min−1) KASP M (app) (µM) 5 mM 29.2 ± 0.58 36 ± 1.9 10 mM 34.8 ± 1.5 16.8 ± 2.4 25 mM 44.7 ± 2.9 5.5 ± 1.5 50 mM 45.5 ± 2.3 1.2 ± 0.3 100 mM 52.1 ± 2.6 0.70 ± 0.16 200 mM 54.2 ± 4.1 0.58 ± 0.23 300 mM 53.7 ± 3.1 0.70 ± 0.22 Values are derived from the data presented in Table 1 and Fig. 1a. Each uptake rate represents the average of three independent biological replicates, each with two technical replicates. The standard error of the mean is indicated. Table 2 vAsp max (app) and KAsp M (app) values for aspartate dependent uptakes obtained at constant [Na+]. E $ k1 Naþ ½ k1 ENa f g $ k2 Naþ ½ k2 ENaNa f g $ k3 Asp ½ k3 ENaNaAsp $ k4 Naþ ½ k4 ENaNaAspNa ! kcat ð1Þ i hi h E d i Gl E $ k1 Naþ ½ k1 ENa f g $ k2 Naþ ½ k2 ENaNa f g $ k3 Asp ½ k3 ENaNaAsp $ k4 Naþ ½ k4 ENaNaAspNa ! kcat ð1Þ in which E designates GltTk. maxima, instead of being reached asymptotically, and KM values may be undeﬁned. Such complex relations often arise when steps occur in the mechanism where two different substrates (corre- sponding to aspartate and Na+ in the case of GltTk) bind randomly38. In our data, the concentration dependences of the in which E designates GltTk. While we base our further analysis on the mechanism shown in Eq. (1), we will show in the discussion section that the main conclusions hold for any mechanism in which at least one sodium ion binds after aspartate. Results T d Values are derived from the data presented in Table 1 and Fig. 1b. Each uptake rate represents the average of three independent biological replicates, each with two technical replicates. The standard error of the mean is indicated. p Equation (2) can also be rearranged to derive an expression for KAsp M (app): Table 4 vAsp maxp (app) and KAsp M (app) values for aspartate dependent uptakes obtained at constant [Na+] in the presence of 750 nM sybody on the outside of the liposomes. [Na+] vAsp max (app) (min−1) KAsp M (app) (µM) 10 mM 9.8 ± 1.0 4.9 ± 1.4 100 mM 18.3 ± 2.0 0.97 ± 0.36 200 mM 17.0 ± 1.0 0.65 ± 0.13 300 mM 17.2 ± 1.5 0.63 ± 0.17 Each uptake rate represents the average of two independent biological replicates, each with two technical replicates. The standard error of the mean is indicated. Table 4 vAsp maxp (app) and KAsp M (app) values for aspartate dependent uptakes obtained at constant [Na+] in the presence of 750 nM sybody on the outside of the liposomes. KAsp M app ¼ a4½Na3 þ a5½Na2 þ a6 Na ½ þ a7 a1½Na3 þ a2½Na2 þ a3 Na ½ ð12Þ ð12Þ From Eq. (12) the values for KAsp M (app) that are reached in the low and high [Na+] regimes can be found: lim Na ½ ! KAsp M app ¼ a4 a1 ¼ kcat k3 ð13Þ lim Na ½ !0 KAsp M app ¼ a7 a3 Na ½ ð14Þ lim Na ½ ! KAsp M app ¼ a4 a1 ¼ kcat k3 ð13Þ ð13Þ lim Na ½ !0 KAsp M app ¼ a7 a3 Na ½ ð14Þ ð14Þ Equation (13) predicts that in the high concentration limit a constant value for KAsp M (app) is reached, which equals the ratio between two rate constants kcat/k3. The data presented in Fig. 2 and Table 2 show that the value of KAsp M (app) levels off to ~0.7 μΜ at high Na+ concentration. Since the turnover number kcat is ~0.9 s−1 (Fig. 3 and Table 2), a value of ~1.3 × 106 M−1 s−1 for k3 is found. COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio Results T d COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio 4 COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.n 4 ARTICLE COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y Table 3 vNa max (app), KNa M (app) and nHill (app) values for Na+ dependent uptakes obtained at constant [Asp]. [L-Asp] vNa max (app) (min−1) KNa M (app) (mM) nHill (app) 0.05 µM 3.7 ± 0.1 36.7 ± 1.8 2.1 0.19 0.1 µM 8.6 ± 0.1 37.0 ± 0.81 2.0 ± 0.08 0.5 µM 28.5 ± 0.15 40.9 ± 0.42 1.70 ± 0.025 1 µM 34.9 ± 1.0 37.3 ± 2.0 1.7 ± 0.13 5 µM 40.6 ± 1.3 24.7 ± 2.1 1.6 ± 0.19 10 µM 51.4 ± 2.7 21.8 ± 3.2 1.5 ± 0.29 50 µM 54.6 ± 1.9 10.9 ± 1.2 1.1 ± 0.14 100 µM 72.34 ± 10.5 14.8 ± 7.9 0.72 ± 0.21 Values are derived from the data presented in Table 1 and Fig. 1b. Each uptake rate represents the average of three independent biological replicates, each with two technical replicates. The standard error of the mean is indicated. Table 3 vNa max (app), KNa M (app) and nHill (app) values for Na+ dependent uptakes obtained at constant [Asp]. a6 ¼ k1k2k3kcat þ k1k2k4kcat þ k1k2k3k4 ð8Þ a7 ¼ k1k2k3kcat þ k1k2k3k4 ð9Þ ð9Þ Equation (2) can be rearranged to derive expressions for vNa max (app) and vAsp max (app): vAsp max app ¼ vmax* a1½Na2 a1½Na2 þ a2 Na ½ þ a3 ð10Þ vNa max app ¼ vmax* Asp a4 a1 þ Asp ¼ vmax* Asp kcat k3 þ Asp ð11Þ ð10Þ ð11Þ Thus, the model predicts that both vNa max (app) and vAsp max (app) are dependent on the concentration of the other substrate, which is fully consistent with the data presented in Fig. 3a, b. Moreover, Eq. (11) describes a rectangular hyperbolic relation between vNa max (app) and [Asp]. By ﬁtting the Michaelis-Menten equation to the data (Fig. 3a), we found a value for kcat/k3 of 0.7 μΜ. Since the turnover number kcat is known from the vmax data (~0.9 s−1 (~54 min−1), see Fig. 3, Table 2), a value of ~1.3 × 106 M−1 s−1 for k3 is derived, remarkably similar to the value of 1.2 × 106 M−1 s−1 that was found for GltPh, obtained in pre-equilibrium binding experiments36. Discussion Reconstitution of puriﬁed membrane proteins in liposomes often leads to mixed-orientation in lipid bilayers. For secondary active transporters, which can readily operate in both directions, the co- existence of right-side-out and inside-out oriented proteins is problematic for kinetic analysis. The work presented here shows that inactivation of transporter from one side of the membrane using a synthetic nanobody (sybody) is an effective way to deal with a mixed-orientation upon reconstitution in liposomes39,42. Synthetic nanobodies are membrane impermeable, and highly speciﬁc for the binding epitope. While also natural nanobodies could be used for this purpose, sybodies offer a major advantage, because the selection can be carried out under deﬁned buffer conditions, which may be used to steer the selection towards binders of a speciﬁc state. Also, the immobilization method used in the ribosome and phage display steps can be used to increase the chance of ﬁnding binders to the external or internal surface of the transporters. Finally, the generation of sybodies is quicker compared to nanobody generation, requires less protein, and does not require animal handling. To determine to which extent the kinetic mechanism depends on the orientation of GltTk in liposomes, we set out to inactivate either the right-side-out or the inside-out oriented transporters. Inhibition of the transporter from only one side of the membrane by modiﬁcation of cysteine mutants did not work for GltTk, possibly because of the one-gate nature of the elevator mechanism, in which the identical binding site residues and gating elements are alternately exposed to either side of the membrane13. Therefore, we chose to explore an alternative method by using synthetic nanobodies (sybodies)39,42. Since sybodies recognize water-exposed surface epitopes and are membrane-impermeable, they are expected to be suitable for orientation-speciﬁc inhibition. We selected 42 unique sybodies against GltTk, using an established platform, which included ribosome display, two rounds of phage display, and ELISA39. q g Here, the inactivation of the population of right-side out oriented GltTk by sybody binding made it possible to analyze uptake catalyzed by inside-out oriented proteins. The aspartate transport rates obtained in the presence of the sybody can be compared directly with a previous pre-equilibrium binding study on detergent-solubilized GltPh. Discussion In the latter study, the transporter was ﬁxed in the inward-oriented state by Hg2+-crosslinking of a double cysteine mutant, which allowed for the determination of the kinetic mechanism of binding of sodium ions and aspartate, and estimation of rate constants for association and dissociation36. In our study, where we used a sybody to inactivate the population of right-side-out oriented GltTk transporters in proteoliposomes, we measured the kinetics of the reversed transport step, which includes binding of Na+ and aspartate to the inward-oriented state similar to the pre-equilibrium binding study. In the Na+ concentration range between 5 and 300 mM, the two studies are fully congruent and consistent with a mechanism in which two sodium ions bind ﬁrst, followed by aspartate, and ﬁnally the last sodium ion. The kinetic analysis presented here shows that the rate constant for association of aspartate (k3 in Eq. (1)), can be obtained using Eq. (13) when KAsp M (App) is determined in the limit of high Na+ concentra- tions (as shown in Fig. 2) and the turnover number kcat is taken directly from the maximally attainable rate at high Na+ and aspartate concentrations (vmax). The value for k3 derived in this way was 1.3 × 106 M−1 s−1, which closely matches the value of 1.2 × 106 M−1 s−1 for GltPh derived from pre-equilibrium bind- ing experiments36. p y p g p y One of these sybodies (sybody 1) completely blocked aspartate transport by GltTk when added from both sides of the membrane (in the lumen and in the external solution), but inhibited partially when added only on the outside of the proteoliposomes (Fig. 4a). It is important to note that the procedure to load sybodies in the liposome lumen includes an extrusion step in which all GltTk molecules in the sample are exposed to the sybody. Therefore, it is not possible to do the opposite experiment, with the sybody exclusively included in the liposome lumen. Regardless of this limitation, the experiment conclusively showed that the sybody causes the sought-after sidedness of inhibition, and the result suggests that GltTk had reconstituted in both orientations in the proteoliposomes, similar to what was shown for GltPh before. To explain the inhibitory properties of sybody 1, we solved a structure of the sybody-GltTk complex using single-particle cryo- EM. The sybody binds on the extracellular surface of GltTk at the interface between the transport and scaffold domain. ARTICLE ARTICLE COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y Side speciﬁc inhibition with sybody. Although the analysis presented above is internally consistent, as well as consistent with existing kinetic data for binding on GltPh, there is a potential complication caused by the proteoliposome system used, because the reconstitution procedure usually results in a mix of inside- out- and right-side-out-oriented proteins in the bilayer. For instance, it has been demonstrated that GltPh reconstitutes in the two orientations with equal probability10. If the oppositely oriented proteins take up aspartate via different kinetic mechanisms (the equivalence of different mechanisms for for- ward and reverse transport in vivo), the results of the kinetic analysis could be convoluted and potentially lead to misinterpretation. parameters of the kinetic mechanism of the right-side-out and inside-out oriented proteins were very similar. Results T d Thus, two approaches (analysis of vNa max (app) and KAsp M (app)), reveal a value for k3 that agrees well with the value of 1.2 × 106 M−1 s−1 that was found for GltPh. To derive a rate equation for this kinetic model of Eq. (1), we used the King-Altman method38,41: is found. Thus, two approaches (analysis of vNa max (app) and KAsp M (app)), reveal a value for k3 that agrees well with the value of 1.2 × 106 M−1 s−1 that was found for GltPh. v0 vmax ¼ a1½Na3 Asp a1½Na3 Asp þ a2½Na2 Asp þ a3 Na ½ Asp þ a4½Na3 þ a5½Na2 þ a6 Na ½ þ a7 ð2Þ When the apparent afﬁnity constants are plotted in a double logarithmic plot against the concentration of Na+, linear relations are approached in both the high and low Na+ concentration extremes (Fig. 2). The slope is zero at high Na+ concentration (because KAsp M (app) levels off to kcat/k3), and a slope of −1.4 is found in the low Na+ concentration regime, which deviates from the slope of −1 predicted by the model. This discrepancy may indicate that lower concentrations of sodium should have been used to meet the conditions for Eq. (14) to be valid, something which was impossible for technical reasons, as discussed above. Alternatively, the deviation might be caused by the experimental error inherent to the transport measurements at low Na+ concentrations. The slope of the log(KM) plot in the low Na+ concentration regime is approximately twofold shallower than that of the log(KD) plot, which is consistent with the binding model36. ð2Þ in which vmax is the maximal attainable rate of transport at high Na+ and L-aspartate concentrations, and a1– a7 are expressions of rate constants: a1 ¼ k1k2k3k4 ð3Þ a2 ¼ k2k3k4kcat þ k1k3k4kcat þ k1k2k3kcat þ k1k2k3k4 ð4Þ a3 ¼ k1k3k4kcat ð5Þ a4¼ k1k2k4kcat ð6Þ a5 ¼ k1k2k4kcat þ k1k2k3kcat þ k1k2k3k4 ð7Þ COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio 5 5 Discussion Left: cartoon representation of the GltTk protomer (transport domain in blue and the scaffold in yellow) that has the sybody bound (red). The surface of the protein is shown in transparent representation, and the approximate location of the membrane boundaries is indicated with dashed lines. Right: sliced-through representation highlighting that the sybody likely blocks movement of the transport domain along with the scaffold domain. The transport domain is in the intermediate-outward state, as described in ref. 13 c, d Same as in Fig. 1a, b for a selection of conditions (using the same color coding), but now in the presence of 750 nM sybody on the outside of the liposomes. Table 5 Cryo-EM data collection, reﬁnement, and validation. Data collection and processing Voltage (kV) 200 Electron exposure (e−/Å2) 53.3 Defocus range (μm) −0.5 to −2.0 Pixel size (Å) 1.012 Symmetry imposed C1 Initial dataset (# of particles) 109217 Final dataset (# of particles) 53983 Map resolution (Å) FSC0.143 3.5 Reﬁnement Initial model used PDB 6XWQ Model composition Nonhydrogen atoms 10357 Protein residues 1378 Ligands 3 Mean B factors (Å2) Protein 160.2 Ligand 165.8 Rms deviations Bond lengths (Å) 0.005 Bond angles (°) 0.635 Validation MolProbity score 2.09 Clash score 17.7 Poor rotamers (%) 0.00 Ramachandran plot (%) Favored 95.07 Allowed 4.93 Outliers 0.00 Model to map ﬁt CC 0.86 For comparison, if kcat is derived directly from the vmax value of 3.4 nmol × mg−1 × min−1, a value of 2.6 × 10−3 s−1 is found43. The huge discrepancy between the two values is probably caused by inaccurate protein concentration determination in the proteoliposomes, or loss of activity during the reconstitution process. Table 5 Cryo-EM data collection, reﬁnement, and validation. Table 5 Cryo-EM data collection, reﬁnement, and validation. p It is noteworthy that in the experiments on GltTk presented here, the k3 value of 1.3 × 106 M−1 s−1 that we determined from the measure- ment of KAsp M (app) and kcat, was remarkably similar to the experi- mentally determined value of k3 for GltPh of 1.2 × 106 M−1 s−1 36. The virtually identical rates of aspartate association step in GltPh and GltTk are consistent with the structures of the binding sites in the two proteins, which are essentially the same7–9,13,15,16,20,44. Therefore, we believe that the GltTk protein concentration, used to derive kcat, was reasonably accurate in this case. Discussion The bound sybody thereby makes the elevator movement impossible, which prevents transport (Fig. 4b). The sybody thus inactivates the GltTk molecules with right-side-out orientation, and the residual uptake activity can be attributed to the inside-out oriented proteins. y We chose to use this sybody, added only on the outside of the proteoliposomes, to repeat a subset of the experiments described above. First, we tested whether vNa max (app) still depended on the aspartate concentration and whether vAsp max (app) still depended on the Na+ concentration when the right-side-out oriented mole- cules were inactivated by the sybody. Indeed, both vmax (app) values still varied with the concentrations of the co-substrate (Fig. 4c, d), consistent with the kinetic mechanism (Eqs. (10) and (11)). Second, we determined whether a constant value for KAsp M (app) was still reached in the high Na+ concentration limit, as predicted by Eq. (13). Indeed, a constant value of ~0.6 μΜ was found above 100 mM Na+, which intriguingly did not deviate signiﬁcantly from the value in the dual-population proteolipo- somes (Table 5). Assuming that the kcat value of the active population of inside-out oriented proteins was still ~0.9 s−1, the value of k3 remained unaltered. In other words, the measurable More importantly, using the same analysis, it is also possible to determine the turnover number kcat, if k3 is known from binding experiments (which is the case for GltPh). This notion is relevant, because quantiﬁcation of the amount of active transport protein present in membranes is often difﬁcult, making a direct deter- mination of kcat from vmax values notoriously error-prone, which can be easily illustrated by the available data on GltPh. For aspartate transport by GltPh, the value for KAsp M (app) at 100 mM Na+ has been determined accurately (120 nM)43. Assuming that this Na+ concentration is sufﬁciently high to represent the limit where KAsp M (app) has become constant and using the k3 value of 1.2 × 106 M−1 s−1 determined by pre-equilibrium bind- ing, from Eq. (13) a kcat value of 0.14 s−1 is calculated. This value has been derived without using the protein concentration. COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio Discussion Sybody Slice through Membrane Out In Sybody Scaffol f d domain Transportr domai d n b 0 50 100 150 0 10 20 30 Time (s) Asp uptake (mol Asp (mol GltTk)-1) No Sybody Sybody on outside Sybody on both sides a b a Sybody Sybody 0 100 200 300 0 5 10 15 [Na+] (mM) v0 (mol Asp (mol GltTk)-1 min-1) 0.05 M Asp 10 M Asp d 0 20 40 60 80 100 0 5 10 15 20 [Asp] (µM) v0 (mol Asp (mol GltTk)-1 min-1) 10 mM Na+ 200 mM Na+ c d c ransport by the sybody. a Uptake of aspartate in liposomes reconstituted with GltTk using aspartate and Na+ Fig. 4 Inhibition of aspartate transport by the sybody. a Uptake of aspartate in liposomes reconstituted with GltTk using aspartate and Na+ concentrations of 1 μM and 300 mM respectively. Uptake traces in the absence of sybody (green), in the presence of 750 nM sybody on the outside (orange), or with sybody present on both sides of the membrane (red). Each data point represents a triplicate measurement (n = 3), and the standard error of the mean is shown. b Cryo-EM structure of the GltTk-sybody complex. Left: cartoon representation of the GltTk protomer (transport domain in blue and the scaffold in yellow) that has the sybody bound (red). The surface of the protein is shown in transparent representation, and the approximate location of the membrane boundaries is indicated with dashed lines. Right: sliced-through representation highlighting that the sybody likely blocks movement of the transport domain along with the scaffold domain. The transport domain is in the intermediate-outward state, as described in ref. 13 c, d Same as in Fig. 1a, b for a selection of conditions (using the same color coding), but now in the presence of 750 nM sybody on the outside of the liposomes. Fig. 4 Inhibition of aspartate transport by the sybody. a Uptake of aspartate in liposomes reconstituted with GltTk using aspartate and Na+ concentrations of 1 μM and 300 mM respectively. Uptake traces in the absence of sybody (green), in the presence of 750 nM sybody on the outside (orange), or with sybody present on both sides of the membrane (red). Each data point represents a triplicate measurement (n = 3), and the standard error of the mean is shown. b Cryo-EM structure of the GltTk-sybody complex. Discussion COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio 6 6 ARTICLE COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y 0 50 100 150 0 10 20 30 Time (s) Asp uptake (mol Asp (mol GltTk)-1) No Sybody Sybody on outside Sybody on both sides a Sybody Slice through Membrane Out In Sybody Scaffol f d domain Transportr domai d n b 0 20 40 60 80 100 0 5 10 15 20 [Asp] (µM) v0 (mol Asp (mol GltTk)-1 min-1) 10 mM Na+ 200 mM Na+ 0 100 200 300 0 5 10 15 [Na+] (mM) v0 (mol Asp (mol GltTk)-1 min-1) 0.05 M Asp 10 M Asp c d Fig. 4 Inhibition of aspartate transport by the sybody. a Uptake of aspartate in liposomes reconstituted with GltTk using aspartate and Na+ concentrations of 1 μM and 300 mM respectively. Uptake traces in the absence of sybody (green), in the presence of 750 nM sybody on the outside (orange), or with sybody present on both sides of the membrane (red). Each data point represents a triplicate measurement (n = 3), and the standard error of the mean is shown. b Cryo-EM structure of the GltTk-sybody complex. Left: cartoon representation of the GltTk protomer (transport domain in blue and the scaffold in yellow) that has the sybody bound (red). The surface of the protein is shown in transparent representation, and the approximate location of the membrane boundaries is indicated with dashed lines. Right: sliced-through representation highlighting that the sybody likely blocks movement of the transport domain along with the scaffold domain. The transport domain is in the intermediate-outward state, as described in ref. 13 c, d Same as in Fig. 1a, b for a selection of conditions (using the same color coding), but now in the presence of 750 nM sybody on the outside of the liposomes. COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y mechanistic differences between the mammalian and archaeal transporters reﬂect evolutionary pressure to support different physiological needs is not clear. Alternatively, it is possible that the apparent differences in kinetic mechanism between GltTk and EAAC1 are caused by differences in the readout of transport. While transport was measured directly (using radioactivity) for GltTk, for EAAC1 the substrate-induced chloride conductance was used as a readout. In addition, different ranges of co-ion concentrations were used in the study of EAAC1 compared to the work presented here, which also may make the two studies not directly comparable. of the empty binding site in the apo-state, movement of the transport domain in the occluded apo-state, the opening of the binding site externally. The latter step also includes the rate constant for reshaping the binding sites for the ﬁrst two sodium ions, leading to conformational selection that was shown to occur in GltPh34. Therefore, differences between GltPh and GltTk that affect any of these steps may affect kcat, which is observable in the value for KAsp M (app) at high sodium concentrations. M pp g The origin of the differences between GltPh and GltTk may be similar to that of differences between wild-type GltPh and a faster “unlocked” mutant18,43. This mutant has the same value for k3 as the wild-type36, consistent with the observation that the structure of the binding site is identical to the wild-type. Therefore, determination of KAsp M (app) in the limit of high sodium ion concentration allows for the derivation of accurate values for kcat. The values for KAsp M (app) have been determined at 100 mM Na+ for both wild-type GltPh and the fast mutant (120 nM and 406 nM respectively)43. If we again assume that this Na+ concentration is high enough to represent the limit where KAsp M (app) becomes constant for both proteins, then kcat is predicted to be ~3.5 times higher in the fast mutant, which is in good agreement with bulk transport data that showed ~4.5 fold difference in vmax values (although the absolute numbers are both far off for the reasons discussed above)43. y p In conclusion, analysis of the kinetic mechanism of sodium- coupled aspartate transport by GltTk and GltPh provides a way to determine accurate turnover numbers from KAsp M (app) values, without the need to use error-prone protein quantiﬁcation. Methods Glt iﬁ GltTk puriﬁcation and reconstitution in proteoliposomes. GltTk was produced in Escherichia coli strain MC1061 with the L-arabinose inducible vector pBad24 as described in Arkhipova et al. 14. The cells were grown in LB media supplemented with 100 mg/L ampicillin. The expression was induced by the addition of 0.05% L- arabinose when the culture reached 0.8 OD600. Three hours after induction the cells were harvested by centrifugation (7000 RPM, 15′, 4 °C Beckman JLA 9.1000) and resuspended in ice-cold 20 mM Tris-HCl pH 8. The cells were lysed by means of a cell disruptor cooled to 4 °C and operated at 25 PSI. The lysate went through an intermediate centrifugation (7500 RPM, 20′, 4 °C, Beckman JA25.50) step to remove cell debris, the supernatant was ﬁnally ultracentrifuged (40000 RPM 150′, 4 °C, Beckman 50.1 Ti) and the pellet was resuspended in 20 mM Tris-HCl pH 8 before storing the membrane vesicles at −80 °C. While the kinetic mechanism presented here is valid only for the reverse transport reaction (because we inactivated the forward-operating transporters by the sybody), it is well possible that the same mechanism is also used for the forward transport reaction. The similarity of the forward and reverse kinetic mechanisms is indeed supported by the essentially identical kinetic parameters in the presence and absence of the sybody. From a structural point of view, this similarity can be explained because the transporter uses the same gating element on both sides of the membrane (“one-gate elevator”), and the binding site geometry and access path in the inward- and outward-oriented states are essentially the same13,20,46,47. In future experiments, the kinetic mechanism for forwarding transport by GltTk may be tested if we manage to inactivate the inside-out oriented popu- lation of GltTk molecules by a suitable sybody. g The membrane vesicles were then added to solubilization buffer (50 mM Tris- HCl pH8, 300 mM KCl, 1% DDM), incubated for 45′ on a rocking platform at 4 °C, and ﬁnally centrifuged (55,000 RPM, 30′, 4 °C, Beckman MLA 55) to separate the insoluble fraction from the solubilized protein. The supernatant was supplemented with 15 mM imidazole pH 8 and with 0.5 mL of Ni-Sepharose slurry pre-equilibrated with 50 mM Tris-HCl pH 8, 300 mM KCl. After 1 h of incubation the mixture was loaded onto a Poly-Prep column and unbound protein was allowed to ﬂow through. COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y This is of great use for instance when analyzing the effects of muta- tions. Depending on the details of the kinetic mechanism, it may also be possible to determine turnover numbers in a similar way for other transport proteins. To test whether kinetic mechanisms different from the one analyzed here (Eq. (1)) would also allow for a similar determination of the turnover number, we used the King-Altman method38,41 to derive rate equations for all possible kinetic mechanisms leading to the coupled transport of three sodium ions and aspartate40. It turns out that if at least one sodium ion binds after aspartate, KAsp M (app) values in the limit of high sodium concentration equal the ratio between kcat and the on-rate for aspartate binding (as in Eq. (13)). It is also noteworthy that this even holds if multiple sodium ions bind randomly (for instance if steps 1 and 2 in Eq. (1) would be random). Therefore, it is likely that more transporters can be analyzed in the same way as presented here, and we conclude that systematic analysis of transport rates, and derivation of the rate equation, are essential steps in the elucidation of transport mechanisms. It is also possible to determine turnover numbers using a recently established transport assay at the single-molecule level45, but combining turnover numbers from single-molecule transport measurements with rate constants determined in pre-equilibrium binding experiments is not straightforward. In pre-equilibrium binding experiments33,34,36,37 and bulk transport experiments (presented here), the entire ensemble contributes to the measured rate constants, and therefore data can be combined to describe the ensemble properties. In contrast, in single-molecule transport experiments, the kcat is determined for only a fraction of the ensemble, and therefore it is difﬁcult to quantitatively extract ensemble behavior, which would require accounting for the contribution of all subpopulations of (slightly) differently behaving individual proteins that together make up the ensemble45. Conversely, ensemble measurements also cannot predict the single-molecule behavior, but the observation that the kcat value determined by the ensemble measurements is lower than the maximal kcat value determined in single-molecule measurements is consistent with the heterogeneity at the single- molecule level. ARTICLE ARTICLE Discussion The sixfold difference in turnover number kcat between GltPh and GltTk (0.14 s−1 and 0.9 s−1 respectively) could be caused by (small) structural differences away from the binding site. kcat is not a single rate constant but is composed of contributions from all steps that take place after binding of the last sodium ion in the catalytic cycle until the binding of the ﬁrst sodium ion in the next round of catalysis. These steps include movement of the fully- loaded transport domain between outward- and inward-oriented states, the opening of the binding site towards the lumen of the liposomes, the release of the sodium ions and aspartate, occlusion 7 7 COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y suspension on the side of a test tube containing 200 μL of reaction buffer and subsequently ﬂushing them in the reaction buffer with 2 mL of ice-cold quenching buffer (10 mM potassium phosphate (pH 7), 300 mM KCl). BioBeads in three steps: First 15 mg/mL BioBeads were added followed by incubation for 60′ at 4 °C, then 19 mg/mL BioBeads were added followed by overnight incubation at 4 °C. Finally, 29 mg/mL BioBeads were added followed by 120′ incubation at 4 °C. BioBeads were then removed and the proteoliposomes were pelleted (80,000 RPM, 25′, 4 °C, Beckman MLA80) and resuspended in 50 mM potassium phosphate buffer (pH 7) to a ﬁnal lipid concentration of 20 mg/ml. The proteoliposomes were subjected to three cycles of freeze-thawing using liquid nitrogen and stored until use. p p p p The value for each uptake rate represents the average and standard error of three independent biological replicates (different batches of expressed, puriﬁed, and reconstituted protein), each constituted by two or three technical replicates. The substrate-dependent uptake rates obtained at a ﬁxed concentration of Na+ were plotted as a function of L-aspartate, and the Michaelis-Menten equation was ﬁtted to the data to obtain apparent KM (KAsp M (app)) and vmax (vAsp max (app)) values for different [Na+]. The co-ion-dependent uptake rates obtained at a ﬁxed concentration of L-aspartate were plotted as a function of Na+, and the Hill equation was ﬁtted to the data for obtaining KM (KAsp D (app)), vmax (vNa max (app)), and nHill values for different [L-Asp]. The statistical analysis of the data was executed in GraphPad Prism 9. Sybody selection. Sybodies were selected against two GltTk cysteine mutants (298 C and 367 C), which while biotinylated and immobilized during the selection procedures would make extracellular and intracellular epitopes accessible for binding, respectively. Selection was done in the presence of 50 µM L-aspartate, 150 mM NaCl, and 0.15% DM according to an established in vitro selection platform that included ribosome display, two rounds of phage display, and ELISA (Zim- mermann et al.39,42). During ELISA, every single clone was analyzed for binding against GltTk in the presence and absence of L-aspartate. Sequencing of 48 ELISA positive hits resulted in 42 unique sybody sequences (20 for the 298 C mutant and 22 for the C367 mutant). Single particle Cryo-EM. COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y The structure of GltTk in complex with sybody 1 (molar ratio 3:1) was determined using essentially the same protocol as described in Arkhipova et al.13, in the presence of 300 mM Na+ and 50 μM L-Asp. The puriﬁed complex at the concentration of 0.5–1 mg/ml was applied onto freshly glow- discharged Quantifoil grids (Au R1.2/1.3, 300 mesh) at 22 °C and 100% humidity and plunged-frozen in liquid ethane. The Cryo-EM data were collected using 200- keV Talos Arctica microscope (Thermo Fisher). Cryo-EM image processing was performed using cryoSPARC software49. Sybody expression and puriﬁcation. Each of 42 sybodies was expressed in E. coli, puriﬁed from the periplasm using Ni2+-afﬁnity chromatography, and analyzed by size exclusion chromatography (SEC). Based on the quality of the SEC proﬁles (absence of aggregates, no interactions with column material, high yield) we selected 33 puriﬁed sybodies (14 for the 298 C mutant and 19 for the 367 C mutant), which were tested for their ability to inhibit GltTk transport of aspartate in uptake assays. For large scale puriﬁcation of inhibitory sybody 1, a preculture of E. coli MC1061 transformed with pSB_initSB1 was used to inoculate 50 mL of TB medium supplemented with 25 µg/ml chloramphenicol. The culture was grown for 2 h at 37 °C while shaking at 200 rpm, the temperature was then lowered to 22 °C and let grow until OD ~0.8. The In brief, 824 micrographs were selected for the processing after motion correction and CTF estimation. The template for particle picking was generated from 100 manually picked particles. Template-based picking identiﬁed 109,217 particles. Subsequent 2D classiﬁcation reduced the number of particles to 67,498 and subsequently 53,983 particles were left in the selected ab initio class. Final non- uniform 3D reﬁnement resulted in a 3.5 Å map (with C1 symmetry applied), which was sharpened using Autosharpen Map procedure in Phenix50 and used for model building using Coot51. The reﬁnement of the coordinates was performed in the realspace reﬁne module of Phenix52. The data collection and reﬁnement statistics are shown in Table 5. Visualization and structure interpretation was carried out in UCSF Chimera53 and PyMol (Schrödinger, LLC). expression was induced by adding L-arabinose to a ﬁnal concentration of 0.02% and let express overnight at 22 °C while shaking. Data availability Data supporting the ﬁndings of this manuscript are available from the corresponding authors upon reasonable request. A reporting summary for this Article is available as a Supplementary Information ﬁle. The source data underlying Figs. 1–4 and Tables 1–4 are provided as a Supplementary Data 1. The three-dimensional cryo-EM density map of the glutamate transporter homolog GltTk bound to the sybody has been deposited in the Electron Microscopy Data Bank under accession number EMD-12314 (https://www.ebi. ac.uk/pdbe/emdb/). Coordinates of the corresponding ﬁve models have been deposited in the Protein Data Bank under the accession number 7NGH (https://www.rcsb.org/). Transport assay. The lumenal buffer in each proteoliposome preparation was changed to 10 mM potassium phosphate buffer (pH 7) and 300 mM KCl. For this, the proteoliposomes were ﬁrst pelleted (80,000 RPM, 25′, 4 °C, Beckman MLA80) and then resuspended in the lumenal buffer. After three freeze-thaw cycles, the suspension was extruded 11 times through a polycarbonate ﬁlter with 400-nm pore size in order to obtain homogeneously sized unilamellar vesicles which were pel- leted (80,000 RPM, 25′, 4 °C, Beckman TLA100.3) and resuspended to a ﬁnal lipid concentration of 100 mg/mL. To completely inhibit aspartate transport of GltTk by sybody 1 (on both sides of the membrane) the lumenal solution was supplemented with 150 µM of sybody 1. After performing 6 freeze-thaw cycles, and extrusion (11 times) through a To completely inhibit aspartate transport of GltTk by sybody 1 (on both sides of the membrane) the lumenal solution was supplemented with 150 µM of sybody 1. After performing 6 freeze-thaw cycles, and extrusion (11 times) through a polycarbonate ﬁlter with 400-nm pore size, the homogeneous solution was pelleted (80,000 RPM, 20′, 4 °C, Beckman TLA100.1) and resuspended to a ﬁnal lipid concentration of 100 mg/mL in a solution containing 75 µM of sybody 1, 10 mM potassium phosphate buffer (pH 7) and 300 mM KCl. Received: 23 February 2021; Accepted: 26 May 2021; Received: 23 February 2021; Accepted: 26 May 2021; polycarbonate ﬁlter with 400-nm pore size, the homogeneous solution was pelleted (80,000 RPM, 20′, 4 °C, Beckman TLA100.1) and resuspended to a ﬁnal lipid concentration of 100 mg/mL in a solution containing 75 µM of sybody 1, 10 mM potassium phosphate buffer (pH 7) and 300 mM KCl. References 1. Danbolt, N. C. Glutamate uptake. Prog. Neurobiol. 65, 1–105 (2001). 2 K B I & Sh I A i f L l b b 2. Kanner, B. I. & Sharon, I. Active transport of L-glutamate by membrane vesicles isolated from rat brain. Biochemistry https://doi.org/10.1021/ bi00612a011 (1978). 2. Kanner, B. I. & Sharon, I. Active transport of L-glutamate by membrane vesicles isolated from rat brain. Biochemistry https://doi.org/10.1021/ bi00612a011 (1978). To start the transport the proteoliposome suspension was diluted 100 fold into external buffer while stirring. The external buffer contained 10 mM potassium phosphate buffer (pH 7), 3 μM valinomycin, 1–300 mM NaCl, 0.05–100 μM L- aspartate; to balance the osmotic strength with that of the lumenal solution, choline chloride was added (Table 1). In case the sybody was present, the 100-fold dilution resulted in a ﬁnal external concentration of sybody of 750 nM. 3. Freidman, N. et al. Amino acid transporters and exchangers from the SLC1A family: structure, mechanism and roles in physiology and cancer. Neurochem. Res. https://doi.org/10.1007/s11064-019-02934-x (2020). 4. Jensen, A. A., Fahlke, C., Bjørn-Yoshimoto, W. E. & Bunch, L. Excitatory amino acid transporters: Recent insights into molecular mechanisms, novel modes of modulation and new therapeutic possibilities. Curr. Opin. Pharmacol. 20, 116–123 (2015). After the indicated incubation period (20 s for the data in Figs. 1 and 4c, d), 2 mL of ice-cold quenching buffer (10 mM potassium phosphate (pH 7), 300 mM KCl) was added. The content of the tube was then poured onto a BA 45 nitrocellulose ﬁlter which was then washed with 2 mL of quenching buffer. The ﬁlters were ﬁnally dissolved in scintillation cocktail Ultima Gold (Perkin Elmer) and the β-decay from the radiolabeled substrate was counted. The time-point zero measurements for each condition was measured by pipetting the liposome After the indicated incubation period (20 s for the data in Figs. 1 and 4c, d), 2 mL of ice-cold quenching buffer (10 mM potassium phosphate (pH 7), 300 mM KCl) was added. The content of the tube was then poured onto a BA 45 nitrocellulose ﬁlter which was then washed with 2 mL of quenching buffer. The ﬁlters were ﬁnally dissolved in scintillation cocktail Ultima Gold (Perkin Elmer) and the β-decay from the radiolabeled substrate was counted. The time-point zero measurements for each condition was measured by pipetting the liposome 5. Zerangue, N. & Kavanaugh, M. P. Flux coupling in a neuronal glutamate transporter. COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y Cells were pelleted and resuspended in 5 mL periplasmic extraction buffer (20% sucrose (w/v) 50 mM Tris-HCl (pH 8.0), 0.5 mM EDTA and 0.5 µg/ml lysozyme) and incubated on ice for 30 min, after incubation 20 mL of TBS (20 mM Tris-HCl (pH 7.4) 150 mM NaCl) supplemented with 1 mM MgCl2 were added. The lysate was centrifuged at 4000 g for 20 min and the super- natant was transferred in a tube containing 500 μL of Ni-sepharose pre-equilibrated in TBS and supplemented with imidazole to a ﬁnal concentration of 15 mM. After 1 h incubation on a shaking platform, the solution containing the SyBody was applied to a polyprep gravity column and the unbound fraction was let ﬂow through. The resin was then washed with 10 CV of TBS supplemented with 30 mM imidazole and eluted in 500 µl of TBS supplemented with 300 mM imidazole. The sybody solution was then passed through a NAP-10 column equilibrated with the internal uptake buffer (10 mM potassium phosphate buffer (pH 7), 300 mM KCl) and stored at −80 °C. sybody 1 was selected from the library created by the mutant 298 C. Statistics and reproducibility. Each uptake rate represents the average of three independent biological replicates (separate puriﬁcations and reconstitutions), each constituted by two technical replicates, and the standard error of the mean is shown in the ﬁgures and tables. Reporting summary. Further information on research design is available in the Nature Research Reporting Summary linked to this article. Data availability To inhibit only the right-side-out fraction of GltTk a homogeneous solution of proteoliposomes was prepared as described above (with lumenal buffer devoid of sybody) and after pelleting by ultracentrifugation it was resuspended to a ﬁnal lipid concentration of 100 mg/mL with a solution containing 75 µM of sybody 1, 10 mM potassium phosphate buffer (pH 7) and 300 mM KCl. Methods Glt iﬁ The column was washed with 20 column volumes of washing buffer (50 mM Tris-HCl, 300 mM KCl, 60 mM imidazole, 0.15% DM) and ﬁnally eluted in three fractions of 300, 800, and 400 μL respectively using elution buffer (50 mM Tris-HCl, 300 mM KCl, 500 mM imidazole, 0.15% DM). The second fraction was loaded onto a Superdex-200I gel ﬁltration column equilibrated with 10 mM HEPES pH 8, 100 mM KCl and 0.15%DM. The ﬁnal concentration of the puriﬁed protein was determined by measuring the absorbance at 280 nm (GltTk ε = 37,360). The lipids used to reconstitute GltTk contained a 3:1 mixture of E. coli lipid polar extract and egg phosphatidylcholine (PC) (Avanti). Liposomes were homogenized by extruding 11 times through a 400 nm pore size polycarbonate ﬁlter and subsequently diluted to 5 mg/mL in 50 mM potassium phosphate buffer (pH 7.0). To allow the insertion of the protein into the bilayer, the lipids were destabilized by step-wise addition of 10% Triton X-100 while scattering was followed at a wavelength of 540 nm. The titration was stopped once the absorption signal decreased to about 60% of the maximum value reached. Puriﬁed protein was added at a protein:lipid ratio (w/w) of 1:1600. The protein-lipid mixture was incubated for 30′ at RT, and then the detergent was removed by addition of In contrast to what we found for GltTk, differences in the kinetic mechanisms of the forward and reverse reactions have been reported for the mammalian glutamate transporter EAAC148. The analysis of EAAC1 transport was based on the rapid-equilibrium assumption, which may be valid in this case, although it has not been tested systematically. Whether potential COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio 8 COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio ARTICLE COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y ARTICLE COMMUNICATIONS BIOLOGY \| https://doi.org/10.1038/s42003-021-02267-y 7. Yernool, D., Boudker, O., Jin, Y. & Gouaux, E. Structure of a glutamate transporter homologue from Pyrococcus horikoshii. Nature 431, 811–818 (2004). 33. Ewers, D., Becher, T., Machtens, J. P., Weyand, I. & Fahlke, C. Induced ﬁt substrate binding to an archeal glutamate transporter homologue. Proc. Natl Acad. Sci. USA 110, 12486–12491 (2013). 34. Alleva, C. et al. Na+-dependent gate dynamics and electrostatic attraction ensure substrate coupling in glutamate transporters. Sci. Adv. 6, eaba9854 (2020). 8. Jensen, S., Guskov, A., Rempel, S., Hänelt, I. & Slotboom, D. J. Crystal structure of a substrate-free aspartate transporter. Nat. Struct. Mol. Biol. 20, 1224–1227 (2013). g g 35. Reyes, N., Oh, S. & Boudker, O. Binding thermodynamics of a glutamate transporter homolog. Nat. Struct. Mol. Biol. 20, 634–640 (2013). 9. Boudker, O., Ryan, R. M., Yernool, D., Shimamoto, K. & Gouaux, E. Coupling substrate and ion binding to extracellular gate of a sodium-dependent aspartate transporter. Nature 445, 387–393 (2007). 36. Oh, S. & Boudker, O. Kinetic mechanism of coupled binding in sodium- aspartate symporter GltPh. Elife 7, 1–20 (2018). p y p f 37. Hänelt, I., Jensen, S., Wunnicke, D. & Slotboom, D. J. Low afﬁnity and slow Na+ binding precedes high afﬁnity aspartate binding in the secondary-active transporter GltPh. J. Biol. Chem. 290, 15962–15972 (2015). p p 10. Ryan, R. M., Compton, E. L. R. & Mindell, J. A. Functional characterization of a Na+-dependent aspartate transporter from Pyrococcus horikoshii. J. Biol. Chem. 284, 17540–17548 (2009). p 38. Segel, I. H. Enzyme kinetics. in Encyclopedia of Biological Chemistry: Second Edition https://doi.org/10.1016/B978-0-12-378630-2.00012-8 (2013). 11. Kortzak, D. et al. Allosteric gate modulation confers K+ coupling in glutamate transporters. EMBO J. 38, e101468 (2019). 12. Arkhipova, V. et al. Structural aspects of photopharmacology: insight into the binding of photoswitchable and photocaged inhibitors to the glutamate transporter homologue. J. Am. Chem. Soc. https://doi.org/10.1021/jacs.0c11336 (2021). 39. Zimmermann, I. et al. Synthetic single domain antibodies for the conformational trapping of membrane proteins. Elife https://doi.org/10.7554/ eLife.34317 (2018). 40. Lolkema, J. S. & Slotboom, D. J. Models to determine the kinetic mechanisms of ioncoupled transporters. J. Gen. Physiol. 151, 369–380 (2019). 13. Arkhipova, V., Guskov, A. & Slotboom, D. J. Structural ensemble of a glutamate transporter homologue in lipid nanodisc environment. Nat. Commun. 11, 998 (2020). 41. King, E. L. & Altman, C. A schematic method of deriving the rate laws for enzyme-catalyzed reactions. J. Phys. ARTICLE Chem. https://doi.org/10.1021/j150544a010 (1956). 14. Arkhipova, V. et al. Binding and transport of D-aspartate by the glutamate transporter homolog GltTk. Elife 8, 1–12 (2019). 42. Zimmermann, I. et al. Generation of synthetic nanobodies against delicate proteins. Nat. Protoc. https://doi.org/10.1038/s41596-020-0304-x (2020). 15. Guskov, A., Jensen, S., Faustino, I., Marrink, S. J. & Slotboom, D. J. Coupled binding mechanism of three sodium ions and aspartate in the glutamate transporter homologue GltTk. Nat. Commun. 7, 1–6 (2016). 43. Ryan, R. M., Kortt, N. C., Sirivanta, T. & Vandenberg, R. J. The position of an arginine residue inﬂuences substrate afﬁnity and K+ coupling in the human glutamate transporter, EAAT1. J. Neurochem. https://doi.org/10.1111/j.1471- 4159.2010.06796.x (2010). p g 16. Reyes, N., Ginter, C. & Boudker, O. Transport mechanism of a bacterial homologue of glutamate transporters. Nature 462, 880–885 (2009). 44. Verdon, G., Oh, S. C., Serio, R. & Boudker, O. Coupled ion binding and structural transitions along the transport cycle of glutamate transporters. Elife 2014, 1–23 (2014). 17. Georgieva, E. R., Borbat, P. P., Ginter, C., Freed, J. H. & Boudker, O. Conformational ensemble of the sodium-coupled aspartate transporter. Nat. Struct. Mol. Biol. https://doi.org/10.1038/nsmb.2494 (2013). 18. Akyuz, N. et al. Transport domain unlocking sets the uptake rate of an aspartate transporter. Nature https://doi.org/10.1038/nature14158 (2015). 45. Ciftci, D. et al. Single-molecule transport kinetics of a glutamate transporter homolog shows static disorder. Sci. Adv. https://doi.org/10.1126/sciadv. aaz1949 (2020). 19. Ruan, Y. et al. Direct visualization of glutamate transporter elevator mechanism by high-speed AFM. Proc. Natl Acad. Sci. USA 114, 1584–1588 (2017). 46. Garaeva, A. A., Guskov, A., Slotboom, D. J. & Paulino, C. A one-gate elevator mechanism for the human neutral amino acid transporter ASCT2. Nat. Commun. https://doi.org/10.1038/s41467-019-11363-x (2019). 20. Wang, X. & Boudker, O. Large domain movements through the lipid bilayer mediate substrate release and inhibition of glutamate transporters. Elife https://doi.org/10.7554/eLife.58417 (2020). 47. Garaeva, A. A. & Slotboom, D. J. Elevator-type mechanisms of membrane transport. Biochem. Soc. Trans. 48, 1227–1241 (2020). p g 21. Erkens, G. B., Hänelt, I., Goudsmits, J. M. H. H., Slotboom, D. J. & Van Oijen, A. M. Unsynchronised subunit motion in single trimeric sodium-coupled aspartate transporters. Nature 502, 119–123 (2013). 48. Zhang, Z. et al. Transport direction determines the kinetics of substrate transport by the glutamate transporter EAAC1. Proc. Natl Acad. Sci. USA 104, 18025–18030 (2007). p p 22. Hänelt, I., Wunnicke, D., Bordignon, E., Steinhoff, H. J. & Slotboom, D. J. ARTICLE Conformational heterogeneity of the aspartate transporter GltPh. Nat. Struct. Mol. Biol. 20, 210–214 (2013). 49. Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. CryoSPARC: Algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods https://doi.org/10.1038/nmeth.4169 (2017). 23. Tanui, R., Tao, Z., Silverstein, N., Kanner, B. & Grewer, C. Electrogenic steps associated with substrate binding to the neuronal glutamate transporter EAAC1. J. Biol. Chem. 291, 11852–11864 (2016). 50. Terwilliger, T. C., Sobolev, O. V., Afonine, P. V. & Adams, P. D. Automated map sharpening by maximization of detail and connectivity. Acta Crystallogr. Sect. D Struct. Biol. https://doi.org/10.1107/S2059798318004655 (2018). 24. Grewer, C. et al. Individual subunits of the glutamate transporter EAAC1 homotrimer function independently of each other. Biochemistry https://doi. org/10.1021/bi050987n (2005). p g 51. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. Sect. D Biol. Crystallogr. https://doi.org/10.1107/ S0907444910007493 (2010). g 25. Leary, G. P., Stone, E. F., Holley, D. C. & Kavanaugh, M. P. The glutamate and chloride permeation pathways are colocalized in individual neuronal glutamate transporter subunits. J. Neurosci. https://doi.org/10.1523/ JNEUROSCI.4851-06.2007 (2007). 52. Afonine, P. V. et al. Real-space reﬁnement in PHENIX for cryo-EM and crystallography. Acta Crystallogr. Sect. D Struct. Biol. https://doi.org/10.1107/ S2059798318006551 (2018). 53. Pettersen, E. F. et al. UCSF Chimera—a visualization system for exploratory research and analysis. J. Comput. Chem. https://doi.org/10.1002/jcc.20084 (2004). 26. Koch, H. P., Brown, R. L. & Larsson, H. P. The glutamate-activated anion conductance in excitatory amino acid transporters is gated independently by the individual subunits. J. Neurosci. https://doi.org/10.1523/ JNEUROSCI.0118-07.2007 (2007). 27. Akyuz, N., Altman, R. B., Blanchard, S. C. & Boudker, O. Transport dynamics in a glutamate transporter homologue. Nature 502, 114–118 (2013). Acknowledgements We thank Jan Rheinberger for help with cryo-EM sample preparation and data collec- tion. This work was supported by the Dutch Research Council (NWO TOP grant 714 018 003 to DJS) and EMBO (short term fellowship to AAG) We thank Jan Rheinberger for help with cryo-EM sample preparation and data collec- tion. This work was supported by the Dutch Research Council (NWO TOP grant 714 018 003 t DJS) d EMBO ( h t t f ll hi t AAG) 28. Stolzenberg, S., Khelashvili, G. & Weinstein, H. Structural intermediates in a model of the substrate translocation path of the bacterial glutamate transporter homologue GltPh. J. Phys. Chem. B 116, 5372–5383 (2012). pp y 714.018.003 to DJS) and EMBO (short-term fellowship to AAG). 29. Jiang, J., Shrivastava, I. H., Watts, S. D., Bahar, I. & Amara, S. G. Large collective motions regulate the functional properties of glutamate transporter trimers. Proc. Natl. Acad. Sci. USA https://doi.org/10.1073/pnas.1112216108 (2011). References Nature 383, 634–637 (1996). 6. Owe, S. G., Marcaggi, P. & Attwell, D. The ionic stoichiometry of the GLAST glutamate transporter in salamander retinal glia. J. Physiol. 577, 591–599 (2006). 6. Owe, S. G., Marcaggi, P. & Attwell, D. The ionic stoichiometry of the GLAST glutamate transporter in salamander retinal glia. J. Physiol. 577, 591–599 (2006). 9 Author contributions G.T. performed all experiments with the exception of the sybody selection (done by A.A. G. with help and supervision from C.A.J.H. and M.A.S.) and the cryo-EM sample pre- paration (done by V.A.) and structure determination (V.A. and A.G.). All authors designed experiments and analyzed data. D.J.S. and G.T. wrote the manuscript with input from all other authors. 30. Groeneveld, M. & Slotboom, D. J. Rigidity of the subunit interfaces of the trimeric glutamate transporter GltT during translocation. J. Mol. Biol. 372, 565–570 (2007). 31. Verdon, G. & Boudker, O. Crystal structure of an asymmetric trimer of a bacterial glutamate transporter homolog. Nat. Struct. Mol. Biol. https://doi. org/10.1038/nsmb.2233 (2012). Competing interests g 32. Groeneveld, M. & Slotboom, D. J. Na+:aspartate coupling stoichiometry in the glutamate transporter homologue GltPh. Biochemistry 49, 3511–3513 (2010). Competing interests The authors declare no competing interests. p g The authors declare no competing interests 10 COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio Additional information Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Supplementary information The online version contains supplementary material available at https://doi.org/10.1038/s42003-021-02267-y. Correspondence and requests for materials should be addressed to D Peer review information Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Anam Akhtar. Peer reviewer reports are available. Reprints and permission information is available at http://www.nature.com/reprints © The Author(s) 2021 Reprints and permission information is available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. © The Author(s) 2021 11 11 COMMUNICATIONS BIOLOGY \| (2021) 4:751 \| https://doi.org/10.1038/s42003-021-02267-y \| www.nature.com/commsbio
https://openalex.org/W2935405257	https://vestnik.rsmu.press/files/issues/vestnik.rsmu.press/2019/1/2019-1-11_en.pdf?lang=en	Latin	null	Injectable collagen in correction of age-related skin changes: experimental and clinical parallels	Bulletin of Russian State Medical University/Bulletin of RSMU	2,019	cc-by	6,882	Manturova NE1, Stenko AG2, Petinati YaA3, Chaikovskaya EA2 , Bolgarina AA4 Author contribution: Manturova NE — research planning and data interpretation; Stenko AG, Petinati YaA — selection of participants, conducting research, data interpretation; Chaikovskaya EA — research planning, data collection and interpretation, manuscript preparation; Bolgarina AA — literature analysis, research planning. Compliance with ethical standards: the study was approved by the Society of Aesthetic Medicine ethical committee (Protocol №2 of September 07, 2017). Registration number at www.ClinicalTrials.gov: NCT03677258. Correspondence should be addressed: Ekaterina A. Chaikovskaya Olkhovskaya, 27, Moscow, 105066; ktchaikovskaya@yandex.ru Received: 27.09.2018 Accepted: 25.02.2019 Published online: 09.03.2019 Received: 27.09.2018 Accepted: 25.02.2019 Published online: 09.03.2019 DOI: 10.24075/brsmu.2019.010 Manturova NE1, Stenko AG2, Petinati YaA3, Chaikovskaya EA2 , Bolgarina AA4 Manturova NE1, Stenko AG2, Petinati YaA3, Chaikovskaya EA2 , Bolgarina AA4 1 Pirogov Russian National Research Medical University, Moscow, Russia 1 Pirogov Russian National Research Medical University, Moscow, Russia 2 Institute of plastic surgery and cosmetology, Moscow, Russia 3 Izmerov Research Institute of Occupational Health, Moscow, Russia 4 LLC Nearmedic PLUS Moscow Russia 3 Izmerov Research Institute of Occupational Health, Moscow, Russia LLC Nearmedic PLUS, Moscow, Russia To a large extent, age-related facial skin changes, wrinkles and flabbiness, are attributed to the structural alterations in dermis, including of collagen fibers fragmentation and disorganization. There are various cosmetological correction methods that aim to activate neocollagenesis and dermal remodeling. From this perspective, intradermal injections of exogenous collagen preparations seem logical. This study aimed to investigate the efficacy and safety of Collost 7% collagen complex applied to correct the age-related facial skin changes, as well as clarify the possible mechanisms of skin rejuvenation resulting from a course of intradermal injections. 35 participants entered the study, 30 of them finished it. A set of indicators describing age-related skin changes was assessed with the help of clinical scales; the assessment revealed a pronounced improvement in the quality of the patients' skin, including smoothed relief in the area of localization of fine wrinkles. The therapy resulted in a statistically significant improvement of the skin's elasticity, which, combined with the changes discovered through US scanning (greater dermis thickness and echodensity), is an indirect indication of skin restructuring associated with accumulation of fibrous protein structures. These results allow parallels with the experimental data that shows activation of neocollagenesis in the skin of laboratory animals after a course of Collost 7% gel. The research revealed no serious adverse events. A course of collagen administered intradermally can be recommended as an aesthetic correction procedure, as well as means of prevention of atrophy that has a significant effect on skin's appearance and health status. Keywords: skin ageing, skin rejuvenation, collagen injection Keywords: skin ageing, skin rejuvenation, collagen injection Acknowledgements: the authors would like to thank “Nearmedic Plus” (Moscow) for the support of the clinical observational study, as well as Natalya Indilova, Galina Sofinskaya, Evgenia Ikonnikova, medical researchers of the Institute of Plastic Surgery and Cosmetology, and Olga Bondareva and Olesya Sidorova, doctors with Cosmoprodtest SPC, for the assistance they provided in the context of this study. Н. Е. Мантурова1, А. Г. Стенько2, Я. А. Петинати3, Е. А. Чайковская2 , А. А. Болгарина4 1 Российский национальный исследовательский медицинский университет имени Н. И. Пирогова, Москва, Росс 3 Научно-исследовательский институт медицины труда имени академика Н. Ф. Измерова, Москва, Россия аучно-исследовательский институт медицины труда имени академика Н. Ф. Измерова, Москва, Россия Возрастные изменения лица в виде морщин, дряблости кожи во многом связаны со структурными изменениями дермы, в том числе с фрагментацией и дезорганизацией коллагеновых волокон. Различные методы косметологической коррекции направлены на активизацию неоколлагенеза и ремоделирование дермы. С этой точки зрения логичным видится проведение внутрикожных инъекций препаратов экзогенного коллагена. Целью исследования было изучить эффективность и безопасность применения коллагенового комплекса Коллост 7% в коррекции возрастных изменений кожи лица, а также уточнить возможные механизмы развития эффекта омоложения кожи после курса внутрикожных инъекций. В исследование были включены 34 участницы, завершили его 30 участниц. Оценка комплекса показателей возрастных изменений кожи с помощью клинических шкал продемонстрировала выраженное улучшение качества кожи и разглаживание ее рельефа, особенно в области локализации тонких морщин. После завершения курса лечения достоверно повысилась эластичность кожи лица, что вкупе с изменениями, выявленными при УЗ-сканировании (повышение толщины дермы и ее акустической плотности), косвенно свидетельствует о структурной перестройке кожи с накоплением белковых волокнистых структур. Полученные результаты позволяют проводить параллели с данными экспериментальных исследований, показывающими активизацию неоколлагенеза в коже лабораторных животных после курсового введения геля Коллост 7%. Серьезных нежелательных явлений при проведении исследования не выявлено. Курсовое внутрикожное введение коллагена можно рекомендовать как процедуру эстетической коррекции, а также в качестве профилактики развития атрофических процессов, которые существенно сказываются на внешнем виде и здоровье кожи. Ключевые слова: старение кожи, омоложение кожи, инъекции коллагена Благодарности: авторы благодарят компанию ООО «Ниармедик Плюс» (Москва) за поддержку клинического наблюдательного исследования, а также благодарят Наталью Индилову, Галину Софинскую, Евгению Иконникову — врачей-исследователей АО «Институт пластической хирургии и косметологии»; Ольгу Бондареву, Олесю Сидорову — врачей НПЦ «Космопродтест» за помощь в проведении исследования. Информация о вкладе авторов: Н. Е. Мантурова — планирование исследования и интерпретация данных; А. Г. Стенько, Я. А. Петинати — подбор участников, проведение исследования, интерпретация данных; Е. А. Чайковская — планирование исследования, сбор и интерпретация данных, подготовка рукописи; А. А. Болгарина — анализ литературы, планирование исследования. Соблюдение этических стандартов: исследование одобрено этическим комитетом Общества эстетической медицины (протокол № 2 от 7 сентября 2017 г.). Регистрационный номер на портале www.ClinicalTrials.gov — NCT03677258. Для корреспонденции: Екатерина Александровна Чайковская ул. Ольховская, д. 27, г. INJECTABLE COLLAGEN IN CORRECTION OF AGE-RELATED SKIN CHANGES: EXPERIMENTAL AND CLINICAL PARALLELS Manturova NE1, Stenko AG2, Petinati YaA3, Chaikovskaya EA2 , Bolgarina AA4 ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ Intradermal injections of the material instantly eliminate the soft tissue volume deficiency; administered as a course, it guarantees lasting results [31, 32]. A test for hypersensitivity to animal collagen is a mandatory measure before any procedures involving Collost gels. Collagen is the basic protein of any connective tissue, and half of its total amount resides in surface tissues, making up about 70% of the skin proteins [6]. The main structural components of the dermis are fiber-forming collagens of types I and III. Their complex architectonics define skin's sturdiness, elasticity, ability to repair [1, 5, 7, 8]. With age, collagen matrix changes significantly. In part, the change is the results of the predetermined normal ageing processes, but there are other factors that speed up these processes, including diseases, endocrine profile alterations, smoking, unhealthy nutritional habits, UV radiation, pollutants found in the environment [1, 9–12]. The studies conducted to date show that the content of collagen in dermis decreases progressively as people age. After 40, a woman loses approximately 20% of her skin's collagen every 10 years [7]. Different research contain different specific data [13–16], but the trend is the same: as people age, proliferative and synthesizing functions of fibroblasts fade, their pool diminishes, cells convert to an inactive "collapsed" phenotype [17]. The quality of collagen fibers changes together with the quantity: they become compacted due to the development of additional covalent cross-links of polypeptide chains. Such structures are more resistant to the action of matrix metalloproteinases, which catabolize proteins and thus incite their renewal. Randomly positioned fragmented collagen fibers that lose focal contacts with fibroblasts accumulate in extracellular matrix. Cells lose structural and functional contact with the matrix and convert to an inactive phenotype [9, 12, 17–19]. Thus forms the vicious circle: the changes of the skin's structure that draw the clinical picture of an "aged face" consolidate. This thesis is confirmed by the fact that in systemic scleroderma patients, whose skin is thicker and contains a greater amount of collagen, age-related facial changes — wrinkles, folds and enlarged pores — manifest themselves much later [20]. Metabolism and biochemical effects of a Collost 7% gel implanted into the skin were detailed in the context of recent experimental studies [21]. Laboratory animals (rats) were injected with the material intradermally twice. ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ network. The same purpose drives research of such methods as intracutaneous injection of active peptides, microgranules of polycaprolactone, implantation of polycaprolactone threads, exposure to radio frequencies and microfocused ultrasound, mechanical microperforation of skin, etc. Injecting the dermis proper with collagen appears etiologically sound. Aesthetic collagen therapy has been officially recognized about 40 years ago: in 1981, FDA (Food and Drug Administration) registered the first injection material based on this substance. Medical products developed around cow skin collagen are safe and show good biocompatibility [27]. The assumed pattern of action of collagen materials administered intradermally includes temporary hydration of dermis, optimization of the extracellular matrix properties, and stimulating effect of short peptides (matricins), a product of biodegradation of exogenous collagen [8, 28, 29]. Via the biological feedback mechanism, protein fragments accumulation per se stimulates synthesis of protein de novo [30]. In the body, skin is multifunctional: it is a barrier and thermal regulator, immune defense forefront and producer of a wide range of hormones. Lately, its social function has been drawing much attention, since appearance of the face, its aesthetic qualities and signs of diseases thereon, largely define the person's "external" age. Youthfulness and attractiveness of the face help people communicate, socialize and maintain the desired level of social life, which, one way or another, affects their quality of life and determines personal and professional success [1]. Therefore, medical researchers continue to study all the aspects of correction of age-related facial skin changes, with a certain number of such efforts addressing the topics of efficacy and safety of the products and methods developed. Assessing external age, people primarily focus on how noticeable facial wrinkles and skin folds are [2, 3]. The wrinkles and folds depend on biomechanical properties of skin [4, 5], which, in turn, are a reflection of the state of dermis' extracellular matrix, or connective tissue layer. Collost 7% and Collost 15% collagen material in the form of sterile gel products were developed and are produced in Russia by BioPHARMAHOLDING; the gels find use in various situations related to medical treatment, including cosmetological procedures. The collagen for Collost is derived from cattle hides; it is purified to remove the impurities without breaking the native fibrous structure and corrupting properties of the protein. The material is registered as an injectable medical product (FSR 2008/02112 of February 26, 2016). In cosmetology, Collost gels are prescribed to counter skin ageing and atrophy. ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ Throughout the entire observation period, which lasted 37 weeks, the researchers noted constant growth of the levels of total collagen and its soluble fraction. According to the researchers, within the several few days after the injection (2nd and 7th days) collagen level grows due to the presence of exogenous protein, whereas after a considerable period of time (21st and 37th days) the same trend signals of activation of neocollagenesis in the skin. This hypothesis is also confirmed by the fact that within the same timespans there was registered a significant intensification of incorporation of radiomarked C14 amino acids into the skin protein. Through the seven days of observation, collagenalytic tissue activity was accelerating, which may be the result of the foreign protein biodegradation. The discovered processes suggest that exogenous collagen that has its native structure preserved stimulates synthesis of the skin's native protein while biodegrading. The above justifies the need for further examination of various cosmetic methods to restore the collagen framework of dermis. Along with the aesthetic aspect, it is very important to work on restoring the skin's reparative resource: as an organized fiber network, collagen participates in regulation of migration, proliferation, differentiation of cells and their interaction [8, 21]. One of the objectives of the prospective observational study "Investigation of efficacy and safety of Collost 7% collagen recovery complex applied to correct age-related skin changes" was to clarify, applying non-invasive diagnostic methods, the possible mechanisms of skin rejuvenation during and after a course of intradermal injections. Topical retinoid preparations, laser irradiation and intradermal injection of fillers (gels of stabilized hyaluronic acid, polylactic acid, calcium hydroxyapatite particles) were proven to reliably induce neocollagenesis [22–26]. The action mechanisms of these means differ, but the result of their application is the same: affected skin develops a dense fibrous collagen Н. Е. Мантурова1, А. Г. Стенько2, Я. А. Петинати3, Е. А. Чайковская2 , А. А. Болгарина4 Москва, 105066; ktchaikovskaya@yandex.ru Статья получена: 27.09.2018 Статья принята к печати: 25.02.2019 Опубликована онлайн: 09.03.2019 Статья получена: 27.09.2018 Статья принята к печати: 25.02.2019 Опубликована онлайн: 09.03.2019 DOI: 10 24075/vrgmu 2019 010 Статья получена: 27.09.2018 Статья принята к печати: 25.02.2019 Опубликована онлайн: 09.03.2019 DOI: 10.24075/vrgmu.2019.010 BULLETIN OF RSMU 1, 2019 VESTNIKRGMU.RU \| \| 71 ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ ORIGINAL RESEARCH DERMATOLOGY four participants were screened for fitness to accept a course of Collost injections, 30 were found fit. four participants were screened for fitness to accept a course of Collost injections, 30 were found fit. The results were documented using the LifeViz system (QuantifiCare; France) that makes and enables analysis of 3D images of the face. Inclusion criteria: female; age 35–65 years; skin phototypes I–III; signs of facial skin age-related changes; no chronic diseases at the stage of decompensation; abstention from any cosmetic procedures during the study. Exclusion criteria: pregnancy; lactation; infection; dermatosis; malignant skin neoplasms in the supposed correction zone; systemic connective tissue diseases that damage skin and subcutaneous tissue (systemic lupus erythematosus, ring granuloma, discoid lupus erythematosus, scleroderma, dermatomyositis, etc.); exacerbation or decompensation of chronic somatic diseases; infectious, oncological diseases; coagulation disorders, including iatrogenic caused by drugs taken; use of isotretinoin within the previous 6 months; propensity hypertrophic and keloid scars; use of antihistamines, glucocorticoids, NSAIDs, immunosuppressants and other drugs that affect skin reactivity; hypersensitivity to the components of the studied medical product. To assess the safety of Collost intradermal injections, each procedure was followed by registration of adverse events, immediate and delayed, that could be connected to the individual response to collagen material and/or injection itself. The study design implied five visits, screening based on inclusion and exclusion criteria, allergy testing as prescribed by the Collost gel material leaflet. Allergy history of the participants and the result of their allergy tests were scrutinized by an immunologist. The participants received intradermal injections of Collost 7% three times with an interval of 3 weeks between each. To make the procedures comfortable, they had Acriol Pro anesthetic cream applied to their skin 1 hour prior to injection. The injected amount was 2 ml stored in prefilled syringes that were preheated in a thermostat at 40 °C. The skin was cleaned and disinfected with the help of 0.05% chlorhexidine digluconate solution; Traumeel gel was applied to the patients' facial skin at the end of each procedure. The participants underwent full examination before the course, prior to each procedure (evaluating the results of the previous ones) and 3 weeks after the course was over. The participants were 36 to 64 years old, (average age — 48.50; 25th and 75th percentiles — 43.00 and 57.00 years), Caucasian, with a body weight of 52–80 kg and body mass index less than 30. Safety profile analysis During the study, all participants exhibited the expected local reactions in the form of skin reddening, slight edemas, palpable pimples at the injection sites. The reactions occurred after the procedures and lasted several hours to several days. In spite of the injection being considerably traumatic, in most cases bleeding took form of petechiae (point hemorrhages), which may be explained by the hemostatic properties of collagen. Following a detailed examination, the research program included a description of the overall condition of facial skin using 8 indicators (scored against the 5-point Gejnic, Alexiades-Armenakas scale [33, 34]) calculated into an integral skin condition indicator, as well as assessment of wrinkles in 7 zones applying the 5-point MAS (Merz Aesthetic Scale) validated grading scale [35]. PATIENTS AND METHODS The study lasted from October 2017 to June 2018; it was conducted at the Institute of Plastic Surgery and Cosmetology and Izmerov Research Institute of Occupational Health. Thirty- ВЕСТНИК РГМУ 1, 2019 VESTNIKRGMU.RU \| \| 72 ORIGINAL RESEARCH DERMATOLOGY ORIGINAL RESEARCH DERMATOLOGY To avoid skewing the clinical picture of facial skin ageing, the participants maintained a stable body weight (fluctuations of 2 kg max.) through the entire observation period. No allergies were found in the histories of 93.33% of participants; all of them had no allergic reaction to local anesthetics. The majority of patients (24 people, 80.00%) were non-smoking, all adhered to a mixed diet. The skin of 16.67% of them was highly sensitive; the thickness and sensitivity of skin of other participants was normal. The "average" participant of this study could be described as follows: a woman about 50 years old with fair skin, photoaging picture fitting her age, with signs of age-related deformational changes on her face that are largely the result of progressing flabbiness, i. e., changes in the skin's biomechanical properties. Statistica 6.0 (Statsoft; Russia) software was used to perform statistical analysis; we used nonparametric methods. The data are presented as medians (Me), 25th and 75th percentiles. The two groups of dependent markers were compared through the Wilcoxon test; the differences were considered statistically significant at p < 0.05. ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ employed non-invasive diagnostic methods exclusively, which enabled thorough examination of all participants. collected after the third procedure) (Table 1): the reduction median reflecting the overall visibility of wrinkles on the MAS scale was 28.57%. The detailed analysis of wrinkles in certain zones showed that the most visible changes (skin relief smoothening) occurred on the forehead, paraorbital areas ("crow's feet" wrinkles), lower eyelid and upper lip (Fig. 1). The criteria suggested by Gejnic et al, Alexiades-Armenakas [33, 34] were used to assess the overall condition of skin, including photodamage / photoaging symptoms (modified assessment pattern). The integrated indicator combined scores reflecting the state of micro- and macro-relief of the skin, its color, texture, pigmentation, vascular pattern, elastosis and keratosis. Post-course, the indicator has decreased significantly; detailing the results, the visible changes at the skin relief and color levels (wrinkle and folds reduction) should be noted. Two patients with thin skin around eyelids saw the blues spots under their eyes disappear, which was perceived extremely positively by the patients themselves and their close ones. Overall, the skin of most participants, acquired a pink hue and a healthier look. Skin elasticity in 77% of participants has improved significantly (Table 2): the indicator growth median was 30.90% for forehead, 19.10% for paraorbital area and 15.75% for cheeks. All changes are significant (p < 0.05; screening data compared to data collected after the third procedure). According to the results of the sonography, the following indicators have improved significantly in all participants post- course: there was an increase in epidermis thickness (by 30.7– 39.3%), dermis thickness (by 18.7–22.3%), dermis acoustic density (by 28.7–44.5%, all or some zones of the face) (Table 2, Fig. 2 and 3) (p < 0.05; screening data compared to data collected after the third procedure). In the forehead, paraorbital and upper lip zones wrinkles smoothed out the best. These zones have superficial wrinkles associated with the dermal skin layer atrophy/alterations [36]. Restoration of the collagen framework of dermis enables effective correction of skin relief in these areas. However, correcting deep folds (e. g., nasolabial furrows) requires intervention at the level of subcutaneous fat. All (100%) participants considered the COllost 7% gel injection course to be effective. Applying the GAIS scale, half of them called the result satisfactory and expressed a wish to have it improved, while another half was fully satisfied with the result (Fig. 4). ОРИГИНАЛЬНОЕ ИССЛЕДОВАНИЕ ДЕРМАТОЛОГИЯ From a medical doctor's point of view, a third of participants had the result at the optimal level for them, whereas the majority did not enjoy a full correction but only a considerable improvement. In 2 participants, the improvement was insignificant, an opinion shared by the participants themselves and the medical researcher. Most researchers see a connection between biomechanical properties of the skin and quality of the dermis' fibrous collagen framework. It is logical to assume that the significant skin elasticity improvement registered in the participants may be the result of neocollagenesis, generation of mature fibers with intact structure and sufficient hydrophilicity. The other indirect confirmations of the skin's restructurisation are the increased thickness of dermis and growth of its acoustic density, which were detected by ultrasound scanning. According to the earlier research [37, 38], reduction of dermis thickness and acoustic Efficacy analysis The equipment-aided tests were: The equipment-aided tests were: • skin elasticity test (cutometry) on forehead, paraorbital zone, cheek, the devices enabling the test were Multi Skin Test Center MC-900 and Cutometer Dual MPA-580 (Courage + Khazaka electronic GmbH; Germany); Statistical analysis showed a significant improvement of the clinical and aesthetic properties of facial skin and face on the whole. The integral indicator used to assess the changes decreased significantly (p < 0.05; screening data compared to data collected after the third procedure) (Table 1): the reduction median was 24.26%. It should be explained here that the integral indicator grows together with the level of contribution of specific markers (wrinkles, folds, rough skin pattern, telangiectasia, etc.) to the "aged face" picture. Thus, when this indicator goes down, the face rejuvenates. • epidermis and dermis thickness test, dermis acoustic density test, implying ultrasound scanning of the equivalent collateral patches of skin using DUB SkinScanner (Taberna pro medicum; Germany) and DermaScan C-System (Cortex Technology; Denmark) featuring 20–22 MHz sensors. Both medical researchers and participants of the study assessed the efficacy of cosmetological correction course. The latter applied the universal Global Aesthetic Improvement Scale (GAIS) [35]. We have registered significant positive trend in the wrinkle depth change (p < 0.05; screening data compared to data Table 1. Skin quality indicators, screening and post-course (p < 0.05) Note: * — this Table and Table 2 provide median values, brackets contain values of 25th and 75th percentiles. Indicator Screening Post-course Integral skin condition indicator, points 14 (10; 17)* 11.5 (8; 13) Wrinkle visibility (consolidated indicator, 7 zones), points 12 (11; 16) 8 (7; 11) Table 1. Skin quality indicators, screening and post-course (p < 0.05) Table 1. Skin quality indicators, screening and post-course (p < 0.05) BULLETIN OF RSMU 1, 2019 VESTNIKRGMU.RU \| \| 73 DISCUSSION This clinical study confirmed the efficacy of Collost 7% as a correction product for involutional skin changes. The study This clinical study confirmed the efficacy of Collost 7% as a correction product for involutional skin changes. The study Fig. 1. Wrinkle visibility change dynamics by face zones (screening data compared to post-course data) Table 2. Structural and functional skin indicators, screening and post-course (p < 0.05) Indicators Screening Post-course Zones Zones Forehead Paraorbital area Cheek Forehead Paraorbital area Cheek Epidermis thickness (US scan), mm 0.128 (0.105; 0.158) 0.110 (0.105; 0.133) 0.132 (0.108; 0.158) 0.177 (0.151; 0.192) 0.163 (0.150; 0.184) 0.186 (0.162; 0.199) Dermis thickness (US scan), mm 1.330 (1.195; 1.531) 1.206 (1.086; 1.376) 1.592 (1.256; 1.869) 1.628 (1.411; 1.790) 1.496 (1.344; 1.672) 1.855 (1.685; 2.094) Dermis acoustic density (US scan), c. u. 5.102 (3.741; 13.190) 5.738 (3.778; 13.470) 5.794 (2.877; 11.906) 7.236 (5.672; 13.540) 9.071 (6.026; 17.010) 6.611 (4.739; 13.230) Skin elasticity (cutometry), c. u. 52.63 (46.00; 61.45) 55.08 (49.85; 62.65) 55.06 (49.30; 62.00) 66.75 (56.03; 93.39) 65.65 (58.00; 74.16) 63.25 (55.50; 61.50) Forehead -33% -42% -12% -49% -36% -22% -12% Intercilium Paraorbital areas Lower eyelid Upper lip Nasolabial furrows Labiomental folds -33% -42% -12% -49% -36% -22% -12% Paraorbital areas Nasolabial furrows Fig. 1. Wrinkle visibility change dynamics by face zones (screening data compared to post-course data) Fig. 1. Wrinkle visibility change dynamics by face zones (screening data compared to post-course data) Table 2. Structural and functional skin indicators, screening and post-course (p < 0.05) Indicators Screening Post-course Zones Zones Forehead Paraorbital area Cheek Forehead Paraorbital area Cheek Epidermis thickness (US scan), mm 0.128 (0.105; 0.158) 0.110 (0.105; 0.133) 0.132 (0.108; 0.158) 0.177 (0.151; 0.192) 0.163 (0.150; 0.184) 0.186 (0.162; 0.199) Dermis thickness (US scan), mm 1.330 (1.195; 1.531) 1.206 (1.086; 1.376) 1.592 (1.256; 1.869) 1.628 (1.411; 1.790) 1.496 (1.344; 1.672) 1.855 (1.685; 2.094) Dermis acoustic density (US scan), c. u. 5.102 (3.741; 13.190) 5.738 (3.778; 13.470) 5.794 (2.877; 11.906) 7.236 (5.672; 13.540) 9.071 (6.026; 17.010) 6.611 (4.739; 13.230) Skin elasticity (cutometry), c. u. 52.63 (46.00; 61.45) 55.08 (49.85; 62.65) 55.06 (49.30; 62.00) 66.75 (56.03; 93.39) 65.65 (58.00; 74.16) 63.25 (55.50; 61.50) Table 2. DISCUSSION Structural and functional skin indicators, screening and post-course (p < 0.05) ВЕСТНИК РГМУ 1, 2019 VESTNIKRGMU.RU \| \| 74 ORIGINAL RESEARCH DERMATOLOGY Fig 3 US scans of forehead skin: screening (A) post course (B) А B А B density is an age-related phenomenon, associated with diseases or steroid-induced atrophy that can be caused by the decreasing content of structural fibrillar proteins. Thus, the growing thickness and acoustic density reflect accumulation of fibrous structures, restructuring of the skin [39, 40]. According to our research, the manifestation of these changes was most evident 3 weeks post-course, which allows a parallel with the afore-mentioned experimental studies that registered the peak of collagen content in the laboratory animals' skin 2–4 weeks after the final Collost 7% injection. B А It should be noted that improvement of the skin's biomechanical properties, the structural changes observed after a course Collost 7% injections do not only enhance the appearance but help to prevent the age-related changes associated with atrophy and contribute to healing of the skin. References References 1. Rittié L, Fisher GJ. Natural and sun-induced aging of human skin. Cold Spring Harb Perspect Med. 2015; 5 (1): a015370. 22. Kruglikov I. Neocollagenesis in Non-Invasive Aesthetic Treatments. J Cosmet Dermatol Sci Appl. 2013; 3 (1A): 1–5. 23. Shao Y, He T, Fisher GJ, Voorhees JJ, Quan T. Molecular basis of retinol anti-ageing properties in naturally aged human skin in vivo. Int J Cosmet Sci. 2017; 39 (1): 56–65. 2. Gunn DA, Rexbye H, Griffiths CE, Murray PG, Fereday A, Catt SD, et al. Why Some Women Look Young for Their Age. Tregenza T, ed. PLoS ONE. 2009; 4 (12): e8021. 24. Wang F, Garza LA, Kang S, Varani J, Orringer JS, Fisher GJ, et al. In vivo stimulation of de novo collagen production caused by cross- linked hyaluronic acid dermal filler injections in photodamaged human skin. Arch Dermatol. 2007; 143 (2): 155–63. 3. Nkengne A, Bertin C, Stamatas GN, Giron A, Rossi A, Issachar N, et al. Influence of facial skin attributes on the perceived age of Caucasian women. J Eur Acad Dermatol Venereol. 2008: 22 (8): 982–91. 4. Hussain SH, Limthongkul B, Humphreys TR. The biomechanical properties of the skin. Dermatol Surg. 2013; 39 (2): 193–203. properties of the skin. Dermatol Surg. 2013; 39 (2): 193–203 25. Yutskovskaya Y, Kogan E, Leshunov E. A randomized, split- face, histomorphologic study comparing a volumetric calcium hydroxylapatite and a hyaluronic acid-based dermal filler. J Drugs Dermatol. 2014; 13 (9): 1047–52. 5. Aziz J, Shezali H, Radzi Z, Yahya NA, Abu Kassim NH, Czernuszka J, et al. Molecular Mechanisms of Stress-Responsive Changes in Collagen and Elastin Networks in Skin. Skin Pharmacol Physiol. 2016; 29 (4): 190–203. Dermatol. 2014; 13 (9): 1047–52. ( ) 26. Stein P, Vitavska O, Kind P, Hoppe W, Wieczorek H, Schürer NY. The biological basis for poly-L-lactic acid-induced augmentation. J Dermatol Sci. 2015; 78 (1): 26–33. ( ) 6. Czekalla C, Schönborn KH, Döge N, Jung S, Darvin ME, Lademann J, et al. Impact of Body Site, Age, and Gender on the Collagen/Elastin Index by Noninvasive in vivo Vertical Two-Photon Microscopy. Skin Pharmacol Physiol. 2017; 30 (5): 260–7. 27. Rao KP. Recent developments of collagen-based materials for medical applications and drug delivery systems. J Biomater Sci Polymer Ed. 1995; 7 (7): 623–45. Microscopy. Skin Pharmacol Physiol. 2017; 30 (5): 260–7. 7. References Baroni Edo R, Biondo-Simões Mde L, Auersvald A, Auersvald LA, Montemor Netto MR, Ortolan MC, et al. Influence of aging on the quality of the skin of white women: the role of collagen. Acta Cir Bras. 2012; 27 (10): 736–40. 28. Katayama K, Armendariz-Borunda J, Raghow R, Kang AH, Seyer JM. A pentapeptide from type I procollagen promotes extracellular matrix production. J Biol Chem. 1993; 268 (14): 9941–4. 8. Kapuler O, Selskaja B, Galeeva A, Kamilov F. Metabolizm kollagenovyh volokon na fone vozrastnyh izmenenij. Vrach. 2015; (8): 64–9. 29. Maquart FX, Pasco S, Ramont L, Hornebeck W, Monboisse JC. An introduction to matrikines: extracellular matrix-derived peptides which regulate cell activity. Implication in tumor invasion. Crit Rev Oncol Hematol. 2004; 49 (3): 199–202. 9. Quan T, Fisher GJ. Role of Age-Associated Alterations of the Dermal Extracellular Matrix Microenvironment in Human Skin Aging: A Mini-Review. Gerontology. 2015; 61 (5): 427–34. 30. Abojanc RK, Istranov LP, Istranova EV, Rudenko TG. Plasticheskie materialy napravlennogo dejstvija na osnove kollagena. Jelektronnyj sbornik nauchnyh trudov "Zdorov'e i obrazovanie v XXI veke". 2011; (4): 184. 10. Calleja-Agius J, Brincat M, Borg M. Skin connective tissue and ageing. Best Pract Res Clin Obstet Gynaecol. 2013; 27 (5): 727–40. g g 11. Danby FW. Nutrition and aging skin: sugar and glycation. Clin Dermatol. 2010; 28 (4): 409–11. 31. Kubanova AA, Smoljannikova VA, Sluzhaeva NG. Starenie kozhi i vozmozhnosti korrekcii preparatom kollagena. Vestnik dermatologii i venerologii. 2007; (5): 70–3. 12. Ahmed T, Nash A, Clark KE, Ghibaudo M, de Leeuw NH, Potter A, et al. Combining nano-physical and computational investigations to understand the nature of "aging" in dermal collagen. Int J Nanomedicine. 2017; (12): 3303–14. 32. Kapuler OM, Kuramshina ER. Prikladnye aspekty kollagenoterapii v jesteticheskoj medicine. Jeksperimental'naja i klinicheskaja dermatokosmetologija. 2013; (5): 40–3. 13. Moragas A, Garcia-Bonafé M, Sans M, Torán N, Huguet P, Martin-Plata C. Image analysis of dermal collagen changes during skin aging. Analyt Quant Cytol Histol. 1998; (20): 493–9. 33. Gejnic AV, Kiani A, Okushko SS. Novye vozmozhnosti primenenija glubokoj i poverhnostnoj frakcionnoj abljacii v anti-age terapii. Plasticheskaja hirurgija i kosmetologija. 2013; (4): 625–32. 14. El-Domyati M, Attia S, Saleh F, Brown D, Birk DE, Gasparro F, et al. Intrinsic aging VS photoaging: a comparative histopathological, immunohistochemical and ultrastructural study of skin. Exp Dermatol. 2002; (11): 398–405. 34. Alexiades-Armenakas M, Newman J, Willey A, Kilmer S, Goldberg D, Garden J, et al. 1. Rittié L, Fisher GJ. Natural and sun-induced aging of human skin. Cold Spring Harb Perspect Med. 2015; 5 (1): a015370. CONCLUSIONS Fig. 3. US scans of forehead skin: screening (A), post-course (B) collagen framework remodeling process. Thus, Collost 7% intradermal injections can be considered an effective and safe method of aesthetic correction. Perhaps, the results of this study will contribute to the development of etiologically and pathogenetically substantiated protocols of combined correction of age-related skin changes that would include administration of HA, polylactic acids and high-energy methods. According to our study, the course of intradermal injections of Collost 7% improves the aesthetic appearance of facial skin, reliably normalizes its color, reduces visibility of wrinkles. Moreover, the positive clinical dynamics occur against the background of improving biomechanical properties of the skin (better elasticity) and structural changes identified by US scanning, which indirectly signal of the dermis Fig. 2. Dynamics of alteration of skin structure indicators, US scans (post-course median against screening data) 44.50% 35.80% 28.70% 21.80% 22.30% 18.70% 39.30% 36% 30.70% Epidermis thickness (mm) Forehead Paraorbital area Cheek Dermis thickness (mm) Dermis acoustic density (c. u.) Fig. 4. Injection course efficacy assessment, GAIS, applied by participants and medical researchers 7% 40% 53% 7% 60% 33% Participants Insignificant improvement Significant improvement Optimal result Doctors Fig. 2. Dynamics of alteration of skin structure indicators, US scans (post-course median against screening data) 44.50% 35.80% 28.70% 21.80% 22.30% 18.70% 39.30% 36% 30.70% Epidermis thickness (mm) Forehead Paraorbital area Cheek Dermis thickness (mm) Dermis acoustic density (c. u.) Fig. 2. Dynamics of alteration of skin structure indicators, US scans (post-course median against screening data) Fig. 4. Injection course efficacy assessment, GAIS, applied by participants and medical researchers 7% 40% 53% 7% 60% 33% Participants Insignificant improvement Significant improvement Optimal result Doctors BULLETIN OF RSMU 1, 2019 VESTNIKRGMU.RU \| \| 75 Gunn DA, Rexbye H, Griffiths CE, Murray PG, Fereday A, Catt SD, et al. Why Some Women Look Young for Their Age. Tregenza T, References Prospective multicenter clinical trial of a minimally invasive temperature-controlled bipolar fractional radiofrequency system for rhytid and laxity treatment. Dermatol Surg. 2013; 39 (2): 263–73. 15. Yaar M, Eller MS, Gilchrest BA. Fifty years of skin aging. J Invest Dermatol. 2002; (7): 51–8. 35. Carruthers A, Carruthers J. A validated facial grading scale: the future of facial ageing measurement tools? J Cosmet Laser Ther. 2010; 12 (5): 235–41. 16. Cheng W, Yan-hua R, Fang-gang N, Guo-an Z. The content and ratio of type I and III collagen in skin differ with age and injury. African J Biotechnol. 2011; 10 (13): 2524–9. 17. Fisher GJ, Varani J, Voorhees JJ. Looking older: fibroblast collapse and therapeutic implications. Arch Dermatol. 2008; 144 (5): 666–72. 36. Tsukahara K, Tamatsu Y, Sugawara Y, Shimada K. The relationship between wrinkle depth and dermal thickness in the forehead and lateral canthal region. Arch Dermatol. 2011; 147 (7): 822–8. 18. Cole MA, Quan T, Voorhees JJ, Fisher GJ. Extracellular matrix regulation of fibroblast function: redefining our perspective on skin aging. J Cell Commun Signal. 2018; 12 (1): 35–43. 37. Waller JM, Maibach HI. Age and skin structure and function, a quantitative approach (I): blood flow, pH, thickness, and ultrasound echogenicity. Skin Res Technol. 2005; (11): 221–35. 19. Varani J, Spearman D, Perone P, Fligiel SE, Datta SC, Wang ZQ, et al. Inhibition of type I procollagen synthesis by damaged collagen in photoaged skin and by collagenase-degraded collagen in vitro. Am J Pathol. 2001; 158 (3): 931–42. 38. Jasaitiene D, Valiukeviciene S, Linkeviciute G, et al. Principles of high-frequency ultrasonography for investigation of skin pathology. J Eur Acad Dermatol Venereol. 2011; (25): 375–82. 20. Sawamura S, Jinnin M, Kajihara I, Makino K, Aoi J, Ichihara A, et al. Do scleroderma patients look young? Evaluation by using facial imaging system. Drug Discov Ther. 2017; 11 (6): 342–5. 39. Kozarova A, Kozar M, Minarikova E, Pappova T. Identification of the age related skin changes using high-frequency ultrasound. Acta Medica Martiniana. 2017; 17 (1): 15–20. 40. Lacarrubba F, Tedeschi A, Nardone B, Micali G. Mesotherapy for skin rejuvenation: assessment of the subepidermal low- echogenic band by ultrasound evaluation with cross-sectional B-mode scanning. Dermatol Ther. 2008; 21 (Suppl 3): 1–5. 21. Kamilov FH, Selskaja BN, Danilova OV, Kapuler OM. Metabolizm kollagena v kozhe jeksperimental'nyh zhivotnyh pri intradermal'noj in’ekcii nemodificirovannogo bych'ego kollagena tipa I. Vestnik Udmurdskogo universiteta. 2017; 27 (3): 356–61. References A pentapeptide from type I procollagen promotes extracellular matrix production. J Biol Chem. 1993; 268 (14): 9941–4. 8. Капулер О., Сельская Б., Галеева А., Камилов Ф. Метаболизм коллагеновых волокон на фоне возрастных изменений. Врач. 2015; (8): 64–9. 29. Maquart FX, Pasco S, Ramont L, Hornebeck W, Monboisse JC. An introduction to matrikines: extracellular matrix–derived peptides which regulate cell activity. Implication in tumor invasion. Crit Rev Oncol Hematol. 2004; 49 (3): 199–202. 9. Quan T, Fisher GJ. Role of Age-Associated Alterations of the Dermal Extracellular Matrix Microenvironment in Human Skin Aging: A Mini-Review. Gerontology. 2015; 61 (5): 427–34. 30. Абоянц Р. К., Истранов Л. П., Истранова Е. В., Руденко Т. Г. Пластические материалы направленного действия на основе коллагена. Электронный сборник научных трудов «Здоровье и образование в XXI веке». 2011; (4): 184. 10. Calleja-Agius J, Brincat M, Borg M. Skin connective tissue and ageing. Best Pract Res Clin Obstet Gynaecol. 2013; 27 (5): 727–40. g g y ( ) 11. Danby FW. Nutrition and aging skin: sugar and glycation. Clin Dermatol. 2010; 28 (4): 409–11. 31. Кубанова А. А., Смольянникова В. А., Служаева Н. Г. Старение кожи и возможности коррекции препаратом коллагена. Вестник дерматологии и венерологии. 2007; (5): 70–3. 12. Ahmed T, Nash A, Clark KE, Ghibaudo M, de Leeuw NH, Potter A, et al. Combining nano-physical and computational investigations to understand the nature of "aging" in dermal collagen. Int J Nanomedicine. 2017; (12): 3303–14. 32. Капулер О. М., Курамшина Е. Р. Прикладные аспекты коллагенотерапии в эстетической медицине. Экспериментальная и клиническая дерматокосметология. 2013; (5): 40–3. 13. Moragas A, Garcia-Bonafé M, Sans M, Torán N, Huguet P, Martin-Plata C. Image analysis of dermal collagen changes during skin aging. Analyt Quant Cytol Histol. 1998; (20): 493–9. 33. Гейниц А. В., Киани А., Окушко С. С. Новые возможности применения глубокой и поверхностной фракционной абляции в anti-age терапии. Пластическая хирургия и косметология. 2013; (4): 625–32. 14. El-Domyati M, Attia S, Saleh F, Brown D, Birk DE, Gasparro F, et al. Intrinsic aging VS photoaging: a comparative histopathological, immunohistochemical and ultrastructural study of skin. Exp Dermatol. 2002; (11): 398–405. 34. Alexiades-Armenakas M, Newman J, Willey A, Kilmer S, Goldberg D, Garden J, et al. Prospective multicenter clinical trial of a minimally invasive temperature-controlled bipolar fractional radiofrequency system for rhytid and laxity treatment. Dermatol Surg. 2013; 39 (2): 263–73. 15. Yaar M, Eller MS, Gilchrest BA. Fifty years of skin aging. J Invest Dermatol. 2002; (7): 51–8. 16. References Литература Gunn DA, Rexbye H, Griffiths CE, Murray PG, Fereday A, Catt SD, et al. Why Some Women Look Young for Their Age. Tregenza T, ВЕСТНИК РГМУ 1, 2019 VESTNIKRGMU.RU \| \| 76 ORIGINAL RESEARCH DERMATOLOGY 23. Shao Y, He T, Fisher GJ, Voorhees JJ, Quan T. Molecular basis of retinol anti-ageing properties in naturally aged human skin in vivo. Int J Cosmet Sci. 2017; 39 (1): 56–65. ed. PLoS ONE. 2009; 4 (12): e8021. 3. 3. Nkengne A, Bertin C, Stamatas GN, Giron A, Rossi A, Issachar N, et al. Influence of facial skin attributes on the perceived age of Caucasian women. J Eur Acad Dermatol Venereol. 2008: 22 (8): 982–91. 24. Wang F, Garza LA, Kang S, Varani J, Orringer JS, Fisher GJ, et al. In vivo stimulation of de novo collagen production caused by cross- linked hyaluronic acid dermal filler injections in photodamaged human skin. Arch Dermatol. 2007; 143 (2): 155–63. 4. Hussain SH, Limthongkul B, Humphreys TR. The biomechanical properties of the skin. Dermatol Surg. 2013; 39 (2): 193–203. 5. Aziz J, Shezali H, Radzi Z, Yahya NA, Abu Kassim NH, Czernuszka J, et al. Molecular Mechanisms of Stress Responsive Changes in Collagen and Elastin Networks in Skin. Skin Pharmacol Physiol. 2016; 29 (4): 190–203. 25. Yutskovskaya Y, Kogan E, Leshunov E. A randomized, split- face, histomorphologic study comparing a volumetric calcium hydroxylapatite and a hyaluronic acid-based dermal filler. J Drugs Dermatol. 2014; 13 (9): 1047–52. Dermatol. 2014; 13 (9): 1047–52. ( ) 26. Stein P, Vitavska O, Kind P, Hoppe W, Wieczorek H, Schürer NY. The biological basis for poly-L-lactic acid-induced augmentation. J Dermatol Sci. 2015; 78 (1): 26–33. ( ) 6. Czekalla C, Schönborn KH, Döge N, Jung S, Darvin ME, Lademann J, et al. Impact of Body Site, Age, and Gender on the Collagen/Elastin Index by Noninvasive in vivo Vertical Two-Photon Microscopy. Skin Pharmacol Physiol. 2017; 30 (5): 260–7. 27. Rao KP. Recent developments of collagen-based materials for medical applications and drug delivery systems. J Biomater Sci Polymer Ed. 1995; 7 (7): 623–45. Microscopy. Skin Pharmacol Physiol. 2017; 30 (5): 260–7. 7. Baroni Edo R, Biondo-Simões Mde L, Auersvald A, Auersvald LA, Montemor Netto MR, Ortolan MC, et al. Influence of aging on the quality of the skin of white women: the role of collagen. Acta Cir Bras. 2012; 27 (10): 736–40. 28. Katayama K, Armendariz-Borunda J, Raghow R, Kang AH, Seyer JM. References Cheng W, Yan-hua R, Fang-gang N, Guo-an Z. The content and ratio of type I and III collagen in skin differ with age and injury. African J Biotechnol. 2011; 10 (13): 2524–9. 35. Carruthers A, Carruthers J. A validated facial grading scale: the future of facial ageing measurement tools? J Cosmet Laser Ther. 2010; 12 (5): 235–41. 17. Fisher GJ, Varani J, Voorhees JJ. Looking older: fibroblast collapse and therapeutic implications. Arch Dermatol. 2008; 144 (5): 666–72. 36. Tsukahara K, Tamatsu Y, Sugawara Y, Shimada K. The relationship between wrinkle depth and dermal thickness in the forehead and lateral canthal region. Arch Dermatol. 2011; 147 (7): 822–8. 18. Cole MA, Quan T, Voorhees JJ, Fisher GJ. Extracellular matrix regulation of fibroblast function: redefining our perspective on skin aging. J Cell Commun Signal. 2018; 12 (1): 35–43. 19. Varani J, Spearman D, Perone P, Fligiel SE, Datta SC, Wang ZQ, et al. Inhibition of type I procollagen synthesis by damaged collagen in photoaged skin and by collagenase-degraded collagen in vitro. Am J Pathol. 2001; 158 (3): 931–42. 37. Waller JM, Maibach HI. Age and skin structure and function, a quantitative approach (I): blood flow, pH, thickness, and ultrasound echogenicity. Skin Res Technol. 2005; (11): 221–35. 38. Jasaitiene D, Valiukeviciene S, Linkeviciute G, et al. Principles of high-frequency ultrasonography for investigation of skin pathology. J Eur Acad Dermatol Venereol. 2011; (25): 375–82. 20. Sawamura S, Jinnin M, Kajihara I, Makino K, Aoi J, Ichihara A, et al. Do scleroderma patients look young? Evaluation by using facial imaging system. Drug Discov Ther. 2017; 11 (6): 342–5. 39. Kozarova A, Kozar M, Minarikova E, Pappova T. Identification of the age related skin changes using high-frequency ultrasound. Acta Medica Martiniana. 2017; 17 (1): 15–20. 21. Камилов Ф. Х., Сельская Б. Н., Данилова О. В., Капулер О. М. Метаболизм коллагена в коже экспериментальных животных при интрадермальной инъекции немодифицированного бычьего коллагена типа I. Вестник Удмуртского университета. 2017; 27 (3): 356–61. 40. Lacarrubba F, Tedeschi A, Nardone B, Micali G. Mesotherapy for skin rejuvenation: assessment of the subepidermal low- echogenic band by ultrasound evaluation with cross-sectional B-mode scanning. Dermatol Ther. 2008; 21 (Suppl 3): 1–5. 22. Kruglikov I, Neocollagenesis in Non-Invasive Aesthetic Treatments. J Cosmet Dermatol Sci Appl. 2013; 3 (1A): 1–5. BULLETIN OF RSMU 1, 2019 VESTNIKRGMU.RU \| \| 77
https://openalex.org/W1631236554	https://asr.copernicus.org/articles/12/187/2015/asr-12-187-2015.pdf	English	null	Comparison of regional and global reanalysis near-surface winds with station observations over Germany	Advances in science and research	2,015	cc-by	8,131	Germany A. K. Kaiser-Weiss1, F. Kaspar1, V. Heene1, M. Borsche1, D. G. H. Tan2, P. Poli2, A. Obregon1, and H. Gregow3 1Deutscher Wetterdienst, Frankfurter Straße 135, 63067 Offenbach, Germany 2European Centre for Medium-Range Weather Forecasts, Shinﬁeld Park, Reading, RG2 9AX, UK 3Finnish Meteorological Institute, P.O. Box 503, 00101 Helsinki, Finland Correspondence to: A. K. Kaiser-Weiss (andrea.kaiser-weiss@dwd.de) Received: 24 January 2015 – Revised: 9 May 2015 – Accepted: 1 July 2015 – Published: 29 July 2015 A. K. Kaiser-Weiss1, F. Kaspar1, V. Heene1, M. Borsche1, D. G. H. Tan2, P. Poli2, A. Obregon1, and H. Gregow3 A. K. Kaiser-Weiss1, F. Kaspar1, V. Heene1, M. Borsche1, D. G. H. Tan2, P. Poli2, A. Obregon1, and H. Gregow3 1Deutscher Wetterdienst, Frankfurter Straße 135, 63067 Offenbach, Germany 2European Centre for Medium-Range Weather Forecasts, Shinﬁeld Park, Reading, RG2 9AX, UK 3Finnish Meteorological Institute, P.O. Box 503, 00101 Helsinki, Finland Correspondence to: A. K. Kaiser-Weiss (andrea.kaiser-weiss@dwd.de) Received: 24 January 2015 – Revised: 9 May 2015 – Accepted: 1 July 2015 – Published: 29 July 2015 an Conference on Applied Climatology (ECAC) Correspondence to: A. K. Kaiser-Weiss (andrea.kaiser-weiss@dwd.de) Abstract. Reanalysis near-surface wind ﬁelds from multiple reanalyses are potentially an important informa- tion source for wind energy applications. Inter-comparing reanalyses via employing independent observations can help to guide users to useful spatio-temporal scales. Here we compare the statistical properties of wind speeds observed at 210 traditional meteorological stations over Germany with the reanalyses’ near-surface ﬁelds, con- ﬁning the analysis to the recent years (2007 to 2010). In this period, the station time series in Germany can be expected to be mostly homogeneous. We compare with a regional reanalysis (COSMO-REA6) and two global reanalyses, ERA-Interim and ERA-20C. We show that for the majority of the stations, the Weibull parameters of the daily mean wind speed frequency distribution match remarkably well with the ones derived from the re- analysis ﬁelds. High correlations (larger than 0.9) can be found between stations and reanalysis monthly mean wind speeds all over Germany. Generally, the correlation between the higher resolved COSMO-REA6 wind ﬁelds and station observations is highest, for both assimilated and non-assimilated (i.e., independent) observa- tions. As expected from the lower spatial resolution and reduced amount of data assimilated into ERA-20C, the correlation of monthly means decreases somewhat relative to the other reanalyses (in our investigated period of 2007 to 2010). 14th EMS Annual Meeting & 10th European Conference on Applied Climatology (ECAC) 14th EMS Annual Meeting & 10th European Conference on Applied Climatology (ECAC) 14th EMS Annual Meeting & 10th European Con Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ doi:10.5194/asr-12-187-2015 © Author(s) 2015. CC Attribution 3.0 Lice nnual Meeting & 10th European Conference on Applied Clima Comparison of regional and global reanalysis near-surface winds with station observations over Germany ting & 10th European Conference on Applied Climatology (EC 1 Introduction Atmospheric reanalysis is “a consistent reprocessing of archived weather observations using a modern forecasting system” (Dee et al., 2014). The past two decades have brought forth remarkable advances in the quality and quan- tity of the resulting reanalysis datasets (Dee et al., 2014; Bosilovich et al., 2013; Hartmann et al., 2013). These ad- vances have been accompanied (and in many respects catal- ysed) by rapid growth in uptake of reanalysis datasets for a diverse and growing range of applications. But it remains challenging for individual users to know which (if any) of the available datasets (whether from reanaly- sis or other procedures) are appropriate for their applica- tions (see, e.g., Gregow et al., 2015). Assessing “ﬁtness-for- purpose” requires synthesising knowledge about (a) the rel- ative strengths/limitations of the available datasets, (b) the extent to which such characteristics affect the ﬁdelity of derived results, and (c) the user’s own application-speciﬁc tolerance of uncertainty; it is important to note that ﬁtness for one purpose does not guarantee ﬁtness for all purposes. Here, we describe the inter-comparisons that were under- taken to elucidate the relative properties of the near-surface winds over Germany in two global and one regional reanaly- sis datasets on the time scale of days to several years. The inter-comparisons employ independent wind observations from German weather stations and also illustrate how re- analyses can support quality control of the observations. The COSMO-REA6 regional reanalysis is driven by the global ERA-Interim reanalysis, and covers Europe with increased spatial and temporal resolution. We demonstrate that inter- comparison of reanalyses can help to guide users to useful scales of variability. Reanalysis wind ﬁelds have been found to differ from sta- tion records on the multi-decadal scale. The so-called “still- ing effect”, i.e., a decrease of wind speeds deduced from sta- tion measurements since the 1970s, is reported in many pa- pers on mid-latitude wind observations (see McVicar et al., 2012, for a comprehensive review). This is in general con- trast to various reanalyses, which do not show a stilling ef- fect (Smits et al., 2005; McVicar et al., 2008, 2012, and ref- erences therein). Several reasons have been suggested. Long- term changes in surface roughness could explain, at least partly, the difference (Vautard et al., 2010; Wever, 2012). Changes in aerosols, sea surface temperature and greenhouse gas concentrations were found to be unlikely causes (Bichet et al., 2012). A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 188 a state-of-the-art NWP model, constrained by the whole ob- serving system. While larger wind energy companies might be able to build their own wind measurement masts and de- velop sophisticated analyses, other users like smaller enter- prises, district managers, government agencies, or interested individual citizens rely on products based on station mea- surements at 10 m above ground, and statistical models build upon those, like e.g., wind climatologies at levels of interest for wind energy production (Gerth and Christoffer, 1994). All these users might draw beneﬁt from our study. 1 Introduction There still remain the possibilities that measure- ment artefacts, or processes not modelled in the reanalysis, may contribute to the discrepancies. In this work we do not investigate the multi-decadal still- ing effect directly, but report on a preparatory multi-annual study. We focus on the area of Germany and the period 2007 to 2010. Although the dataset is short, it is best ﬁtted for the inter-comparison since all stations are known to operate au- tomatically, instrumental problems are unlikely, and changes in surface roughness are negligible. This allows us to char- acterize the recent variability captured in regional and global reanalysis near surface wind ﬁelds, and to compare their sta- tistical characteristics with the traditional station measure- ments. Several approaches have been identiﬁed as useful for reanalysis comparisons (see Fig. 1) and applied below for the near-surface wind ﬁelds, namely direct comparisons (Sect. 3.1 and 3.2), thematic comparisons (Sect. 3.3) and in- ternal metrics comparisons (Sect. 3.4). For the purpose of wind energy production, for example, there are dedicated wind mast measurements, which mea- sure at a certain point over a short period of time and at heights typical for wind energy production. This is highly valuable for wind energy applications, but it remains de- sirable to add to this short-term information an estimate of the inter-annual temporal variability, as is typically avail- able in the time series of station measurements made for traditional weather/climate applications (approximately 10 m above ground). In Germany, these are relatively abundant and evenly distributed and cover many decades. On the other hand, observation practices, instruments, and the height of the sensor above ground might have changed over time. This information should ideally be found in the metadata accom- panying the historical station records. It is also valuable to track changes in the surroundings (like growing trees or changes in land cover) and measurement or processing er- rors, but this is not uniformly available worldwide given his- torical variations in the rigour of observing practice around the globe. Germany Still, the inter-annual variability connected to the North Atlantic Oscillation (NAO) found in the reanalysis surface wind anomalies is in accordance with the anomalies recorded by the stations. ce on Applied Climatology (ECAC) ed Climatology (ECAC) We discuss some typical examples where differences are found, e.g., where the mean wind distributions differ (probably related to either height or model topography differences) and where the correlations break down (be- cause of unresolved local topography) which applies to a minority of stations. We also identiﬁed stations with homogeneity problems in the reported station values, demonstrating how reanalyses can be applied to support quality control for the observed station data. Finally, as a demonstration of concept, we discuss how comparing feedback ﬁles of the different reanalyses can guide users to useful scales of variability. Published by Copernicus Publications. 2 Data We used the 10 m wind ﬁelds and feedback ﬁles containing the radiosonde data from the global reanalysis ERA-Interim (Dee et al., 2011). The data of this reanalysis are available from the ECMWF website http://apps.ecmwf.int/datasets/. ERA-Interim is based on the ECMWF Integrated Forecast System, IFS (Cy31r2), 4-dimensional variational analysis (4D-Var) with a 12-h analysis window, and assimilation of a wide range of surface and upper-air observational data from both in-situ and satellite instruments. The spatial resolution of the data set is approximately 80 km (T255 spectral) on 60 Reanalyses also provide wind information, here derived from a physically consistent state of the atmosphere using Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observatio 189 Figure 1. Useful approaches for intercomparing reanalyses include: (1) descriptive, e.g., a comparison table of characteristics, (2) direct comparisons with in-situ, ground-based remote sensing or satel- lite observations, (3) thematic comparison, and (4) internal metrics comparison. tation and measurement time series. From this archive (Au- gust 2014, version v002), the example station observations below were chosen. The station observations of wind speeds are not assimilated within ERA-Interim and ERA-20C, i.e., are independent measurements. For COSMO-REA6, station data below 100 m are assimilated, above 100 m they are not, i.e., the former may be dependent, the latter are independent measurements. The North Atlantic Oscillation (NAO) index was taken from http://www.cru.uea.ac.uk/~timo/datapages/naoi. htm which is an extension of Jones et al. (1997). For our comparison, we took the grid cells matching the location of the station from the regional reanalysis, and from the global reanalyses interpolated by 0.125◦(provided by ECMWF). Because it cannot be expected that processes on the sub-daily scale can be resolved in the archived global reanalysis datasets, we compare frequency distributions of daily means. The correlation coefﬁcients were calculated with monthly means, which were calculated from daily means, to avoid the issue of calculating correlations with in- complete time series. Here we restrict ourselves to the time period 2007 to 2010, which is the overlap period of COSMO- REA6 at our disposal. 2 Data In this time span, we are relatively conﬁdent no large changes in surroundings of the stations occurred, the method of measurements remained constant, and no gross errors are expected. As users would especially be interested in information about the past spanning many decades, we include ERA-20C into our preparatory compar- isons because of its length. ERA-Interim is used as a bench- mark. Based on the comparison, we discuss the potential of reanalyses to enhance the traditional approaches for assess- ing wind variability based on the 10 m station winds. Figure 1. Useful approaches for intercomparing reanalyses include: (1) descriptive, e.g., a comparison table of characteristics, (2) direct comparisons with in-situ, ground-based remote sensing or satel- lite observations, (3) thematic comparison, and (4) internal metrics comparison. vertical levels from the surface up to 0.1 hPa. For compari- son with station data, the interpolated ﬁelds (interpolated to 0.125◦) as obtained from the ECMWF website are taken. We also used the ERA-20C reanalysis (Poli et al., 2013; Dee et al., 2014), also accessed through http://apps.ecmwf. int/datasets/. This reanalysis provides a global realization at spectral resolution T159 which corresponds to the horizontal resolution of the reduced Gaussian grid of ∼125 km. The data assimilation method applied is an updated version of the 4D-Var scheme (IFS Cy38r1) and used a 24-h assimila- tion window but it assimilated observations of surface pres- sure and surface marine winds only. From ERA-20C, the 10 m winds as pre-processed by ECMWF and interpolated to 0.125◦horizontal resolution were taken for this comparison study. www.adv-sci-res.net/12/187/2015/ Adv. Sci. Res., 12, 187–198, 2015 3 Results For an illustration of spatial variability, the 10 m surface winds from the regional reanalysis COSMO-REA6 and the interpolated global reanalysis ERA-20C for February 2007 are shown in Fig. 2. The higher wind speeds over the North and Baltic Seas are prominent in both reanalyses as well as the sharp decline at the coast line. As expected, due to the much ﬁner horizontal resolution of the COSMO-REA6 re- gional reanalysis, higher spatial variability is present here. Note the coincidences of wind features with topographical features, e.g., mountain ridges in the South of East Germany (compare to Fig. 3). The features related to topography are more pronounced in the regional reanalysis, e.g., the land- sea transition and the wind speed variability over the hetero- geneous low mountain ranges in Mid-Germany. Note, how- ever, the close match of the average absolute values of both reanalysis datasets. The third reanalysis product used is the COSMO-based re- gional reanalysis of DWD’s Hans-Ertel Centre for Weather Research (HErZ), University of Bonn (Bollmeyer et al., 2014, http://www.herz-tb4.uni-bonn.de/) called COSMO- REA6. The regional reanalysis COSMO-REA6 is driven via the global reanalysis boundary conditions (ERA-Interim) ev- ery 3 h. It is based on the COSMO-model and applies nudg- ing as the data assimilation technique. Here we use data from a conﬁguration that was run for the European CORDEX domain at a nominal resolution of 0.055◦. From COSMO- REA6, the wind speed at the ground layer (10 m) as well as feedback ﬁles containing radiosonde data were used in the work below. The station data and associated metadata were taken from the freely accessible climate data archive of Deutscher Wet- terdienst (DWD) which is available through ftp://ftp-cdc. dwd.de/pub/CDC/observations_germany/climate/ (Kaspar et al., 2013). Available are the station data and associated meta- data describing the particularities of the station instrumen- Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ er-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 190 Figure 2. 10 m surface winds from the COSMO-REA6 regional reanalysis (left) and the ERA-20C interpolated wind ﬁelds (right) illustrate the different spatial resolution for the month of February 2007. Figure 2. 10 m surface winds from the COSMO-REA6 regional reanalysis (left) and the ERA-20C interpolated wind ﬁelds (right) illustrate the different spatial resolution for the month of February 2007. ences in temporal sampling of the reanalyses (3 vs. 6 h for ERA-20C and ERA-Interim respectively). Figure 3. 3 Results Topography of Germany with colour-coded height above sea level, and location of selected stations. For most station locations, the frequency distributions de- rived from the reanalyses match well with the one from the observations (e.g., Nuremberg, see Fig. 4). This might come as a surprise, given the rather coarse grid resolutions, and that the 10 m winds are expected to be strongly inﬂuenced by the local scale topography. Some stations show an offset in wind speed, due to the model height differing from the real topographic height, e.g., mountain stations like Feldberg- Schwarzwald (Fig. 5). For some applications, ﬁtness-for- purpose would therefore be improved by determining a more representative model height in cases of complex topography. Local inﬂuences of topography can result in pronounced dif- ferences in the frequency distributions. As an extreme ex- ample, the station Garmisch-Partenkirchen (Fig. 6) is shown, which is located in a valley for which the restricted repre- sentativity of the station cannot match the scales resolved in the reanalyses. Generally, only few histogram mismatches are found (and less for the higher resolved COSMO-REA6 reanalysis). An unexpected outcome was that some dubious station data could be identiﬁed: Fig. 7 shows an unusually large proportion of zero wind-speed values; it seems that these were in fact reported on occasions where no observa- tions were performed. This illustrates the use of reanalysis data to support quality control for observed wind data. Wind energy applications often require Weibull parame- ters, thus Weibull distributions with two parameters had been ﬁtted to the frequency distributions: Figure 3. Topography of Germany with colour-coded height above sea level, and location of selected stations. P (ν) = k c ν c k−1 e−( ν c )k, (1) (1) 3.1 Frequency distributions We compare the frequency distributions of daily-mean wind speed from the reanalyses with the distribution from the sta- tion observations, using the respective interpolated reanalysis grid cells containing the geographical location of the station. Daily-mean diagnostics were chosen to accommodate differ- where v is the wind speed (ν ≥0), k is the shape parame- ter (k > 0), and c is the scale parameter of the distribution (c > 0). The Weibull parameters and error estimates are cal- culated with the maximum likelihood method implemented in the Cran R-project package ﬁtdistrplus (Venables and Rip- Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 191 Figure 4. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the station Nuremberg (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Figure 4. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the station Nuremberg (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Figure 5. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the station Feldberg-Schwarzwald (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Figure 5. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the station Feldberg-Schwarzwald (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ Weiss et al.: Comparison of reanalysis near-surface winds with station observations 192 Figure 6. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the example station Garmisch-Partenkirchen (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Figure 6. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the example station Garmisch-Partenkirchen (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Figure 7. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the station Selb-Oberfranken (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Figure 7. Frequency distribution of daily mean wind speeds in the period 2007 to 2010 for the station Selb-Oberfranken (top left) and matching grid cells of ERA-20C (top right), COSMO-REA6 (bottom left), and ERA-Interim wind ﬁelds (bottom right). Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 193 Figure 8. Pearson correlation coefﬁcient of monthly mean wind speeds (calculated from daily wind speeds) between station observations and reanalyses: ERA-20C (left), ERA-Interim (middle), and COSMO-REA6 (right). Stations not assimilated in COSMO-REA6 are highlighted with grey circles. Figure 8. Pearson correlation coefﬁcient of monthly mean wind speeds (calculated from daily wind speeds) between station observations and reanalyses: ERA-20C (left), ERA-Interim (middle), and COSMO-REA6 (right). Stations not assimilated in COSMO-REA6 are highlighted with grey circles. Figure 9. Scatterplots of Pearson correlation coefﬁcients for stations that are not assimilated in COSMO-REA6 (left) and stations that are assimilated (right). Correlations are calculated between monthly mean wind speeds of station observations and COSMO-REA6 (y axis) and compared to correlations between station observations and ERA-Interim (x axis), respectively. Three of the total 210 stations are outliers (correlation below 0.5) and are not shown. Figure 9. Scatterplots of Pearson correlation coefﬁcients for stations that are not assimilated in COSMO-REA6 (left) and stations that are assimilated (right). Correlations are calculated between monthly mean wind speeds of station observations and COSMO-REA6 (y axis) and compared to correlations between station observations and ERA-Interim (x axis), respectively. Three of the total 210 stations are outliers (correlation below 0.5) and are not shown. ley, 2010; Delignette-Muller et al., 2014). Weiss et al.: Comparison of reanalysis near-surface winds with station observations The Weibull pa- rameters for selected stations are compared in Tables 1 and 2, with the dependence on station altitude (listed in Table 3) clearly prominent. For Hamburg-Fuhlsbüttel, the Weibull pa- rameters ks (shape) and cs (scale) of the station observa- tions match well with the ones from reanalyses, also for Potsdam and Nuremberg. Mountain stations like Feldberg- Schwarzwald have signiﬁcantly different c (scale) parame- ters. Here a more representative height would be needed to be determined for each reanalysis. The situations in Garmisch- Partenkirchen and Selb-Oberfranken are different: The sus- pect histogram of the station data from Selb-Oberfranken (Fig. 7) causes a deviation of ks (shape) and cs (scale) due to issues with the missing values of the station recordings. For stations in complex topography, we expect signiﬁcant dif- ferences in Weibull parameters due to the difference of real topography and model topography and the coarse reanaly- sis resolution. A close match of Weibull parameters (derived from stations and from reanalyses) could still be due to a favourable match of height above ground and above model ground. Thus we need to examine the correlation coefﬁcients to judge whether reanalysis winds and station observations are in accordance. For instance, the Weibull parameters of the station Garmisch-Partenkirchen are not close, but still comparable to the ones from reanalysis (see Tables 1 and 2), but there is no correlation found between the time series of monthly means (see Table 3) which can be explained by its location in a valley. 3.2 Correlation of reanalysis data with monthly station data A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observation Still, the correlations over time do hold (see Table 3) for several exposed mountain stations, like Feldberg-Schwarzwald. 82 % of stations show a correlation coefﬁcient greater than or equal to 0.8 and 47 % greater than or equal to 0.9. For ERA- Interim, these correlations are achieved at more stations (89 and 66 %, respectively), and yet more in COSMO-REA6 (96 and 80 %, respectively). The COSMO-REA6 reanalysis as- similates station observations below 100 m height; whether locations contribute assimilated observations or not can be distinguished in Figs. 8 and 9. In our case, 130 stations are not assimilated, i.e., are independent from COSMO-REA6. These independent stations show generally an improvement of correlation with COSMO-REA6 (Fig. 9). In Fig. 8 it can be seen that the regional reanalysis improves monthly cor- relations where it can be expected, namely in the areas with more complex topography (South of Germany, towards the Alps and Bayrischer Wald in the South-West). Three sta- tions in complex terrain (Garmisch-Partenkirchen is one of them) have no correlation with the reanalyses and remain outside the scales of Fig. 9. Still, the correlations over time do hold (see Table 3) for several exposed mountain stations, like Feldberg-Schwarzwald. mate and its spatial extent are of special interest. The North Atlantic Oscillation (NAO) (Walker and Bliss, 1932; Hurell, 1995), is the main synoptic mode of atmospheric circulation and climate variability for Germany. Thus it is of interest, what the magnitude of the NAO-related effect is for the sta- tion data and to what extend the NAO-related effect is cap- tured in the surface ﬁelds of the reanalysis over Germany. For the recent years, the correlations of the stations and reanalysis ﬁelds are high. The question is, whether the corre- lation caused by NAO-related effects holds also for the past, keeping in mind the reanalysis systems might have long- term variability connected to changes of the observing sys- tem (Dee, 2005; Dee and Uppala, 2009). To describe the magnitude of the effect, we select seasonal averages for se- lected seasons corresponding to high, neutral, and low values of NAO index, and calculate the anomalies relative to the av- erage of the period 2007 to 2010. This is done with the inter- polated ERA-20C near-surface ﬁelds. The station anomalies are calculated analogously, i.e., their reference is also derived from the station averages. In Fig. A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observation 10, we show examples for seasons from 2000 and 2002. The NAO-related anomalies show a west-east gradient (left), or a north-south gradient (middle and right) over Germany. The ﬁrst panel (left) corre- sponds to a neutral NAO index autumn with higher than usual wind speeds over the western part and lower wind speeds in the East. With a negative NAO index, the average wind pat- tern is shifted southwards, leaving the North Sea area with lower wind speeds on average (middle panel). For the case of a positive NAO index, the wind pattern is shifted towards Single stations only sometimes show statistically signif- icant differences between the correlations with the three re- analyses, but the combined effect of the 210 stations becomes clear in Fig. 8, and especially in Fig. 9. A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observation 194 Table 1. Weibull parameters k (shape) derived from selected station daily mean wind observations (ks), and from ERA-20C, ERA-Interim, and COSMO-REA6 interpolated grid cells matching the station locations, together with their 1 σ standard deviations (SD). Table 1. Weibull parameters k (shape) derived from selected station daily mean wind observations (ks), and from ERA-20C, ERA-Interim, and COSMO-REA6 interpolated grid cells matching the station locations, together with their 1 σ standard deviations (SD). ks k_ERA-20C k_ERA-Interim k_COSMO-REA6 (SD) (SD) (SD) (SD) Hamburg-Fuhlsbüttel 2.65 (0.05) 2.61 (0.05) 2.56 (0.05) 2.62 (0.05) Potsdam 2.89 (0.05) 2.56 (0.05) 2.48 (0.05) 2.73 (0.05) Nuremberg 2.43 (0.04) 2.32 (0.04) 2.25 (0.04) 2.35 (0.04) Selb-Oberfranken 1.68 (0.03) 2.36 (0.04) 2.26 (0.04) 2.56 (0.05) Garmisch-Partenkirchen 3.40 (0.06) 3.26 (0.06) 2.94 (0.05) 2.17 (0.04) Feldberg-Schwarzwald 2.26 (0.05) 2.55 (0.05) 2.07 (0.04) 2.07 (0.04) Table 2. Weibull parameters c (scale) derived from selected station daily mean wind observations (cs), and from ERA-20C, ERA-Interim, and COSMO-REA6 interpolated grid cells matching the station locations, together with their 1 σ standard deviations (SD). cs c_ERA-20C c_ERA-Interim c_COSMO-REA6 (SD) (SD) (SD) (SD) Hamburg-Fuhlsbüttel 4.46 (0.05) 4.49 (0.05) 4.77 (0.05) 3.83 (0.04) Potsdam 4.78 (0.05) 3.90 (0.04) 4.44 (0.05) 3.40 (0.03) Nuremberg 3.39 (0.04) 3.50 (0.04) 3.99 (0.05) 2.79 (0.03) Selb-Oberfranken 1.72 (0.03) 3.66 (0.04) 4.17 (0.05) 3.73 (0.04) Garmisch-Partenkirchen 1.73 (0.01) 2.03 (0.02) 2.49 (0.02) 1.92 (0.02) Feldberg-Schwarzwald 9.50 (0.12) 2.80 (0.03) 3.40 (0.05) 3.69 (0.05) 82 % of stations show a correlation coefﬁcient greater than or equal to 0.8 and 47 % greater than or equal to 0.9. For ERA- Interim, these correlations are achieved at more stations (89 and 66 %, respectively), and yet more in COSMO-REA6 (96 and 80 %, respectively). The COSMO-REA6 reanalysis as- similates station observations below 100 m height; whether locations contribute assimilated observations or not can be distinguished in Figs. 8 and 9. In our case, 130 stations are not assimilated, i.e., are independent from COSMO-REA6. These independent stations show generally an improvement of correlation with COSMO-REA6 (Fig. 9). In Fig. 8 it can be seen that the regional reanalysis improves monthly cor- relations where it can be expected, namely in the areas with more complex topography (South of Germany, towards the Alps and Bayrischer Wald in the South-West). Three sta- tions in complex terrain (Garmisch-Partenkirchen is one of them) have no correlation with the reanalyses and remain outside the scales of Fig. 9. 3.2 Correlation of reanalysis data with monthly station data The correlation of monthly wind speed from the ERA-20C, ERA-Interim, and COSMO-REA6 reanalyses with the 210 stations over Germany were calculated with the Pearson cor- relation for the period 2007 to 2010. In case of ERA-20C, www.adv-sci-res.net/12/187/2015/ Adv. Sci. Res., 12, 187–198, 2015 K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations Our con- clusion here is that the inter-annual variability in the near sur- face winds captured in the reanalyses is consistent with the station time series, considering a time scale of several years. We can detect this common NAO-pattern (in both reanaly- sis and station data) in recent years. To extend our studies to longer time scales, we would have to separate the effect the North (left panel). Note that, allowing for some scat- ter among the stations, they generally match the spatial ex- tent and the magnitude of the anomaly present in the ERA- 20C ﬁelds. The examples picked here have months with a similar NAO-index, and a rather pronounced NAO anomaly. Of course, the inter-annual wind variability is dependent not only on the NAO strength, but also on details of the jet stream position, and the individual storm tracks, when moving to shorter time scales and selected station positions. Our con- clusion here is that the inter-annual variability in the near sur- face winds captured in the reanalyses is consistent with the station time series, considering a time scale of several years. We can detect this common NAO-pattern (in both reanaly- sis and station data) in recent years. To extend our studies to longer time scales, we would have to separate the effect of instrumental changes at the stations from possible effects of NAO-related variability. For instance, Welker and Martius (2015) found a positive correlation between the NAO-index and occurrence of high wind speeds over Switzerland, de- duced from the global 20CR reanalysis (Compo et al., 2011) on the inter-annual time scale. Welker and Martius (2015) found an increase of the correlation at the decadal time scale; and point to a change of strength of this correlation over time due to an eastward shift of the NAO pattern. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 195 Table 3. Pearson correlation coefﬁcient together with its 1 σ-conﬁdence interval, between mean monthly wind speeds from selected station observations and matching grid cells from interpolated ERA-20C, interpolated ERA-Interim, and COSMO-REA6. Note that for ERA-20C and ERA-Interim, all station observations are independent. For COSMO-REA6, station observations are assimilated (nudged) as indicated in the last column. Table 3. Pearson correlation coefﬁcient together with its 1 σ-conﬁdence interval, between mean monthly wind speeds from selected station observations and matching grid cells from interpolated ERA-20C, interpolated ERA-Interim, and COSMO-REA6. Note that for ERA-20C and ERA-Interim, all station observations are independent. For COSMO-REA6, station observations are assimilated (nudged) as indicated in the last column. Station Altitude r_ERA-20C r_ERA-Interim r_COSMO-REA6 nudged (conf. int.) (conf. int.) (conf. int.) Hamburg-Fuhlsbüttel 11 0.93 0.96 0.98 yes (0.90, 0.94) (0.95, 0.97) (0.97, 0.98) Potsdam 81 0.94 0.96 0.98 yes (0.92, 0.96) (0.95, 0.97) (0.98, 0.99) Nuremberg 314 0.93 0.92 0.97 no (0.90, 0.94) (0.89, 0.94) (0.96, 0.98) Selb-Oberfranken 535 0.86 0.89 0.89 no (0.82, 0.90) (0.86, 0.92) (0.85, 0.91) Garmisch-Partenkirchen 719 −0.23 −0.11 −0.04 no (−0.36, −0.09) (−0.38, 0.18) (−0.18, 0.12) Feldberg-Schwarzwald 1490 0.90 0.89 0.93 no (0.87, 0.92) (0.85, 0.91) (0.91, 0.95) Figure 10. Seasonal anomalies of the ERA-20C 10 m wind speeds compared to the reference period 2007 to 2010 for September, October, November (SON) of 2000 at neutral NAO-index (left panel), SON2002 at negative NAO-index (middle), and December, January, February (DJF) 1999/2000 at positive NAO-index (right). Figure 10. Seasonal anomalies of the ERA-20C 10 m wind speeds compared to the reference period 2007 to 2010 for September, October, November (SON) of 2000 at neutral NAO-index (left panel), SON2002 at negative NAO-index (middle), and December, January, February (DJF) 1999/2000 at positive NAO-index (right). the North (left panel). Note that, allowing for some scat- ter among the stations, they generally match the spatial ex- tent and the magnitude of the anomaly present in the ERA- 20C ﬁelds. The examples picked here have months with a similar NAO-index, and a rather pronounced NAO anomaly. Of course, the inter-annual wind variability is dependent not only on the NAO strength, but also on details of the jet stream position, and the individual storm tracks, when moving to shorter time scales and selected station positions. 3.3 Interannual variability related to NAO For many users, notably renewable energy users, the spatial characteristics and inter-annual variability of weather and cli- www.adv-sci-res.net/12/187/2015/ Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/ 3.4 Comparing feedback statistics The opportunities to compare reanalyses against stable time series of fully-independent observations covering a long time-period are limited. Fortunately, the reanalysis process www.adv-sci-res.net/12/187/2015/ Adv. Sci. Res., 12, 187–198, 2015 Adv. Sci. Res., 12, 187–198, 2015 A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 196 Figure 11. Root mean square errors of analysis departures (o-a, solid line) and background departures (o-b, dashed line) for Lin- denberg radiosonde observations of wind speed at standard pres- sure levels, in January 2011: feedback statistics for the ERA-Interim (black) and COSMO-REA6 (red). agated for a lead time of 3 h. In addition, the sounding data that have inﬂuenced the initial conditions for the background forecasts in both systems are comparable. At leading order, the regional and global reanalyses both match the sound- ing observations to the same extent, for the statistical sam- ple investigated here. The match is not uniform, for example ERA-Interim o-b exhibits large values around 400 hPa while COSMO o-b exceeds ERA-Interim o-b between 1000 and 700 hPa. While the limited sample shown here does not per- mit such discrepancies to be deﬁnitively traced to speciﬁc system deﬁciencies, this example highlights a more general issue: namely, that relatively few users have the tools and skills to identify such discrepancies, or to make informed as- sessments of any consequent impacts in their particular ap- plication, and hence lack clarity on whether a dataset is ﬁt for their speciﬁc purpose. We believe that these are important is- sues that warrant attention in the context of current efforts to develop climate services. Figure 11. Root mean square errors of analysis departures (o-a, solid line) and background departures (o-b, dashed line) for Lin- denberg radiosonde observations of wind speed at standard pres- sure levels, in January 2011: feedback statistics for the ERA-Interim (black) and COSMO-REA6 (red). To draw conclusion on the performance of the different systems, a comparison of observations to background (o-b) is easier to interpret than observations to analysis (o-a), be- cause the background can be regarded, in ﬁrst approximation (in the absence of time-correlated errors), as rather indepen- dent from the observations. From Fig. 11 we can see that the background of ERA-Interim and COSMO-REA6 are similar, but differ to up to 25 %. Both analyses are drawn closer to the observations, which shows the effect of the data assimilation for this example. 3.4 Comparing feedback statistics Comparing the resulting analyses with each other is harder, because the analysis depends on the observa- tions to a degree that varies between the different assimila- tion systems. For more conclusions, longer periods of feed- back statistics with comparable forecast lead times should be analysed, which were not available at the time of writing. provides opportunities to use the assimilated observations in a quasi-independent manner, by comparing these observa- tions against the free forecasts (or background ﬁelds) which are started from the preceding (re-)analysis window. These so-called feedback statistics are in fact routinely produced by the data assimilation system, and relate assimilated observa- tions, free forecasts (i.e., background ﬁelds), analysis incre- ments, and analysis ﬁelds to each other. They yield valuable additional information, e.g., estimates of the analysis error or diagnostics on systematic changes in increments which, if traced to biases in observations or model (or both), indicate deﬁciencies in the system. Favourable statistics may show that the frequency distribution and time series of observed and reanalysed parameters are matching. Thus, it is poten- tially of high practical value for the user to take into account the results of feedback statistics. www.adv-sci-res.net/12/187/2015/ 4 Summary and conclusions Here we inter-compared the frequency distributions, Weibull parameters and monthly to inter-annual variability of re- gional and global reanalysis near surface wind ﬁelds of the recent years (2007–2010). We illustrated three approaches, namely direct comparison (with independent station mea- surements), thematic comparison (the magnitude and spatial extend of NAO-related mean wind speed anomalies) and in- ternal metrics (feedback statistics). As a demonstration of concept, here we analyse the differences between observations (o), background (b), and (re-)analysis (a) of wind speed for a global (ERA-Interim) and the COSMO-REA6 regional reanalysis, for the Linden- berg radiosondes launched at 06Z (see Fig. 11). Care must be taken to make the differences arising from different re- analysis systems as comparable as possible. In particular, the uncertainties of the background ﬁelds from the reanal- ysis systems should be made as similar as possible. These uncertainties depend on the initial condition uncertainty and the growth of this uncertainty with forecast lead time. The forecast lead times of both systems should be comparable. In our example, COSMO background ﬁelds valid at 06Z are taken from forecasts initialized at 0Z and propagated for a lead time of 6 h. The most comparable ERA-Interim back- ground ﬁelds are from forecasts initialized at 3Z and prop- The 2 σ conﬁdence intervals of the Weibull parameters de- rived from the reanalysis wind ﬁelds overlap the ones de- rived from the stations in most cases for the Weibull shape parameter (k), whereas the Weibull scale parameters (c) dif- fer more. With respect to correlations, generally the ERA- 20C ﬁts the station data well, ERA-Interim ﬁts better, and the regional COSMO-REA6 reanalysis is closest to the sta- tion data. This is valid for the daily, monthly, and seasonal scale, for the period investigated here (2007 to 2010). It has to be kept in mind that the station observations situated be- www.adv-sci-res.net/12/187/2015/ Adv. Sci. Res., 12, 187–198, 2015 A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 197 low 100 m a.s.l. are assimilated within COSMO-REA6. The increased correlation with COSMO-REA6 holds true regard- less whether or not a station contributes observations that are assimilated in COSMO-REA6. References Bichet, A., Wild, M., Folini, D., and Schär, C.: Causes for decadal variations of wind speed over land: Sensitivity studies with a global climate model. Geophys. Res. Lett., 39, L11701, doi:10.1029/2012GL051685, 2012. The correlation holds for ERA-20C, even though fewer observations are assimilated than in ERA-Interim and the spatial resolution is signiﬁcantly reduced. The correlation with COSMO-REA6 reanalysis is highest, because of sev- eral, and possibly combined, reasons: the higher resolution of the model, the regional data assimilation (nudging), and possibly also the higher temporal resolution of the output. It should be kept in mind that COSMO-REA6 is forced ev- ery 3 h by ERA-Interim, and we compare aggregated daily means from hourly COSMO-REA6 output with daily means from 6-hourly ERA-Interim output. While skill is inherited from ERA-Interim via boundary conditions, we get the addi- tional beneﬁt from COSMO-REA6, possibly also from the temporal resolution. Particular the stations with low cor- relations show a strong improvement, pointing to the ex- pected difference in the reanalysis performance, namely that the smaller scales which are resolved with COSMO-REA6 are of importance for the monthly mean near surface wind speeds, i.e., hinting to that a simple scaling of global reanal- ysis would not give the same information (on the daily to monthly time scale). Bollmeyer, C., Keller, J. D., Ohlwein, C., Wahl, S., Crewell, S., Friederichs, P., Hense, A., Keune, J., Kneifel, S., Pscheidt, I., Redl, S., and Steinke, S.: Towards a high-resolution regional re- analysis for the European CORDEX domain, Q. J. Roy. Meteo- rol. Soc., 141, 1–15, doi:10.1002/qj.2486, 2014. Bosilovich, M. G., Kennedy, J., Dee, D., Allan, R., and O’Neill, A.: On the Reprocessing and Reanalysis of Observations for Cli- mate. Climate Science for Serving Society: Research, Modelling and Prediction Priorities, edited by: Asrar, G. and Hurrell, J. W., Springer Netherlands, 51–71, 2013. Compo, G. P., Whitaker, J. S., Sardeshmukh, P. D., Matsui, N., Al- lan, R. J., Yin, X., Gleason, B. E., Vose, R. S., Rutledge, G., Bessemoulin, P., Brönnimann, S., Brunet, M., Crouthamel, R. I., Grant, A. N., Groisman, P. Y., Jones, P. D., Kruk, M. C., Kruger, A. C., Marshall, G. J., Maugeri, M., Mok, H. Y., Nordli, Ø., Ross, T. F., Trigo, R. M., Wang, X. L., Woodruff, S. D., and Worley, S. J.: The Twentieth Century Reanalysis Project. Q. J. Roy. Meteo- rol. Soc., 137, 1–28, doi:10.1002/qj.776, 2011. Dee, D. P.: Bias and data assimilation, Q. 4 Summary and conclusions Noticeable improvement for COSMO-REA6 correlations are found across Germany, for instance at stations in Southern Germany, closer to the Alps (not assimilated in COSMO-REA6), and at the coastline (as- similated in COSMO-REA6). Author contributions. The ideas and methodologies were devel- oped by Andrea Kaiser-Weiss, Frank Kaspar, David Tan, Paul Poli and Hilppa Gregow during the CORE-CLIMAX project. Analy- ses in the manuscript were conducted by Andrea Kaiser-Weiss, Frank Kaspar, Vera Heene, Michael Borsche and Andre Obregon. Andrea Kaiser-Weiss prepared the manuscript with contribution from all co-authors. As expected from the fact that the regional COSMO- REA6 reanalysis is driven by the global ERA-Interim reanal- ysis via boundary condition, the feedback statistics exhibit a similar ﬁt to the Lindenberg radio soundings for both reanal- ysis systems for our example month of January 2011. Acknowledgements. Andre Obregon and Vera Heene were supported through the CORE-CLIMAX (grant no. 313085 within the EU Seventh Framework Programme). Michael Borsche is supported by the UERRA project (grant no. 607193 within the EU Seventh Framework Programme). We thank Christoph Bollmeyer and Liselotte Bach from University of Bonn for the provision of the COSMO-REA6 feedback ﬁles. We thank Karsten Friedrich (DWD) for his help with the ﬁgures. We would like to acknowledge helpful discussions with Hermann Mächel (DWD). We thank our two reviewers for their constructive criticism which improved the manuscript. We demonstrated that although local and regional effects can be expected to determine the variability in wind ﬁelds measured at the stations (10 m over ground), the frequency distributions of mean wind speeds match quite well with the ones of the reanalysis ﬁelds for which such effects are at the sub-grid scale. The correlation of ERA-20C, ERA-Interim and COSMO-REA6 monthly means with station observa- tions is high (> 0.8) for the majority of the German stations. Thus we conclude that the monthly and seasonal anomalies recorded at these stations can be understood as representa- tive for a wider spatial area, comparable to the resolution of the reanalyses, at least for the recent years. Due to the short- ness of period, we cannot make such a statement concerning inter-annual variability, though the NAO-related anomalies indicate there is also coherence at the inter-annual scale. Edited by: E. Bazile y Reviewed by: T. Landelius and one anonymous referee www.adv-sci-res.net/12/187/2015/ References Soc., 95, 1235–1248, doi:10.1175/BAMS-D-13-00043.1, 2014. Delignette-Muller, M., Pouillot, R., Denis, J., and Dutang, C.: _tdis- trplus: Help to Fit of a Parametric Distribution to Non-Censored or Censored Data, R package version 1.0-2, available at: http: //cran.r-project.org (last access: 1 April 2015), 2014. Poli, P., Hersbach, H., Tan, D., Dee, D., Thépaut, J.-N., Simmons, A., Peubey, C., Laloyaux, P., Komori, T., Berrisford, P., Dragani, R., Trémolet, Y., Holm, E., Bonavita, M., Isaksen, L., and Fisher, M.: The data assimilation system and initial performance evalu- ation of the ECMWF pilot reanalysis of the 20th-century assim- ilating surface observations only (ERA-20C), ERA report series No. 14, European Centre for Medium-Range Weather Forecasts, Reading, England, 2013. Gerth, W. P. and Christoffer, J.: Windkarten von Deutschland, Me- teorolog. Zeitschrift, N.F., 3, 67–77, 1994. Gregow, H., Poli, P., Mäkelä, H. M., Jylhä, K., Kaiser-Weiss, A. K., Obregon, A., Tan, D. G. H., Kekki, S., and Kaspar, F.: User awareness concerning feedback data and input observations used in reanalysis systems, Adv. Sci. Res., 12, 63–67, doi:10.5194/asr- 12-63-2015, 2015. Smits, A., Klein-Tank, A. M. G., and Können, G. P.: Trends in storminess over the Netherlands, 1962–2002, Int. J. Climatol., 25, 1331–1344, doi:10.1002/joc.1195, 2005. Hartmann, D. L., Klein Tank, A. M. G., Rusticucci, M., Alexan- der, L. V., Brönnimann, S., Charabi, Y., Dentener, F. J., Dlugo- kencky, E. J., Easterling, D. R., Kaplan, A., Soden, B. J., Thorne, P. W., Wild, M., and Zhai, P. M.: Observations: Atmosphere and Surface, in: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2013. Vautard, R., Cattiaux, J., Yiou, P., Thépaut, J.-N., and Ciais, P.: Northern Hemisphere atmospheric stilling partly attributed to an increase in surface roughness, Nature GeoScience, 3, 756–761, 2010. Venables, W. N. and Ripley, B. D.: Modern Applied Statistics with S, 4th Edn., Springer-Verlag, 2010. Walker, G. T. and Bliss, E. W.: World Weather V, Mem. R. Meteorol. Soc., 4, 53–84, 1932. Hurrell, J. W.: Decadal trends in the North Atlantic oscillation: regional temperature and precipitation, Science, 269, 676–697, 1995. Welker, C. and Martius, O.: Large-scale atmospheric ﬂow con- ditions and sea surface temperatures associated with haz- ardous winds in Switzerland, Clim. Dynam., 44, 1857–1869, doi:10.1007/s00382-014-2404-1, 2015. Jones, P. References J. Roy. Meteorol. Soc., 131, 3323–3343, doi:10.1256/qj.05.137, 2005. Dee, D. P. and Uppala, S.: Variational bias correction of satellite radiance data in the ERA-Interim reanalysis, Q. J. Roy. Meteorol. Soc., 135, 1830–1841, doi:10.1002/qj.493, 2009. Further analysis would be needed to identify under which circumstances the differences are most pronounced, and to what extent the low-frequency information from global re- analyses is represented or improved with regional reanalyses for regional-scale parameters. Dee, D. P., Uppala, S. M., Simmons, A. J., Berrisford, P., Poli, P., Kobayashi, S., Andrae, U., Balmaseda, M. A., Balsamo, G., Bauer, P., Bechtold, P., Beljaars, A. C. M., van de Berg, L., Bid- lot, J., Bormann, N., Delsol, C., Dragani, R., Fuentes, M., Geer, A. J., Haimberger, L., Healy, S. B., Hersbach, H., Hólm, E. V., www.adv-sci-res.net/12/187/2015/ Adv. Sci. Res., 12, 187–198, 2015 A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations A. K. Kaiser-Weiss et al.: Comparison of reanalysis near-surface winds with station observations 198 Isaksen, L., Kållberg, P., Köhler, M., Matricardi, M., McNally, A. P., Monge-Sanz, B. M., Morcrette, J.-J., Park, B.-K., Peubey, C., de Rosnay, P., Tavolato, C., Thépaut, J.-N., and Vitart, F.: The ERA-Interim reanalysis: conﬁguration and performance of the data assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553– 597 doi:10 1002/qj 828 2011 Isaksen, L., Kållberg, P., Köhler, M., Matricardi, M., McNally, A. P., Monge-Sanz, B. M., Morcrette, J.-J., Park, B.-K., Peubey, McVicar, T. R., Van Niel, T. G., Li, L. T., Roderick, M. L., Rayner, D. P., Ricciardulli, L., and Donohue R. J., Wind speed climatology and trends for Australia, 1975–2006: Cap- turing the stilling phenomenon and comparison with near- surface reanalysis output, Geophys. Res. Lett., 35, L20403, doi:10.1029/2008GL035627, 2008. C., de Rosnay, P., Tavolato, C., Thépaut, J.-N., and Vitart, F.: The ERA-Interim reanalysis: conﬁguration and performance of the data assimilation system, Q. J. Roy. Meteorol. Soc., 137, 553– 597, doi:10.1002/qj.828, 2011. McVicar, T. R., Roderick, M. L., Donohue, R. J., Li, L. T., Van Niel, T. G., Thomas, A., Grieser, J., Jhajharia, D., Himri, Y., Ma- howald, N. M., Mescherskaya, A. V., Kruger, A. C., Rehman, S., and Dinpashoh, Y.: Global review and synthesis of trends in observed terrestrial near-surface wind speeds: Implications for evaporation, J. Hydrol., 416–417, 182–205, 2012. Dee, D. P., Balmaseda, M., Balsamo, G., Engelen, R., Simmons, A. J., and Thepaut, J.-N.: Towards a consistent reanalysis of the climate system, B. Am. Meteorol. www.adv-sci-res.net/12/187/2015/ References D., Jónsson, T., and Wheeler, D.: Extension to the North Atlantic Oscillation using early instrumental pressure observa- tions from Gibraltar and South-West Iceland, Int. J. Climatol., 17, 1433–1450, 1997. Wever, N.: Quantifying trends in surface roughness and the effect on surface wind speed observations, J. Geophys. Res., 117, D11104, doi:10.1029/2011JD017118, 2012. Kaspar, F., Müller-Westermeier, G., Penda, E., Mächel, H., Zim- mermann, K., Kaiser-Weiss, A., and Deutschländer, T.: Monitor- ing of climate change in Germany – data, products and services of Germany’s National Climate Data Centre, Adv. Sci. Res., 10, 99–106, doi:10.5194/asr-10-99-2013, 2013. Adv. Sci. Res., 12, 187–198, 2015 www.adv-sci-res.net/12/187/2015/
https://openalex.org/W2802115176	https://biolres.biomedcentral.com/track/pdf/10.1186/s40659-018-0158-7	English	null	Independent prognostic genes and mechanism investigation for colon cancer	Biological Research	2,018	cc-by	6,375	Propose: We aimed to explore the potential molecular mechanism and independent prognostic genes for colon cancer (CC). Propose: We aimed to explore the potential molecular mechanism and independent prognostic genes for colon cancer (CC). Propose: We aimed to explore the potential molecular mechanism and independent prognostic genes for colon cancer (CC). Methods: Microarray datasets GSE17536 and GSE39582 were downloaded from Gene Expression Omnibus. Mean- while, the whole CC-related dataset were downloaded from The Cancer Genome Atlas (TCGA) database. Differentially expressed mRNA (DEMs) were identified between cancer tissue samples and para-carcinoma tissue samples in TCGA dataset, followed by the KEGG pathway and GO function analyses. Furthermore, the clinical prognostic analysis includ- ing overall survival (OS) and disease-free survival (DFS) were performed in all three datasets. Results: A total of 633 up- and 321 down-regulated mRNAs were revealed in TCGA dataset. The up-regulated mRNAs were mainly assembled in functions including extracellular matrix and pathways including Wnt signaling. The down- regulated mRNAs were mainly assembled in functions like Digestion and pathways like Drug metabolism. Further- more, up-regulation of UL16-binding protein 2 (ULBP2) was associated with OS in CC patients. A total of 12 DEMs including Surfactant Associated 2 (SFTA2) were potential DFS prognostic genes in CC patients. Meanwhile, the GRP and Transmembrane Protein 37 (TMEM37) were two outstanding independent DFS prognostic genes in CC. Results: A total of 633 up- and 321 down-regulated mRNAs were revealed in TCGA dataset. The up-regulated mRNAs were mainly assembled in functions including extracellular matrix and pathways including Wnt signaling. The down- regulated mRNAs were mainly assembled in functions like Digestion and pathways like Drug metabolism. Further- more, up-regulation of UL16-binding protein 2 (ULBP2) was associated with OS in CC patients. A total of 12 DEMs including Surfactant Associated 2 (SFTA2) were potential DFS prognostic genes in CC patients. Meanwhile, the GRP and Transmembrane Protein 37 (TMEM37) were two outstanding independent DFS prognostic genes in CC. Conclusions: ULBP2 might be a potential novel OS prognostic biomarker in CC, while GRP and TMEM37 could be served as the independent DFS prognostic genes in CC. Furthermore, functions including extracellular matrix and digestion, as well as pathways including Wnt signaling and drug metabolism might play important roles in the pro- cess of CC. Conclusions: ULBP2 might be a potential novel OS prognostic biomarker in CC, while GRP and TMEM37 could be served as the independent DFS prognostic genes in CC. © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Propose: We aimed to explore the potential molecular mechanism and independent prognostic genes for colon cancer (CC). Furthermore, functions including extracellular matrix and digestion, as well as pathways including Wnt signaling and drug metabolism might play important roles in the pro- cess of CC. Keywords: Colon cancer, Differentially expressed microRNAs, Function and pathway analysis, Independent prognostic gene, Overall survival, Disease-free survival Keywords: Colon cancer, Differentially expressed microRNAs, Function and pathway analysis, Independent prognostic gene, Overall survival, Disease-free survival Chunsheng Li1, Zhen Shen1, Yangyang Zhou2 and Wei Yu1* Chunsheng Li1, Zhen Shen1, Yangyang Zhou2 and Wei Yu1* Chunsheng Li1, Zhen Shen1, Yangyang Zhou2 and Wei Yu1* Li et al. Biol Res (2018) 51:10 https://doi.org/10.1186/s40659-018-0158-7 Li et al. Biol Res (2018) 51:10 https://doi.org/10.1186/s40659-018-0158-7 Biological Research Background not suffice as a preventative measure because of the high risk of rectal cancer if the rectum remains [6]. Colon cancer (CC) is one of the best-understood neo- plasms from a genetic perspective [1]. Globally, CC is the third most common type of cancer making up about 10% of all cases [2]. There are over 1.4 million new cases and 694,000 deaths from the CC in 1 year worldwide [3]. Treatments for CC may include some combination of surgery, radiation therapy, chemotherapy and targeted therapy [4, 5]. Although the integrated surgical strategies increased the survival rate, the removal of the colon may Numerous data indicate that the aberrant accumula- tion of genetic changes functions as vital roles in initia- tion and development of colon and rectal cancer [7, 8]. Messenger RNA (mRNAs) are important regulatory mol- ecules which can affect a variety of cellular and molecular targets in various cancers including CC [9]. A previous study shows that different types of mRNA can be used as tissue- and exosome-based diagnostic biomarkers for human CC [10]. The high expression of mRNA such as interleukin-6 can be used as a predictor of relapse in CC [11]. Furthermore, certain kinds of mRNA can be used to predict survival in CC patients [12]. A previous study shows that Ephrin-A1 mRNA is associated with poor Correspondence: Warner_Y792@hotmail.com 1 Gastrointestinal Colorectal and Anal surgery, China-Japan Union Hospital of Jilin University, No. 126 Xiantai Street, Changchun, Jilin 130033, China Full list of author information is available at the end of the article Correspondence: Warner_Y792@hotmail.com 1 Gastrointestinal Colorectal and Anal surgery, China-Japan Union Hospital of Jilin University, No. 126 Xiantai Street, Changchun, Jilin 130033, China Full list of author information is available at the end of the article Li et al. Biol Res (2018) 51:10 Page 2 of 9 prognosis of CC patients based on short disease-free survival (DFS) data [13]. Alexopoulou et al. showed that Kallikrein Related Peptidase 11 mRNA expression could predict poor DSF and overall survival (OS) in colorec- tal adenocarcinoma patients [14]. Thus, a better knowl- edge of the molecular mechanisms and cancer associated gene is vital for the early diagnosis and personalized care of CC patients. However, the independent prognostic gene associated with death and recurrence of CC is still unclear. prognosis of CC patients based on short disease-free survival (DFS) data [13]. Alexopoulou et al. Background showed that Kallikrein Related Peptidase 11 mRNA expression could predict poor DSF and overall survival (OS) in colorec- tal adenocarcinoma patients [14]. Thus, a better knowl- edge of the molecular mechanisms and cancer associated gene is vital for the early diagnosis and personalized care of CC patients. However, the independent prognostic gene associated with death and recurrence of CC is still unclear. [20] and Log2 fold change [21] using Robust Multi-array Average (RMA) algorithm [22] in Affy software [23]. Finally, a total of 22,844 and 22,854 genes were obtained from GSE39582 and GSE17536 respectively after data processing. The DEMs were identified between cancer tissue sam- ples and para-carcinoma tissue samples in TCGA data- set based on Student’s t test [24]. P value < 0.05 and fold change > 2 (or < 1/2) were defined to be statistically significant. In previous studies, Smith et al. [15, 16] and Marisa et al. [17] tried to predict the mRNAs associated with the recurrence and death in CC patients based on gene expression profiles. Although some valuable biologi- cal markers for prognosis of CC has been revealed, the limited sample size and survival evaluation in these stud- ies are not benefit for the investigation of independent prognostic genes. Based on the previous results from Smith et al. and Marisa et al. The Cancer Genome Atlas (TCGA) dataset associated with CC were added in the current bioinformatics study. By comparing the data between cancer tissue samples and para-carcinoma tis- sue samples in TCGA dataset, the differentially expressed mRNAs (DEMs) investigation, functional and pathway enrichment analysis were performed. Furthermore, prog- nostic analysis including overall survival and disease- free survival rate investigation was performed based on data in all three datasets. We hoped to explore molecular mechanism of CC, and identify candidate independent prognostic genes for CC prognosis. Clinical prognostic analysis Based on the DEMs obtained in TCGA dataset, the clini- cal prognostic analyses including overall survival (OS) and disease-free survival (DFS) analysis were performed on the datasets which have clinical prognostic informa- tion. All the three datasets including TCGA, GSE17536 and GSE39582 were used for the overall survival analysis in the present study. Meanwhile, two datasets including GSE17536 and GSE39582 were used for the disease-free survival analysis. The mRNAs in all datasets were divided into high expression group (H group) and low expression group (L group) according to the mean value of DEMs. The survival estimation and survival curve examination were performed using Kaplan–Meier method [29] and log-rank test [30], respectively. The hazard ratio (HR) was estimated with the single variable Cox proportional risk regression model [31]. The independent analysis of prog- nostic factor was performed based on multivariable Cox proportional risk regression model [32]. Based on the Cox proportional risk regression model, the outstanding DEMs of the single variable in each dataset were consid- ered as the corrected variables. P < 0.05 was considered statistically significant. Functional annotation and pathway enrichment analysis of DEMsh The Database for Annotation, Visualization and Inte- grated Discovery (DAVID) [25] is a gene functional classification tool that provides a comprehensive set of functional annotation tools for investigators to under- stand biological meaning behind large list of genes. By DAVID software, the Gene Ontology (GO, http:// www.geneontology.org) functional annotation [26] and Kyoto Encyclopedia of Genes and Genomes (KEGG) (http://www.genome.jp/kegg/pathway.html) [27] path- way analyses were performed on DEMs in TCGA data- set. P < 0.05 was chosen as the cut-off criterion for the enrichment analysis. The results of GO function and KEGG pathway analysis were visualized by Erichment- Map [28] software. The mRNA microarray data Microarray dataset GSE17536 [15, 16] and GSE39582 [17] were downloaded from Gene Expression Omnibus (GEO, http://www.ncbi.nlm.nih.gov/geo/) database. A total of 177 invasive murine CC cells sample were included in dataset GSE17536. Meanwhile, a total of 566 CC samples were included in GSE39582. The gene expression pro- file data of GSE17536 and GSE39582 were all generated based on the platform of Affymetrix Human Genome U133 Plus 2.0 Array (GPL570 [HG-U133_Plus_2]). Fur- thermore, the whole CC-related dataset in TCGA data- base (https://cancergenome.nih.gov/) were downloaded. DEMs investigation With P value < 0.05 and fold change > 2 (or < 1/2), a total of 954 DEMs including 633 up- and 321 down-regulated mRNA were revealed in TCGA dataset (Fig. 1). Investigation of DFS prognostic geneh g p g g The relationship between DMEs and DFS prognostic genes associated with CC in each dataset was showed in Fig. 5. The results showed that a total of 12 DEMs were outstanding in both GSE17536 and GSE39582, includ- ing Surfactant Associated 2 (SFTA2), LEM Domain Con- taining 1 (LEMD1), Cartilage Oligomeric Matrix Protein (COMP), Kinesin Family Member 26B (KIF26B), Kal- likrein Related Peptidase 10 (KLK10), Matrix Metallo- peptidase 11 (MMP11), Gastrin Releasing Peptide (GRP), Twist Family BHLH Transcription Factor 1 (TWIST1), Regulator Of G Protein Signaling 16 (RGS16), Collagen Type VIII Alpha 1 Chain (COL8A1), Transmembrane Protein 37 (TMEM37), Rho GTPase Activating Protein 44 (ARHGAP44) (Table 1). Notably, the relation between Investigation of CC associated OS prognostic genesh Investigation of CC associated OS prognostic genes The relationship between DMEs and OS in each data- set was showed in Fig. 3. The results showed that UL16- binding protein 2 (ULBP2) was the outstanding gene in all three dataset including TCGA [Log-rank P 0.0030, HR: 0.552, 95% confidence interval (CI) 0.37–0.82], GSE17536 (Log-rank P 0.0005, HR: 0.380, 95% CI 0.22– 0.67) and GSE39582 (Log-rank P 0.0092, HR: 0.685, 95% CI 0.51–0.91). Furthermore, up-regulation of ULBP2 gene was associated with shorter OS of CC patients. The detail information was showed in Fig. 4. Independent DFS prognostic genes analysis The independent DFS prognostic genes in CC were explored in GSE17536 and GSE39582. The results showed that GRP and TMEM37 were outstanding in totally 12 DEMs after the multivariate Cox proportional risk regression (Table 2). Conclusion In sum, ULBP2 might be a potential novel OS prognos- tic biomarker in CC, while GRP and TMEM37 could be served as the independent DFS prognostic genes in CC. Furthermore, functions including extracellular matrix and digestion, as well as pathways including Wnt signal- ing and drug metabolism may play important roles in the process of CC. Data preprocessing and DEMs identification Normalized RNA-seq data (including 24,991 genes) of TCGA dataset were downloaded for the further inves- tigation. Meanwhile, the CEL source files of GSE17536 and GSE39582 were processed into background adjust- ment [18], quantile normalization [19], summarization Li et al. Biol Res (2018) 51:10 Page 3 of 9 Results DEMs investigation With P value < 0.05 and fold change > 2 (or < 1/2), a total of 954 DEMs including 633 up- and 321 down-regulated mRNA were revealed in TCGA dataset (Fig. 1). Fig. 1 Volcano plot showing differentially expressed mRNAs. The X-axis represents the log fold-change (FC) values; the Y-axis repre- sents the—log P values; green point represents the down-regulated mRNAs; the red point represents the up-regulated mRNAs; those points having a FC larger than 2 (or FC < 1/2) are shown in gray Function annotation and pathway enrichment investigation With P < 0.05, the GO function and KEGG pathway of DEMs were investigated, followed by visualized using ErichmentMap software. As showed in Fig. 2a, extra- cellular matrix (GO:0031012), chromatin assembly or disassembly (GO:0006333) and endopeptidase activity (GO:0004175) were outstanding functions assembled with up-regulated DEMs. Meanwhile, the Systemic lupus erythematosus (hsa05322) and Wnt signaling pathway (hsa04310) were two outstanding pathways enriched by up-regulated DEMs (Fig. 2a). Fig. 1 Volcano plot showing differentially expressed mRNAs. The X-axis represents the log fold-change (FC) values; the Y-axis repre- sents the—log P values; green point represents the down-regulated mRNAs; the red point represents the up-regulated mRNAs; those points having a FC larger than 2 (or FC < 1/2) are shown in gray As showed in Fig. 2b, Microsome (GO:0005792), Vesic- ular fraction (GO:0042598) and Digestion (GO: 0007586) were outstanding functions assembled with the down- regulated DEMs, while the Drug metabolism (hsa00983), Androgen and estrogen metabolism (hsa00150, Genes) and Retinol metabolism (hsa00830) were outstanding pathways enriched by down-regulated DMEs (Fig. 2b). up-regulation of SFTA2 and short DFS in CC patients was showed in Fig. 6. Discussionh The current informatics study revealed the potential independent prognostic genes for survival of CC, as well as the molecular mechanism during CC process. A total of 633 up- and 321 down-regulated DEMs were revealed between cancer tissue samples and para-carcinoma tis- sue samples. The up-regulated mRNAs were mainly assembled in functions including Extracellular matrix and pathways including Wnt signaling. The down-reg- ulated mRNAs were mainly assembled in functions like Page 4 of 9 Li et al. Biol Res (2018) 51:10 Fig. 2 The GO function and KEGG pathways enrichment analysis for the up- and down-regulated mRNAs. a the network constructed by the up-regulated mRNAs; b the network constructed by the down-regulated mRNAs; GO represents Gene Ontology; KEGG represents Kyoto Encyclo- pedia of Genes and Genomes; the red point represents a certain name of GO function or KEGG pathway; the green line represents the interaction between two points; the thickness of the line indicates the number of overlapping genes between the different gene sets Fig. 3 Cox proportional risk regression plot analyses between differentially expressed microRNAs and overall survival in colon cancer. X-axis repre- sents the logarithm of hazard ratio (HR), the positive expression is associated with better prognosis, and negative expression of gene expression is associated with poor prognosis; Y-axis represents logarithm of P value for the log-rank test Fig. 2 The GO function and KEGG pathways enrichment analysis for the up- and down-regulated mRNAs. a the network constructed by the up-regulated mRNAs; b the network constructed by the down-regulated mRNAs; GO represents Gene Ontology; KEGG represents Kyoto Encyclo- pedia of Genes and Genomes; the red point represents a certain name of GO function or KEGG pathway; the green line represents the interaction between two points; the thickness of the line indicates the number of overlapping genes between the different gene sets Fig. 2 The GO function and KEGG pathways enrichment analysis for the up- and down-regulated mRNAs. a the network constructed by the up-regulated mRNAs; b the network constructed by the down-regulated mRNAs; GO represents Gene Ontology; KEGG represents Kyoto Encyclo- pedia of Genes and Genomes; the red point represents a certain name of GO function or KEGG pathway; the green line represents the interaction between two points; the thickness of the line indicates the number of overlapping genes between the different gene sets Fig. Discussionh 3 Cox proportional risk regression plot analyses between differentially expressed microRNAs and overall survival in colon cancer. X-axis repre- sents the logarithm of hazard ratio (HR), the positive expression is associated with better prognosis, and negative expression of gene expression is associated with poor prognosis; Y-axis represents logarithm of P value for the log-rank test Fig. 3 Cox proportional risk regression plot analyses between differentially expressed microRNAs and overall survival in colon cancer. X-axis repre- sents the logarithm of hazard ratio (HR), the positive expression is associated with better prognosis, and negative expression of gene expression is associated with poor prognosis; Y-axis represents logarithm of P value for the log-rank test Fig. 4 Kaplan-Meier analyses for the relation between shorter overall survival and UL16 Binding Protein 2 gene in colon cancer. a The TCGA dataset; b the GSE17536 dataset; c the GSE39582 dataset Fig. 4 Kaplan-Meier analyses for the relation between shorter overall survival and UL16 Binding Protein 2 gene in colon cancer. a The TCGA dataset; b the GSE17536 dataset; c the GSE39582 dataset Li et al. Biol Res (2018) 51:10 Page 5 of 9 Fig. 5 Cox proportional risk regression plot analyses between differentially expressed microRNAs and disease-free survival in colon cancer. X-axis represents the logarithm of hazard ratio (HR), the positive expression is associated with better prognosis, and negative expression of gene expres- sion is associated with poor prognosis; Y-axis represents logarithm of P value for the log-rank test Fig. 5 Cox proportional risk regression plot analyses between differentially expressed microRNAs and disease-free survival in colon cancer. Discussionh X-axis represents the logarithm of hazard ratio (HR), the positive expression is associated with better prognosis, and negative expression of gene expres- sion is associated with poor prognosis; Y-axis represents logarithm of P value for the log-rank test Table 1 The outstanding disease-free prognostic genes for colon cancer FC fold change, HR hazard ratio, CI confidence interval P < 0.05 was considered as significant different Gene FC P GSE17536 GSE39582 Log-rank P HR 95% CI Log-rank P HR 95% CI SFTA2 63.45 1.6E−91 8.87E−03 0.488 0.49 (0.28–0.84) 2.29E−02 0.708 0.71 (0.53–0.95) LEMD1 32.47 4.3E−98 1.32E−02 0.506 0.51 (0.29–0.88) 1.47E−02 0.692 0.69 (0.51–0.93) COMP 14.33 4.3E−82 3.70E−03 0.444 0.44 (0.25–0.78) 1.48E−02 0.691 0.69 (0.51–0.93) KIF26B 6.86 2.4E−64 5.85E−03 0.466 0.47 (0.27–0.81) 1.09E−02 0.679 0.68 (0.5–0.92) KLK10 6.06 1.1E−99 5.28E−03 0.462 0.46 (0.26–0.81) 7.06E−03 0.666 0.67 (0.49–0.9) MMP11 3.74 6.8E−127 2.96E−03 0.44 0.44 (0.25–0.77) 3.68E−02 0.731 0.73 (0.54–0.98) GRP 3.22 2.4E−15 1.43E−02 0.515 0.52 (0.3–0.88) 1.16E−04 0.554 0.55 (0.41–0.75) TWIST1 2.59 1.3E−28 4.07E−02 0.562 0.56 (0.32–0.98) 4.29E−02 0.736 0.74 (0.55–0.99) RGS16 2.29 2.1E−25 1.90E−04 0.346 0.35 (0.19–0.62) 3.48E−02 0.726 0.73 (0.54–0.98) COL8A1 2.01 9.6E−15 1.68E−04 0.338 0.34 (0.19–0.61) 6.21E−03 0.661 0.66 (0.49–0.89) TMEM37 0.47 1.3E−23 1.41E−02 1.986 1.99 (1.14–3.47) 1.28E−02 1.457 1.46 (1.08–1.96) ARHGAP44 0.45 4.8E−20 4.88E−02 1.717 1.72 (1–2.96) 4.58E−02 1.349 1.35 (1–1.82) Table 1 The outstanding disease-free prognostic genes for colon cancer shows that matrix metalloproteinase-9 is an important marker for analysis of the postoperative prognosis and risk of metastases in patients with colorectal cancer [35]. Recently, Wang et al. showed that the extracellular matrix protein mindin attenuated colon cancer progres- sion by blocking angiogenesis via Egr-1-mediated regu- lation [36]. Despite of extracellular matrix, the digestion function is closed related to cancer preventative activity [37]. The microflora in digestion system degrades a wide variety of organic compounds including food additives, drugs, bile salts and cholesterol which may be relevant shows that matrix metalloproteinase-9 is an important marker for analysis of the postoperative prognosis and risk of metastases in patients with colorectal cancer [35]. Recently, Wang et al. showed that the extracellular matrix protein mindin attenuated colon cancer progres- sion by blocking angiogenesis via Egr-1-mediated regu- lation [36]. Despite of extracellular matrix, the digestion function is closed related to cancer preventative activity [37]. Discussionh The microflora in digestion system degrades a wide variety of organic compounds including food additives, drugs, bile salts and cholesterol which may be relevant Digestion and pathways like Drug metabolism. Further- more, up-regulation of ULBP2 was associated with OS in CC patients. A total of 12 genes including SFTA2 were potential DFS prognostic genes in CC patients. Mean- while, the GRP and TMEM37 were two outstanding independent DFS prognostic genes in CC.h The extracellular matrix is a collection of extracellu- lar molecules secreted by cells that provides structural and biochemical support to the surrounding cells [33]. Extracellular matrix plays an important role in cancer progression as a dynamic niche [34]. A previous study Li et al. Biol Res (2018) 51:10 Page 6 of 9 Fig. 6 Kaplan-Meier analyses for the relation between shorter disease-free survival and Surfactant Associated 2 gene in colon cancer. a The GSE17536 dataset; b the GSE39582 dataset e survival and Surfactant Associated 2 gene in colon cancer. a The Fig. 6 Kaplan-Meier analyses for the relation between shorter disease-free survival and Surfactant Associated 2 gene in colon cancer. a The GSE17536 dataset; b the GSE39582 dataset Table 2 The potential disease-free survival independent prognostic gene for colon cancer HR hazard ratio, HRse hazard ratio standard error P < 0.05 was considered as significant different Gene GSE17536 GSE39582 P HR HRse P HR HRse SFTA2 0.07 0.542 0.28–1.051 0.114 0.775 0.565–1.063 LEMD1 0.465 0.789 0.417–1.491 0.156 0.796 0.581–1.091 COMP 0.871 0.943 0.466–1.91 0.953 0.988 0.658–1.483 KIF26B 0.915 1.042 0.491–2.211 0.808 0.955 0.656–1.39 KLK10 0.319 0.705 0.355–1.401 0.364 0.861 0.623–1.19 MMP11 0.566 0.807 0.388–1.679 0.498 1.148 0.77–1.711 GRP 0.275 0.707 0.38–1.317 0.011 0.612 0.418–0.895 TWIST1 0.215 1.706 0.734–3.967 0.932 1.017 0.686–1.509 RGS16 0.184 0.613 0.298–1.262 0.917 0.982 0.696–1.385 COL8A1 0.088 0.445 0.175–1.128 0.467 0.856 0.563–1.301 TMEM37 0.012 2.177 1.183–4.007 0.114 1.289 0.941–1.766 ARHGAP44 0.931 0.973 0.524–1.807 0.642 1.08 0.782–1.49 Table 2 The potential disease-free survival independent prognostic gene for colon cancer to the development of CC [38]. A previous study shows that peptides derived from in vitro gastrointestinal diges- tion can inhibit human colon cancer cells proliferation and inflammation [39]. In the present study, GO func- tion analysis showed that the up- and down-regulated mRNAs were mainly assembled in extracellular matrix and digestion. Thus, our results reveal that the abnormal of extracellular matrix and digestion function may play vital roles in the progression of CC. Abbreviations CC: colon cancer; TCGA: The Cancer Genome Atlas; OS: overall survival; DFS: disease-free survival. CC: colon cancer; TCGA: The Cancer Genome Atlas; OS: overall survival; DFS: disease-free survival. g g g Prognostic biomarkers for cancer have the power to change the course of a disease if they add value beyond knew prognostic factors [45]. ULBP2, which located on the chromosome 6, is a gene that encoding the cell sur- face glycoprotein [46]. A previous study shows that ULBP2 is a novel prognostic biomarker for CC [47]. Demirkol et al. indicated that ULBP2 was a mRNA based stage-independent prognostic marker to prognos- ticate CC in vivo [48]. In the analysis of pancreatic can- cer, researchers have proved that a high level of soluble ULBP2 is deemed an independent indicator for OS [49]. In this study, ULBP2 was identified as the unique mRNA outstanding in all three datasets. Thus, we speculate that ULBP2 may be used as the OS prognostic biomarker in CC. Furthermore, GRP is a regulatory molecule that has been implicated in a number of physiological and patho- physiological processes [50]. A Previous study shows that GRP can better predict the prognosis of patients with colorectal cancer and distant metastasis, and has good sensitivity and specificity [51]. Matkowskyj et al. con- firmed that GRP and its receptor’s co-expression had the function of differentiation, with highest levels observed in well-differentiated CC cells [52]. Moreover, the trans- membrane protein is a type of integral membrane pro- tein that spans the entirety of the biological membrane to which it is permanently attached [53]. Based on the results of previous studies, various transmembrane pro- teins such as MutL Homolog 1 (MLH1) and Bone Mar- row Stromal Cell Antigen 2 (BST2) are associated with the progression of CC [54, 55]. Unfortunately, there is no related study based on transmembrane protein TMEM37 and CC. In the current study, independent DFS prognos- tic genes analysis showed that GRP and TMEM37 were two most outstanding mRNAs, which might be used as the independent DFS prognostic genes. However, a fur- ther clinical investigation based on a large scale of sample size is still needed to confirm the thesis speculation. References 1. O’Brien CA, Pollett A, Gallinger S, Dick JE. A human colon cancer cell capable of initiating tumour growth in immunodeficient mice. Nature. 2007;445(7123):106. 2. Stewart BW, Kleihues P. World cancer report. World cancer report, vol. 45. Lyon: IARCPress; 2014. p. 12–351. 3. Peto R, Lopez AD, Pan H, Thun MJ. World cancer report. Lyon: IARCPress; 2014. Discussionh to the development of CC [38]. A previous study shows that peptides derived from in vitro gastrointestinal diges- tion can inhibit human colon cancer cells proliferation and inflammation [39]. In the present study, GO func- tion analysis showed that the up- and down-regulated mRNAs were mainly assembled in extracellular matrix and digestion. Thus, our results reveal that the abnormal of extracellular matrix and digestion function may play vital roles in the progression of CC. including colorectal cancer [40]. A previous study shows that alteration in the Wnt signaling pathway is frequently observed in colorectal cancer with microsatellite instabil- ity [41]. Inhibiting the Wnt signaling pathway may be a fruitful strategy for targeting chemotherapy-resistant CC cells [42]. Moreover, drug metabolism is the metabolic breakdown of drugs by living organisms [43]. Landmann indicated that drug metabolism determined resistance of colorectal cancer to resorcinol-based heat shock pro- tein 90 inhibitors [44]. In the present study, Wnt signal- ing and drug metabolism were two outstanding pathways Furthermore, the clinical importance of Wnt signal- ing pathway has been demonstrated in various diseases Li et al. Biol Res (2018) 51:10 Page 7 of 9 enriched by the up- and down-regulated mRNAs. Thus, we speculate that the DEMs may take part in the CC pro- cess via Wnt signaling and drug metabolism. Authors’ contributions ZS carried out the Conception and design of the research, participated in the Acquisition of data and drafted the manuscript. YZ carried out the Analysis and interpretation of data. WY participated in the design of the study and performed the statistical analysis. CL conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript. Funding Funding None. Ethics approval and consent to participate Ethics approval and consent to participate This study was approved by Ethics Committee of China-Japan Union Hospital of Jilin University and The First Hospital of Jilin University. pp p p This study was approved by Ethics Committee of China-Japan Union Hospital of Jilin University and The First Hospital of Jilin University. Availability of data and materials y Not applicable. This study was only the primary research, and further study has been in progress. Consent for publication Not applicable Ethics approval and consent to participate This study was approved by Ethics Committee of China-Japan Union Hospital of Jilin University and The First Hospital of Jilin University. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations. Received: 8 January 2018 Accepted: 29 March 2018 Author details 1 1 Gastrointestinal Colorectal and Anal surgery, China-Japan Union Hospital of Jilin University, No. 126 Xiantai Street, Changchun, Jilin 130033, China. 2 Department of Neurology, The First Hospital of Jilin University, Changchun, Jilin 130021, China. Highlights Key enzymes of the extracellular matrix in colorectal can- cer. Bull Exp Biol Med. 2009;147(3):353. 36. Wang LF, Liu YS, Yang B, Li P, Cheng XS, Xiao CX, Liu JJ, Li S, Ren JL, Guleng B. The extracellular matrix protein mindin attenuates colon cancer progression by blocking angiogenesis via Egr-1-mediated regulation. Oncogene. 2017;37:601 (Epub ahead of print). 12. Lassmann S, Schuster I, Walch A, Göbel H, Jütting U, Makowiec F, Hopt U, Werner M. STAT3 mRNA and protein expression in colorectal cancer: effects on STAT3-inducible targets linked to cell survival and proliferation. J Clin Pathol. 2007;60(2):173. 37. Ferruzzi MG, Blakeslee J. Digestion, absorption, and cancer preventative activity of dietary chlorophyll derivatives. Nutr Res. 2007;27(1):1–12. 13. Nielsen BS, Jørgensen S, Fog JU, Søkilde R, Christensen IJ, Hansen U, Brün- ner N, Baker A, Møller S, Nielsen HJ. High levels of microRNA-21 in the stroma of colorectal cancers predict short disease-free survival in stage II colon cancer patients. Clin Exp Metas. 2011;28(1):27–38. 38. Cummings JG. The colon: absorptive, seccretory and metabolic functions. Digestion. 1975;13(4):232–40. g 39. González-Montoya M, Hernández-Ledesma B, Silván JM, Mora-Escobedo R, Martínez-Villaluenga C. Peptides derived from in vitro gastrointestinal digestion of germinated soybean proteins inhibit human colon cancer cells proliferation and inflammation. Food Chem. 2017;242:75–82. 14. Alexopoulou DK, Kontos CK, Spyridon C, Papadopoulos IN, Andreas S. KLK11 mRNA expression predicts poor disease-free and overall survival in colorectal adenocarcinoma patients. Biomark Med. 2014;8(5):671–85. 15. Smith JJ, Deane NG, Fei WU, Merchant NB, Zhang B, Jiang A, Pengcheng LU, Johnson JC, Schmidt C, Bailey CE. Experimentally derived metastasis gene expression profile predicts recurrence and death in patients with colon cancer. Gastroenterology. 2010;138(3):958. 40. Godman CA, Joshi R, Tierney BR, Greenspan E, Rasmussen TP, Wang H, Shin DG, Rosenberg DW, Giardina C. HDAC3 impacts multiple oncogenic pathways in colon cancer cells with effects on Wnt and vitamin D signal- ing. Cancer Biol Ther. 2008;7(10):1570–80. 16. Freeman TJ, Smith JJ, Chen X, Washington MK, Roland JT, Means AL, Eschrich SA, Yeatman TJ, Deane NG, Beauchamp RD. Smad4-mediated signaling inhibits intestinal neoplasia by inhibiting expression of β-catenin. Gastroenterology. 2012;142(3):562. 41. Shimizu Y, Ikeda S, Fujimori M, Kodama S, Nakahara M, Okajima M, Asahara T. Frequent alterations in the Wnt signaling pathway in colo- rectal cancer with microsatellite instability. Genes Chromosom Cancer. 2002;33(1):73–81. 17. Marisa L, De RA, Duval A, Selves J, Gaub MP, Vescovo L, Etiennegrimaldi MC, Schiappa R, Guenot D, Ayadi M. Highlights Gene expression classification of colon cancer into molecular subtypes: characterization, validation, and prognostic value. Plos Med. 2013;10(5):e1001453. 42. Chikazawa N, Tanaka H, Tasaka T, Nakamura M, Tanaka M, Onishi H, Katano M. Inhibition of Wnt signaling pathway decreases chemo- therapy-resistant side-population colon cancer cells. Anticancer Res. 2010;30(6):2041–8. 18. Wu Z, Irizarry RA, Gentleman R, Martinezmurillo F, Spencer F. A model based background adjustment for oligonucleotide expression arrays. Johns Hopkins University Dept of Biostatistics Working Paper. 2004; 99(468):909–17. 43. Wilkinson GR. Drug metabolism and variability among patients in drug response. N Engl J Med. 2005;352(21):2211–21. 44. Landmann H. Drug metabolism determines resistance of colorectal cancer to resorcinol-based HSP90 inhibitors. Göttingen: University Library Göttingen; 2014. 19. Hu J, He X. Enhanced quantile normalization of microarray data to reduce loss of information in gene expression profiles. Biometrics. 2007;63(1):50–9. 45. Guo Z, Shu Y, Zhou H, Zhang W. Identification of diagnostic and prog- nostic biomarkers for cancer: focusing on genetic variations in microRNA regulatory pathways (Review). Mol Med Rep. 2016;13(3):1943–52. 20. Mishra R, Bian J, Fiszman M, Weir CR, Jonnalagadda S, Mostafa J, Fiol GD. Text summarization in the biomedical domain: a systematic review of recent research. J Biomed Inform. 2014;52:457–67. 46. Radosavljevic M, Cuillerier B, Wilson MJ, Clément O, Wicker S, Gilfillan S, Beck S, Trowsdale J, Bahram S. A cluster of ten novel MHC class I related genes on human chromosome 6q24.2-q25.3. Genomics. 2002;79(1):114–23. 21. Benoît M, Bernadette G, Annie R, Bertrand B, Jérôme A, Jean-Luc G. Impact of the spotted microarray preprocessing method on fold-change compression and variance stability. BMC Bioinform. 2011;12(1):1–12. 47. Gomceli I. A combined ULBP2 and SEMA5A expression signature as a prognostic and predictive biomarker for colon cancer. J Cancer. 2017;8(7):1113–22. 22. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics. 2003;4(2):249–64. 23. Gautier L, Cope L, Bolstad BM, Irizarry RA. affy–analysis of Affymetrix GeneChip data at the probe level. Bioinformatics. 2004;20(3):307–15. 48. Demirkol S, Gomceli I, Isbilen M, Dayanc BE, Tez M, Bostanci EB, Akoglu M, Ozyerli E, Durdu S, Konu O. Expression of SEMA5A and ULBP2 are predic- tors of prognosis and chemosensitivity in colon cancer. In: International symposium on abdominal metastatic cancer: 2015; 2015. 24. Gooch JW. Student’s t. Encyclopedic dictionary of polymers, vol. 81(5). Berlin: Springer; 2011. p. 707. 25. Highlights 1. ULBP2 was a novel overall survival prognostic gene in CC. 4. Dlamini Z, Khoza T, Hull R, Choene M, Mkhize-Kwitshana Z. Current immunotherapeutic treatments in colon cancer. Rijeka: Intech; 2016. 5. Dienstmann R, Salazar R, Tabernero J. Personalizing colon cancer adjuvant therapy: selecting optimal treatments for individual patients. J Clin Oncol Off J Am Soc Clin Oncol. 2015;33(16):1787. 2. GRP was an independent disease-free survival prog- nostic gene in CC. 3. TMEM37 was an independent disease-free survival prognostic gene in CC. f 6. Abdalla M, Landerholm K, Andersson P, Andersson RE, Myrelid P. Risk of rectal cancer after colectomy for patients with ulcerative colitis-a national cohort study. Clin Gastroenterol Hepatol Off Clin Pract J Am Gastroen- terol Assoc. 2016;15:1055–60. 4. Extracellular matrix and digestion functions were important for CC. 7. Grady WM, Markowitz SD. Genetic and epigenetic alterations in colon cancer. Annu Rev Genomics Hum Genet. 2002;3(3):101–28. 5. Wnt signaling and drug metabolism pathways were vital for CC. Page 8 of 9 Li et al. Biol Res (2018) 51:10 8. Zhang J, Pan M. Genetic or non-genetic prognostic factors in colon cancer. In: International conference on fuzzy systems and knowledge discovery: 2016; 2016. p. 1146–50. 31. Persson I. Bias of the cox model hazard ratio. J Mod Appl Stat Methods Jmasm. 2005;4(1):90–9. 32. He J, Xiao L, Tan J, Dai Z, Wang Q, Surgery DO. The risk factors analysis of prognosis of patients with intracerebral hemorrhage by Cox proportional hazard regression model. Chin J Pract Neruous Dis. 2015;2:1–3. 9. Vishnubalaji R, Hamam R, Abdulla MH, Mohammed MAV, Kassem M, Alobeed O, Aldahmash A, Alajez NM. Genome-wide mRNA and miRNA expression profiling reveal multiple regulatory networks in colorectal cancer. Cell Death Dis. 2015;6(1):e1614. 33. Michel GT, Tonon T, Scornet D, Cock JM, Kloareg B. The cell wall polysac- charide metabolism of the brown alga Ectocarpus siliculosus. Insights into the evolution of extracellular matrix polysaccharides in Eukaryotes. New Phytol. 2010;188(1):82–97. 10. Amirkhah R, Schmitz U, Linnebacher M, Wolkenhauer O, Farazmand A. MicroRNA–mRNA interactions in colorectal cancer and their role in tumor progression. Genes Chromosom Cancer. 2015;54(3):129–41. y 34. Lu P, Weaver VM, Werb Z. The extracellular matrix: a dynamic niche in cancer progression. J Cell Biol. 2012;196(4):395–406. 11. Olsen J, Kirkeby LT, Olsen J, Eiholm S, Jess P, Gögenur I, Troelsen JT. High interleukin-6 mRNA expression is a predictor of relapse in colon cancer. Anticancer Res. 2015;35(4):2235. 35. Golovkov DA. 53. Almén MS, Nordström KJ, Fredriksson R, Schiöth HB. Mapping the human membrane proteome: a majority of the human membrane proteins can be classified according to function and evolutionary origin. BMC Biol. 2009;7(1):50. is associated with poor survival of patients with esophageal, gastric, or colorectal cancer. Ann Surg Oncol. 2017;24(2):1–9. 55. Hepzibah W, John William S, Balaji M. Molecular modeling and identifica- tion of transmembrane protein (MLH1) related to colon cancer using Insilico techniques. Int J Novel Trends Pharm Sci. 2016;6(3):3. 54. Mukai S, Oue N, Oshima T, Mukai R, Tatsumoto Y, Sakamoto N, Sentani K, Tanabe K, Egi H, Hinoi T. Overexpression of transmembrane protein BST2 Highlights Dennis G Jr, Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. DAVID: database for annotation, visualization, and integrated discov- ery. Genome Biol. 2003;4(5):P3. 49. Chen J, Zhu XX, Xu H, Fang HZ, Zhao JQ. Expression and prognostic significance of unique ULBPs in pancreatic cancer. Oncotargets Ther. 2016;9:5271. y 26. Harris MA. The gene ontology (GO) database and informatics resource. New York City: WCB/McGraw-Hill; 2004. 50. Pansky A, Weerth AD, Fasler-Kan E, Boulay JL, Schulz M, Ketterer S, Selck C, Beglinger C, Schrenck TV, Hildebrand P. Gastrin releasing peptide. J Am Soc Nephrol Jasn. 2000;11(8):1409–18. 27. Arakawa K, Kono N, Yamada Y, Mori H, Tomita M. KEGG-based pathway visualization tool for complex omics data. Silico Biol. 2005;5(4):419–23. p 51. Liao JT, Li Y, Chen YY. Gastrin-releasing peptide and its receptors expres- sion in colorectal cancer. Guide China Med. 2015;30:7–10. 28. Merico D, Isserlin R, Stueker O, Emili A, Bader GD. Enrichment map: a network-based method for gene-set enrichment visualization and inter- pretation. PLoS ONE. 2010;5(11):e13984. 52. Matkowskyj KA, Keller K, Glover S, Kornberg L, Transontay R, Benya RV. Expression of GRP and its receptor in well-differentiated colon cancer cells correlates with the presence of focal adhesion kinase phosphoryl- ated at tyrosines 397 and 407. J Histochem Cytochem Off J Histochem Soc. 2003;51(8):1041. 29. Bland JM, Altman DG. Survival probabilities (the Kaplan–Meier method). BMJ. 1998;317(7172):1572. 30. Alberti C, Timsit JF, Chevret S. Survival analysis—the log rank test. Rev Mal Respir. 2005;22(5 Pt 1):829–32. Li et al. Biol Res (2018) 51:10 Li et al. Biol Res (2018) 51:10 Page 9 of 9 • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research ? Choose BMC and benefit from:
https://openalex.org/W3040248990	https://europepmc.org/articles/pmc7331752?pdf=render	English	null	Biological properties of almond proteins produced by aqueous and enzyme-assisted aqueous extraction processes from almond cake	Scientific reports	2,020	cc-by	11,496	Biological properties of almond proteins produced by aqueous and enzyme‑assisted aqueous extraction processes from almond cake Thaiza S. P. de Souza1,2, Fernanda F. G. Dias1, Joana Paula S. Oliveira2, Juliana M. L. N. de Moura Bell1,3 & Maria Gabriela B. Koblitz2* The almond cake is a protein-rich residue generated by the mechanical expression of the almond oil. The effects of the aqueous (AEP) and enzyme-assisted aqueous extraction processes (EAEP) on the biological properties of the almond cake protein were evaluated. Total phenolic content (TPC), antioxidant capacity, inhibitory effects against crucial enzymes related to metabolic syndrome, antimicrobial potential, and in vitro protein digestibility profile were assessed. EAEP provided the best results for antioxidant capacity by both ORAC (397.2 µmol TE per g) and ABTS (650.5 µmol TE per g) methods and also showed a high (~ 98%) potential for α-glucosidase inhibition. The AEP resulted in protein extracts with the highest lipase inhibition (~ 70%) in a dose-dependent way. Enzymatic kinetic analyses revealed that EAEP generated uncompetitive inhibitors against α-glucosidase, while EAEP, AEP, and HEX-AEP (used as control) generated the same kind of inhibitors against lipase. No protein extract was effective against any of the bacteria strains tested at antimicrobial assays. An in silico theoretical hydrolysis of amandin subunits corroborated with the results found for antioxidant capacity, enzyme inhibitory experiments, and antimicrobial activity. Digestibility results indicated that the digestive proteases used were efficient in hydrolyzing almond proteins, regardless of the extraction applied and that HEX-AEP presented the highest digestibility (85%). In summary, EAEP and AEP skim proteins have the potential to be used as a nutraceutical ingredient. The biological properties observed in these extracts could help mitigate the development of metabolic syndrome where EAEP and AEP skim proteins could be potentially used as a prophylactic therapy for diabetes and obesity, respectively. The consumption of almonds (Prunus dulcis, syn. Prunus amygdalus) has been associated with various health benefits. Its antioxidant activity is usually attributed to the presence of α-tocopherols and polyphenols, with the latter being associated with reduced risk of metabolic syndrome—through the regulation of postprandial hyperglycemia and reducing the occurrence of diabetes mellitus type II—and potential antimicrobial effect1. Almond polyphenols are mainly found at the lipid interface and contribute to improving whole almonds shelf- life due to their antioxidant and antimicrobial activities2. In addition, almonds are considered a good source of protein, although methionine, lysine, and threonine are essential limiting amino acids. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Biological properties of almond proteins produced by aqueous and enzyme‑assisted aqueous extraction processes from almond cake Almond’s major storage protein is called amandin and represents about 70% of the total soluble proteins in the seed. Amandin is formed by two subunits: prunin-1 and prunin-2, which are composed of basic (20–22 kDa) and acidic (42–46 kDa) polypeptides3. 1Department of Food Science and Technology, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA. 2Department of Food Science, Food and Nutrition Graduate Program, Federal University of the State of Rio de Janeiro, 296, Pasteur Avenue, Urca, Rio de Janeiro, RJ 29622290‑240, Brazil. 3Biological and Agricultural Engineering, University of California, Davis, One Shields Avenue, Davis, CA 95616, USA. *email: maria.koblitz@unirio.br \| https://doi.org/10.1038/s41598-020-67682-3 Scientific Reports \| (2020) 10:10873 www.nature.com/scientificreports/ Because almonds are a good source of lipids and proteins, prior removal of oil is needed to produce defatted protein fractions. The mechanical expression of almond oil generates a protein-rich cake which is commonly used as animal feed. While the composition of the cake is intrinsically related to the pressing conditions, which in turn dictates the extraction efficiency, oil and protein contents of 16 and 37% have been observed for the almond cake, respectively4. The residual oil in the protein-rich cake is commonly extracted by the use of hexane, a practice that has raised environmental, safety (flammability), and health (neurotoxic effects) concerns5. These concerns, along with increasingly restrictive regulations, have prompted the search for more environmentally friendly extraction approaches6. Aqueous (AEP) and enzyme-assisted aqueous extraction processes (EAEP) are environmentally friendly strategies that can replace the use of hexane and enable the simultaneous extraction of oil, protein, and carbohydrate from many oil-bearing materials7. In addition, the EAEP offers the possibility of generating frac- tions with improved functionality and biological properties8. Research has shown that protein hydrolysis may generate bioactive peptide, improving the functional and biological properties of the original proteins9. Bioactive peptides have been associated with various biological properties such as antioxidant, antihypertensive, antimi- crobial, antithrombotic, hypocholesterolemic, and immunomodulatory functions10. Valorization of the almond cake, as many other food byproducts, could be achieved by the tailored extraction of its major constituents (oil, proteins, and bioactive compounds) for subsequent use in food, feed, fuel, and nutraceutical applications. These compounds may be used in the formulation of functional foods or to improve the nutritional characteristics of other food products while contributing to the reduction of food waste11. Materials and methods Materials. Cold pressed almond cake was supplied by Blue Diamond (Sacramento, CA, USA). For EAEP, an endoprotease (FoodPro Alkaline Protease from Bacillus licheniformis; Danisco—NY, USA) was used. Hex- ane, Folin-Ciocalteu reagent, gallic acid, 2,20–Azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) (ABTS), potassium persulfate, Trolox, fluorescein, 2,2′-Azobis (2 methylptopionamidine) dihydrochloride (AAPH), α-glucosidase, p-nitrophenyl-α-D-glucopyranoside (p-NPG), lipase from porcine pancreas type II (EC 3.1.3), 4-nitrophenyl laurate (p-NP-Laurate) were purchased from Sigma-Aldrich (St. Louis, USA). The pathogenic strains tested were supplied by Fiocruz (Rio de Janeiro, Brazil). Aqueous (AEP) and enzyme‑assisted aqueous extraction processes (EAEP). Optimum extrac- tion conditions identified by Souza et al.4 were scaled up for the AEP (aqueous extraction process), EAEP (enzyme-assisted aqueous extraction process), and HEX-AEP (solvent aqueous extraction process). Briefly, the almond cake (oil 16.25 g.100 g−1, protein 37.20 g.100 g−1, and moisture 9.04 g.100 g−1) was dispersed in dis- tilled water in a ratio of 1:12.8 (w/v) into a 10-L jacketed glass reactor (Chemglass, NJ, USA). For HEX-AEP, the almond cake was previously defatted with hexane in a Soxhlet device for 6 h at 68 °C. Process conditions were: 50 °C, pH 9.0, 120 rpm, reaction time—2 h (AEP and HEX-AEP), or 1 h and addition of alkaline pro- tease 0.85 g.100 g−1 (EAEP). The slurry was centrifuged at 3,000 × g for 30 min at 25 °C and the supernatant was allowed to cool down overnight at 4 °C to separate skim from cream fraction when the latter was formed. Approximately 16 kg of each skim fraction was freeze-dried (Lyophilizer—Virtis, CA, EUA) generating about 650 g of freeze-dried sample, as each extraction was done in duplicate. Total phenolic content (TPC). Samples of 1 g of the freeze-dried skim proteins were mixed with 10 mL of distilled water, ethanol (100%), or methanol (100%), stirred for 1 h and centrifuged (5,000 × g/ 5 min)11. TPC was quantified by the Folin-Ciocalteu reagent method as modified by Singleton et al.13. Readings were per- formed in a microplate reader (FlexStation 3; Molecular Devices, CA, USA) and the absorbance was recorded at 750 nm. TPC was calculated from the equation of a standard curve of gallic acid ranging from 5 to 130 µg. mL−1 (R2 = 0.9982, y = 0.0102x—0.0215). TPC was expressed as mg of gallic acid equivalents (GAE) per 100 g of dry sample. Antioxidant capacity. ABTS assay. ABTS assay was based on Ngoh and Gan10 with slight modifications. Biological properties of almond proteins produced by aqueous and enzyme‑assisted aqueous extraction processes from almond cake The search for bioactive compounds (i.e., bioactive peptides) in agricultural and industrial byproducts has been increasing, as they bear substances with properties of interest to the food and pharmaceutical industries12.hf The overall goal of this study was to evaluate the effects of the aqueous (AEP) and enzyme-assisted aqueous extraction processes (EAEP) on the biological properties of the skim fraction (protein-rich fraction) produced from the almond cake. The specific objectives of this study were to evaluate the in vitro protein digestibility, the total phenolic content, and the bioactive properties (antioxidant capacity, inhibitory effect against α-glucosidase, and pancreatic lipase, and antimicrobial activity) of the skim fractions generated from the AEP and EAEP of almond cake. These properties were compared with the skim fraction produced by the AEP of the hexane-defatted almond cake (used as a control). Therefore, this work aimed to evaluate the effects of two environmentally friendly extraction processes (AEP and EAEP) to generate proteins with improved biological functions from an underutilized food byproduct, ultimately leading to potential health benefits. www.nature.com/scientificreports/ with 80 µL of fluorescein (78 nM) followed by 40 µL of APPH (221 mM). The fluorescence was measured every minute for 90 min (excitation—485 nm; emission—535 nm). A calibration curve using Trolox solutions (10– 120 µM) (R2 = 0.9891, y = 62.444x + 734.12) was used and the antiradical activity was expressed as µMol of Trolox equivalent (TE) per g of sample using Eq. 1. with 80 µL of fluorescein (78 nM) followed by 40 µL of APPH (221 mM). The fluorescence was measured every minute for 90 min (excitation—485 nm; emission—535 nm). A calibration curve using Trolox solutions (10– 120 µM) (R2 = 0.9891, y = 62.444x + 734.12) was used and the antiradical activity was expressed as µMol of Trolox equivalent (TE) per g of sample using Eq. 1. (1) ORAC µM TE g−1 = (AUCs −AUCb) −b a C , (1) where AUCs and AUCb are the areas under the curves of the sample and blank, respectively; a is the intercept and b is the slope from the standard curve, and C is the sample concentration tested in the experiment. Enzymatic inhibition. α‑Glucosidase inhibition assay. The α-glucosidase inhibitory activity was evalu- ated according to Ibrahim et al.15, with slight modifications. Samples diluted in water (50 μL) at final concentra- tions of 0.5 to 40 mg mL−1 or ultrapure water (control) were incubated with 25 μL of 0.5 U.mL−1 α-glucosidase solution in PBS (100 mM, pH 6.8), at 37 °C, for 1 h. After pre-incubation, 25 μL of p-NPG substrate solution (5 mM) in PBS (100 mM, pH 6.8) was added, and the mixture was incubated at 37 °C, for 30 min. One hundred microliters of glycine–NaOH buffer (pH 10.0) were added to stop the reaction. The absorbance was measured at 405 nm in a microplate reader. The inhibitory activity was expressed as a percentage of a control sample (without the inhibitors). The α-glucosidase inhibitory activity was calculated by using Eq. 2. (2) α −glucosidase inhibitory activity (%) = 1 −As Ac × 100 , (2) where As and Ac are absorbances of sample and control, respectively. Pancreatic lipase inhibition assay. The pancreatic lipase inhibitory activity was evaluated based on the method described by McDougall et al.16, adapted to a 96-well microplate. www.nature.com/scientificreports/ Samples (15 μL) at final concentrations of 5 to 10 mg mL−1 or ultrapure water (control) were incubated with 60 μL of pancreatic lipase (10 mg mL−1) solution in Tris–HCl buffer (100 Mm, pH 8.2), at 37 °C, for 30 min. After pre-incubation, 135 μL of p-NP-laurate (2.5 mM, in 5 mM sodium acetate buffer, pH 5.0, with 1% Triton X-100) was added to start the reaction and was incubated at 37 °C, for 2 h. The absorbance was measured at 405 nm using a microplate reader. The inhibitory activity was expressed as a percentage of a control sample (without the inhibitors). The pancreatic lipase inhibitory activity was calculated by using Eq. 3. (3) pancreatic lipase inhibitory activity (% ) = 1 −As Ac × 100, (3) where As and Ac are absorbances of sample and control, respectively. where As and Ac are absorbances of sample and control, respectively. Materials and methods Samples (10 µL), at concentrations ranging from 0.5 to 5 mg mL−1 (in distilled water) were mixed with 190 µL of diluted ABTS reagent. Absorbance was recorded at 720 nm, in a microplate reader, at 37 °C, in triplicate. Antioxidant activity was calculated from the equation of a standard curve of Trolox ranging from 50 to 750 µM (R2 = 0.979, y = -0.0004x—0.0117). The antiradical activity was expressed as µMol of Trolox equivalent per g of sample. ORAC assay. The ORAC method was performed according to Zulueta et al.14. Samples (80 µL) at concentra- tions ranging from 0.025 to 0.200 mg mL−1 (in distilled water), PBS (blank) or Trolox (standard) were mixed Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ stains, in the range of 0.5 to 70 mg mL−1 (final concentration in well) were added in each well. The microplates were incubated at 37 °C, for 24 h. Bacterial growth was estimated by absorbance readings at 660 nm. Table 1. Composition of the simulated fluids. Simulated Saliva Fluid (SSF) Final concentration (mg mL−1) Final pH Mucin 1.0 7.0 NaCl 0.117 KCl 0.149 NaHCO3 2.1 Simulated Gastric Fluid (SGF) Final concentration (mg mL−1) Final pH Pepsin 0.75 (2000 U/mL) 1.8–2.0 Gastric mucin 1.5 NaCl 8.78 Simulated Intestinal Fluid (SIF) Final concentration (mg mL−1) Final pH Pancreatin 8.0 (800 U/mL) 7.0 Bile extract 10.0 NaHCO3 16.8 Table 1. Composition of the simulated fluids. Simulated Saliva Fluid (SSF) Final concentration (mg mL−1) Final pH Mucin 1.0 7.0 NaCl 0.117 KCl 0.149 NaHCO3 2.1 Simulated Gastric Fluid (SGF) Final concentration (mg mL−1) Final pH Pepsin 0.75 (2000 U/mL) 1.8–2.0 Gastric mucin 1.5 NaCl 8.78 Simulated Intestinal Fluid (SIF) Final concentration (mg mL−1) Final pH Pancreatin 8.0 (800 U/mL) 7.0 Bile extract 10.0 NaHCO3 16.8 Table 1. Composition of the simulated fluids. Table 1. Composition of the simulated fluids. tains, in the range of 0.5 to 70 mg mL−1 (final concentration in well) were added in each well. The microplates were incubated at 37 °C, for 24 h. Bacterial growth was estimated by absorbance readings at 660 nm. Digestibility. In vitro protein digestibility. Protein digestibility was measured as described by Roman et al.21 and Bornhorst and Singh22. The composition of the digestive solutions is presented in Table 1. Five grams of liquid skim fractions were mixed with 3.33 mL of SSF (Simulated Saliva Fluid) and vortexed. Subsequently, 6.66 mL of SGF (Simulated Gastric Fluid) was added. Afterward, the pH was adjusted to 3.0 and the samples were placed into a water-bath (37 °C, 140 rpm, 2 h). Then, 10 mL of SIF (Simulated Intestinal Fluid) was added, and the pH was adjusted to 7.0. The samples were incubated into a water-bath at 37 °C, 140 rpm, for 2 h. To stop the digestion, samples were heated in a water bath at 85 °C for 3 min. TCA (12 g.100 g−1) was added in a 1:1 (v/v) proportion and the samples were centrifuged at 4,000 rpm for 30 min at 4 °C. www.nature.com/scientificreports/ Total nitrogen (NT) and nonpro- tein nitrogen (NPN)—soluble fraction after TCA (12 g.100 g−1) precipitation—were measured in the samples, by the Dumas method using a conversion factor of 5.18 (Vario MAX cube, HE, DE) before and after the digestion. In vitro protein digestibility was calculated by using Eq. 423. (4) D(%) = NPNafter −NPNbefore −NPNenzymes(blank) NTbefore −NPNbefore x100 (4) where NPNafter = protein after digestion, NPNbefore = protein before digestion, NPNenzyme = enzyme blank and NTbefore = total protein before digestion. where NPNafter = protein after digestion, NPNbefore = protein before digestion, NPNenzyme = enzyme blank and NTbefore = total protein before digestion. In vitro digestibility evaluation by SDS‑page. SDS-PAGE was used to evaluate different stages during the diges- tion of each skim. The protein profile was assessed as described by Laemmli24 with few modifications. Thirty micrograms of protein were loaded onto the precast 12% acrylamide Criterion TGX Precast gel. A low range SDS-PAGE standard (14.4–97.4 kDa) was used as a molecular mass marker. The gel was imaged using a Gel Doc EZ Imager system and Image Lab software (Bio-Rad Laboratories, CA, USA). Statistical analyses. The experiments were performed at least in triplicate, and the results were expressed as mean ± standard deviation (SD) of the replicates. Analyses of variance (ANOVA) were performed, followed by the Tukey test using Graph Pad Prism 5.0 (version 5.04, GraphPad Software, CA, USA). Associations between antioxidant capacity and phenolic compounds were assessed by Pearson correlation. Significant differences were established at p < 0.05. where As and Ac are absorbances of sample and control, respectively. Kinetics of enzyme inhibition. To identify the type of inhibition exerted by the protein extracts on α-glucosidase and pancreatic lipase, an enzyme inhibition kinetic experiment was performed according to Ibrahim et al.15. For the α-glucosidase and pancreatic lipase inhibition assays, a range of concentrations from 0.15 to 5.0 mM of p-NPG and 0.05 to 2.5 mM of p-NP-Laurate was used. Lineweaver–Burk plots were used to determine the kinetic constants, Km (Michaelis constant) and Vmax (maximum velocity). In silico theoretical hydrolysis. In silico digestion of prunin-1 and prunin-2 sequences by the subtilisin enzyme (EC 3.4.21.62) was conducted using the enzymatic action tool incorporated into the BIOPEP platform17. Prunin-1 and prunin-2 sequences were taken from the UniProt platform (accession number Q43607 and E3SH29, respec- tively). The antioxidant peptides were evaluated by the BIOPEP biological activity database limiting the activities evaluated to antioxidants. For α-glucosidase and pancreatic lipase inhibitory peptide profiles, the criteria sug- gested by Ibrahim et al.18 and Ngoh and Gan10 was used. Antimicrobial assay. For antimicrobial assays, the gram-positive bacterial strains Staphylococcus aureus (NCQS 00,402), Bacillus cereus (NCQS 00,445) and Listeria monocytogenes (NCQS 00,673) and the gram-negative bac- teria strains Escherichia coli (NCQS 00,595) and Salmonella enterica subsp. enterica (NCQS 00,236). The bac- terial strains were consecutively sub-cultured with 24 h intervals. After the activation, were made a bacterial suspension, which was adjusted to 0.5 McFarland scale. Agar disk diffusion method. The antimicrobial effects were firstly determined by the agar disk diffusion method according to Balouiri et al.19 and Kim et al.20, with modifications. Filter paper discs were placed on the tryptic soy agar surface. Ten μL of the diluted sample in water (20 mg mL−1), the standard (amoxicillin, 2 mg mL−1) or the blank (saline solution) were added on top of discs. Petri dishes were incubated for 24 h and observed. A clear zone (halo) characterizes a positive result and no halo formed means negative result for bacterial growth inhibition. Broth dilution method. The broth microdilution test according to Balouiri et al.19 and Kim et al.20 was per- formed with modifications. Serial two-fold micro dilutions were prepared in a 96-well microplate. Fifty μL of the sample (diluted in water) or the standard (amoxicillin) and the same volume of overnight bacterial suspension Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 Results and discussion Different capital letters indicate significant difference (p < 0.05) among the solvents (water, ethanol, and methanol) for the same skim and different lowercase letters indicate significant difference (p < 0.05) between the different skims (AEP, EAEP, and HEX-AEP) within the same solvent; (b) Antioxidant capacity evaluated by the ORAC and by the ABTS methods. Different capital letters indicate a significant difference (p < 0.05) between the antioxidant activity analyses (ORAC and ABTS) for the same skim and different lowercase letters indicate a significant difference (p < 0.05) between the different skims for the same antioxidant activity evaluation method. Data represent the mean ± one standard deviation of triplicates. the higher the extractability (Polarity Index: water = 9.0; methanol = 6.6 and ethanol = 5.2)28. A similar trend was also found by Gomaa29, where higher TPC yields in bitter almonds and sweet apricots were achieved when using water as a solvent, compared to methanolic and ethanolic extracts. In a previous review regarding almond polyphenols, Bolling et al.26 showed that the most abundant polyphenols were proanthocyanidins (epicatechin and catechin), hydrolyzable tannins (gallotannins, ellagitannins, and phlorotannins), and flavonoids (anthocyani- dins, flavan-3-ols, flavonols, flavanones, and biflavone). Proanthocyanidins have also been identified as primary phenolic compounds in whole almonds by other authors30. Zam et al.31 evaluated TPC at pomegranate’s peel and observed that water archived the highest extraction yield of polyphenols in general and of proanthocyanins in particular. The high extractability when using water as solvent can be related to the weakening of the hydrogen bonds in aqueous solutions32. This observation may explain the higher solubility of the phenolic compounds from the almond cake skim fraction in water than in the other solvents.l In addition to the solvent used, the use of enzyme or not during the extraction also influenced the total recov- ery of phenolic compounds from the skims. The EAEP skim showed the highest yields of TPC compared with the AEP, which is likely due to the effect of the protease action during the extraction. Phenolic compounds can bind to proteins and carbohydrates, therefore hydrolysis of these complexes by proteases might help the release of entrapped phenolics33. Pinelo et al.33 studied the release of phenolic compounds from apple skin and observed that proteases favored their release. Results and discussion In the present study, the phenolic compounds that were inaccessible to the AEP and HEX-AEP skims were possibly released after the addition of the protease in the EAEP, enabling their solubilization into the solvent. Among the sample, HEX-AEP presented the lowest phenolic content in water. This might be related to the previous fat removal of HEX-AEP, where the sample was exposed to a mild heat treatment at ~ 68 °C and extraction temperature above 65 °C may lead to phenolics degradation34. Antioxidant activity. The antioxidant capacity was assessed in the extracted proteins using the ABTS and ORAC assays (Fig. 1b). EAEP skim showed the highest antioxidant capacity for both ORAC (397.2 µmol TE per g) and ABTS (650.5 µmol TE per g) methods. HEX-AEP (163.5 µmol TE per g) showed higher antioxidant capacity followed by AEP (108.7 µmol TE per g) by the ORAC method, while HEX-AEP (226.7 µmol TE per g) and AEP (221.7 µmol TE per g) showed no significant difference by the ABTS method. Although both methods followed the same trend and presented an extremely high (R2 = 0.98) correlation coefficient by Pearson’s evalu- ation, in general, the values found by the ABTS assay were higher than by the ORAC method. This variation is probably related to the differences in the antioxidant capacities measured by the two methods. ABTS or TEAC (Trolox equivalent antioxidant capacity) is a method based on the measurement of the electron transfer capacity of the antioxidant evaluated whereas ORAC (oxygen radical antioxidant capacity) is an assay that is based on the quantification of the hydrogen atom transfer capacity of the antioxidant evaluated14. Thus, the same antioxidant mixture may have a high capacity to transfer electrons and a low capacity to transfer hydrogen atoms or vice versa, resulting in significantly different values for the two methods tested. EAEP showed 3.6 and 2.4 times more antioxidant capacity than AEP and HEX-AEP skins respectively when the ORAC method was applied, whereas this feature was 2.9 times higher than AEP and HEX-AEP when the ABTS assay is considered. g y Higher antioxidant capacity of EAEP skim can be in part explained by the highest TPC in this sample (Fig. 1a), as proanthocyanidins, already identified in whole almonds, are recognized for increasing antioxidant capacity26,30 and, according to Pearson’s correlation coefficient, ABTS (R2 = 0.94) and ORAC (R2 = 0.87) assays strongly correlated with the TPC for each sample. Results and discussion Total phenolic content (TPC). The effects of different extraction conditions on the content of phenolic compounds in the skims are shown in Fig. 1a. It can be observed that the EAEP generated a skim fraction with the highest TPC, regardless of the solvent used to extract the phenolics. Among the solvents evaluated, higher phenolic extraction was achieved when using water as a solvent. TPC in the skims ranged from 7.7 to 342 mg GAE/100 g dry basis, depending on the solvent used and the type of extraction used. Using water as a solvent, TPC of 342; 219.7; and 161.4 mg GAE per 100 g dry sample were observed for the EAEP, AEP, and HEX-AEP skim fractions, respectively. In the present study, the cold pressed cake used was produced from whole natu- ral almonds. When comparing with the literature, our results are in agreement with the values obtained by Mil- bury et al.25, who observed concentrations from 126.8 to 240.8 mg GAE per 100 g, depending on the almond variety; and with Bolling26, that reported 190 mg GAE per 100 g, both studies on the whole almond.hff The solvent polarity plays an important role in the recovery of polyphenols from different matrices, as it affects the solubility of the phenolic compounds27 and, according to Abarca-Vargas et al.28, water, ethanol, methanol, and their mixtures are the most frequently solvents used for TPC extraction. In this study, TPC yield related to the solvent as follows: water > methanol > ethanol. The results suggested that the higher the polarity of the solvent, https://doi.org/10.1038/s41598-020-67682-3 Scientific Reports \| (2020) 10:10873 \| www.nature.com/scientificreports/ Figure 1. (a) Total phenolic content extracted by water, ethanol, and methanol. Different capital letters indicate significant difference (p < 0.05) among the solvents (water, ethanol, and methanol) for the same skim and different lowercase letters indicate significant difference (p < 0.05) between the different skims (AEP, EAEP, and HEX-AEP) within the same solvent; (b) Antioxidant capacity evaluated by the ORAC and by the ABTS methods. Different capital letters indicate a significant difference (p < 0.05) between the antioxidant activity analyses (ORAC and ABTS) for the same skim and different lowercase letters indicate a significant difference (p < 0.05) between the different skims for the same antioxidant activity evaluation method. Data represent the mean ± one standard deviation of triplicates. Figure 1. (a) Total phenolic content extracted by water, ethanol, and methanol. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Figure 2. (a) α-glucosidase inhibitory activity of EAEP skim. Different letters indicate a significant difference (p < 0.05) among the concentrations evaluated. (b) Lipase inhibitory activity of AEP, EAEP, and HEX-AEP skims. Different capital letters indicate a significant difference (p < 0.05) among the concentrations evaluated (5, 10 and 20 mg mL−1) for the same skim and different lowercase letters indicate a significant difference (p < 0.05) among AEP, EAEP, and HEX-AEP skims for the same concentration. Data represent the mean ± one standard deviation of triplicates. Figure 2. (a) α-glucosidase inhibitory activity of EAEP skim. Different letters indicate a significant difference (p < 0.05) among the concentrations evaluated. (b) Lipase inhibitory activity of AEP, EAEP, and HEX-AEP skims. Different capital letters indicate a significant difference (p < 0.05) among the concentrations evaluated (5, 10 and 20 mg mL−1) for the same skim and different lowercase letters indicate a significant difference (p < 0.05) among AEP, EAEP, and HEX-AEP skims for the same concentration. Data represent the mean ± one standard deviation of triplicates. The protease used for EAEP processing might have hydrolyzed the proteins present in the almond cake and later in the skim fraction, generating peptides with antioxidant activities. In fact, the theoretical hydrolysis (Supplementary Table S1) of the most important storage protein of almonds (amandin) revealed the formation of 4 different antioxidant peptides (HL, IY, VY, PHW). Some studies have identified antioxidant peptides from several plant-based proteins, such as peanuts8 and okara35. These studies correlated the antioxidant capacity with the degree of hydrolysis (DH %), whereas the higher DH %, the higher the antioxidant capacity. The DH % of the skims were evaluated in a previous study from our group and 23, 1.8, and 1.3% were achieved for the EAEP, AEP, and HEX-AEP skim fractions. The higher DH % of the EAEP skim may have contributed to the higher antioxidant capacity observed, which is in agreement with the studies above cited. AEP and HEX-AEP skims showed no significant differences for DH %, nor for TCP, when methanol and ethanol were applied, or for the antioxidant capacity measured by the ABTS assay.if To the best of our knowledge, this is the first study to investigate the effects of extraction conditions on the antioxidant properties of almond cake protein extracts. Results and discussion Thus, another possible explanation for the higher activity in EAEP skim may be the formation of potential antioxidant peptides during the enzymatic extraction (Fig. 4). Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 www.nature.com/scientificreports/ Interestingly, our results showed the use of an enzyme in the EAEP can improve the antioxidant capacity of the extracted protein, as revealed by the ABTS and ORAC assay results. α‑glucosidase and pancreatic lipase inhibition. A therapeutic approach to prevent diabetes and obe- sity is to inhibit digestive enzymes such as α-glucosidase and pancreatic lipase. Inhibitors targeting these enzymes prevent the uptake of glucose from complex dietary carbohydrates and free fatty acids from triacylglycerols to be absorbed into the body9,15. Preliminary tests evaluated potential inhibitory effects of the extracted proteins against α-glucosidase (Fig. 2a) or pancreatic lipase (Fig. 2b) as well as the lowest concentration that exerted the highest inhibition. It seems evident that, among the samples tested, only the EAEP skim presented α-glucosidase inhibitory activity (Fig. 2a). In addition to the concentrations evaluated in Fig. 2a, higher concentrations of AEP and HEX-AEP skims (10, 20, and 40 mg mL−1) were also evaluated but showed no activity. The EAEP skim, on the other hand, showed high inhibition, even in the lowest concentration tested: 0.5 mg mL−1 inhibited 84.6% of the α-glucosidase activity while 2.0 mg mL−1 inhibited 97.8% of the α-glucosidase activity, with no significant improvement for higher concentrations (5.0 mg mL−1—98.9%). Unlike α-glucosidase, the pancreatic lipase was inhibited by all the samples tested (Fig. 2b). AEP skim showed the highest dose-dependent inhibitory activity followed by HEX-AEP skim, which showed little over half of AEP’s inhibitory activity. EAEP skims showed the lowest inhibition of the pancreatic lipase, which remained unchanged (p > 0.05) regardless of the concentration tested. A kinetic study was performed to verify the type of inhibition taking place in the skim fraction, applying the least concentration of each sample that provided the highest inhibition (Fig. 3). h g g The enzymes tested in this study—α-glucosidase and pancreatic lipase—are a part of the digestion of car- bohydrates and triacylglycerols in the digestive tract35. Alpha-glucosidase is involved in the digestion of starch, catalyzing the hydrolysis of oligo- and disaccharides to release free molecules of glucose9,15. Therefore inhibi- tors of α-glucosidase aid to control postprandial hyperglycemia by modulating the digestion rate of complex carbohydrates such as starch and may be employed in the prevention of diabetes and other associated diseases15. www.nature.com/scientificreports/ Figure 3. (a) Lineweaver–Burk plots of α-glucosidase inhibition for control (no inhibitor) and EAEP skim (2 mg mL−1); (b) Lineweaver–Burk plots of lipase inhibition for control (no inhibitor) and AEP (20 mg mL−1); (c) EAEP (5 mg mL−1) and (d) HEX-AEP (20 mg mL−1) skims. Figure 3. (a) Lineweaver–Burk plots of α-glucosidase inhibition for control (no inhibitor) and EAEP skim (2 mg mL−1); (b) Lineweaver–Burk plots of lipase inhibition for control (no inhibitor) and AEP (20 mg mL−1); (c) EAEP (5 mg mL−1) and (d) HEX-AEP (20 mg mL−1) skims. Ibrahim et al.18 identified the structural and physicochemical requirements to design an active α-glucosidase inhibitory peptide. According to their findings, α-glucosidase inhibitory activity is related to short peptides, with 3 to 6 amino acid residues, containing a hydroxyl or basic side-chain amino acid at the N-terminus and a proline residue closer to the C-terminus with methionine or alanine occupying the final C-terminal position. The theoretical hydrolysis of amandin subunits (Supplementary Table S2) showed the formation of 13 different peptides with at least one of the above characteristics. It is therefore possible that, after proteolysis, some bioac- tive peptides were released from the almond cake proteins, leading to the high inhibitory activity of EAEP skim against α-glucosidase. Awosika and Aluko9 reported that 20 mg mL−1 of yellow field pea peptides inhibited up to 53.3% α-glucosidase activity and Oseguera-Toledo et al.37 evaluated peptide fractions from pinto Durango beans and observed inhibitory activity of 76.4% against α-glucosidase. This study is, however, the first report of such activity for almond protein hydrolysates from the almond cake. y p y y Pancreatic lipase is responsible for the digestion and absorption of dietary fat through the hydrolysis of the triacylglycerols to glycerol and fatty acids9,36. The inhibition of pancreatic lipase prevents the breakdown of dietary fat into fatty acids, therefore reducing their absorption in the gut, which may be a viable approach to the control of hyperlipidemia and obesity36. Some in vitro studies reported plants with a high concentration of fat or high content of tannins as pancreatic lipase inhibitors16,36. Proanthocyanidins, already reported in almonds30, are among the primary active tannins and their activity was attributed to their ability to bind proteins, leading to the complexation and precipitation of the enzymes38. The same authors, however, claim that the mechanism of polyphenolic compounds on pancreatic lipase inhibition remains unclear. www.nature.com/scientificreports/ In this study, 2 mg mL−1 of EAEP skim inhibited ~ 98% of the α-glucosidase activity, while AEP and HEX-fh The enzymes tested in this study—α-glucosidase and pancreatic lipase—are a part of the digestion of car- bohydrates and triacylglycerols in the digestive tract35. Alpha-glucosidase is involved in the digestion of starch, catalyzing the hydrolysis of oligo- and disaccharides to release free molecules of glucose9,15. Therefore inhibi- tors of α-glucosidase aid to control postprandial hyperglycemia by modulating the digestion rate of complex carbohydrates such as starch and may be employed in the prevention of diabetes and other associated diseases15. In this study, 2 mg mL−1 of EAEP skim inhibited ~ 98% of the α-glucosidase activity, while AEP and HEX- AEP skims did not present an inhibitory effect. These results may be explained by the generation of bioactive peptides, through the alkaline protease activity in the enzymatic extraction (EAEP) compared with the aqueous extraction process (AEP nor HEX-AEP) where no enzyme was used during the extraction. Bioactive peptides are encrypted within the primary structure of proteins and may be released by proteolysis. These peptides may act as inhibitors of metabolic enzymes and present potential use as therapeutic agents against specific diseases. y y p y p In this study, 2 mg mL−1 of EAEP skim inhibited ~ 98% of the α-glucosidase activity, while AEP and HEX- AEP skims did not present an inhibitory effect. These results may be explained by the generation of bioactive peptides, through the alkaline protease activity in the enzymatic extraction (EAEP) compared with the aqueous extraction process (AEP nor HEX-AEP) where no enzyme was used during the extraction. Bioactive peptides are encrypted within the primary structure of proteins and may be released by proteolysis. These peptides may act as inhibitors of metabolic enzymes and present potential use as therapeutic agents against specific diseases. Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 www.nature.com/scientificreports/ www.nature.com/scientificreports/ In the present study, all skim frac- tions showed some degree of lipase inhibition: AEP skim exhibited the highest inhibitory percentage, followed by HEX-AEP and EAEP skims. The TPC in this same samples did not follow this trend; on the contrary, EAEP skim showed the highest TPC of all samples (Fig. 1a). g p ( g ) Information about lipase inhibition by proteins or peptides is scarce in the literature, and there is even less information on pancreatic lipase activity inhibition by food protein-derived peptides9. However, a few recent studies reported peptides as pancreatic lipase inhibitors: Ngoh and Gan10 found bioactive peptides in pinto beans that inhibited the lipase activity in a range between 23 to 87% and Stefanucc et al.39, who discovered novel tripeptides as lipase inhibitors, observed inhibitions from 50 to 100 mg Orlistat (i.e., standard drug to treat obesity) equivalent per g of sample, depending on the peptide sequences. Ngoh and Gan10 verified, through docking analysis, that most of the amino acid residues of the peptides involved in pancreatic lipase inhibition were hydrophobic amino acids such as proline (P), leucine (L), glycine (G), phenylalanine (F), alanine (A) and Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 www.nature.com/scientificreports/ Fi 4 SDS PAGE (12% C it i TGX P t G l ) l i f th diff t t f i it di ti Figure 4. SDS-PAGE (12% Criterion TGX Precast Gels) analysis of the different stages of in vitro digestion of EAEP, AEP, and HEX-AEP. Samples derive from the same experiment and the gel was processed in parallel. The gel was imaged with a Gel Doc EZ (Bio-Rad, USA) using the Image Lab software. lane (1) Molecular mass marker standards (14.4–97.4 kDa); lane (2) AEP total protein; lane (3) AEP plus SSF and SGF; lane (4) AEP plus SSF, and SGF after 2 h reaction; lane (5) AEP plus SSF, SGF, and SIF after 2 h reaction; lane (6) EAEP total protein; lane (7) EAEP plus SSF and SGF; lane (8) EAEP plus SSF, and SGF after 2 h reaction; lane (9) EAEP plus SSF, SGF, and SIF after 2 h reaction; lane (10) HEX-AEP total protein; lane (11) HEX-AEP plus SSF and SGF; lane (12) HEX-AEP plus SSF, and SGF after 2 h reaction; lane (13) HEX-AEP plus SSF, SGF, and SIF after 2 h reaction; lane (14) MW marker standards (14.4–97.4 kDa). Figure 4. www.nature.com/scientificreports/ SDS-PAGE (12% Criterion TGX Precast Gels) analysis of the different stages of in vitro digestion of EAEP, AEP, and HEX-AEP. Samples derive from the same experiment and the gel was processed in parallel. The gel was imaged with a Gel Doc EZ (Bio-Rad, USA) using the Image Lab software. lane (1) Molecular mass marker standards (14.4–97.4 kDa); lane (2) AEP total protein; lane (3) AEP plus SSF and SGF; lane (4) AEP plus SSF, and SGF after 2 h reaction; lane (5) AEP plus SSF, SGF, and SIF after 2 h reaction; lane (6) EAEP total protein; lane (7) EAEP plus SSF and SGF; lane (8) EAEP plus SSF, and SGF after 2 h reaction; lane (9) EAEP plus SSF, SGF, and SIF after 2 h reaction; lane (10) HEX-AEP total protein; lane (11) HEX-AEP plus SSF and SGF; lane (12) HEX-AEP plus SSF, and SGF after 2 h reaction; lane (13) HEX-AEP plus SSF, SGF, and SIF after 2 h reaction; lane (14) MW marker standards (14.4–97.4 kDa). methionine (M). The theoretical hydrolysis of amandin (Supplementary Table S3) showed the formation of 17 highly hydrophobic peptides, due to the presence of these amino acid residues. Awosika and Aluko9 evaluated yellow field pea protein hydrolysates and observed that trypsin and alcalase generated peptides with higher lipase inhibition capacity than chymotrypsin and pepsin. Alcalase is a subtilisin similar to the alkaline protease used in the EAEP. Although the EAEP skim showed the highest TPC and peptide content, this sample was not the most efficient for lipase inhibition. This inefficiency may have occurred due to the high degree of hydrolysis (DH = 23%) of the EAEP skim, which possibly generated peptides with a different profile from those naturally present in the AEP and HEX-AEP skims (Fig. 4). These results indicate that the AEP skim is a good candidate for use as a pancreatic lipase inhibitor, while the EAEP skim could be used as a highly efficient α-glucosidase inhibitor. Kinetics of enzyme inhibition. Kinetics experiments were performed to reveal the mechanisms of action of the different skim samples on α-glucosidase and pancreatic lipase inhibition, applying the least concentration of each sample that provided the highest inhibition (Fig. 3): for α-glucosidase, 2 mg mL−1 of the EAEP skim was applied; for lipase 20 mg mL−1 of the AEP or of HEX-AEP skims and 5 mg mL−1 of EAEP skim were tested. www.nature.com/scientificreports/ Lineweaver–Burk plots revealed that EAEP skim behaved as an uncompetitive inhibitor for both, α-glucosidase and pancreatic lipase (Table 2). Likewise, AEP and HEX-AEP skims also behaved as uncompetitive inhibitors of pancreatic lipase (Table 2). The results showed that the lines (control vs. inhibitors) intersected at both y-axis Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 www.nature.com/scientificreports/ Table 2. Apparent Km (Michaelis constant) and Vmax (maximum velocity) and mode of inhibition of α-glucosidase and pancreatic lipase. Different letters indicate a significant difference (p < 0.05) among the different samples and control for Km or Vmax. Data represent the mean ± standard deviation of triplicates. α-glucosidase Km Vmax Mode of inhibition Control 8.67 ± 6.58a 2.81 ± 1.45a – EAEP skim (2 mg mL−1) 0.17 ± 0.02b 0.02 ± 0.004b Uncompetitive Lipase Km Vmax Mode of inhibition Control 0.24 ± 0.01a 0.68 ± 0.01a – AEP skim (20 mg mL−1) 0.17 ± 0.01b 0.62 ± 0.02b Uncompetitive EAEP skim (5 mg mL−1) 0.19 ± 0.02b 0.59 ± 0.02b Uncompetitive HEX-AEP skim (20 mg mL−1) 0.19 ± 0.01b 0.61 ± 0.02b Uncompetitive Table 2. Apparent Km (Michaelis constant) and Vmax (maximum velocity) and mode of inhibition of α-glucosidase and pancreatic lipase. Different letters indicate a significant difference (p < 0.05) among the different samples and control for Km or Vmax. Data represent the mean ± standard deviation of triplicates. α-glucosidase Km Vmax Mode of inhibition Control 8.67 ± 6.58a 2.81 ± 1.45a – EAEP skim (2 mg mL−1) 0.17 ± 0.02b 0.02 ± 0.004b Uncompetitive Lipase Km Vmax Mode of inhibition Control 0.24 ± 0.01a 0.68 ± 0.01a – AEP skim (20 mg mL−1) 0.17 ± 0.01b 0.62 ± 0.02b Uncompetitive EAEP skim (5 mg mL−1) 0.19 ± 0.02b 0.59 ± 0.02b Uncompetitive HEX-AEP skim (20 mg mL−1) 0.19 ± 0.01b 0.61 ± 0.02b Uncompetitive Table 2. Apparent Km (Michaelis constant) and Vmax (maximum velocity) and mode of inhibition of α-glucosidase and pancreatic lipase. Different letters indicate a significant difference (p < 0.05) among the different samples and control for Km or Vmax. Data represent the mean ± standard deviation of triplicates. Table 2. Apparent Km (Michaelis constant) and Vmax (maximum velocity) and mode of inhibition of α-glucosidase and pancreatic lipase. Different letters indicate a significant difference (p < 0.05) among the different samples and control for Km or Vmax. Data represent the mean ± standard deviation of triplicates. www.nature.com/scientificreports/ and x-axis, at different points, in the Lineweaver–Burk plots (Fig. 3). This indicates that, in the presence of the inhibitor, both the Michaelis constant (Km) and the maximum velocity (Vmax) decreased when compared to the control. These results were confirmed by calculating the apparent Km and Vmax (Table 2). Inhibition of pan- creatic lipase by AEP and HEX-AEP, however, showed inhibition graphs that cross before reaching the y-axis. This kind of behavior was related by Park et al.40 to competitive inhibition, although without the corroboration of Km or Vmax values. For uncompetitive inhibition, the inhibitor binds the enzyme–substrate complex in an allosteric site and has no substrate-like structure. The uncompetitive inhibitor causes a structural distortion of the active and allosteric sites of the complexed enzyme, which prevents the catalysis to occur and results in a decrease in Km and Vmax values41. According to Leskovac42, uncompetitive inhibition works better in the presence of high concentra- tions of substrate, a condition likely to occur in the digestive tract during digestion. EAEP skim inhibition of α-glucosidase produced a Km over 50 times lower and a Vmax around 140 times lower than the control. These results indicate that the peptides generated by the proteolysis of almond protein showed a strong ability to bind to the α-glucosidase-substrate complex. Thereby, EAEP skim may be an excellent pool of peptides to delay the breakdown of dietary carbohydrates and consequently reduce the rate of glucose release at the small intestine15. Pancreatic lipase inhibition at AEP showed Vmax approximately 10% lower than the control and Km up to 40% lower than the control, showing a weak binding between inhibitors and enzyme–substrate complex.hi g g y p The uncompetitive inhibition of α-glucosidase by peptides was also observed by Ibrahim et al15, who identified two active α-glucosidase inhibitory peptides—SVPA and SEPA. SVPA acted as an uncompetitive inhibitor and SEPA was a non-competitive inhibitor against α-glucosidase. In contrast, Awosika and Aluko9 studied yellow field pea protein-derived peptides and found a non-competitive inhibition of α-glucosidase. The uncompetitive inhibition is considered as a rare type of inhibition that may occur in multimeric enzymes41. To the best of our knowledge, the first study on inhibition of pancreatic lipase by food-derived bioactive peptides was published by Ngoh and Gan10 and dealt with bean protein hydrolysates. Therefore, there is scant information relating peptides and pancreatic lipase inhibition, especially on the enzyme kinetics. www.nature.com/scientificreports/ At this stage, all samples showed similar profiles, with the higher molecular mass bands probably derived from the different proteins in the pancreatin solution added. The results indicate that both pepsin and pancreatic proteases were efficient in hydrolyzing almond cake proteins independent of the extraction method.h y y g p p Simulated digestion includes an oral, a gastric and an intestinal phase. These methods involve the use of digestive enzymes and their concentrations as well as the pH, digestion time, and salt concentrations to simulate physiological conditions in vivo46. After each digestion phase, samples must be analyzed to quantify the amount of the initial protein that was hydrolyzed by the digestive proteases added. There are several methods for chemi- cally determining the hydrolysis of proteins such as pH–stat, o-phthaldialdehyde (OPA) and trichloroacetic acid soluble nitrogen (SN-TCA) methods. The pH–stat method is based on the amount of protons released during hydrolysis; the OPA method is based on the measurement of amino groups generated by the hydrolysis and the SN-TCA method measures the amount of TCA-soluble nitrogen released by the hydrolysis47. The pH–stat and OPA methods are considered quick and simple methods for quantifying the degree of hydrolysis (DH) of protein hydrolysates and allow for real-time monitoring of the hydrolysis reaction as it proceeds. However, the accuracy of DH values obtained by the pH–stat and OPA methods depends on the type of enzyme activity and on the protein substrate, respectively. The pH–stat method may underestimate DH if the enzyme shows high exopepti- dase activity. In the OPA method, unstable products are formed by reacting with cysteine and proline residues and this method also requires a derivatization step47,48. The SN-TCA method does not determine the number of peptide bonds cleaved, so this method does not determine the DH directly, it measures the TCA-soluble nitrogen generated by the hydrolysis, which is supposed to consist only of amino acids and small peptides. The SN-TCA method is simple and rapid to perform, however, the assumption that TCA-soluble nitrogen is composed only by small peptides and amino acids may not be correct, since not all intact proteins are precipitated by TCA53. Considered a suitable method by several authors53,49−51, the SN-TCA method was applied in this study to compare the effect of either EAEP or AEP on the digestibility of the protein (skim fraction) extracted. www.nature.com/scientificreports/ dilution method. Nevertheless, the samples did not exhibit antimicrobial activity against the five strains tested. The theoretical in silico hydrolysis and BIOPEP search did not reveal any antibacterial, antiviral or antifungal peptides, corroborating the in vitro assays.f p p g y Different extracts (water, methanol, and ethanol) of bitter almonds were tested against human pathogenic bacteria by Gomaa29. The authors observed inhibition activity against only 4 of the 11 microorganisms tested. They attributed the significant inhibitory effect to the phenolic compounds in the extracts. The antimicrobial activity of almond skin43, almond oil44, and cold press edible oil byproduct11 has already been tested. According to the literature, a plant extract depends on which part of the plant was evaluated, the method and solvent used for extraction and finally the concentration tested to be a potent antimicrobial extract11. In vitro protein digestibility. The profile of each skim protein fraction during oral, gastric, and intesti- nal digestion was evaluated by SDS-PAGE (Fig. 4). SDS-PAGE showed similar initial protein profiles for AEP and HEX-AEP skims, presenting proteins with a molecular mass between 21.5 and 45 kDa, although some low molecular weight protein bands may be also observed (Fig. 4—Dig 01). On the other hand, for the EAEP skim, a concentration of low molecular mass proteins and peptides at 14.4 kDa and below became evident. Dig 02 (Fig. 4) shows the profile for samples that received salivary solution (SSF) plus gastric solution (SGF), immediately after addition, Dig 03 shows SSF plus SGF after 2 h. Overall, there was little difference in the same sample, between Dig 02 and Dig 03, indicating fast digestion of the extracted almond cake proteins by pep- sin. In a previous study regarding almond protein digestibility45, pepsin was able to hydrolyze the proteins in 5 min, indicating rapid hydrolysis capability. Although EAEP showed initially an already digested profile due to the partial hydrolysis caused by the use of protease during the extraction, the addition of pepsin seemed to reduce the peptide cluster at the bottom of the gel—below 14.4 kDa, indicating further digestion and formation of lower mass peptides. Dig 04 (Fig. 4) presents the profile of the samples containing salivary solution (SSF), gastric solution (SGF) and also the intestinal solution (SIF), after another 2 h reaction. www.nature.com/scientificreports/ f g y p Among the tested samples, the highest digestibility was found for the HEX-AEP (85%) skim, followed by the AEP skim (73%) and the EAEP skim (64%). The high digestibility regardless of the extraction is consistent with a previous True Protein Digestibility (% TPD) in vivo test from Ahrens et al.52, who found digestibility from 82 to 92% depending on the varieties of the whole almonds. Moreover, Sathe53 and Sze-Tao and Sathe45 also reported high in vitro almond protein digestibility. The high digestibility of HEX-AEP skim may be explained by the removal of the residual oil by hexane extraction (in a Soxhlet). During the hexane extraction, the almond cake was subjected to mild temperatures (68–70 °C) which might have resulted in protein denaturation. In some situations, some degree of protein denaturation might improve enzyme accessibility to the protein thus improving the digestibility54. Conversely, the EAEP skim presented the lowest digestibility despite its higher initial DH. Typi- cally, digestibility increases with protein hydrolysis55. However, the extensive hydrolysis achieved by the EAEP (DH = 23%) entailed fewer attack sites available to the digestion enzymes (pepsin and pancreatin), which led to an underestimation of this parameter55. Similar results were found by Betancur-Ancona et al.55, who reported higher digestibility in the raw materials (Phaseolus lunatus) than for the hydrolyzates produced. www.nature.com/scientificreports/ Most of the studies regarding lipase kinetics are related to phenolic compounds, such as the study from, Park et al.40 who identified a non-competitive or competitive inhibition against lipase at flavonol-3-O-glycosides and flavonol aglycones in Polygonum aviculare L., respectively. p y Overall, the EAEP skim presented a high inhibitory activity against α-glucosidase and AEP, EAEP, and HEX- AEP skims showed varying degrees of inhibitory activity against lipase. Unlike the α-glucosidase inhibitory effect, AEP skim showed higher inhibitory activity against pancreatic lipase compared with EAEP skim. However, it must be taken into account that the concentration of the EAEP skim (5 mg mL−1) was 4 times lower than the active concentration of the AEP skim (20 mg mL−1). The lower concentration needed by the EAEP skim can be addressed to its chemical reaction order. EAEP skim presented a zero-order behavior, which means a constant rate of inhibition, independent of the concentration tested. The zero-order kinetics occurs at the limit where the enzyme is saturated with substrate and an increase in the initial substrate concentration will have no effect on the rate of reaction, as no free enzyme is available42. The result indicates that the protein fraction extracted from the almond cake by the EAEP has the potential to be used as a source of bioactive protein/peptides to control hyperglycemia and obesity. EAEP skim can retard the release of glucose from complex dietary carbohydrates and could thus be subsequently evaluated regarding potential hyperglycemia reductions. All skim samples may partially suppress and delay the triacylglycerol digestion and consequently help to control hyperlipidemia and obesity15,36. Antimicrobial activity. The antimicrobial potency of the extracts against five bacteria strains was evalu- ated by agar disk-diffusion, which measures the formation of a halo, known as the zone of inhibition, and also by broth microdilution, that indicates the minimum inhibitory concentration (MIC). The samples were tested by agar disk-diffusion at a concentration of 20 mg mL−1. None of the skim fractions showed inhibition by this method, as they did not cause the formation of inhibition zones (halos). Although the positive control (amoxicil- lin) did. To eliminate the hypothesis that the concentration tested was too low to exert some antimicrobial activ- ity, all samples were tested against the same bacterial strains at the concentration of 70 mg mL−1, by the broth https://doi.org/10.1038/s41598-020-67682-3 Scientific Reports \| (2020) 10:10873 \| www.nature.com/scientificreports/ www.nature.com/scientificreports/ skim presented a similar profile as the HEX-AEP skim regarding their antioxidant capacity, TPC and digestibility, but provided the highest lipase inhibition potential, and therefore could be used in the prevention of obesity. Although further studies are required to characterize the active compounds and the mechanisms of action associ- ated with the observed bioactivities, the present study demonstrated that the almond cake can be transformed into value-added health-promoting products by the application of environmentally friendly extraction processes. skim presented a similar profile as the HEX-AEP skim regarding their antioxidant capacity, TPC and digestibility, but provided the highest lipase inhibition potential, and therefore could be used in the prevention of obesity. Although further studies are required to characterize the active compounds and the mechanisms of action associ- ated with the observed bioactivities, the present study demonstrated that the almond cake can be transformed into value-added health-promoting products by the application of environmentally friendly extraction processes. Received: 16 January 2020; Accepted: 8 June 2020 References 1. Kamil, A. & Chen, C.-Y.O. Health benefits of almonds beyond cholesterol reduction. J. Agric. Food Chem. 60, 6694–6702 (2012). 2. Bolling, B. W., Dolnikowski, G., Blumberg, J. B. & Chen, C.-Y.O. Polyphenol content and antioxidant activity of California almonds depend on cultivar and harvest year. Food Chem. 122, 819–825 (2010). 2. Bolling, B. W., Dolnikowski, G., Blumberg, J. B. & Chen, C.-Y.O. Polyphenol content and antioxidant activity of California alm depend on cultivar and harvest year. Food Chem. 122, 819–825 (2010). p y 3. Sathe, S. K. et al. Biochemical characterization of amandin, the major storage protein in almond (Prunus dulcis L.). J. Agric. Food Chem. 50, 4333–4341 (2002). 4. Souza, T. S. P., Dias, F. F. G., Koblitz, M. G. B. & M. L. N. de M. Bell, J. Aqueous and enzymatic extraction of oil and protein from almond cake: a comparative study. Processes 7, 472 (2019).h p y 5. Li, Z., Smith, K. H. & Stevens, G. W. The use of environmentally sustainable bio-derived solvents in solvent extraction applications a review. Chin. J. Chem. 24, 215–220 (2016). 6. Chemat, et al. Review of alternative solvents for green extraction of food and natural products: panorama, principles, applications and prospects. Molecules 24, 3007 (2019). p p 7. Balvardi, M., Rezaei, K., Mendiola, J. A. & Ibáñez, E. Optimization of the aqueous enzymatic extraction of oil from Iranian wild almond. J. Am. Oil Chem. Soc. 92, 985–992 (2015).l 8. Jamdar, S. N. et al. Influence of degree of hydrolysis on functional properties, antioxidant activity and ACE inhibitory activity o peanut protein hydrolysate. Food Chem. 121, 178–184 (2010).i 8. Jamdar, S. N. et al. Influence of degree of hydrolysis on functi l g y y peanut protein hydrolysate. Food Chem. 121, 178–184 (2010) p p y y 9. Awosika, T. O. & Aluko, R. E. Inhibition of the in vitro activities of α-amylase, α-glucosidase and pancreatic lipase by yellow field pea (Pisum sativum L.) protein hydrolysates. Int. J. Food Sci. Technol. 54, 2021–2034 (2019).i 9. Awosika, T. O. & Aluko, R. E. Inhibition of the in vitro activiti y y 10. Ngoh, Y. & Gan, C. Enzyme-assisted extraction and identification of antioxidative and a-amylase inhibitory peptides from Pinto beans. Food Chem. 190, 331–337 (2016). 11. Karaman, S. et al. Recovery potential of cold press by products obtained from the edible oil industry: physicochemical, bioactive, and antimicrobial properties. J. Agric. Food Chem. References 24 L li U K Cl f t t l t i d i th bl f th h d f b t i h T4 N t 227 680 685 (1970) tions and their relationship to in vitro and in vivo protein digestibility. J. Agric. Food Chem. 63, 6787–6798 (2015). 24 Laemmli U K Clea age of structural proteins during the assembl of the head of bacteriophage T4 Nature 227 680 685 (1970) p p g y J g , ( ) 24. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685 (1970). p p g y g 4. Laemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685 (1970) lb h l k k l b f fl d d h l d h d b aemmli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685 (1970). Milbury, P. E., Chen, C.-Y., Dolnikowski, G. G. & Blumberg, J. B. Determination of flavonoids and phenolics and their distribution mli, U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685 (1970). ury, P. E., Chen, C.-Y., Dolnikowski, G. G. & Blumberg, J. B. Determination of flavonoids and phenolics and their distribution monds. J. Agric. Food Chem. 54, 5027–5033 (2006). g p g y p g ( ) 5. Milbury, P. E., Chen, C.-Y., Dolnikowski, G. G. & Blumberg, J. B. Determination of flavonoids and phenolics and their distribution in almonds. J. Agric. Food Chem. 54, 5027–5033 (2006). 26. Bolling, B. W. Almond polyphenols: methods of analysis, contribution to food quality, and health promotion: almond polyphenols. Compr. Rev. Food Sci. Food Saf. 16, 346–368 (2017). p f 7. Romero-Díez, R. et al. Phenolic characterization of aging wine lees: correlation with antioxidant activities. Food Chem. 259 188–195 (2018). ( ) 8. Abarca-Vargas, R., Peña Malacara, C. & Petricevich, V. Characterization of chemical compounds with antioxidant and cytotoxic activities in bougainvillea x buttiana holttum and standl, (var. Rose) extracts. Antioxidants 5, 45 (2016). d b l d f b l d d ( 29. Gomaa, E. Z. In vitro antioxidant, antimicrobial, and antitumor activities of bitter almond and sweet apricot (Prunus armeniaca L.) kernels. Food Sci. Biotechnol. 22, 455–463 (2013). 30. Smeriglio, A., Barreca, D., Bellocco, E. Conclusionhf The effects of the aqueous (AEP) and enzyme-assisted aqueous extraction processes (EAEP) on the biological properties of the AEP and EAEP skims were evaluated. The use of an enzyme to assist the extraction of the almond cake resulted in the production of a skim fraction with higher TPC and the presence of bioactive pep- tides associated with increased antioxidant activity and inhibitory effects against α-glucosidases. The in silico theoretical hydrolysis indicated the presence of 4 antioxidant peptides, 13 peptides with structural requirements for α-glucosidase inhibition and 16 highly hydrophobic peptides, likely to inhibit pancreatic lipase. The AEP Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 www.nature.com/scientificreports/ Received: 16 January 2020; Accepted: 8 June 2020 References 63, 2305–2313 (2015). p p g 2. Tatullo, M. et al. Antioxidant and antitumor activity of a bioactive polyphenolic fraction isolated from the brewing process. Sci Rep. 6, 36042 (2016). p 3. Singleton, V. L., Orthofer, R. & Lamuela-Raventós, R. M. Analysis of total phenols and other oxidation substrates and antioxidants by means of folin-ciocalteu reagent. Method enzymol (Elsevier, London, 1999). 14. Zulueta, A., Esteve, M. J. & Frígola, A. ORAC and TEAC assays comparison to measure the antioxidant capacity of food products. Food Chem. 114, 310–316 (2009). 5. Ibrahim, M. A., Bester, M. J., Neitz, A. W. & Gaspar, A. R. M. Rational in silico design of novel α-glucosidase inhibitory peptides and in vitro evaluation of promising candidates. Biomed. Pharmacothermc. 107, 234–242 (2018). 6. McDougall, G. J., Kulkarni, N. N. & Stewart, D. Berry polyphenols inhibit pancreatic lipase activity in vitro. Food Chem. 115 193–199 (2009). 17. Minkiewicz, P., Iwaniak, A. & Darewicz, M. BIOPEP-UWM database of bioactive peptides: current opportunities. Int. J. Mol. Sci. 20, 2–23 (2019). 8. Ibrahim, M. A., Bester, M. J., Neitz, A. W. H. & Gaspar, A. R. M. Structural properties of bioactive peptides with α-glucosidase inhibitory activity. Chem. Biol. Drug Des. 91, 370–379 (2018). y y g 9. Balouiri, M., Sadiki, M. & Ibnsouda, S. K. Methods for in vitro evaluating antimicrobial activity: a review. J. Pharm. Anal. 6, 71–79 (2016). 0. Kim, S.-J., Cho, A. R. & Han, J. Antioxidant and antimicrobial activities of leafy green vegetable extracts and their applications to meat product preservation. Food Control 29, 112–120 (2013).i 1. Roman, M. J., Burri, B. J. & Singh, R. P. Release and bioaccessibility of β-carotene from fortified almond butter during in vitro digestion. Agric. Food Chem. 60, 9659–9666 (2012). g g 22. Bornhorst, G. M. & Singh, R. P. Kinetics of in vitro bread bolus digestion with varying oral and gastric digestion parameters. Food Biophys. 8, 50–59 (2013).fii p y , ( ) 3. Wada, Y. & Lönnerdal, B. Effects of industrial heating processes of milk-based enteral formulas on site-specific protein modifica- tions and their relationship to in vitro and in vivo protein digestibility. J. Agric. Food Chem. 63, 6787–6798 (2015). 23. Wada, Y. & Lönnerdal, B. Effects of industrial heating processes of milk based enteral formulas on site specific protein modifica tions and their relationship to in vitro and in vivo protein digestibility. J. Agric. Food Chem. 63, 6787–6798 (2015). www.nature.com/scientificreports/ 113, 1138–1144 (2011). 45. Sze-Tao, K. W. C. & Sathe, S. K. Functional properties and in vitro digestibility of almond (Prunus dulcis L.) protein isolate. Food Chem. 69, 153–160 (2000). 6. Minekus, M. et al. A standardised static in vitro digestion method suitable for food: an international consensus. Food Funct. 5 1113–1124 (2014). 7. Rutherfurd, S. M. Methodology for determining degree of hydrolysis of proteins in hydrolysates: a review. Food Comp. Addit. 93 1515–1522 (2010). 48. Spellman, D., McEvoy, E., O’Cuinn, G. & FitzGerald, R. J. Proteinase and exopeptidase hydrolysis of whey protein: comparison of the TNBS, OPA and pH stat methods for quantification of degree of hydrolysis. Int. Dairy J. 13, 447–453 (2003).l i 49. Xia, N. et al. Characterization and In Vitro digestibility of rice protein prepared by enzyme-assisted microfluidization: comparison to alkaline extraction. J. Cereal Sci. 56, 482–489 (2012).f to alkaline extraction. J. Cereal Sci. 56, 482–489 (2012). 0. Zhong, C. et al. Functional properties of protein isolates from Caragana korshinskii Kom. Extracted by three different methods. J Agric. Food Chem. 60, 10337–10342 (2012). 51. de Almeida, N. M., de Moura Bell, J. M. L. N. & Johnson, L. A. Properties of soy protein produced by countercurrent, two-stage, enzyme-assisted aqueous extraction. J. Am. Oil Chem. Soc. 91, 1077–1085 (2014). 52. Ahrens, S., Venkatachalam, M., Mistry, A. M., Lapsley, K. & Sathe, S. K. Almond (Prunus dulcis L.) protein quality. Plant Foods Hum. Nutr. 60, 123–128 (2005). 3. Sathe, S. K. Solubilization, electrophoretic characterization and in vitro digestibility of almond (Prunus amygdalus) proteins. J Food Biochem. 16, 249–264 (1992).f 4. Alonso, R., Aguirre, A. & Marzo, F. Effects of extrusion and traditional processing methods on antinutrients and in vitro digest- ibility of protein and starch in faba and kidney beans. Food Chem. 68, 159–165 (2000). 55. Betancur-Ancona, D. et al. Functional properties of hydrolysates from Phaseolus lunatus seeds. Int. J. Food Sci. Technol. 44, 128–137 (2009). Acknowledgementsh Acknowledgements The authors are thankful for the support from CAPES (Coordination for the Improvement of Higher Education ersonnel Program) for supplying a fellowship and CAPPS (Center for Advanced Processing and Packaging g The authors are thankful for the support from CAPES (Coordination for the Improvement of Higher Education P l P ) f l f ll h d CAPPS (C f Ad d P d P k The authors are thankful for the support from CAPES (Coordination for the Improvement of Higher Education e aut o s a e t a u o t e suppo t o C S (Coo d at o o t e p ove e t o g e ducat o ersonnel Program) for supplying a fellowship and CAPPS (Center for Advanced Processing and Packaging tudies) that funded the research. www.nature.com/scientificreports/ www.nature.com/scientificreports/ 36. Zhang, B. et al. Phenolic profiles of 20 Canadian lentil cultivars and their contribution to antioxidant activity and inhibitory effects on α-glucosidase and pancreatic lipase. Food Chem. 172, 862–872 (2015). g p p 37. Oseguera-Toledo, M. E., Gonzalez de Mejia, E. & Amaya-Llano, S. L. Hard-to-cook bean (Phaseolus vulgaris L.) proteins hydrol by alcalase and bromelain produced bioactive peptide fractions that inhibit targets of type-2 diabetes and oxidative stress. Res. Int. 76, 839–851 (2015). by alcalase and bromelain produced bioactive peptide fractions that inhibit targets of type-2 diabetes and oxidative stress. Food Res. Int. 76, 839–851 (2015). 38. Sergent, T., Vanderstraeten, J., Winand, J., Beguin, P. & Schneider, Y.-J. Phenolic compounds and plant extracts as potential natural i b i b F d Ch 13 68 3 (2012) Res. Int. 76, 839–851 (2015). 38. Sergent, T., Vanderstraeten, J., Winand, J., Beguin, P. & Schneider, Y.-J. Phenolic compounds and plant extracts as potential natural anti-obesity substances Food Chem 135 68–73 (2012) 38. Sergent, T., Vanderstraeten, J., Winand, J., Beguin, P. & Schneider, Y.-J. Phenolic compounds and plant extracts as potential na anti-obesity substances. Food Chem. 135, 68–73 (2012). y 39. Stefanucci, A. et al. Discovery of arginine-containing tripeptides as a new class of pancreatic lipase inhibitors. Future Med. Chem. 11, 5–19 (2019).l ( ) 0. Park, J.-Y., Kim, C. S., Park, K.-M. & Chang, P.-S. Inhibitory characteristics of flavonol-3-O-glycosides from Polygonum aviculare L. (common knotgrass) against porcine pancreatic lipase. Sci. Rep. 9, 18080 (2019). ( g ) g p p p p , ( ) 41. Sharma, R. R. Enzyme inhibition: mechanisms and scope. Enzyme inhibition and bioapplications, 3–37 (In tech, New York, 2012). 41. Sharma, R. R. Enzyme inhibition: mechanisms and scope. Enzyme inhibition and bioapplications, 3 41. Sharma, R. R. Enzyme inhibition: mechanisms and scope. Enzyme inhibition and bioapplications, 3–37 (In tech, New York, 20 42. Leskovac, V. Comprehensive enzyme kinetics (Springer, New York, 2003). c, V. Comprehensive enzyme kinetics (Springer, New York, 2003). . Leskovac, V. Comprehensive enzyme kinetics (Springer, New Yor p y g 43. Smeriglio, A. et al. Polyphenolic content and biological properties of Avola almond (Prunus dulcis Mill. D. A. Webb) skin an industrial byproducts. Ind. Crops Prod. 83, 283–293 (2016). yp p 4. Tian, H., Zhang, H., Zhan, P. & Tian, F. Composition and antioxidant and antimicrobial activities of white apricot almond (Amyg‑ dalus communis L.) oil. Eur. J. Lipid Sci. Technol. Competing interests h The authors declare no competing interests. References & Trombetta, D. Proanthocyanidins and hydrolysable tannins: occurrence, dietary intake and pharmacological effects: pharmacological aspects of tannins. Br. J. Pharmacol. 174, 1244–1262 (2017).i p gf p g p 1. Zam, W., Bashour, G., Abdelwahed, W. & Khayata, W. Separation and purification of proanthocyanidins extracted from pomegran- ate’s peels (Punica Granatum). Int. J. Pharm. Sci. Nanotech. 5, 1808–1813 (2002).ff p 32. Aggarwal, B., Katara, A., Raghav, H. & Verma, S. K. Ichnocarpus frutescens stem extract: effect of different concentrations of solvent on total phenol content, total flavonoid content and on its antioxidant activity. Am. J. Pharmacol. Sci. 7, 1–4 (2019). l 33. Pinelo, M., Zornoza, B. & Meyer, A. S. Selective release of phenols from apple skin: mass transfer kinetics during solvent and enzyme-assisted extraction. Sep. Purif. Technol. 63, 620–627 (2008). y p f ( ) 4. Iglesias-Carres, L. et al. Optimization and characterization of Royal Dawn cherry (Prunus avium) phenolics extraction. Sci. Rep 9, 17626 (2019).l 35. Sbroggio, M. F., Montilha, M. S., de Figueiredo, V. R. G., Georgetti, S. R. & Kurozawa, L. E. Influence of the degree of hydrolysis and type of enzyme on antioxidant activity of okara protein hydrolysates. Food Sci. Technol. 36, 375–381 (2016). Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3 Author contributions T.S.P.S. and M.G.B.K. wrote the main manuscript text. J.P.S.O. ran the in silico experiments. T.S.P.S. and F.G.D. ran the in vitro experiments. J.M.N.L.M.B. and M.G.B.K. designed and supervised all experiments. All authors reviewed the manuscript. Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-020-67682-3. Supplementary information is available for this paper at https://doi.org/10.1038 Supplementary information is available for this paper at https://doi.org/10.1038/s41598-020-67682-3. Correspondence and requests for materials should be addressed to M.G.B.K. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and nstitutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2020 Scientific Reports \| (2020) 10:10873 \| https://doi.org/10.1038/s41598-020-67682-3
https://openalex.org/W1863129893	https://europepmc.org/articles/pmc2970983?pdf=render	Latin	null	The zwitterion (23′<i>E</i>)-(23<i>R</i>,25<i>S</i>)-23-[1-(oxidoiminio)ethyl]-5β-spirostan-3β-yl acetate	Acta crystallographica. Section E	2,009	cc-by	5,429	organic compounds Acta Crystallographica Section E Structure Reports Online ISSN 1600-5368 Acta Crystallographica Section E Structure Reports Online ISSN 1600-5368 Table 1 Data collection: XSCANS (Siemens, 1994); cell reﬁnement: XSCANS; data reduction: XSCANS; program(s) used to solve structure: SHELXTL (Sheldrick, 2008); program(s) used to reﬁne structure: SHELXTL; molecular graphics: SHELXTL; software used to prepare material for publication: SHELXTL. Related literature For the synthesis of (23R)-23-acetylsarsasapogenin acetate used as starting material, see: Meza-Reyes et al. (2005). For the full spectroscopic characterization of the tautomers of the title compound, see: Herna´ndez-Linares (2005). For tautomerism between oximes and imine N-oxides, see: Ferna´ndez et al. (1994). For related zwitterionic oximes and hydrochloride oximes characterized by X-ray diffraction, see: Witte et al. (1984); Ferna´ndez et al. (1994); Gurkova et al. (1988); Laus et al. (2008); Forgan et al. (2008). This work was supported by CONACyT (grant 83049). Supplementary data and ﬁgures for this paper are available from the IUCr electronic archives (Reference: NK2010). Marı´a-Guadalupe Herna´ndez Linares,a Jesu´s Sandoval Ramı´rez,b Socorro Meza Reyes,b Sara Montiel Smithb and Sylvain Berne`sc* V = 2949.2 (10) A˚ 3 Z = 4 Mo K radiation = 0.08 mm1 T = 298 K 0.60 0.60 0.35 mm aEscuela de Ingenierı´a Quı´mica, Universidad del Istmo, Ciudad Universitaria s/n, 70760 Sto. Domingo Tehuantepec, Oax., Mexico, bFacultad de Ciencias Quı´micas, Beneme´rita Universidad Auto´noma de Puebla, Ciudad Universitaria, San Manuel, 72000 Puebla, Pue., Mexico, and cDEP Facultad de Ciencias Quı´micas, UANL, Guerrero y Progreso S/N, Col. Trevin˜o, 64570 Monterrey, N.L., Mexico Correspondence e-mail: sylvain_bernes@hotmail.com Ferna´ndez, M. J., Huertas, R., Ga´lvez, E., Sanz-Aparicio, J., Fonseca, I. & Bellanato, J. (1994). J. Mol. Struct. 323, 85–91. Forgan, R. S., Davidson, J. E., Galbraith, S. G., Henderson, D. K., Parsons, S., Tasker, P. A. & White, F. J. (2008). Chem. Commun. pp. 4049–4051. Gurkova, S. N., Gusev, A. I., Alekseev, N. V. & Lakhtin, V. G. (1988). Zh. Strukt. Khim. 29, 184–186. Experimental Experimental Crystal data C31H49NO5 Mr = 515.71 Orthorhombic, P212121 a = 8.385 (2) A˚ b = 11.5627 (16) A˚ c = 30.420 (5) A˚ Experimental Crystal data C31H49NO5 Mr = 515.71 Orthorhombic, P212121 a = 8.385 (2) A˚ b = 11.5627 (16) A˚ c = 30.420 (5) A˚ Marı´a-Guadalupe Herna´ndez Linares,a Jesu´s Sandoval Ramı´rez,b Socorro Meza Reyes,b Sara Montiel Smithb and Sylvain Berne`sc* Supplementary data and ﬁgures for this paper are available from the IUCr electronic archives (Reference: NK2010). Reﬁnement The title steroidal compound, C31H49NO5, resulted from the selective oximation of (23R)-23-acetylsarsasapogenin acetate. One- and two-dimensional 1H and 13C NMR spectra, as well as IR data, are in agreement with the presence of a ketoxime group at C-23. However, recrystallization in slightly acidic media affords the title compound in the rare zwitterionic oxime form, as a consequence of migration of the hydroxy H atom to the N atom in the oxime group. This H atom is clearly detected and its position was reﬁned from X-ray data. The geometry for the C N+(H)—O group features long C N and short N—O bond lengths compared to non-zwitterionic oximes. The ketoxime is stabilized with the E conﬁguration, avoiding steric hindrance between the oxime O atom and H atom at C-23. The sum of the angles around the oxime N atom is 359.6, giving a planar conﬁguration for that atom, as expected for sp2 hybridization. The title steroidal compound, C31H49NO5, resulted from the selective oximation of (23R)-23-acetylsarsasapogenin acetate. Table 1 Selected geometric parameters (A˚ , ). C32—C33 1.477 (5) C32—N34 1.361 (6) N34—H34 0.980 (18) N34—O35 1.367 (5) C23—C32—C33 122.0 (3) N34—C32—C23 111.6 (3) N34—C32—C33 126.3 (3) C32—N34—O35 109.0 (4) C32—N34—H34 132.0 (16) O35—N34—H34 118.6 (16) Data collection Rint = 0.041 3 standard reﬂections every 97 reﬂections intensity decay: 3% Received 20 October 2009; accepted 27 October 2009 Key indicators: single-crystal X-ray study; T = 298 K; mean (C–C) = 0.005 A˚; R factor = 0.048; wR factor = 0.129; data-to-parameter ratio = 8.6. Herna´ndez-Linares, G. (2005). PhD thesis, Universidad Auto´noma de Puebla, Mexico. Laus, G., Schwa¨rzler, A., Bentivoglio, G., Hummel, M., Kahlenberg, V., Wurst, K., Kristeva, E., Schutz, J., Kopacka, H., Kreutz, C., Bonn, G., Andriyko, Y., Nauer, G. & Schottenberger, H. (2008). Z. Naturforsch. Teil B, 63, 447–464. Meza-Reyes, S., Sandoval-Ramı´rez, J., Montiel-Smith, S., Herna´ndez-Linares, G., Vin˜as-Bravo, O., Martı´nez-Pascual, R., Ferna´ndez-Herrera, M. A., Vega- References Ferna´ndez, M. J., Huertas, R., Ga´lvez, E., Sanz-Aparicio, J., Fonseca, I. & Bellanato, J. (1994). J. Mol. Struct. 323, 85–91. Tasker, P. A. & White, F. J. (2008). Chem. Commun. pp. 4049–4051. Gurkova, S. N., Gusev, A. I., Alekseev, N. V. & Lakhtin, V. G. (1988). Zh. Strukt. Khim. 29, 184–186. o2954 Herna´ndez Linares et al. Acta Cryst. (2009). E65, o2954–o2955 doi:10.1107/S1600536809044651 Ba´ez, J. L., Merino-Montiel, P., Santilla´n, R. L., Farfa´n, N., Rinco´n, S. & del Rı´o, R. E. (2005). Arkivoc, vi, 307–320. Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122. Siemens (1994). XSCANS. Siemens Analytical X-ray Instruments Inc., Madison, Wisconsin, USA. Witte, E. G., Schwochau, K. S., Henkel, G. & Krebs, B. (1984). Inorg. Chim. Acta, 94, 323–331. Acta Cryst. (2009). E65, o2954–o2955 Herna´ndez Linares et al. C31H49NO5 o2955 Witte, E. G., Schwochau, K. S., Henkel, G. & Krebs, B. (1984). Inorg. Chim. Acta, 94, 323–331. organic compounds Ba´ez, J. L., Merino-Montiel, P., Santilla´n, R. L., Farfa´n, N., Rinco´n, S. & del Rı´o, R. E. (2005). Arkivoc, vi, 307–320. Siemens (1994). XSCANS. Siemens Analytical X-ray Instruments Inc., Madison, Wisconsin, USA. Witte, E. G., Schwochau, K. S., Henkel, G. & Krebs, B. (1984). Inorg. Chim. Acta, 94, 323–331. Herna´ndez Linares et al. C31H49NO5 o2955 Comment Synthetic routes are available for the large scale preparation of (23R)-23-acetylsarsasapogenin acetate (Meza-Reyes et al., 2005). During attempts to functionalize the 23 position in sarsasapogenin acetate, we have found that the oximation of (23R)-23-acetylsarsasapogenin acetate using NH2OH.HCl afforded the expected 23-hydroxyimino derivative. IR spectrum for that compound exhibits a strong vibration at 3400 cm-1 (νO—H), a weak vibration at 1636 cm-1 (νC=N), and a vibration at 951 cm-1 (νN—O), characteristics of ketoximes. One-dimensional and two-dimensional 1H and 13C NMR data are also in full agreement with that assignment (Hernández-Linares, 2005, and archived CIF). For instance, 31 signals are detected in 13C-NMR, one of which, at 159.3 p.p.m., is characteristic of the C231 atom substituted by an hydroxyimino group. However, recrystallization of the ketoxime in a slightly acidic medium (MeOH/CH2Cl2, 98:2) afforded a solid with a significantly different melting point (459 K for the oxime, 516 K for the recrystallized material). On the other hand, the high m.p. solid presents a new 1H-NMR broad signal at 9.05 p.p.m., and the stretching vibration for OH group in the IR spectrum no longer appears, while a new strong vibration is observed at 2300 cm-1. In order to rationalize these dramatic changes, the X-ray analysis of this material was carried out. The molecular structure (Fig. 1) resolves the above mentioned problem: a difference map shows a strong residual in the vicinity of atom N34, while no H atom is found bonded to O35. The oxime is thus stabilized in the solid-state in a rather rare zwittterionic form. The residual close to N34 refines well as an H atom. The N—H bond accounts for the IR vibration at 2300 cm-1 and H34 is detected in 1H-NMR at 9.05 p.p.m.. Refinement of the proposed model was however not so straightforward. Alternative models assuming a non-zwitterionic oxime were probed, with disordered N and O sites, which resulted in oxime groups with geometry far from expectation and physically unreasonable intermolecular contacts. On the other hand, refinements carried out with a free H34 atom converged to a geometry where the oxime N atom deviates significantly from the expected planar trigonal arrangement, although the N34—H34 bond length lies in the expected range. Considering that the structure is based on room-temperature data, we eventually refined the structural model with a pair of soft restraints for non-bonding separations involving H34 (see Experimental). Acta Cryst. (2009). E65, o2954-o2955 [ doi:10.1107/S1600536809044651 ] Acta Cryst. (2009). E65, o2954-o2955 [ doi:10.1107/S1600536809044651 ] supplementary materials supplementary materials supplementary materials compared to common non-zwitterionic oximes (ca. 1.28 and 1.41 Å, respectively). The sum of angles around N2 is very close to 360°, as expected for a sp2 hybridized atom. Therefore, delocalization is extended on the whole hydroxyimino group, C32/N34/O35, which is almost planar. The imine group presents the E configuration, which avoids steric hindrance between the oxime O atom and methine H atom H23C. This geometry is close to that previously reported for the few zwitterionic oximes or hydrochloride oximes characterized by X-ray diffraction (Fernández et al., 1994; Gurkova et al., 1988; Laus et al., 2008). The imine N-oxide tautomer has also been used as a ligand (Witte et al., 1984; Forgan et al., 2008). compared to common non-zwitterionic oximes (ca. 1.28 and 1.41 Å, respectively). The sum of angles around N2 is very close to 360°, as expected for a sp2 hybridized atom. Therefore, delocalization is extended on the whole hydroxyimino group, C32/N34/O35, which is almost planar. The imine group presents the E configuration, which avoids steric hindrance between the oxime O atom and methine H atom H23C. This geometry is close to that previously reported for the few zwitterionic oximes or hydrochloride oximes characterized by X-ray diffraction (Fernández et al., 1994; Gurkova et al., 1988; Laus et al., 2008). The imine N-oxide tautomer has also been used as a ligand (Witte et al., 1984; Forgan et al., 2008). Experimental In a 100 ml round bottom flask was dissolved 2 mmol (1 g) of (23R)-23-acetylsarsasapogenin acetate in ethanol (30 ml). Pyridine (1 ml) and 2 mmol of NH2OH.HCl (0.139 g) were added. The mixture was refluxed for 3 h, following the reaction by TLC. Solvent was then eliminated under reduced pressure, affording the crude oxime. The crude was recrystallized in a mixture of MeOH/CH2Cl2 (98:2), yielding the iminium olate zwitterion (96%). When the crude product was neutralized and then crystallized, the hydroxyimino derivative was obtained Anal. found (calc. for C31H49NO5): C 72.20 (72.20), H 9.56 (9.57), N 2.71 (2.71%). Comment Finally, a supplementary concern is about the terminal O atom O35, which is separated by 2.757 (5) Å from the carbonyl O30 atom of a symmetry-related molecule (symmetry code: 2 - x, -1/2 + y, 1/2 - z). Such an arrangement would be suitable for a stabilizing hydrogen bond, which is not observed because of the migration of the oxime H atom to N34. With the currently available data, it is however difficult to decide if this situation resulted from the actual migration of the H atom in the solid-sate. Definitively, more data are required in order to fully characterize this zwitterion, for example low temperature neutron diffraction data, and the X-ray structure of the non-zwitterionic title molecule. The tautomerism between oximes and imine N-oxides seems to be poorly documented (Fernández et al., 1994). In the case of the title compound, the C═N bond length is long, 1.361 (6) Å, and the N—O bond length is short, 1.367 (5) Å, sup-1 Refinement H34 was found in a difference map and refined with free coordinates, although geometry around N34 was restrained through soft restraints in order to avoid significant deviations from trigonal geometry. The X-ray structure of acetone oxime hydro- chloride was used as target (Gurkova et al., 1988): H34···C32 and H34···O35 separations were restrained to 2.10 (2) Å. The isotropic displacement parameter for H34 was fixed to Uiso(H34) = 1.5Ueq(N34). Other H atoms were placed in ideal- ized positions and refined using a riding model, with fixed C—H bond lengths: 0.96 (methyl), 0.97 (methylene) or 0.98 Å (methine). Isotropic displacement parameters were computed as Uiso(H) = 1.5Ueq(carrier C) for methyl groups and Uiso(H) = 1.2Ueq(carrier C) for other H atoms. Methyl groups were allowed to rotate about their C—C bonds. Figures Fig. 1. The title molecule with displacement ellipsoids for non-H atoms shown at the 30% probability level. Figures Fig. 1. The title molecule with displacement ellipsoids for non-H atoms shown at the 30% probability level. Figures Fig. 1. The title molecule with displacement ellipsoids for non-H atoms shown at the 30% probability level. (23'E)-(23R,25S)-23-[1-(oxidoiminio)ethyl]-5β- spirostan-3β-yl acetate (23'E)-(23R,25S)-23-[1-(oxidoiminio)ethyl]-5β- spirostan-3β-yl acetate Crystal data C31H49NO5 Mr = 515.71 Orthorhombic, P212121 Hall symbol: P 2ac 2ab a = 8.385 (2) Å b = 11.5627 (16) Å Dx = 1.161 Mg m−3 Melting point: 516 K Mo Kα radiation, λ = 0.71073 Å Cell parameters from 44 reflections θ = 4.7–12.5º µ = 0.08 mm−1 sup-2 supplementary materials c = 30.420 (5) Å T = 298 K V = 2949.2 (10) Å3 Irregular, colourless Z = 4 0.60 × 0.60 × 0.35 mm F000 = 1128 c = 30.420 (5) Å T = 298 K V = 2949.2 (10) Å3 Irregular, colourless Z = 4 0.60 × 0.60 × 0.35 mm F000 = 1128 = 30.420 (5) Å T = 298 K = 2949.2 (10) Å3 Irregular, colourless = 4 0.60 × 0.60 × 0.35 mm T = 298 K Irregular, colourless 0.60 × 0.60 × 0.35 mm Data collection Bruker P4 diffractometer Rint = 0.041 Radiation source: fine-focus sealed tube θmax = 25.0º Monochromator: graphite θmin = 1.9º T = 298 K h = −9→2 ω scans k = −13→1 Absorption correction: none l = −36→1 4182 measured reflections 3 standard refle 2956 independent reflections every 97 reflect Rint = 0.041 θmax = 25.0º θmin = 1.9º h = −9→2 k = −13→1 l = −36→1 3 standard reflections every 97 reflections intensity decay: 3% Secondary atom site location: difference Fourier map Hydrogen site location: inferred from neighbouring sites H atoms treated by a mixture of independent and constrained refinement H atoms treated by a mixture of independent and constrained refinement w = 1/[σ2(Fo 2) + (0.0651P)2 + 1.0057P] where P = (Fo 2 + 2Fc 2)/3 (Δ/σ)max < 0.001 Δρmax = 0.46 e Å−3 Δρmin = −0.25 e Å−3 w = 1/[σ2(Fo 2) + (0.0651P)2 + 1.0057P] where P = (Fo 2 + 2Fc 2)/3 (Δ/σ)max < 0.001 Δρmax = 0.46 e Å−3 Δρmin = −0.25 e Å−3 Extinction correction: SHELXTL (Sheldrick, 2008), Fc=kFc[1+0.001xFc2λ3/sin(2θ)]-1/4 Extinction correction: SHELXTL (Sheldrick, 2008), Fc=kFc[1+0.001xFc2λ3/sin(2θ)]-1/4 Extinction coefficient: 0.0095 (18) Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2) Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å2) x y z Uiso/Ueq C1 1.0856 (4) 0.6448 (3) 0.33730 (11) 0.0516 (9) H1A 1.0915 0.5917 0.3619 0.062 H1B 1.1393 0.6086 0.3126 0.062* C2 1.1766 (4) 0.7539 (4) 0.34940 (12) 0.0560 (9) H2D 1.1862 0.8030 0.3237 0.067* H2E 1.2832 0.7333 0.3589 0.067* C3 1.0929 (5) 0.8205 (3) 0.38583 (11) 0.0554 (9) H3A 1.1486 0.8934 0.3917 0.066* C4 0.9219 (5) 0.8431 (3) 0.37427 (11) 0.0531 (9) H4A 0.9182 0.8964 0.3497 0.064* sup-3 supplementary materials supplementary materials sup-4 H4B 0.8701 0.8802 0.3991 0.064* C5 0.8285 (4) 0.7338 (3) 0.36217 (10) 0.0485 (8) H5A 0.8271 0.6848 0.3885 0.058* C6 0.6548 (4) 0.7627 (4) 0.35160 (11) 0.0599 (11) H6A 0.6116 0.8113 0.3747 0.072* H6B 0.5929 0.6919 0.3508 0.072* C7 0.6396 (5) 0.8252 (4) 0.30744 (11) 0.0587 (10) H7B 0.5277 0.8390 0.3012 0.070* H7C 0.6925 0.8996 0.3093 0.070* C8 0.7128 (4) 0.7551 (3) 0.27006 (10) 0.0447 (8) H8B 0.6522 0.6831 0.2670 0.054* C9 0.8885 (4) 0.7238 (3) 0.27973 (10) 0.0421 (7) H9B 0.9461 0.7972 0.2821 0.051* C10 0.9083 (4) 0.6615 (3) 0.32507 (10) 0.0441 (8) C11 0.9635 (5) 0.6585 (3) 0.24098 (10) 0.0552 (9) H11B 0.9145 0.5827 0.2389 0.066* H11C 1.0761 0.6473 0.2469 0.066* C12 0.9452 (4) 0.7207 (3) 0.19660 (11) 0.0519 (9) H12A 1.0078 0.7913 0.1970 0.062* H12B 0.9866 0.6716 0.1734 0.062* C13 0.7726 (4) 0.7502 (3) 0.18680 (10) 0.0410 (7) C14 0.7059 (4) 0.8196 (3) 0.22627 (10) 0.0434 (8) H14B 0.7764 0.8867 0.2296 0.052* C15 0.5478 (4) 0.8673 (4) 0.20928 (10) 0.0531 (9) H15B 0.5123 0.9329 0.2266 0.064* H15C 0.4655 0.8084 0.2093 0.064* C16 0.5917 (4) 0.9040 (3) 0.16224 (10) 0.0448 (8) H16B 0.6064 0.9880 0.1607 0.054* C17 0.7482 (4) 0.8411 (3) 0.14966 (10) 0.0405 (7) H17C 0.8361 0.8968 0.1508 0.049* C18 0.6769 (5) 0.6401 (3) 0.17818 (12) 0.0576 (9) H18C 0.6759 0.5932 0.2042 0.086* H18D 0.7251 0.5976 0.1545 0.086* H18E 0.5695 0.6602 0.1704 0.086* C19 0.8329 (5) 0.5408 (3) 0.32368 (13) 0.0615 (10) H19A 0.8376 0.5066 0.3524 0.092* H19B 0.8903 0.4932 0.3032 0.092* H19C 0.7237 0.5471 0.3145 0.092* C20 0.7218 (4) 0.8046 (3) 0.10144 (10) 0.0435 (8) H20A 0.7122 0.7202 0.1007 0.052* C21 0.8573 (4) 0.8390 (4) 0.07093 (12) 0.0606 (10) H21B 0.8276 0.8230 0.0411 0.091* H21C 0.9512 0.7956 0.0783 0.091* H21D 0.8787 0.9201 0.0741 0.091* C22 0.5588 (4) 0.8561 (3) 0.08961 (10) 0.0391 (7) O22 0.4759 (2) 0.86704 (19) 0.13038 (6) 0.0415 (5) C23 0.4554 (4) 0.7843 (3) 0.05791 (10) 0.0430 (8) H23C 0.5201 0.7697 0.0317 0.052* C24 0.3086 (4) 0.8530 (3) 0.04293 (11) 0.0511 (8) sup-4 supplementary materials supplementary materia H24C 0.2537 0.8100 0.0201 0.061* H24D 0.2359 0.8616 0.0675 0.061* C25 0.3526 (5) 0.9721 (3) 0.02546 (11) 0.0541 (9) H25C 0.2538 1.0157 0.0209 0.065* C26 0.4476 (5) 1.0321 (3) 0.06089 (12) 0.0552 (9) H26D 0.4773 1.1088 0.0509 0.066* H26E 0.3821 1.0405 0.0870 0.066* O26 0.5890 (3) 0.96787 (19) 0.07171 (7) 0.0473 (6) C27 0.4413 (6) 0.9665 (4) −0.01843 (12) 0.0700 (11) H27C 0.4699 1.0432 −0.0275 0.105* H27D 0.3735 0.9319 −0.0402 0.105* H27E 0.5360 0.9207 −0.0151 0.105* O28 1.0885 (3) 0.7483 (2) 0.42577 (7) 0.0540 (6) C29 1.2125 (5) 0.7503 (3) 0.45297 (11) 0.0539 (9) O30 1.3306 (4) 0.8083 (3) 0.44684 (10) 0.0741 (8) C31 1.1864 (5) 0.6731 (4) 0.49129 (12) 0.0710 (12) H31B 1.2539 0.6970 0.5151 0.106* H31C 1.2117 0.5950 0.4833 0.106* H31D 1.0769 0.6775 0.5003 0.106* C32 0.4111 (4) 0.6680 (3) 0.07679 (11) 0.0483 (8) C33 0.2777 (5) 0.6538 (4) 0.10802 (13) 0.0660 (11) H33A 0.3050 0.5962 0.1295 0.099* H33B 0.1839 0.6299 0.0924 0.099* H33C 0.2574 0.7260 0.1225 0.099* N34 0.5075 (6) 0.5826 (3) 0.06127 (13) 0.0854 (13) H34 0.605 (5) 0.585 (2) 0.0434 (17) 0.128* O35 0.4610 (5) 0.4795 (3) 0.07914 (14) 0.1043 (12) Atomic displacement parameters (Å2) U11 U22 U33 U12 U13 U23 C1 0.051 (2) 0.058 (2) 0.0460 (17) 0.0135 (19) 0.0002 (16) 0.0069 (16) C2 0.0428 (18) 0.069 (2) 0.056 (2) −0.003 (2) −0.0010 (16) 0.013 (2) C3 0.060 (2) 0.054 (2) 0.0515 (18) −0.009 (2) −0.0074 (18) 0.0068 (16) C4 0.060 (2) 0.0538 (19) 0.0458 (17) 0.0118 (19) −0.0014 (17) −0.0013 (16) C5 0.0440 (18) 0.059 (2) 0.0422 (17) 0.0067 (18) 0.0023 (15) 0.0068 (16) C6 0.0442 (18) 0.094 (3) 0.0415 (17) 0.019 (2) 0.0072 (16) 0.003 (2) C7 0.048 (2) 0.083 (3) 0.0454 (17) 0.027 (2) 0.0009 (16) −0.0018 (18) C8 0.0405 (17) 0.0530 (18) 0.0405 (16) 0.0062 (18) 0.0007 (14) 0.0011 (15) C9 0.0392 (16) 0.0439 (17) 0.0433 (16) 0.0060 (15) −0.0002 (14) 0.0001 (14) C10 0.0425 (17) 0.0450 (17) 0.0449 (16) 0.0041 (16) −0.0018 (15) 0.0045 (14) C11 0.052 (2) 0.066 (2) 0.0474 (18) 0.019 (2) 0.0044 (16) 0.0054 (17) C12 0.0462 (19) 0.064 (2) 0.0458 (17) 0.0113 (18) 0.0065 (16) 0.0026 (16) C13 0.0413 (17) 0.0433 (16) 0.0385 (15) 0.0079 (16) 0.0020 (14) −0.0003 (14) C14 0.0425 (17) 0.0485 (18) 0.0392 (15) 0.0102 (17) −0.0026 (14) −0.0036 (15) C15 0.052 (2) 0.069 (2) 0.0375 (15) 0.022 (2) 0.0008 (15) −0.0045 (16) C16 0.0460 (18) 0.0464 (17) 0.0421 (16) 0.0087 (17) −0.0037 (15) −0.0033 (14) C17 0.0370 (16) 0.0425 (16) 0.0420 (16) 0.0005 (15) −0.0003 (14) −0.0008 (14) Atomic displacement parameters (Å2) Atomic displacement parameters (Å ) U11 U22 U33 U12 U13 U23 C1 0.051 (2) 0.058 (2) 0.0460 (17) 0.0135 (19) 0.0002 (16) 0.0069 (16) C2 0.0428 (18) 0.069 (2) 0.056 (2) −0.003 (2) −0.0010 (16) 0.013 (2) C3 0.060 (2) 0.054 (2) 0.0515 (18) −0.009 (2) −0.0074 (18) 0.0068 (16) C4 0.060 (2) 0.0538 (19) 0.0458 (17) 0.0118 (19) −0.0014 (17) −0.0013 (16) C5 0.0440 (18) 0.059 (2) 0.0422 (17) 0.0067 (18) 0.0023 (15) 0.0068 (16) C6 0.0442 (18) 0.094 (3) 0.0415 (17) 0.019 (2) 0.0072 (16) 0.003 (2) C7 0.048 (2) 0.083 (3) 0.0454 (17) 0.027 (2) 0.0009 (16) −0.0018 (18) C8 0.0405 (17) 0.0530 (18) 0.0405 (16) 0.0062 (18) 0.0007 (14) 0.0011 (15) C9 0.0392 (16) 0.0439 (17) 0.0433 (16) 0.0060 (15) −0.0002 (14) 0.0001 (14) C10 0.0425 (17) 0.0450 (17) 0.0449 (16) 0.0041 (16) −0.0018 (15) 0.0045 (14) C11 0.052 (2) 0.066 (2) 0.0474 (18) 0.019 (2) 0.0044 (16) 0.0054 (17) C12 0.0462 (19) 0.064 (2) 0.0458 (17) 0.0113 (18) 0.0065 (16) 0.0026 (16) C13 0.0413 (17) 0.0433 (16) 0.0385 (15) 0.0079 (16) 0.0020 (14) −0.0003 (14) C14 0.0425 (17) 0.0485 (18) 0.0392 (15) 0.0102 (17) −0.0026 (14) −0.0036 (15) C15 0.052 (2) 0.069 (2) 0.0375 (15) 0.022 (2) 0.0008 (15) −0.0045 (16) C16 0.0460 (18) 0.0464 (17) 0.0421 (16) 0.0087 (17) −0.0037 (15) −0.0033 (14) C17 0.0370 (16) 0.0425 (16) 0.0420 (16) 0.0005 (15) −0.0003 (14) −0.0008 (14) sup-5 supplementary materials supplementary materials C18 0.068 (2) 0.0508 (19) 0.0541 (19) −0.007 (2) 0.0033 (19) −0.0030 (17) C19 0.069 (3) 0.053 (2) 0.063 (2) −0.005 (2) −0.002 (2) 0.0078 (18) C20 0.0406 (17) 0.0491 (18) 0.0408 (16) 0.0034 (16) 0.0011 (14) −0.0040 (15) C21 0.0439 (19) 0.089 (3) 0.0490 (18) 0.005 (2) 0.0077 (17) 0.000 (2) C22 0.0366 (16) 0.0435 (16) 0.0373 (14) 0.0037 (15) 0.0031 (14) −0.0023 (13) O22 0.0375 (11) 0.0500 (12) 0.0369 (10) 0.0064 (11) 0.0016 (9) −0.0021 (10) C23 0.0383 (17) 0.0507 (18) 0.0399 (16) 0.0009 (15) 0.0032 (15) −0.0040 (14) C24 0.0439 (18) 0.061 (2) 0.0485 (17) 0.0003 (18) −0.0036 (16) −0.0034 (17) C25 0.050 (2) 0.059 (2) 0.0530 (19) 0.0090 (19) −0.0032 (18) 0.0042 (17) C26 0.061 (2) 0.0443 (18) 0.060 (2) 0.0087 (18) 0.001 (2) 0.0056 (16) O26 0.0466 (13) 0.0468 (12) 0.0486 (12) −0.0029 (12) 0.0005 (11) 0.0054 (10) C27 0.079 (3) 0.079 (3) 0.052 (2) 0.004 (3) −0.002 (2) 0.015 (2) O28 0.0521 (13) 0.0620 (14) 0.0480 (12) −0.0073 (14) −0.0077 (12) 0.0084 (12) C29 0.054 (2) 0.059 (2) 0.0481 (18) 0.009 (2) −0.0064 (17) −0.0047 (18) O30 0.0616 (17) 0.0797 (19) 0.0810 (19) −0.0100 (17) −0.0189 (16) 0.0064 (17) C31 0.065 (3) 0.092 (3) 0.056 (2) 0.015 (3) −0.002 (2) 0.010 (2) C32 0.0515 (19) 0.0424 (17) 0.0511 (18) −0.0031 (17) −0.0078 (17) −0.0053 (15) C33 0.064 (2) 0.061 (2) 0.073 (2) −0.018 (2) 0.000 (2) 0.010 (2) N34 0.118 (3) 0.0420 (17) 0.096 (3) −0.013 (2) −0.039 (3) 0.0033 (18) O35 0.110 (3) 0.079 (2) 0.123 (3) −0.018 (2) 0.028 (3) 0.001 (2) Geometric parameters (Å, °) C1—C2 1.519 (5) C17—C20 1.542 (4) C1—C10 1.544 (5) C17—H17C 0.9800 C1—H1A 0.9700 C18—H18C 0.9600 C1—H1B 0.9700 C18—H18D 0.9600 C2—C3 1.520 (5) C18—H18E 0.9600 C2—H2D 0.9700 C19—H19A 0.9600 C2—H2E 0.9700 C19—H19B 0.9600 C3—O28 1.474 (4) C19—H19C 0.9600 C3—C4 1.500 (5) C20—C21 1.520 (5) C3—H3A 0.9800 C20—C22 1.534 (5) C4—C5 1.531 (5) C20—H20A 0.9800 C4—H4A 0.9700 C21—H21B 0.9600 C4—H4B 0.9700 C21—H21C 0.9600 C5—C6 1.529 (5) C21—H21D 0.9600 C5—C10 1.556 (5) C22—O26 1.426 (4) C5—H5A 0.9800 C22—O22 1.427 (4) C6—C7 1.531 (5) C22—C23 1.539 (5) C6—H6A 0.9700 C23—C32 1.509 (5) C6—H6B 0.9700 C23—C24 1.534 (5) C7—C8 1.526 (5) C23—H23C 0.9800 C7—H7B 0.9700 C24—C25 1.522 (5) C7—H7C 0.9700 C24—H24C 0.9700 C8—C14 1.528 (4) C24—H24D 0.9700 C8—C9 1.545 (5) C25—C26 1.509 (5) C8—H8B 0.9800 C25—C27 1.530 (5) C9—C11 1.534 (5) C25—H25C 0.9800 sup-6 supplementary materials supplementary materials C9—C10 1.565 (4) C26—O26 C9—H9B 0.9800 C26—H26D C10—C19 1.532 (5) C26—H26E C11—C12 1.537 (5) C27—H27C C11—H11B 0.9700 C27—H27D C11—H11C 0.9700 C27—H27E C12—C13 1.517 (5) O28—C29 C12—H12A 0.9700 C29—O30 C12—H12B 0.9700 C29—C31 C13—C18 1.528 (5) C31—H31B C13—C14 1.549 (4) C31—H31C C13—C17 1.556 (4) C31—H31D C14—C15 1.526 (5) C32—C33 C14—H14B 0.9800 C32—N34 C15—C16 1.537 (5) C33—H33A C15—H15B 0.9700 C33—H33B C15—H15C 0.9700 C33—H33C C16—O22 1.437 (4) N34—H34 C16—C17 1.548 (4) N34—O35 C16—H16B 0.9800 C2—C1—C10 116.0 (3) O22—C16—H16B C2—C1—H1A 108.3 C15—C16—H16B C10—C1—H1A 108.3 C17—C16—H16B C2—C1—H1B 108.3 C20—C17—C16 C10—C1—H1B 108.3 C20—C17—C13 H1A—C1—H1B 107.4 C16—C17—C13 C1—C2—C3 111.4 (3) C20—C17—H17C C1—C2—H2D 109.3 C16—C17—H17C C3—C2—H2D 109.3 C13—C17—H17C C1—C2—H2E 109.3 C13—C18—H18C C3—C2—H2E 109.3 C13—C18—H18D H2D—C2—H2E 108.0 H18C—C18—H18D O28—C3—C4 105.5 (3) C13—C18—H18E O28—C3—C2 109.0 (3) H18C—C18—H18E C4—C3—C2 111.0 (3) H18D—C18—H18E O28—C3—H3A 110.4 C10—C19—H19A C4—C3—H3A 110.4 C10—C19—H19B C2—C3—H3A 110.4 H19A—C19—H19B C3—C4—C5 113.7 (3) C10—C19—H19C C3—C4—H4A 108.8 H19A—C19—H19C C5—C4—H4A 108.8 H19B—C19—H19C C3—C4—H4B 108.8 C21—C20—C22 C5—C4—H4B 108.8 C21—C20—C17 H4A—C4—H4B 107.7 C22—C20—C17 C6—C5—C4 110.9 (3) C21—C20—H20A C6—C5—C10 112.0 (3) C22—C20—H20A C4—C5—C10 113.4 (3) C17—C20—H20A C6—C5—H5A 106.7 C20—C21—H21B C4—C5—H5A 106.7 C20—C21—H21C 1.565 (4) C26—O26 0.9800 C26—H26D 1.532 (5) C26—H26E 1.537 (5) C27—H27C 0.9700 C27—H27D 0.9700 C27—H27E 1.517 (5) O28—C29 0.9700 C29—O30 0.9700 C29—C31 1.528 (5) C31—H31B 1.549 (4) C31—H31C 1.556 (4) C31—H31D 1.526 (5) C32—C33 0.9800 C32—N34 1.537 (5) C33—H33A 0.9700 C33—H33B 0.9700 C33—H33C 1.437 (4) N34—H34 1.548 (4) N34—O35 0.9800 116.0 (3) O22—C16—H16B 108.3 C15—C16—H16B 108.3 C17—C16—H16B 108.3 C20—C17—C16 108.3 C20—C17—C13 107.4 C16—C17—C13 111.4 (3) C20—C17—H17C 109.3 C16—C17—H17C 109.3 C13—C17—H17C 109.3 C13—C18—H18C 109.3 C13—C18—H18D 108.0 H18C—C18—H18D 105.5 (3) C13—C18—H18E 109.0 (3) H18C—C18—H18E 111.0 (3) H18D—C18—H18E 110.4 C10—C19—H19A 110.4 C10—C19—H19B 110.4 H19A—C19—H19B 113.7 (3) C10—C19—H19C 108.8 H19A—C19—H19C 108.8 H19B—C19—H19C 108.8 C21—C20—C22 108.8 C21—C20—C17 107.7 C22—C20—C17 110.9 (3) C21—C20—H20A 112.0 (3) C22—C20—H20A 113.4 (3) C17—C20—H20A 106.7 C20—C21—H21B 106.7 C20—C21—H21C sup-7 1.565 (4) C26—O26 1.437 (4) 0.9800 C26—H26D 0.9700 1.532 (5) C26—H26E 0.9700 1.537 (5) C27—H27C 0.9600 0.9700 C27—H27D 0.9600 0.9700 C27—H27E 0.9600 1.517 (5) O28—C29 1.328 (4) 0.9700 C29—O30 1.210 (5) 0.9700 C29—C31 1.484 (5) 1.528 (5) C31—H31B 0.9600 1.549 (4) C31—H31C 0.9600 1.556 (4) C31—H31D 0.9600 1.526 (5) C32—C33 1.477 (5) 0.9800 C32—N34 1.361 (6) 1.537 (5) C33—H33A 0.9600 0.9700 C33—H33B 0.9600 0.9700 C33—H33C 0.9600 1.437 (4) N34—H34 0.980 (18) 1.548 (4) N34—O35 1.367 (5) 0.9800 116.0 (3) O22—C16—H16B 110.3 108.3 C15—C16—H16B 110.3 108.3 C17—C16—H16B 110.3 108.3 C20—C17—C16 104.0 (2) 108.3 C20—C17—C13 121.6 (3) 107.4 C16—C17—C13 104.4 (2) 111.4 (3) C20—C17—H17C 108.7 109.3 C16—C17—H17C 108.7 109.3 C13—C17—H17C 108.7 109.3 C13—C18—H18C 109.5 109.3 C13—C18—H18D 109.5 108.0 H18C—C18—H18D 109.5 105.5 (3) C13—C18—H18E 109.5 109.0 (3) H18C—C18—H18E 109.5 111.0 (3) H18D—C18—H18E 109.5 110.4 C10—C19—H19A 109.5 110.4 C10—C19—H19B 109.5 110.4 H19A—C19—H19B 109.5 113.7 (3) C10—C19—H19C 109.5 108.8 H19A—C19—H19C 109.5 108.8 H19B—C19—H19C 109.5 108.8 C21—C20—C22 114.9 (3) 108.8 C21—C20—C17 113.7 (3) 107.7 C22—C20—C17 104.2 (3) 110.9 (3) C21—C20—H20A 107.9 112.0 (3) C22—C20—H20A 107.9 113.4 (3) C17—C20—H20A 107.9 106.7 C20—C21—H21B 109.5 106.7 C20—C21—H21C 109.5 sup-7 supplementary materials supplementary materials 106.7 H21B—C21—H21C 109.5 111.6 (3) C20—C21—H21D 109.5 109.3 H21B—C21—H21D 109.5 109.3 H21C—C21—H21D 109.5 109.3 O26—C22—O22 109.7 (2) 109.3 O26—C22—C20 106.4 (3) 108.0 O22—C22—C20 105.3 (2) 111.7 (3) O26—C22—C23 110.5 (2) 109.3 O22—C22—C23 108.6 (2) 109.3 C20—C22—C23 116.1 (3) 109.3 C22—O22—C16 106.5 (2) 109.3 C32—C23—C24 112.1 (3) 107.9 C32—C23—C22 112.4 (3) 112.0 (3) C24—C23—C22 111.1 (3) 111.5 (3) C32—C23—H23C 107.0 108.4 (3) C24—C23—H23C 107.0 108.3 C22—C23—H23C 107.0 108.3 C25—C24—C23 112.2 (3) 108.3 C25—C24—H24C 109.2 111.1 (3) C23—C24—H24C 109.2 114.0 (3) C25—C24—H24D 109.2 112.2 (3) C23—C24—H24D 109.2 106.3 H24C—C24—H24D 107.9 106.3 C26—C25—C24 107.1 (3) 106.3 C26—C25—C27 112.7 (3) 106.9 (3) C24—C25—C27 112.6 (3) 109.4 (3) C26—C25—H25C 108.1 107.8 (3) C24—C25—H25C 108.1 110.5 (3) C27—C25—H25C 108.1 111.8 (3) O26—C26—C25 111.2 (3) 110.2 (3) O26—C26—H26D 109.4 113.8 (3) C25—C26—H26D 109.4 108.8 O26—C26—H26E 109.4 108.8 C25—C26—H26E 109.4 108.8 H26D—C26—H26E 108.0 108.8 C22—O26—C26 114.2 (3) 107.7 C25—C27—H27C 109.5 111.9 (3) C25—C27—H27D 109.5 109.2 H27C—C27—H27D 109.5 109.2 C25—C27—H27E 109.5 109.2 H27C—C27—H27E 109.5 109.2 H27D—C27—H27E 109.5 107.9 C29—O28—C3 118.9 (3) 110.3 (3) O30—C29—O28 123.6 (3) 108.0 (3) O30—C29—C31 125.0 (4) 112.1 (3) O28—C29—C31 111.3 (4) 114.8 (3) C29—C31—H31B 109.5 111.7 (3) C29—C31—H31C 109.5 99.5 (2) H31B—C31—H31C 109.5 C10—C5—H5A 106.7 H21B—C21—H21C 109.5 C5—C6—C7 111.6 (3) C20—C21—H21D 109.5 C5—C6—H6A 109.3 H21B—C21—H21D 109.5 C7—C6—H6A 109.3 H21C—C21—H21D 109.5 C5—C6—H6B 109.3 O26—C22—O22 109.7 (2) C7—C6—H6B 109.3 O26—C22—C20 106.4 (3) H6A—C6—H6B 108.0 O22—C22—C20 105.3 (2) C8—C7—C6 111.7 (3) O26—C22—C23 110.5 (2) C8—C7—H7B 109.3 O22—C22—C23 108.6 (2) C6—C7—H7B 109.3 C20—C22—C23 116.1 (3) C8—C7—H7C 109.3 C22—O22—C16 106.5 (2) C6—C7—H7C 109.3 C32—C23—C24 112.1 (3) H7B—C7—H7C 107.9 C32—C23—C22 112.4 (3) C7—C8—C14 112.0 (3) C24—C23—C22 111.1 (3) C7—C8—C9 111.5 (3) C32—C23—H23C 107.0 C14—C8—C9 108.4 (3) C24—C23—H23C 107.0 C7—C8—H8B 108.3 C22—C23—H23C 107.0 C14—C8—H8B 108.3 C25—C24—C23 112.2 (3) C9—C8—H8B 108.3 C25—C24—H24C 109.2 C11—C9—C8 111.1 (3) C23—C24—H24C 109.2 C11—C9—C10 114.0 (3) C25—C24—H24D 109.2 C8—C9—C10 112.2 (3) C23—C24—H24D 109.2 C11—C9—H9B 106.3 H24C—C24—H24D 107.9 C8—C9—H9B 106.3 C26—C25—C24 107.1 (3) C10—C9—H9B 106.3 C26—C25—C27 112.7 (3) C19—C10—C1 106.9 (3) C24—C25—C27 112.6 (3) C19—C10—C5 109.4 (3) C26—C25—H25C 108.1 C1—C10—C5 107.8 (3) C24—C25—H25C 108.1 C19—C10—C9 110.5 (3) C27—C25—H25C 108.1 C1—C10—C9 111.8 (3) O26—C26—C25 111.2 (3) C5—C10—C9 110.2 (3) O26—C26—H26D 109.4 C9—C11—C12 113.8 (3) C25—C26—H26D 109.4 C9—C11—H11B 108.8 O26—C26—H26E 109.4 C12—C11—H11B 108.8 C25—C26—H26E 109.4 C9—C11—H11C 108.8 H26D—C26—H26E 108.0 C12—C11—H11C 108.8 C22—O26—C26 114.2 (3) H11B—C11—H11C 107.7 C25—C27—H27C 109.5 C13—C12—C11 111.9 (3) C25—C27—H27D 109.5 C13—C12—H12A 109.2 H27C—C27—H27D 109.5 C11—C12—H12A 109.2 C25—C27—H27E 109.5 C13—C12—H12B 109.2 H27C—C27—H27E 109.5 C11—C12—H12B 109.2 H27D—C27—H27E 109.5 H12A—C12—H12B 107.9 C29—O28—C3 118.9 (3) C12—C13—C18 110.3 (3) O30—C29—O28 123.6 (3) C12—C13—C14 108.0 (3) O30—C29—C31 125.0 (4) C18—C13—C14 112.1 (3) O28—C29—C31 111.3 (4) C12—C13—C17 114.8 (3) C29—C31—H31B 109.5 C18—C13—C17 111.7 (3) C29—C31—H31C 109.5 C14—C13—C17 99.5 (2) H31B—C31—H31C 109.5 C10—C5—H5A 106.7 H21B—C21— C5—C6—C7 111.6 (3) C20—C21—H C5—C6—H6A 109.3 H21B—C21— C7—C6—H6A 109.3 H21C—C21— C5—C6—H6B 109.3 O26—C22—O C7—C6—H6B 109.3 O26—C22—C H6A—C6—H6B 108.0 O22—C22—C C8—C7—C6 111.7 (3) O26—C22—C C8—C7—H7B 109.3 O22—C22—C C6—C7—H7B 109.3 C20—C22—C C8—C7—H7C 109.3 C22—O22—C C6—C7—H7C 109.3 C32—C23—C H7B—C7—H7C 107.9 C32—C23—C C7—C8—C14 112.0 (3) C24—C23—C C7—C8—C9 111.5 (3) C32—C23—H C14—C8—C9 108.4 (3) C24—C23—H C7—C8—H8B 108.3 C22—C23—H C14—C8—H8B 108.3 C25—C24—C C9—C8—H8B 108.3 C25—C24—H C11—C9—C8 111.1 (3) C23—C24—H C11—C9—C10 114.0 (3) C25—C24—H C8—C9—C10 112.2 (3) C23—C24—H C11—C9—H9B 106.3 H24C—C24— C8—C9—H9B 106.3 C26—C25—C C10—C9—H9B 106.3 C26—C25—C C19—C10—C1 106.9 (3) C24—C25—C C19—C10—C5 109.4 (3) C26—C25—H C1—C10—C5 107.8 (3) C24—C25—H C19—C10—C9 110.5 (3) C27—C25—H C1—C10—C9 111.8 (3) O26—C26—C C5—C10—C9 110.2 (3) O26—C26—H C9—C11—C12 113.8 (3) C25—C26—H C9—C11—H11B 108.8 O26—C26—H C12—C11—H11B 108.8 C25—C26—H C9—C11—H11C 108.8 H26D—C26— C12—C11—H11C 108.8 C22—O26—C H11B—C11—H11C 107.7 C25—C27—H C13—C12—C11 111.9 (3) C25—C27—H C13—C12—H12A 109.2 H27C—C27— C11—C12—H12A 109.2 C25—C27—H C13—C12—H12B 109.2 H27C—C27— C11—C12—H12B 109.2 H27D—C27— H12A—C12—H12B 107.9 C29—O28—C C12—C13—C18 110.3 (3) O30—C29—O C12—C13—C14 108.0 (3) O30—C29—C C18—C13—C14 112.1 (3) O28—C29—C C12—C13—C17 114.8 (3) C29—C31—H C18—C13—C17 111.7 (3) C29—C31—H C14—C13—C17 99.5 (2) H31B—C31— sup-8 supplementary materials supplementary materials su 120.3 (3) C29—C31—H31D 109.5 103.8 (2) H31B—C31—H31D 109.5 114.1 (3) H31C—C31—H31D 109.5 105.8 C23—C32—C33 122.0 (3) 105.8 N34—C32—C23 111.6 (3) 105.8 N34—C32—C33 126.3 (3) 101.9 (3) C32—C33—H33A 109.5 111.4 C32—C33—H33B 109.5 111.4 H33A—C33—H33B 109.5 111.4 C32—C33—H33C 109.5 111.4 H33A—C33—H33C 109.5 109.2 H33B—C33—H33C 109.5 112.6 (3) C32—N34—O35 109.0 (4) 105.5 (2) C32—N34—H34 132.0 (16) 107.7 (3) O35—N34—H34 118.6 (16) 54.0 (4) O22—C16—C17—C20 19.0 (3) 62.8 (4) C15—C16—C17—C20 139.4 (3) −53.0 (4) O22—C16—C17—C13 −109.5 (3) −63.8 (4) C15—C16—C17—C13 10.9 (3) 54.2 (4) C12—C13—C17—C20 93.1 (4) 179.0 (3) C18—C13—C17—C20 −33.5 (4) −53.9 (4) C14—C13—C17—C20 −151.9 (3) 72.1 (4) C12—C13—C17—C16 −150.0 (3) −55.8 (5) C18—C13—C17—C16 83.4 (3) 56.3 (5) C14—C13—C17—C16 −35.0 (3) −177.0 (3) C16—C17—C20—C21 129.4 (3) −55.3 (4) C13—C17—C20—C21 −113.5 (3) −177.2 (3) C16—C17—C20—C22 3.6 (3) −53.4 (4) C13—C17—C20—C22 120.6 (3) 53.9 (4) C21—C20—C22—O26 −33.8 (4) 177.7 (3) C17—C20—C22—O26 91.3 (3) −168.4 (3) C21—C20—C22—O22 −150.3 (3) −50.8 (4) C17—C20—C22—O22 −25.2 (3) 70.5 (4) C21—C20—C22—C23 89.5 (4) −68.2 (4) C17—C20—C22—C23 −145.4 (3) 165.2 (3) O26—C22—O22—C16 −75.4 (3) 175.9 (3) C20—C22—O22—C16 38.8 (3) 49.3 (4) C23—C22—O22—C16 163.8 (3) 53.5 (4) C15—C16—O22—C22 −153.5 (3) −73.0 (4) C17—C16—O22—C22 −36.3 (3) −58.8 (4) O26—C22—C23—C32 −176.2 (3) 68.5 (4) O22—C22—C23—C32 −55.8 (3) 60.1 (4) C20—C22—C23—C32 62.6 (3) −172.5 (3) O26—C22—C23—C24 −49.7 (3) −179.9 (3) O22—C22—C23—C24 70.7 (3) −52.5 (4) C20—C22—C23—C24 −171.0 (3) 52.7 (4) C32—C23—C24—C25 177.9 (3) −179.3 (3) C22—C23—C24—C25 51.2 (4) −53.8 (4) C23—C24—C25—C26 −54.9 (4) C15—C14—C8 120.3 (3) C29—C31—H31D C15—C14—C13 103.8 (2) H31B—C31—H31D C8—C14—C13 114.1 (3) H31C—C31—H31D C15—C14—H14B 105.8 C23—C32—C33 C8—C14—H14B 105.8 N34—C32—C23 C13—C14—H14B 105.8 N34—C32—C33 C14—C15—C16 101.9 (3) C32—C33—H33A C14—C15—H15B 111.4 C32—C33—H33B C16—C15—H15B 111.4 H33A—C33—H33B C14—C15—H15C 111.4 C32—C33—H33C C16—C15—H15C 111.4 H33A—C33—H33C H15B—C15—H15C 109.2 H33B—C33—H33C O22—C16—C15 112.6 (3) C32—N34—O35 O22—C16—C17 105.5 (2) C32—N34—H34 C15—C16—C17 107.7 (3) O35—N34—H34 C10—C1—C2—C3 54.0 (4) O22—C16—C17—C20 C1—C2—C3—O28 62.8 (4) C15—C16—C17—C20 C1—C2—C3—C4 −53.0 (4) O22—C16—C17—C13 O28—C3—C4—C5 −63.8 (4) C15—C16—C17—C13 C2—C3—C4—C5 54.2 (4) C12—C13—C17—C20 C3—C4—C5—C6 179.0 (3) C18—C13—C17—C20 C3—C4—C5—C10 −53.9 (4) C14—C13—C17—C20 C4—C5—C6—C7 72.1 (4) C12—C13—C17—C16 C10—C5—C6—C7 −55.8 (5) C18—C13—C17—C16 C5—C6—C7—C8 56.3 (5) C14—C13—C17—C16 C6—C7—C8—C14 −177.0 (3) C16—C17—C20—C21 C6—C7—C8—C9 −55.3 (4) C13—C17—C20—C21 C7—C8—C9—C11 −177.2 (3) C16—C17—C20—C22 C14—C8—C9—C11 −53.4 (4) C13—C17—C20—C22 C7—C8—C9—C10 53.9 (4) C21—C20—C22—O26 C14—C8—C9—C10 177.7 (3) C17—C20—C22—O26 C2—C1—C10—C19 −168.4 (3) C21—C20—C22—O22 C2—C1—C10—C5 −50.8 (4) C17—C20—C22—O22 C2—C1—C10—C9 70.5 (4) C21—C20—C22—C23 C6—C5—C10—C19 −68.2 (4) C17—C20—C22—C23 C4—C5—C10—C19 165.2 (3) O26—C22—O22—C16 C6—C5—C10—C1 175.9 (3) C20—C22—O22—C16 C4—C5—C10—C1 49.3 (4) C23—C22—O22—C16 C6—C5—C10—C9 53.5 (4) C15—C16—O22—C22 C4—C5—C10—C9 −73.0 (4) C17—C16—O22—C22 C11—C9—C10—C19 −58.8 (4) O26—C22—C23—C32 C8—C9—C10—C19 68.5 (4) O22—C22—C23—C32 C11—C9—C10—C1 60.1 (4) C20—C22—C23—C32 C8—C9—C10—C1 −172.5 (3) O26—C22—C23—C24 C11—C9—C10—C5 −179.9 (3) O22—C22—C23—C24 C8—C9—C10—C5 −52.5 (4) C20—C22—C23—C24 C8—C9—C11—C12 52.7 (4) C32—C23—C24—C25 C10—C9—C11—C12 −179.3 (3) C22—C23—C24—C25 C9—C11—C12—C13 −53.8 (4) C23—C24—C25—C26 sup-9 supplementary materials supplementary materials C11—C12—C13—C18 −68.7 (4) C23—C24—C25—C27 69.6 (4) C11—C12—C13—C14 54.1 (4) C24—C25—C26—O26 59.4 (4) C11—C12—C13—C17 164.1 (3) C27—C25—C26—O26 −65.0 (4) C7—C8—C14—C15 −53.3 (4) O22—C22—O26—C26 −63.2 (3) C9—C8—C14—C15 −176.8 (3) C20—C22—O26—C26 −176.7 (3) C7—C8—C14—C13 −177.7 (3) C23—C22—O26—C26 56.4 (3) C9—C8—C14—C13 58.9 (4) C25—C26—O26—C22 −63.0 (4) C12—C13—C14—C15 167.9 (3) C4—C3—O28—C29 −154.6 (3) C18—C13—C14—C15 −70.4 (3) C2—C3—O28—C29 86.1 (4) C17—C13—C14—C15 47.8 (3) C3—O28—C29—O30 −0.2 (5) C12—C13—C14—C8 −59.3 (4) C3—O28—C29—C31 179.7 (3) C18—C13—C14—C8 62.4 (4) C24—C23—C32—N34 135.1 (3) C17—C13—C14—C8 −179.4 (3) C22—C23—C32—N34 −99.0 (3) C8—C14—C15—C16 −170.4 (3) C24—C23—C32—C33 −44.6 (4) C13—C14—C15—C16 −41.2 (3) C22—C23—C32—C33 81.3 (4) C14—C15—C16—O22 134.1 (3) C33—C32—N34—O35 0.2 (6) C14—C15—C16—C17 18.3 (4) C23—C32—N34—O35 −179.5 (3) sup-10 supplementary materials Fig. 1 sup-11
https://openalex.org/W4367310423	https://www.mdpi.com/2077-1312/11/5/948/pdf?version=1683351489	English	null	Artificial Reefs Reduce Morbidity and Mortality of Small Cultured Sea Cucumbers Apostichopus japonicus at High Temperature	Journal of marine science and engineering	2,023	cc-by	8,763	Citation: Wang, H.; Wu, G.; Hu, F.; Tian, R.; Ding, J.; Chang, Y.; Su, Y.; Zhao, C. Artiﬁcial Reefs Reduce Morbidity and Mortality of Small Cultured Sea Cucumbers Apostichopus japonicus at High Temperature. J. Mar. Sci. Eng. 2023, 11, 948. https://doi.org/10.3390/ jmse11050948 Keywords: sea cucumbers; morbidity; mortality; artiﬁcial reef; high temperature Received: 8 March 2023 Revised: 23 April 2023 Accepted: 26 April 2023 Published: 28 April 2023 Journal of Marine Science and Engineering Journal of Marine Science and Engineering Journal of Marine Science and Engineering Journal of Marine Science and Engineering Journal of Marine Science and Engineering Article Artiﬁcial Reefs Reduce Morbidity and Mortality of Small Cultured Sea Cucumbers Apostichopus japonicus at High Temperature n Wang †, Guo Wu †, Fangyuan Hu, Ruihuan Tian, Jun Ding , Yaqing Chang, Yanming Su * h Zh * Key Laboratory of Mariculture and Stock Enhancement in North China′s Sea, Ministry of Agriculture and Rural Affairs, Dalian Ocean University, Dalian 116023, China Key Laboratory of Mariculture and Stock Enhancement in North China′s Sea, Ministry of Agriculture and Rural Affairs, Dalian Ocean University, Dalian 116023, China * Correspondence: suym@dlou.edu.cn (Y.S.); chongzhao@dlou.edu.cn (C.Z.) † These authors contributed equally to this work. † These authors contributed equally to this work. Abstract: Summer mortality and morbidity are serious environment-related problems in cultured sea cucumbers (Apostichopus japonicus). Air exposure probably worsens the impact of high temperature on cultured sea cucumbers. In this present study, two laboratory experiments were designed to investigate the effects of artiﬁcial reefs on mortality, morbidity, crawling, feeding, and adhesion behaviors of small sea cucumbers (~1 g of wet body weight) after air exposure and disease outbreaks at 25 ◦C, respectively. Signiﬁcantly lower mortality and morbidity occurred in the group with artiﬁcial reefs compared with those in the group without artiﬁcial reefs in the two experiments. This present study found that the stressed sea cucumbers cultured inside artiﬁcial reefs showed a signiﬁcantly higher adhesion index, feeding behavior, and crawling frequency than those cultured without artiﬁcial reefs. In disease challenge assays, small sea cucumbers cultured inside the artiﬁcial reefs showed a signiﬁcantly higher adhesion index and crawling frequency than those cultured without artiﬁcial reefs at 25 ◦C. Feeding, crawling, and adhesion behaviors of sea cucumbers cultured outside artiﬁcial reefs were not signiﬁcantly different from those cultured without artiﬁcial reefs. The experimental results indicate that sea cucumbers with good ﬁtness-related behaviors may be less affected by the disease and more likely to move into the crevices of artiﬁcial reefs. Fitness-related behaviors were poor in sea cucumbers cultured outside artiﬁcial reefs, so we considered them as affected individuals. Thus, artiﬁcial reefs provide a place to reduce the physical contact between unaffected and diseased/affected individuals, showing a potential to reduce disease transmission. Our present study establishes a cost-effective approach to increasing the survival of small sea cucumbers in seed production at high temperatures. 1. Introduction The sea cucumber (Apostichopus japonicus) is popular internationally because of its high nutritional and medicinal value [1]. High market demand stimulated the development of sea cucumber aquaculture [2,3]. The annual production of sea cucumber aquaculture was 222,707 tons in 2022 in China, which was 13.3% higher than that in 2021 [4]. Seed production is a common process in aquaculture, in which small sea cucumbers (~1 g of wet body weight) are intensively cultured in land-based factories until they reach the size available for the following pond culture and stock enhancement [5]. During this process, sea cucumbers are commonly challenged by high temperatures in summer [5]. Sea cucumbers are most suitable for survival in the temperature range between 10 and 21 ◦C, and the physiology and growth of sea cucumbers are greatly affected when water temperature is higher than 23 ◦C [6,7]. High temperature (>25 ◦C) damages the immune system and gut microbiota [6,7] and, consequently, causes mass mortality and morbidity Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). https://www.mdpi.com/journal/jmse J. Mar. Sci. Eng. 2023, 11, 948. https://doi.org/10.3390/jmse11050948 J. Mar. Sci. Eng. 2023, 11, 948 2 of 12 of sea cucumbers in summer. It is very expensive to decrease the water temperature in seed production in summer. Therefore, it is essential to establish a cost-effective method to reduce the mortality and morbidity of sea cucumbers at high temperatures. y y g p Changing seawater and removing feces are necessary in seed production. Thus, sea cucumbers are inevitably exposed to the air during this process. Air exposure leads to tissue damage and complicated oxidative stress [8,9], which increases the risk of death and disease in sea cucumbers [10]. Sea cucumbers are more severely affected when high temperature and air exposure happen together [11], which is probably responsible for the high mortality and morbidity of sea cucumbers in seed production in summer. In addition, skin ulcer syndrome (SUS) is one of the most common and severe diseases that occur to sea cucumbers in aquaculture, resulting in high mortality due to the high infectivity [12]. High temperature decreases the immunity of sea cucumbers and increases the transmission of pathogenic bacteria [13]. Therefore, sea cucumbers are more susceptible to disease in summer. Diseased sea cucumbers probably spread the disease to other individuals and increase the frequency of disease outbreaks [14,15], resulting in mass mortality and morbidity in aquaculture [16]. Therefore, it is of primary importance to reduce the mortality and morbidity of sea cucumbers after air exposure and disease outbreaks. Although there is no obvious symptom (e.g., skin ulceration) in the early stages of the disease, sea cucumbers show decreased feeding, movement, and adhesion [16,17]. Thus, ﬁtness-related behaviors are helpful for determining whether sea cucumbers are affected. Decreasing disease transmission is key to reducing mortality and morbidity during the disease outbreak. Our previous study suggests that separation effectively decreases the disease transmission of the sea urchin Strongylocentrotus intermedius [18]. However, it is costly and difﬁcult to separate diseased/affected from unaffected individuals in aquaculture. We attempted to reduce mortality and morbidity by reducing the frequency that unaffected individuals have contact with diseased/affected ones [19]. Sea cucumbers are naturally attracted to artiﬁcial reefs [20], and they hide into crevices formed by rocks or reefs at high temperatures [21]. Diseased and affected individuals probably struggle to move to crevices because of their poor movement ability [16]. Tian et al. [11] found that artiﬁcial reefs improve ﬁtness-related behaviors of stressed sea cucumbers at high temperatures. It is necessary to explore the key role of artiﬁcial reefs, which is beneﬁcial to reduce the mortality and morbidity of sea cucumbers at high temperatures. Thus, we hypothesized that artiﬁcial reefs reduce mortality and morbidity by reducing physical contact between diseased/affected and unaffected sea cucumbers in seed production at high temperatures. g p The main aim of the present study is to investigate whether artiﬁcial reefs improve sea cucumber survival in summer, as detailed below: (1) whether artiﬁcial reefs reduce the morbidity and mortality of sea cucumbers exposed to the air at high temperatures; (2) whether artiﬁcial reefs reduce the morbidity and mortality of sea cucumbers under a disease outbreak at high temperatures; and (3) whether artiﬁcial reefs improve the ﬁtness- related behaviors of sea cucumbers exposed to both air exposure and disease outbreak at high temperatures. 2.2. Experiment I This experiment was carried out to investigate whether artiﬁcial reefs improve the survival and ﬁtness-related behaviors (crawling, feeding, and adhesion) of sea cucumbers exposed to the air at 25 ◦C. The presence or absence of artiﬁcial reefs was the experimental factor. An artiﬁcial reef (length × width × height: 50 mm × 60 mm × 55 mm, Figure 1A) consists of seven plastic cylinders (diameter × length: 15 mm × 50 mm, Figure 1A), in which only one sea cucumber is accessible. The plastic boxes (top length × bottom length × height: 20 cm × 16.5 cm × 14 cm) with and without artiﬁcial reefs were named group O and group C, respectively (n = 6). Twelve plastics boxes were placed in two temperature- controlled tanks (length × width × height: 115 cm × 75 cm × 60 cm) to maintain the water temperature at 25 ◦C. Two groups of sixteen sea cucumbers (~1 g of wet body weight) without ulcerated body walls were randomly selected and placed in each box with aeration (Figure 1B). Groups C and O each had six parallel groups. Sea cucumbers in each box were fed 5 g of a mixture of sea mud and commercial feed (Anyuan Industrial Co., Ltd. Yantai, China) at a 1:4 ratio every day. Sand-ﬁltered seawater was changed daily in plastic boxes. Sea cucumbers were exposed to the air for 30 min after the seawater was removed from each box. Sea cucumbers were cultured in the above experimental conditions for 5 days. p y We identified the sea cucumber with ulcerated skin as a diseased individual (Figure 1D). The numbers of dead and diseased sea cucumbers in each group were counted after the experiment. Sea cucumbers without ulcerated body walls (Figure 1D) were placed in plastics boxes (top length × bottom length × height: 20 cm × 16.5 cm × 14 cm) in group C. Consistently, individuals without ulcerated body walls in group O were divided into two groups, among which sea cucumbers in one group were inside the artiﬁcial reefs and sea cucumbers in the other group were outside the artiﬁcial reefs. Then, the two groups were placed in two different cubic plastic boxes (top length × bottom length × height: 20 cm × 16.5 cm × 14 cm). 2.1. Experimental Animals Small green A. japonicus (~1 g of wet body weight) were transported from Yinhaima Aquatic Products Co., Ltd. (121◦54′ E, 39◦38′ N) to the Key Laboratory of Mariculture and Stock Enhancement in the North China’s Sea and Ministry of Agriculture and Rural Affairs at Dalian Ocean University (121◦56′ E, 38◦87′ N) on 2 March 2022. Small sea cucumbers were cultured in ﬁberglass tanks (length × width × height: 115 cm × 75 cm × 60 cm) at a density of 400 g/m2 [22,23] and with aeration. Water temperature was increased 1 ◦C each day from 8 ◦C to 25 ◦C (ambient temperature) by using a temperature-controlled system (Huixin Co., Dalian, China). Sea cucumbers were maintained at 25 ◦C for two weeks to be acclimatized to the experimental conditions. Sand-ﬁltered seawater was replaced J. Mar. Sci. Eng. 2023, 11, 948 3 of 12 3 of 12 every three days with feces and uneaten food removed. Salinity, pH, and dissolved oxygen were tested weekly using a water quality detector (YSI Incorporated, Yellow Springs, OH, USA). The results were 31.36 ± 0.19, 8.01 ± 0.03 and 9.34 ± 0.21 mg/L, respectively. Small sea cucumbers were fed ad libitum with sea mud and commercial feed (Anyuan Industrial Co., Ltd., Yantai, China) at a ratio of 1:4 until the experiments started in April 2022. 2.4. Mortality and Morbidity Skin ulceration begins with one or more small white ulcerative spots, followed by the appearance of deep and enlarging ulcerative lesions, resulting in the exposure of the underlying muscles and spicules [24] (Figure 1D). Mortality and morbidity were calculated as follows: D Mt = De T × 100% Mb = Di T × 100% Mt = De T × 100% Mb = Di T × 100% where Mt is the mortality, Mb is the morbidity, De is the number of dead sea cucumbers, Di is the number of diseased sea cucumbers, and T is the total number of sea cucumbers. where Mt is the mortality, Mb is the morbidity, De is the number of dead sea cucumbers, Di is the number of diseased sea cucumbers, and T is the total number of sea cucumbers. 2.3. Experiment II 2.3. Experiment II We further investigated whether artiﬁcial reefs improve the survival and ﬁtness- related behaviors (crawling, feeding, and adhesion) of sea cucumbers at 25 ◦C when disease outbreaks. The same artiﬁcial reefs as experiment I were consistently used in experiment II. Sixteen sea cucumbers without ulcerated body walls were placed in each box (Figure 1C). Diseased sea cucumbers were subsequently put in to create a disease outbreak. The plastic boxes (top length × bottom length × height: 20 cm × 16.5 cm × 14 cm) with and without artiﬁcial reefs were named as group O and group C, respectively (n = 6). Twelve plastic boxes (top length × bottom length × height: 20 cm × 16.5 cm × 14 cm) were placed in two temperature-controlled tanks (length × width × height: 115 cm × 75 cm × 60 cm). Each group was daily fed ad libitum with 5 g of a mixture of sea mud and commercial feed (Anyuan Industrial Co., Ltd., Yantai, China) at a 1:4 ratio. Sea cucumbers were cultured for three days without the seawater replacement in the same experimental conditions. The numbers of dead and diseased sea cucumbers were counted after the experi- ment. Consistently, we obtained three groups of sea cucumbers without ulcerated body walls, which were the ones from group C, the ones inside the artiﬁcial reefs in group O (group O—In), and the ones outside the artiﬁcial reefs in group O (group O—Out). Crawl- ing, feeding, and adhesion behaviors of sea cucumbers were recorded following the same procedures as described above for each group. 2.2. Experiment I Finally, we obtained three groups of sea cucumbers without ul- cerated body walls, which were the ones from group C, the ones inside the artiﬁcial reefs in group O (group O—In), and the ones outside the artiﬁcial reefs in group O (group O—Out). Three sea cucumbers were randomly selected from each of the three groups for the mea- surement of ﬁtness-related behaviors (feeding, crawling, and adhesion). Feeding behav- ior and crawling behavior of the sea cucumbers were recorded using a digital camera (FDR-AXP55, Shanghai Suoguang Electronics Co., Ltd., Shanghai, China) ﬁxed above the plastic cubic boxes (length × width × height: 60 mm × 47 mm × 45 mm) within 1 h. Adhesion behavior was measured in cubic plastic boxes (length × width × height: 180 mm × 140 mm × 45 mm) with 18 compartments. We repeated this experiment 6 times using different sea cucumbers for each group (n = 6). 4 of 12 f 12 J. Mar. Sci. Eng. 2023, 11, 948 J. Mar. Sci. Eng. 2023, 11, 94 Figure 1. The diagrams show an artiﬁcial reef (A), the experiment designs for experiments I (B) an II (C), diseased and individuals without ulcerated skin (D), adhesion behavior (E), feeding behavi (F), and crawling behavior (G) of sea cucumbers. Figure 1. The diagrams show an artiﬁcial reef (A), the experiment designs for experiments I (B) and II (C), diseased and individuals without ulcerated skin (D), adhesion behavior (E), feeding behavior (F), and crawling behavior (G) of sea cucumbers. 5 of 12 J. Mar. Sci. Eng. 2023, 11, 948 2.5. Crawling Behavior Crawling behavior of sea cucumbers is composed of ﬁve stages (Figure 1G) [25]. A sea cucumber extends its body and then enters the contraction stage. The sea cu- cumber contracts from the back of the body to the middle and front in sequence. Fi- nally, it returns to a static state [25]. Sea cucumbers complete each crawling cycle as described above, which is counted as one occurrence of crawling behavior. Sea cucum- bers were randomly selected from each group and were placed into cubic plastic boxes (length × width × height: 60 mm × 47 mm × 45 mm). Two grams of mixture of sea mud and commercial feed at a ratio of 1:4 was put in the bottom of the box, and the number of occurrences of crawling behavior was counted within 1 h by using a digital camera (FDR-AXP55, Shanghai Suoguang Electronics Co., Ltd., Shanghai, China) (n = 6). 2.6. Feeding Behavior Feeding behavior refers to the process of sea cucumbers ingesting food through its mouth tentacle activity [26]. Sea cucumbers stretch their tentacles out of their mouths, then extend their tentacles to grab the food, and ﬁnally capture the food into their mouths by contracting their tentacles [26]. One occurrence of tentacle activity was counted if sea cucum- bers completed one tentacle activity cycle. Sea cucumbers were selected from each group and placed into cubic plastic boxes (length × width × height: 60 mm × 47 mm × 45 mm; Figure 1F). Two grams of mixture of sea mud and commercial feed at a ratio of 1:4 was put in the bottom of the box, and the duration of tentacle activity of all sea cucumbers J. Mar. Sci. Eng. 2023, 11, 948 6 of 12 within one hour was recorded by using a digital camera (FDR-AXP55, Shanghai Suoguang Electronics Co., Ltd., Shanghai, China) (n = 6). within one hour was recorded by using a digital camera (FDR-AXP55, Shanghai Suoguang Electronics Co., Ltd., Shanghai, China) (n = 6). 2.7. Adhesion Behavior Movement and ingestion of sea cucumbers greatly relies on the adhesion of mouth tentacles and tube feet. Adhesion behavior was measured according to Tian et al. [11]. Eighteen sea cucumbers from each group were randomly selected and placed in cubic devices (length × width × height: 180 mm × 140 mm × 45 mm; Figure 1E) with eighteen compartments. Each sea cucumber was placed in one compartment. Seawater was added to the devices until the seawater level reached a height of 2 cm. The devices were slowly turned to 180◦after 10 min. We recorded the time that sea cucumbers fell from the top of the devices, since their adhesion abilities were disabled to support their body weight. The adhesion time was recorded as 600 s if sea cucumbers still adhered to the top of the box after 10 min. Wet body weight (g) of each sea cucumber was measured by using an electronic balance (G & G Co., San Diego, CA, USA) after the recording (n = 6). Data are accurate to one decimal place. Adhesion index (Ai) was formulated, as described by Tian et al. [11]: Ai = T W where Ai is the adhesion index, T is the adhesion time (s), and W is the wet weight of an individual sea cucumber (g). 2.8. Statistical Analysis All data were subjected to the analysis of variance distribution and homogeneity of variance using the Kolmogorov–Smirnov test and the Levene test, respectively. The mortality and morbidity of experiments I and II were non-normally distributed and/or heterogeneity of variance. Thus, the data were compared by using the Mann–Whitney U test. Crawling frequency, tentacle activity frequency, and the adhesion index were analyzed by using Kruskal–Wallis H in both experiments. All statistical analyses were performed using SPSS 22.0 statistical software. The level of signiﬁcance was considered as p < 0.05. 3. Results 3.1. Experiment I 3.1.1. Mortality and Morbidity 3.1.2. Crawling Frequency 3.1.2. Crawling Frequency There was no signiﬁcan There was no signiﬁcant difference in crawling frequency between sea cucumbers of group C (4.44 ± 0.88 times) and individuals outside the artiﬁcial reef of group O (3.22 ± 0.50 times; Kruskal–Wallis H = 14.332, p = 1.000; Figure 3A). Crawling frequency was signiﬁcantly higher in sea cucumbers inside the artiﬁcial reefs of group O (10.83 ± 1.52 times) than that in sea cucumbers outside the artiﬁcial reefs of group O (Kruskal–Wallis H = 14.332, p = 0.001) and group C (Kruskal–Wallis H = 14.332, p = 0.008; Figure 3A). There was no signiﬁcant diﬀerence in crawling frequency between sea cucumbers of group C (4.44 ± 0.88 times) and individuals outside the artiﬁcial reef of group O (3.22 ± 0.50 times; Kruskal–Wallis H = 14.332, p = 1.000; Figure 3A). Crawling frequency was sig- niﬁcantly higher in sea cucumbers inside the artiﬁcial reefs of group O (10.83 ± 1.52 times) than that in sea cucumbers outside the artiﬁcial reefs of group O (Kruskal–Wallis H = 14.332, p = 0.001) and group C (Kruskal–Wallis H = 14.332, p = 0.008; Figure 3A). Figure 3. Crawling frequency (A), tentacles activity frequency (B) and adhesion behavior (C) of sea cucumbers. The asterisks , , ** mean p < 0.05, p < 0.01 and p < 0.001, respectively (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the artiﬁcial reefs in group O; Group O-In: sea cucumbers inside the artiﬁcial reefs in group O. Figure 3. Crawling frequency (A), tentacles activity frequency (B) and adhesion behavior (C) of sea cucumbers. The asterisks , , ** mean p < 0.05, p < 0.01 and p < 0.001, respectively (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the artiﬁcial reefs in group O; Group O-In: sea cucumbers inside the artiﬁcial reefs in group O. 3.1.3. Feeding Behavior Figure 3. Crawling frequency (A), tentacles activity frequency (B) and adhesion behavior (C) of sea ucumbers. The asterisks , , ** mean p < 0.05, p < 0.01 and p < 0.001, respectively (mean ± SE, n = ). Group C: sea cucumbers without artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside Figure 3. Crawling frequency (A), tentacles activity frequency (B) and adhesion behavior (C) of sea cucumbers. the artiﬁcial reefs in group O 3.1.3. Feeding Behavior There was no signiﬁcant difference in tentacle activity frequency between sea cu- cumbers of group C (10.44 ± 3.51 times) and individuals outside the artiﬁcial reefs of group O (8.50 ± 1.87 times; Kruskal–Wallis H = 7.705, p = 1.000; Figure 3B). Consistently, tentacle activity frequency was not signiﬁcantly different between sea cucumbers out- side (8.50 ± 1.87 times) and inside (18.72 ± 3.26 times) the artiﬁcial reefs (Kruskal–Wallis H = 7.705, p = 0.067; Figure 3B). Tentacle activity frequency was signiﬁcantly higher in sea cucumbers inside the artiﬁcial reefs in group O than that in individuals of group C (Kruskal–Wallis H = 7.705, p = 0.036; Figure 3B). 3.1.2. Crawling Frequency 3.1.2. Crawling Frequency There was no signiﬁcan The asterisks , , ** mean p < 0.05, p < 0.01 and p < 0.001, respectively (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the artiﬁcial reefs in group O; Group O-In: sea cucumbers inside the artiﬁcial reefs in group O. 3.1.4. Adhesion Behavior The adhesion index was not signiﬁcantly different between sea cucumbers outside the artiﬁcial reefs of group O (16.66 ± 5.50) and individuals of group C (9.75 ± 6.28; Kruskal– Wallis H = 0.561, p = 0.065; Figure 3C). Consistently, the adhesion index was not signiﬁcantly different in sea cucumbers outside and inside the artiﬁcial reefs (283.62 ± 61.4) in group O (Kruskal–Wallis H = 0.561, p = 0.076; Figure 3C). However, sea cucumbers inside the artiﬁcial reefs of group O showed signiﬁcantly higher adhesion index than individuals of group C (Kruskal–Wallis H = 0.561, p < 0.001; Figure 3C). 3.2. Experiment II 3.2.1. Mortality and Morbidity 3.2.1. Mortality and Morbidity 3.1.1. Mortality and Morbidity Mortality was signiﬁcantly higher in group C (43.06 ± 2.56%) than that in group O (5.56 ± 4.12%; Mann–Whitney U = 2.961, p = 0.003; Figure 2A). Consistently, group C showed signiﬁcantly higher morbidity (48.61 ± 1.39%) than group O (6.94 ± 3.98%; Mann– Whitney U = 3.017, p = 0.003; Figure 2B). 7 of 12 igure 2. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks * mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea ucumbers with artiﬁcial reefs. Figure 2. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea cucumbers with artiﬁcial reefs. gure 2. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea cumbers with artiﬁcial reefs. Figure 2. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea cucumbers with artiﬁcial reefs. igure 2. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea ucumbers with artiﬁcial reefs. Figure 2. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks ** mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea cucumbers with artiﬁcial reefs. J. Mar. Sci. Eng. 2023, 11, 948 7 of 12 up O: sea 7 of 12 up O: sea 3.1.2. Crawling Frequency 3.1.2. Crawling Frequency There was no signiﬁcan 3.1.2. Crawling Frequency 3.1.2. Crawling Frequency There was no signiﬁcan ucumbers with artiﬁcial reefs. 3.2.2. Crawling Frequency 2.2. Crawling Frequency There was no signiﬁcant diﬀerence in crawling frequency between sea cucumbers of roup C (1.67 ± 0.35 times) and individuals outside the artiﬁcial reefs in group O (2.56 ± 56 times; Kruskal–Wallis H = 6.908, p = 0.762; Figure 5A). Consistently, crawling frequen- es of sea cucumbers outside (2.56 ± 0.56 times) and inside (3.67 ± 0.60 times) the artiﬁcial eefs in group O (Kruskal–Wallis H = 6.908, p = 0.417; Figure 5A) were not signiﬁcantly There was no signiﬁcant difference in crawling frequency between sea cucumbers of group C (1.67 ± 0.35 times) and individuals outside the artiﬁcial reefs in group O (2.56 ± 0.56 times; Kruskal–Wallis H = 6.908, p = 0.762; Figure 5A). Consistently, crawling frequencies of sea cucumbers outside (2.56 ± 0.56 times) and inside (3.67 ± 0.60 times) the artiﬁcial reefs in group O (Kruskal–Wallis H = 6.908, p = 0.417; Figure 5A) were not signiﬁcantly different in the experiments. Signiﬁcantly higher crawling frequency occurred in sea cucumbers inside the artiﬁcial reefs in group O than individuals in group C (Kruskal– Wallis H = 6.908, p = 0.026; Figure 5A). 9 of 12 diﬀerent in the experiments. Signiﬁcantly higher crawling frequency occurred in sea cu- cumbers inside the artiﬁcial reefs in group O than individuals in group C (Kruskal–Wallis H = 6.908, p = 0.026; Figure 5A). Figure 5. Crawling frequency (A), tentacle activity frequency (B), and adhesion behavior (C) of sea cucumbers. The asterisks , * mean p < 0.05 and p < 0.01, respectively (mean ± SE, n = 6). Group C: sea cucumbers without the artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the ar- tiﬁcial reefs in group O; Group O-In: sea cucumbers inside the artiﬁcial reefs in group O Figure 5. Crawling frequency (A), tentacle activity frequency (B), and adhesion behavior (C) of sea cucumbers. The asterisks , * mean p < 0.05 and p < 0.01, respectively (mean ± SE, n = 6). Group C: sea cucumbers without the artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the artiﬁcial reefs in group O; Group O-In: sea cucumbers inside the artiﬁcial reefs in group O. Figure 5. Crawling frequency (A), tentacle activity frequency (B), and adhesion behavior (C) of sea cucumbers. The asterisks , * mean p < 0.05 and p < 0.01, respectively (mean ± SE, n = 6). ucumbers with artiﬁcial reefs. 3.2.2. Crawling Frequency Group C: sea cucumbers without the artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the ar- Figure 5. Crawling frequency (A), tentacle activity frequency (B), and adhesion behavior (C) of sea cucumbers. The asterisks , * mean p < 0.05 and p < 0.01, respectively (mean ± SE, n = 6). Group C: sea cucumbers without the artiﬁcial reefs in group C; Group O-Out: sea cucumbers outside the artiﬁcial reefs in group O; Group O-In: sea cucumbers inside the artiﬁcial reefs in group O. 3.2.1. Mortality and Morbidity Mortality was signiﬁcantly higher in group C (30.56 ± 3.51%) than that in group O (4.17 ± 2.84%; Mann–Whitney U =2.879 p = 0.004; Figure 4A). Morbidity was signiﬁcantly higher in group C (43.06 ± 2.56%) than that in group O (6.94 ± 4.52%; Mann–Whitney U =2.961, p = 0.003; Figure 4B). J. Mar. Sci. Eng. 2023, 11, 948 8 of 12 Figure 4. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks * mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea ucumbers with artiﬁcial reefs Figure 4. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea cucumbers with artiﬁcial reefs. gure 4. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea b i h iﬁi l f Figure 4. Mortality (A) and morbidity (B) of sea cucumbers between groups C and O. The asterisks mean p < 0.01 (mean ± SE, n = 6). Group C: sea cucumbers without artiﬁcial reefs; Group O: sea cucumbers with artiﬁcial reefs. 4. Discussion The survival of sea cucumbers is constrained by high temperatures. A temperature of 25 ◦C is the threshold for severe heat stress in sea cucumbers, leading to a signiﬁcantly in- creased risk of death and disease in sea cucumbers [6]. Air exposure is unavoidable in seed production, which not only causes mechanical damage [27] but also disrupts the immune system of sea cucumbers [9]. Furthermore, our previous study found that the negative effects of air exposure and high temperature on sea cucumbers were synergistic [11]. This present study found that artiﬁcial reefs signiﬁcantly reduced mortality and morbidity of sea cucumbers exposed to the air at 25 ◦C. Speciﬁcally, in such conditions, the morbidity and mortality of sea cucumbers not cultured in artiﬁcial reefs were 7-fold and 8-fold the morbidity and mortality of those cultured in artiﬁcial reefs, respectively. This suggests that artiﬁcial reefs are an effective approach to improving the survival of sea cucumbers under the combined stress of high temperature and air exposure. Fitness-related behavior is a common indicator for evaluating the condition of sea cucumbers. The typical characteristics of sea cucumbers in poor conditions are anorexia, decreased adhesion capacity, and poor movements ability [17]. Tian et al. [11] found that artiﬁcial reefs improved ﬁtness-related behaviors of sea cucumbers after they were exposed to high temperatures and the air. However, it is unknown how artiﬁcial reefs improve ﬁtness-related behaviors. This present study found that sea cucumbers cultured inside the artiﬁcial reefs had signiﬁcantly better feeding, crawling, and adhesion behaviors than those cultured without artiﬁcial reefs. Sea cucumbers exposed to the air and/or high temperatures reduce feeding or even stop feeding behavior [11,27,28]. Sea cucumbers cultured inside artiﬁcial reefs may acquire more energy by frequent feeding [28] and, consequently, may better cope with combined stressors. The adhesion of the tube feet of sea cucumbers affects the subsequent movement such as crawling behavior [29]. It is important to note that the ﬁtness-related behaviors (feeding, crawling, and adhesion behaviors) of sea cucumbers cultured outside artiﬁcial reefs are not signiﬁcantly different from those of sea cucumbers without artiﬁcial reefs. This suggests that the condition of the sea cucumbers inside the artiﬁcial reefs is better under the combined stressors. Therefore, artiﬁcial reef improves the survival and ﬁtness-related behaviors of sea cucumbers after they are exposed to high temperatures and the air, thereby increasing the survival rate. 3.2.4. Adhesion Behavior 3.2.4. Adhesion Behavior The adhesion index was signiﬁcantly higher in sea cucumbers inside the artiﬁcial reefs of group O (206.44 ± 47.12) than that in the individuals of group C (11.41 ± 5.37; Kruskal–Wallis H = 13.007, p = 0.001; Figure 5C). The adhesion index of sea cucumbers of group C (11.41 ± 5.37) was not signiﬁcantly different from that of those outside the artiﬁcial reefs in group O (105.28 ± 28.26) (Kruskal–Wallis H = 13.007, p = 0.053; Figure 5C). Consistently, no signiﬁcant difference was found in the adhesion indexes between sea cucumbers outside and inside the artiﬁcial reefs in group O (Kruskal–Wallis H = 13.007, p = 0.732; Figure 5C). tiﬁcial reefs in group O; Gro 3.2.3. Feeding Behavior 3.2.3. Feeding Behavior There was no signiﬁcant diﬀerence in feeding behavior of sea cucumbers cultured inside artiﬁcial reefs of group O, outside artiﬁcial reefs of group O, and in group C (p > There was no signiﬁcant difference in feeding behavior of sea cucumbers cultured inside artiﬁcial reefs of group O, outside artiﬁcial reefs of group O, and in group C (p > 0.05; Figure 5B). 9 of 12 J. Mar. Sci. Eng. 2023, 11, 948 4. Discussion Skin ulcer syndrome is one of the common diseases with high infectivity and high mortality in seed production of sea cucumbers [17]. High temperatures reduce the im- munity and disease resistance of sea cucumbers [30,31] and, thus, greatly increases the possibility of infection. The present study found that artiﬁcial reefs signiﬁcantly reduced the mortality and morbidity of sea cucumbers during disease outbreaks at high temperatures. Speciﬁcally, during disease outbreaks at high temperatures, the morbidity and mortality of sea cucumbers not cultured in artiﬁcial reefs were 6-fold and 7-fold the morbidity and mortality of those cultured in artiﬁcial reefs. This suggests that artiﬁcial reefs improve the survival in disease environments and probably affect the spread of skin ulcer syndrome. In disease challenge assays, sea cucumbers cultured inside artiﬁcial reefs had signiﬁcantly better adhesion and crawling behaviors, while there was no signiﬁcant difference in the ﬁtness-related behaviors between the sea cucumbers cultured outside and without arti- ﬁcial reefs. Decreased adhesion and movement ability are the symptoms of skin ulcer J. Mar. Sci. Eng. 2023, 11, 948 10 of 12 10 of 12 syndrome [17]. Hence, we speculated that sea cucumbers cultured outside and without artiﬁcial reefs were more affected by diseased sea cucumbers, although their skins were not ulcerated. Chemosensory cues are important for a variety of fundamental behavioral pro- cesses [32,33]. Lobsters use chemical cues from diseased individuals to determine whether the shelter is safe, and healthy lobsters rarely share the shelters with diseased lobsters [34]. Diseased sea cucumbers emit certain chemical cues for warning [33,35]. Sea cucumbers show escaping behavior after receiving alarm cues from diseased individuals [35,36]. Arti- ﬁcial reefs are possible places for escape due to their attraction to sea cucumbers [20,37]. Frequent physical contact with diseased individuals increases the probability of disease transmission [19]. Diseased sea cucumbers probably release chemical cues that promote the escape of individuals with strong movement to the crevices of artiﬁcial reefs. In the limited space, sea cucumbers that did not enter the reef inevitably contacted with diseased individ- uals, which increased the extent to which they were affected. In the present study, each crevice is accessible to only one sea cucumber. Sea cucumbers cultured with artiﬁcial reefs probably showed less contact with diseased and affected individuals after staying in porous artiﬁcial reefs, which reduced the extent of the affection. 4. Discussion Seed production is an intensive aquaculture mode, and isolation of diseased sea cucumbers is expensive. Therefore, the artiﬁcial reef is a cost-effective approach to increasing the possibility of the survival of sea cucumbers by reducing the frequency of contact with diseased and affected individuals in seed production. Notably, our experiment is based on a laboratory experiment. Therefore, conducting a further ﬁeld experiment is necessary. g p y In conclusion, we encourage aqua-farmers to use artiﬁcial reefs to decrease disease transmission and, thus, reduce the mortality and morbidity of small sea cucumbers in seed production at high temperatures. The present novel ﬁnding provides valuable insights into the improved management for the seed production of sea cucumbers in summer. Author Contributions: Conceptualization, H.W., C.Z. and Y.S.; methodology, H.W., G.W., J.D. and C.Z.; software, H.W. and G.W.; validation, H.W., G.W., F.H. and R.T.; formal analysis, H.W., G.W. and F.H.; investigation, H.W., G.W. and R.T.; data curation, H.W. and G.W.; writing—original draft preparation, H.W. and G.W.; writing—review and editing, C.Z. and Y.S.; visualization, H.W.; supervision, C.Z., Y.S., J.D. and Y.C.; project administration, C.Z. and Y.S.; funding acquisition, C.Z., J.D., Y.S. and Y.C. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the National Natural Science Foundation of Liaoning Province (2022-MS-352), High-level talent support grant for innovation in Dalian (2020RD03), Modern Fisheries Technology Mission in Changhai, Liaoning (2022JH5/10400015) and Liaoning Province “Xingliao Talents Plan” project (XLYC2002107). Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: The data presented in this study is available on request from the corresponding authors. Acknowledgments: We appreciate Wei Tang for her editorial suggestions. We thank Mingfang Yang, Xiaomei Chi, Yushi Yu, and Xiang Li for their assistance in this experiment. Acknowledgments: We appreciate Wei Tang for her editorial suggestions. We thank Mingfang Yang, Xiaomei Chi, Yushi Yu, and Xiang Li for their assistance in this experiment. Conﬂicts of Interest: The authors declare no conﬂict of interest. 3. Jiang, S.H.; Ren, Y.C.; Tang, B.P.; Li, C.F.; Jiang, C.B. Development status and countermeasures of Apostichopus japonicus culture industry in China. J. Agric. Sci. Technol. 2017, 019, 15–23. y J g X.Z. 2022 China Fishery Statistical Yearbook; China Agriculture Press: Beijing, China, 2022; p. 23. (In Chinese) 1. Liu, X.F. Study of Regional Differences in Nutrient Compositions and Bioactivities of Phospholipids in Sea Cucumber (Apostichopus japonicus). Ph.D. Thesis, Ocean University China, Qingdao, China, 2014. (In Chinese with an English Abstract) j p ) y g ( g ) 2. Yang, H.; Hamel, J.F.; Mercier, A. The Sea Cucumber Apostichopus japonicus: History, Biology and Aquaculture; Academic Press: Cambridge, MA, USA, 2015. 4. Liu, X.Z. 2022 China Fishery Statistical Yearbook; China Agriculture Press: Beijing, China, 2022; p. 23. (In Chinese) y g p p p p p japonicus). Ph.D. Thesis, Ocean University China, Qingdao, China, 2014. (In Chinese with an English Abstract) 2. Yang, H.; Hamel, J.F.; Mercier, A. The Sea Cucumber Apostichopus japonicus: History, Biology and Aquaculture; Academic Press: Cambridge, MA, USA, 2015. References 1. Liu, X.F. Study of Regional Differences in Nutrient Compositions and Bioactivities of Phospholipids in Sea Cucumber (Apostichopus japonicus). Ph.D. Thesis, Ocean University China, Qingdao, China, 2014. (In Chinese with an English Abstract) g 3. Jiang, S.H.; Ren, Y.C.; Tang, B.P.; Li, C.F.; Jiang, C.B. Development status and countermeasures of Apostichopus japonicus culture industry in China. J. Agric. Sci. Technol. 2017, 019, 15–23. 3. Jiang, S.H.; Ren, Y.C.; Tang, B.P.; Li, C.F.; Jiang, C.B. Development status and countermeasures of Apostichopus japonicus culture industry in China. J. Agric. Sci. Technol. 2017, 019, 15–23. J. Mar. Sci. Eng. 2023, 11, 948 11 of 12 11 of 12 5. Chang, Y.; Yu, C.; Song, X. Pond Culture of Sea Cucumber, Apostichopus japonicus in Dalian. In Advances in Sea Cucumber Aquaculture and Management; Lovatelli, A., Conand, C., Purcell, S., Uthicke, S., Hamel, J.F., Mercier, A., Eds.; FAO: Rome, Italy, 2004; pp. 269–272. ; pp 6. Li, C.; Feng, H.; Xu, D. Effect of seasonal high temperature on the immune response in Apostichopus ja analysis. Fish Shellﬁsh Immunol. 2019, 92, 765–771. [CrossRef] [PubMed] y ﬁ 7. Wang, C.; Song, S.; Li, D.; Liu, X.; Song, J. Research progress of effects and mechanism of environmental factors on physiology of sea cucumber Apostichopus japonicus. Hebei Fish. 2022, 11, 34–40. (In Chinese with an English Abstract) p p j p ( g ) 8. Hou, S.Y.; Jin, Z.W.; Jiang, W.W.; Chi, L.; Xia, B.; Chen, J.H. Physiological and immunological responses of sea cucumber h p p j p g 8. Hou, S.Y.; Jin, Z.W.; Jiang, W.W.; Chi, L.; Xia, B.; Chen, J.H. Physiological and immunological responses of sea cucumber Apostichopus japonicus during desiccation and subsequent resubmersion. PeerJ 2019, 7, e7427. [CrossRef] Hou, S.Y.; Jin, Z.W.; Jiang, W.W.; Chi, L.; Xia, B.; Chen, J.H. Physiological and immunological respon Apostichopus japonicus during desiccation and subsequent resubmersion. PeerJ 2019, 7, e7427. [CrossRef] 9. Cui, Y.; Hou, Z.; Ren, Y.; Men, X.; Zheng, B.; Liu, P.; Xia, B. Effects of aerial exposure on oxidative stress, antioxidant and non-speciﬁc immune responses of juvenile sea cucumber Apostichopus japonicus under low temperature. Fish Shellﬁsh Immunol. 2020, 101, 58–65. [CrossRef] 10. Saeij, J.; Kemenade, V.V.; Muiswinkel, W.; Wiegertjes, G.F. Daily handling stress reduces resistance of carp to Trypanoplasma borreli: In vitro modulatory effects of cortisol on leukocyte function and apoptosis. Dev. Comp. Immunol. 2003, 27, 233–245. [CrossRef] 11. References Tian, R.; Hu, F.; Wu, G.; Wang, H.; Ding, J.; Chang, Y.; Zhao, C. An effective approach to improving ﬁtness-related behavior and digestive ability of small sea cucumbers Apostichopus japonicus at high temperature: New insights into seed production. Aquaculture 2023, 562, 738755. [CrossRef] q [ ] 12. Wagner-Döbler, I.; Rheims, H.; Felske, A.; Pukall, R.; Tindall, B.J. Jannaschia helgolandensis, gen. nov., sp. nov., a novel abundant member of the marine Roseobacter clade from the North Sea. Int. J. Syst. Evol. Microbiol. 2003, 53, 731–738. [CrossRef] y 13. Wang, L.; Wei, C.; Chang, Y.; Ding, J. Response of bacterial community in sea cucumber Apostichopus japonicus intestine, surrounding water and sediment subjected to high-temperature stress. Aquaculture 2021, 535, 736353. [CrossRef] 14. Eeckhaut, I.; Wayenberghe, K.V.; Nicolas, F.; Delroisse, J. Skin ulcerations in Holothuria scabra can be induced by various types of food. SPC Beche-De-Mer. Inf. Bull. 2019, 39, 31–35. f 15. Delroisse, J.; Wayneberghe, K.V.; Flammang, P.; Gillan, D.; Gerbaux, P.; Opina, N.; Todinanahary, G.G.B.; Eeck-haut, I. Epidemiol- ogy of a Skin Ulceration Disease (SKUD) in the Sea Cucumber Holothuria scabra with a Re-view on the SKUDs in Holothuroidea (Echinodermata). Sci. Rep. 2020, 10, 22150. [CrossRef] [PubMed] ( ) p , , [ ] [ ] 16. Wang, Y.G.; Rong, X.J.; Zhang, C.Y.; Sun, S.F. Main diseases of cultured Apostichopus japonicus: Prevention and treatment. Mar. Sci. 2005, 9, 1–7. (In Chinese with an English Abstract) 17. Deng, H.; He, C.; Zhou, Z.; Liu, C.; Tan, K.; Wang, N.; Jiang, B.; Gao, X.; Liu, W. Isolation and pathogenici ulceration disease and viscera ejection syndrome of the sea cucumber Apostichopus japonicus. Aquaculture 20 17. Deng, H.; He, C.; Zhou, Z.; Liu, C.; Tan, K.; Wang, N.; Jiang, B.; Gao, X.; Liu, W. Isolation and pathogenicity of pathogens from skin ulceration disease and viscera ejection syndrome of the sea cucumber Apostichopus japonicus. Aquaculture 2009, 287, 18–27. [CrossRef] 18. Hu, F.; Zhao, C.; Ding, P.; Li, Y.; Tian, R.; Qiao, Y.; Chang, Y. An effective facility decreases disease transmission and promotes 17. Deng, H.; He, C.; Zhou, Z.; Liu, C.; Tan, K.; Wang, N.; Jiang, B.; Gao, X.; Liu, W. Isolation and pathogenicity of pathogens from skin ulceration disease and viscera ejection syndrome of the sea cucumber Apostichopus japonicus. Aquaculture 2009, 287, 18–27. [CrossRef] 18. Hu, F.; Zhao, C.; Ding, P.; Li, Y.; Tian, R.; Qiao, Y.; Chang, Y. References An effective facility decreases disease transmission and promotes resistance ability of small sea urchins Strongylocentrotus intermedius: A potential application in the longline culture. Aquaculture 2021, 547, 737542. [CrossRef] j y p p j p q 18. Hu, F.; Zhao, C.; Ding, P.; Li, Y.; Tian, R.; Qiao, Y.; Chang, Y. An effective facility decreases disease transmission and promotes resistance ability of small sea urchins Strongylocentrotus intermedius: A potential application in the longline culture. Aquaculture 2021, 547, 737542. [CrossRef] 19. Hu, F.; Zhao, Z.; Tian, R.; Ding, P.; Yin, D.; Li, X.; Qiao, Y.; Chang, Y.; Zhao, C. A cost-effective approach to decreasing the disease transmission of the sea urchin Strongylocentrotus intermedius: New information for seed production and longline culture. Aquaculture 2021, 548, 737569. [CrossRef] q 20. Zhang, J.B. Experimental Studies on the Attractive Effects of Sea Cucumber Artiﬁcial Reefs on Apostichopus japonicus. Master’s Thesis, Ocean University China, Qingdao, China, 2011. (In Chinese with an English Abstract) 21. Minami, K.; Masuda, R.; Takahashi, K.; Sawada, H.; Shirakawa, H.; Yamashita, Y. Seasonal and interannual variation in the density of visible Apostichopus japonicus (Japanese sea cucumber) in relation to sea water temperature. Estuar. Coast. Shelf Sci. 2019, 229, 106384. [CrossRef] 22. Pei, S.; Dong, S.; Wang, F.; Tian, X.; Gao, Q. Effects of density on variation in individual growth and differentiation in endocrine response of Japanese sea cucumber (Apostichopus japonicus Selenka). Aquaculture 2012, 356–357, 398–403. [CrossRef] 23. Xia, B.; Ren, Y.; Wang, Y.; Sun, Y.; Zhang, Z. Effects of feeding frequency and density on growth, energy budget and physiological performance of sea cucumber Apostichopus japonicus (Selenka). Aquaculture 2017, 466, 26–32. [CrossRef] 24. Lv, Z.; Guo, M.; Li, C.; Shao, Y.; Zhao, X.; Zhang, W. Divergent proteomics response of Apostichopus japonicus suffering from skin ulceration syndrome and pathogen infection. Comp. Biochem. 2019, 30, 196–205. [CrossRef] 25. Lin, C.G. Effects of Four Physical Environment Factors on the Movement and Feeding Behavior of Sea Cucumber Apostichopus japonicus (Selenka). Ph.D. Thesis, University of Chinese Academy of Sciences, Beijing, China, 2014. (In Chinese with an English Abstract) ) 26. Sun, J.M.; Zhang, L.B.; Pan, Y.; Lin, C.G.; Wang, F.; Yang, H.S. Feeding behavior and digestive physiology in sea cucumber Apostichopus japonicus. Physiol. Behav. 2015, 139, 336–343. [CrossRef] 27. Yang, M.F.; Li, X.; Hu, F.Y.; Ning, Y.C.; Tian, R.H.; Ding, P.; Zhao, C. 30. Yang, H.; Yuan, X.; Zhou, Y.; Mao, Y.; Zhang, T.; Liu, Y. Effects of body size and water temperature on food consumption and growth in the sea cucumber Apostichopus japonicus (selenka) with special reference to aestivation. Aquac. Res. 2005, 36, 1085–1092. [CrossRef] p p j p ( ) p q [ ] 31. Wang, F.Y.; Yang, H.S.; Gao, F.; Liu, G.B. Effects of acute temperature or salinity stress on the immune response in sea cucumber, Apostichopus japonicus. Comp. Biochem. Physiol. Part A Mol. Integr. Physiol. 2008, 151, 491–498. [CrossRef] 30. Yang, H.; Yuan, X.; Zhou, Y.; Mao, Y.; Zhang, T.; Liu, Y. Effects of body size and water temperature on food consumption and growth in the sea cucumber Apostichopus japonicus (selenka) with special reference to aestivation. Aquac. Res. 2005, 36, 1085–1092. [CrossRef] 31. Wang, F.Y.; Yang, H.S.; Gao, F.; Liu, G.B. Effects of acute temperature or salinity stress on the immune response in sea cucumber, Apostichopus japonicus Comp Biochem Physiol Part A Mol Integr Physiol 2008 151 491 498 [CrossRef] g g y p p g in the sea cucumber Apostichopus japonicus (selenka) with special reference to aestivation. Aquac. Res. 2005, 36, 1085–1092. [CrossRef] 31. Wang, F.Y.; Yang, H.S.; Gao, F.; Liu, G.B. Effects of acute temperature or salinity stress on the immune response in sea cucumber, Apostichopus japonicus. Comp. Biochem. Physiol. Part A Mol. Integr. Physiol. 2008, 151, 491–498. [CrossRef] p p j p p y g y 32. Ross, E.; Behringer, D. Changes in temperature, pH, and salinity affect the sheltering responses of Ca chemosensory cues. Sci. Rep. 2019, 9, 4375. [CrossRef] [PubMed] References Effects of handling stresses on ﬁtness related behaviors of small sea cucumbers Apostichopus japonicus: New insights into seed production. Aquaculture 2022, 546, 737321. [CrossRef] 28. Chen, M.; Sun, S.; Xu, Q.; Gao, F.; Wang, H.; Wang, A. Inﬂuence of water temperature and ﬂow velocity on locomotion behavior in tropical commercially important sea cucumber Stichopus monotuberculatus. Front. Mar. Sci. 2022, 9, 931430. [CrossRef] 29. Clement, J.C.; Schagerström, E.; Dupont, S.; Jutfelt, F.; Ramesh, K. Roll, right, repeat: Short-term repeatability in the self-righting 27. Yang, M.F.; Li, X.; Hu, F.Y.; Ning, Y.C.; Tian, R.H.; Ding, P.; Zhao, C. Effects of handling stresses on ﬁtness related behaviors of small sea cucumbers Apostichopus japonicus: New insights into seed production. Aquaculture 2022, 546, 737321. [CrossRef] 28 Ch M S S X Q G F W H W A I ﬂ f t t t d ﬂ l it l ti b h i 27. Yang, M.F.; Li, X.; Hu, F.Y.; Ning, Y.C.; Tian, R.H.; Ding, P.; Zhao, C. Effects of handling stresses on ﬁtness related behaviors of small sea cucumbers Apostichopus japonicus: New insights into seed production. Aquaculture 2022, 546, 737321. [CrossRef] 28. Chen, M.; Sun, S.; Xu, Q.; Gao, F.; Wang, H.; Wang, A. Inﬂuence of water temperature and ﬂow velocity on locomotion behavior i i l i ll i b Sti h t b l t F t M S i 2022 9 931430 [C R f] in tropical commercially important sea cucumber Stichopus monotuberculatus. Front. Mar. Sci. 2022, 9, 931430. [CrossRef] 29. Clement, J.C.; Schagerström, E.; Dupont, S.; Jutfelt, F.; Ramesh, K. Roll, right, repeat: Short-term repeatability in the self-righting behaviour of a cold-water sea cucumber. J. Mar. Bio Assoc. 2020, 100, 115–120. [CrossRef] 29. Clement, J.C.; Schagerström, E.; Dupont, S.; Jutfelt, F.; Ramesh, K. Roll, right, repeat: Short-term repeatability in the self-righting behaviour of a cold-water sea cucumber. J. Mar. Bio Assoc. 2020, 100, 115–120. [CrossRef] 29. Clement, J.C.; Schagerström, E.; Dupont, S.; Jutfelt, F.; Ramesh, K. Roll, right, repeat: Short-te behaviour of a cold-water sea cucumber. J. Mar. Bio Assoc. 2020, 100, 115–120. [CrossRef] 12 of 12 12 of 12 J. Mar. Sci. Eng. 2023, 11, 948 30. Yang, H.; Yuan, X.; Zhou, Y.; Mao, Y.; Zhang, T.; Liu, Y. Effects of body size and water temperature on food consumption and growth in the sea cucumber Apostichopus japonicus (selenka) with special reference to aestivation. Aquac. Res. 2005, 36, 1085–1092. References [CrossRef] 31. Wang, F.Y.; Yang, H.S.; Gao, F.; Liu, G.B. Effects of acute temperature or salinity stress on the immune response in sea cucumber, Apostichopus japonicus. Comp. Biochem. Physiol. Part A Mol. Integr. Physiol. 2008, 151, 491–498. [CrossRef] 30. Yang, H.; Yuan, X.; Zhou, Y.; Mao, Y.; Zhang, T.; Liu, Y. Effects of body size and water temperature on food consumption and growth in the sea cucumber Apostichopus japonicus (selenka) with special reference to aestivation. Aquac. Res. 2005, 36, 1085–1092. [CrossRef] p p j p ( ) p q [ ] 31. Wang, F.Y.; Yang, H.S.; Gao, F.; Liu, G.B. Effects of acute temperature or salinity stress on the immune response in sea cucumber, Apostichopus japonicus. Comp. Biochem. Physiol. Part A Mol. Integr. Physiol. 2008, 151, 491–498. [CrossRef] p p j p p y g y 32. Ross, E.; Behringer, D. Changes in temperature, pH, and salinity affect the sheltering responses of Caribbean spiny lobsters to chemosensory cues. Sci. Rep. 2019, 9, 4375. [CrossRef] [PubMed] y p 33. Chi, X.; Hu, F.; Qin, C.; Huang, X.; Sun, J.; Cui, Z.; Ding, J.; Yang, M.; Chang, Y.; Zhao, C. Conspecific alarm cues are a potential effective barrier to regulate foraging behavior of the sea urchin Mesocentrotus nudus. Mar. Environ. Res. 2021, 171, 105476. [CrossRef] y p 33. Chi, X.; Hu, F.; Qin, C.; Huang, X.; Sun, J.; Cui, Z.; Ding, J.; Yang, M.; Chang, Y.; Zhao, C. Conspecific alarm cues are a potential effective barrier to regulate foraging behavior of the sea urchin Mesocentrotus nudus. Mar. Environ. Res. 2021, 171, 105476. [CrossRef] 34. Behringer, D.C.; Butler, M.J.; Shields, J.D. Avoidance of disease by social lobsters. Nature 2006, 441, 421. [CrossRef] 35. Sun, J.; Yu, Y.; Zhao, Z.; Tian, R.; Li, X.; Chang, Y.; Zhao, C. Macroalgae and interspeciﬁc alarm cues regulate behavioral interactions between sea urchins and sea cucumbers. Sci. Rep. 2022, 12, 3971. [CrossRef] [PubMed] 36. Vadas, R.L.; Elner, R.W. Responses to predation cues and food in two species of sympatric, tropical sea urchins. Mar. Ecol. 2003, 24, 101–121. [CrossRef] 37. Cui, Y.; Guan, C.; Wang, R.; Tan, J.; Huang, B.; Li, J. The study of attractive effects of artiﬁcial reef models on Apostichopus japonicas. Prog. Mater. Sci. 2010, 31, 109–113. (In Chinese with an English Abstract) 37. Cui, Y.; Guan, C.; Wang, R.; Tan, J.; Huang, B.; Li, J. References The study of attractive effects of artiﬁcial reef models on Apostichopus japonicas. Prog. Mater. Sci. 2010, 31, 109–113. (In Chinese with an English Abstract) 37. Cui, Y.; Guan, C.; Wang, R.; Tan, J.; Huang, B.; Li, J. The study of attractive effects of artiﬁcial reef models on Apostichopus japonicas. Prog. Mater. Sci. 2010, 31, 109–113. (In Chinese with an English Abstract) Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
W4391400952.txt	https://www.researchsquare.com/article/rs-3895501/latest.pdf	en		Tuning surface properties of corn stover through ammonium and sodium sulfite (ASS) pretreatment for enhanced enzymatic hydrolysis	Research Square (Research Square)	2,024	cc-by	5,283	Tuning surface properties of corn stover through ammonium and sodium sulfite (ASS) pretreatment for enhanced enzymatic hydrolysis Xiangxue Chen Nanjing University of Science and Technology Shizhong Yang Nanjing University of Science and Technology Shuiping Ouyang Jiaxing Nanhu University Xinchuan Yuan Nanjing University of Science and Technology Junlin Song Nanjing University of Science and Technology Shuai Ding Nanjing University of Science and Technology Yuanyuan Sha (  shayuanyuan@njust.edu.cn ) Nanjing University of Science and Technology Rui Zhai Nanjing University of Science and Technology Research Article Keywords: Lignocellulosic biomass, Ammonium sodium sulfite, Sulfonation, Low enzyme loading, Enzymatic hydrolysis Posted Date: January 31st, 2024 DOI: https://doi.org/10.21203/rs.3.rs-3895501/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Page 1/19 Abstract The ammonia fiber expansion (AFEX) pretreatment of lignocellulosic biomass offers a significant advantage in terms of obtaining high glucan conversion, with the added benefit of ammonia being fully recyclable. However, despite the high efficiency of AFEX in pretreating lignocellulose, relatively high enzyme loading is still required for effective cellulose conversions. In this study, we have updated the AFEX pretreatment method by supplementing additional sodium sulfite (ASS), which can be used to produce a more digestible substrate. The results demonstrate that ASS-pretreated corn stover (CS) yields a higher fermentable sugar yield compared with AFEX pretreatment, even at lower enzyme loadings. Specifically, at an enzyme loading of 12 mg protein/g glucan, ASS-CS achieved 88.8% glucose and 80.6% xylose yield. Characterization analysis reveals that lignin underwent sulfonation during ASS pretreatment. This modification results in a more negative zeta potential for ASS-CS, indicating a reduction in nonproductive adsorption between lignin and cellulase through increased electrostatic repulsion. Introduction Economic growth and industrial development have led to a rise in the consumption of fossil fuels. The increased consumption of fossil fuels aggravates environmental problems throughout the world including air pollution and global warming. Lignocellulosic biomass attracts global attention as a sustainable alternative to fossil carbon resources to produce biobased chemicals (e.g. bioethanol) [1, 2]. The lignocellulose biomass mainly includes agricultural waste and forestry waste, such as corn stover, etc. Corn stover contains cellulose, hemicellulose and lignin. The three components are associated with each other via chemical or physical bonds to form a recalcitrant structure that limits the enzymatic hydrolysis of cellulose and hemicellulose to produce ethanol [1]. It has been reported that lignin is the major factor limiting enzymatic hydrolysis of cellulose as a steric hindrance while nonproductively absorbing cellulase, resulting in the loss of enzyme activity. In addition, lignin adsorbs enzymes to reduce the adsorption possibility of enzymes on cellulose and hemicellulose [3]. To minimize the inhibitory effect of lignin and improve the conversion of enzymatic hydrolysis, pretreatment can remove or modify lignin and destroy the lignocellulose spatial network structure of lignin, hemicellulose and cellulose to improve the cellulose and hemicellulose accessibility to the enzyme. Many pretreatment methods such as biological, physical and chemical treatments have been developed [4]. AFEX pretreatment is a feasible and effective mild alkaline pretreatment method and minimal inhibitory degradation products are generated during pretreatment [5]. AFEX pretreatment changed the structure and molecular characteristics of the biomass without generating soluble fractions from the biomass, resulting in a near-complete solid recovery. After AFEX pretreatment, the ammonia could be recovered and recycled. To further improve the digestibility of AFEX pretreated corn stover and reduce the amount of enzyme protein during enzymatic hydrolysis of AFEX pretreated corn stover, sodium sulfite can be used to further improve the effectiveness of AFEX pretreatment by modification of lignin properties. It has been reported Page 2/19 that the sodium sulfite pretreatment can remove and modify lignin without degradation of hemicelluloses and cellulose, resulting in more accessible substrates [6]. During sodium sulfite pretreatment, the lignin in the biomass undergoes sulfonation, resulting in sulfonated lignin that contains a significant amount of hydrophilic groups, such as sulfonic acid groups [7]. This reduces the negative impact of lignin on the enzyme. This study investigated the influence of ammonia and sodium sulfite (ASS) pretreatment on corn stover properties and explored the impact of varying enzyme loadings on glucan and xylan conversion in ASSCS. Furthermore, to confirm lignin sulfonation after ASS pretreatment, the chemical and physical characteristics of treated biomass samples were analyzed using Fourier-transform infrared spectroscopy (FT-IR), scanning electron microscopy (SEM), conductometric titration, and zeta potential measurement. Materials and methods 2.1Biomass and enzyme Air-dried CS and poplar used in this study were harvested in 2018 at Lianyungang, Jiangsu province, China. The CS and poplar were knife milled using a laboratory mill with a 4 mm particle size interior sieve. Then, the biomass was stored in plastic containers at 4 ºC for further study. Liquid ammonia was purchased from Nanjing Wenda Special Gas Co., Ltd, China. CTec2 (Novozymes) was purchased from Sigma-Aldrich (Shanghai Warehouse, China). The average protein concentration of these enzymes, as determined by BCA was: 87.31mg/mL (Ctec2). All chemical reagents were purchased from the local company [8]. 2.2. Biomass pretreatment The sodium sulfite with 10% (wt/wt, dry CS or poplar) was prepared [9]. For each test, the CS was homogeneously mixed with an aqueous solution of sodium sulfite, the moisture content was 60% (wt/wt) of dry CS weight. In the case of ASS pretreatment, the heat generated by the exothermic reaction of ammonia and sodium sulfite aqueous solution was used to reach 120°C and 30 minutes, of which 1:1 anhydrous ammonia loads CS into a bench-top stainless steel batch reactor. It took 30 minutes for the reactor to reach 120°C and the reactor maintained temperature for a certain time, then the ammonia was rapidly released, which immediately brought the biomass to room temperature. The residual sodium sulfite was removed and pretreated biomass material was recovered with 500 mL water. AFEX pretreatment in this work was performed in a reactor. AFEX pretreatment conditions were 120 ºC for 30 min at 60% (wt/wt) moisture with the ratio of anhydrous ammonia to biomass (dry weight) equal to 1:1. After the treatment, the pretreated biomass was placed in a fume hood to remove residual ammonia, then was washed with the addition of 500 ml deionized water [8, 10]. 2.3 Enzymatic hydrolysis The washed solid streams from AFEX and ASS pretreated CS or poplar were hydrolyzed at 3% w/w glucan loading in a baffled shake flask. For ASS-CS and AFEX-CS, 4, 8, 12 and 16 mg protein/ g glucan of Page 3/19 CTec 2 were added into the baffled shake flask. For ASS-poplar and AFEX-poplar, 16 mg protein/ g glucan of CTec 2 was added into the baffled shake flask. Hydrolysis was performed for 3 days at pH 4.8, 50 ºC and 250 rpm. Samples were taken at different times and the sugar concentrations were measured by HPLC [11]. 2.4 Measuring protein concentration in liquid streams after enzymatic hydrolysis The protein concentration in liquid streams after enzymatic hydrolysis was determined by the previously reported method [12]. 2.5 Analysis methods Biomass composition was analyzed according to NREL methods and our previous studies [13]. The concentrations of glucose, xylose and ethanol were determined using a Shimadzu HPLC system equipped with an Aminex HPX-87H column maintained at 60°C and a Shimadzu refractive index detector (RID). Sulfuric acid (5 mM) was used as an eluent at 0.6 mL/min and the injection volume was 15 µL [14]. The Fourier transform infrared (FTIR) spectra of samples were obtained ([15]. Twenty scans for each sample were taken with a resolution of 2 cm− 1 ranging from 400 to 4000 cm− 1. The sample was prepared using the method of agate mortar crush before scanning [16]. Zeta potential measurement was performed to determine the charge sign of pretreated CS by Zetasizer instrument (ZS90, Malvern Instruments Ltd). 0.0025 g, 0.005 g and 0.010 g of ASS and AFEX pretreated CS were added to 10 ml of deionized water. Before measuring, the samples were incubated in ultrasonic dispersing for 4 h at a pH of 4.8. The measurements were repeated three times [17]. The surface of ASS and AFEX pretreated CS and native CS was analyzed via field emission scanning electron microscopy (SEM, FEI Quanta 250 FEG) and operated at 5.0 kV. Before SEM, these samples were coated with a thin gold layer (50 nm) using Q150T Sputter Coater (Quorum Technologies, USA). Potentiometric and conductometric titrations of ASS-CS and AFEX-CS were performed according to the methodology described by Fourest et al [18]. 200 mg of native and pretreated CS converted to hydrogen form with 0.1 mol/L hydrochloric acid, then washed with water and dispersed in 100 mL of 1 mmol/L sodium chloride solution. Titration was performed through stepwise addition of 0.1 mol/L sodium hydroxide [19]. Results and discussion 3.1. Chemical compositions of ASS-CS and AFEX-CS The chemical composition of pretreated lignocellulosic biomass is important indicator for evaluating the effectiveness of pretreatment. The insoluble fraction compositions of untreated and pretreated corn Page 4/19 stover are shown in Table 1. As can be seen, AFEX pretreatment is a dry to dry process ( lignocellulosic biomass was added into the reactor dry and leaved the reactor in the dry state), during which minimal cellulose and hemicellulose of CS are degraded and negligible modifications in total carbohydrate composition are seen compared to untreated biomass [5]. The glucan, hemicellulose and lignin contents of untreated CS are 35.1%, 17.1% and 20.3%, respectively. The insoluble fraction compositions of AFEX pretreated CS are 35.7% of glucan, 17.0% of xylan and 20.9% of lignin, respectively. However, the insoluble fraction compositions of ASS pretreated CS are 38.14% of glucan and 18.52% of xylan, respectively. The composition of glucan and xylan slightly increases after ASS-CS compared with native CS. This result shows that more lignin within corn stover is removed during ASS pretreatment, compared with AFEX compared with AFEX. So the content of lignin after ASS-CS was slightly decreased from 20.32% (untreated CS) to 16.99% (Table 1), indicating that lignin structure was destroyed during ASS pretreatment. Table 1 Chemical composition of insoluble fraction compositions in native and treated corn stover. Components of pretreated solids (%) Pretreatment methodology Native CS AFEX-CS ASS-CS Glucan 35.1%±1.2% 35.7%±0.2% 38.1%±1.0% Xylan 17.1%±0.4% 17.0%±1.2% 18.5%±0.0% Lignin 20.3%±1.2% 20.9%±1.0% 17.0%±0.8% ash 7.2%±0.4% 7.8%±0.0% 9.5%±0.6% 3.2. Enzymatic hydrolysis of pretreated lignocellulosic biomass Enzymatic hydrolysis is a key step in the conversion of polysaccharides to fermentable sugars for subsequent fermentation. The raw lignocellulosic biomass exhibits strong recalcitrance to enzymes. Pretreatment can improve the accessibility of enzymes to cellulose via the destruction of lignocellulose structure and modification of lignin structure. Previous work has shown that sulfonation improved the effects of enzymatic hydrolysis of sodium hydroxide pretreated corn stover via mitigating the inhibitory effect of lignin [20]. To investigate the effect of ASS pretreatment on lignocellulosic biomass, enzymatic hydrolysis of ASS and AFEX pretreated biomass were further carried out. The monomeric sugar release for ASS-CS and AFEX-CS at different enzyme loadings are given in Fig. 1 (a-d). Compared with AFEX-CS, the enzymatic hydrolysis efficiency of ASS-CS is higher. In Fig. 1a, the addition of sodium sulfite during AFEX pretreatment (ASS pretreatment) leads to the improvement of the initial hydrolysis rate. At 12 h of enzymatic hydrolysis, the initial rate of glucan to glucose in AFEX-CS is 0.61 g/(L×h) at 4 mg protein/g glucan enzyme loading. Whereas the initial rate of glucan to glucose after ASS-CS is 0.82 g/(L×h) at the Page 5/19 same enzyme loading, which is 35.5% higher than that of the samples of AFEX-CS. Meanwhile, the hydrolysis of ASS-CS generated higher sugar concentration after enzymatic hydrolysis compared to that of AFEX-CS. After enzymatic hydrolysis, higher glucose (23.3 g/L) and xylose (11.6 g/L) concentrations are obtained from ASS-CS than those of AFEX-CS (16.1 g/L glucose and 9.3 g/L xylose) at 4 mg protein/g glucan enzyme loading. As enzyme loading increases, the sugar concentration after enzymatic hydrolysis of ASS-CS and AFEX-CS increases. For example, when the CTec2 enzyme loading increases from 4 to 16 mg protein/g glucan, the glucose concentration increases from 16.1 to 33.4 g/L after enzymatic hydrolysis of AFEX-CS. In addition, High enzyme costs, consuming up to 50% of the operating budget in a biorefinery, pose a major challenge to the industrialization of cellulosic ethanol. Reducing enzyme consumption is thus essential for cost-competitive bioethanol production. In Fig. 1(e), the glucan and xylan conversion of ASSCS at 8 mg protein/g glucan enzyme loading are similar to that of AFEX-CS at 16 mg protein/g glucan enzyme loading after enzymatic hydrolysis, showing that the enzyme loading for ASS-CS can be reduced up to 50% compared with AFEX-CS at similar glucan and xylan conversion. This result suggests that ASS pretreatment has the potential to further decrease enzyme loading for effective cellulose hydrolysis. The low enzyme usage for enzymatic hydrolysis of ASS pretreated biomass is likely linked to the sulfonation of lignin. It was previously reported that lignin was sulfonated with sodium sulfite during alkaline pretreatment [21–23]. Sulfonation occurs through the substitution of a sulfonate group with lignin’s aliphatic hydroxy groups via the addition reaction [24]. In addition, sulfonation of lignin can minimize the nonproductive binding of cellulose to lignin, which is an important limiting factor for enzymatic hydrolysis of cellulose [25, 26]. Nonproductive binding between cellulose and lignin significantly hinders enzymatic hydrolysis in bioethanol production. The adsorption of cellulose to lignin will reduce lignin binding to cellulose, affecting glucan conversion. To reduce nonproductive cellulase binding, one method is to modify functional groups on the surface of lignin. In our work, the lignin of lignocellulosic biomass may be sulfonated during ASS pretreatment. The sulfonated lignin can produce a hydrophilic group on the surface of lignin, which has a low connection to cellulose (Fig. 2). In addition, sulfonic acid groups introduced onto lignin molecules carry negative charges and may repel the negatively charged cellulases. This "like charges repel" phenomenon greatly reduces the cellulase-lignin complex formation. Therefore, the substrate of ASS pretreatment is more easily hydrolyzed compared with AFEX pretreatment. 3.3. Adsorption of cellulase on enzymatic residual lignins Given the established role of nonproductive enzyme binding in lignin inhibition, we further investigated the adsorption properties of enzymatic residual lignin with enzymes to gain deeper insights into this phenomenon. In Fig. 3, the free cellulase concentration is determined during enzymatic hydrolysis of ASSCS and AFEX-CS at different enzyme loadings. The results show that the free cellulase concentration in hydrolyzate of ASS-CS and AFEX-CS increases with the increase of enzyme loading. Increasing the enzyme loading from 4 to 16 mg protein /g glucan increases the free cellulase concentration from 89.5 to Page 6/19 216 mg protein/L in solution in the case of ASS-CS after enzymatic hydrolysis. The concentration of the enzyme (CTec 2) in the supernatant from the enzymatic hydrolysis of ASS-CS is 15.8% higher than that of AFEX-CS at an enzyme loading of 4 mg protein/g glucan. At an enzyme loading of 16 mg protein/g glucan, the enzyme concentration in the supernatant from the enzymatic hydrolysis of ASS-CS was 41.2% higher than that from the hydrolysis of AFEX-CS. These observations suggest that ASS pretreatment enables reduced enzyme adsorption, which is likely related to the reduced adsorption on lignin due to sulfonation. This modification could introduce negatively charged sulfonic acid groups, which could electrostatically repel the negatively charged cellulase. This reduces non-productive binding and allows more enzymes to remain active in the hydrolysate. 3.4. SEM analysis To assess the impact of pretreatment on the accessibility of cellulose for enzymatic hydrolysis, we examined the physicochemical properties of native corn stover (CS), AFEX-pretreated CS (AFEX-CS), and ASS-pretreated CS (ASS-CS) using scanning electron microscopy (SEM) (Fig. 4). Native CS exhibits a smooth surface (Fig. 4a), potentially hindering enzyme adsorption due to limited available binding sites. AFEX pretreatment generates more rough and fibrillated surface (Fig. 4b). In contrast, ASS pretreatment significantly disrupts the lignocellulosic structure, creating a coarser and more porous surface (Fig. 4c). This enhanced porosity is likely due to ammonia's selective interaction with lignin, disrupting its bonds and weakening the fiber structure. Additionally, sodium sulfite's penetration potentially facilitates partial lignin sulfonation. This dramatic structural modification by ASS pretreatment significantly increases the accessibility of polysaccharides to enzymes, beneficial for subsequent enzymatic hydrolysis and potentially higher biofuel yields. 3.5. FT-IR analysis FT-IR spectroscopy is used to study the structural changes of substrates after pretreated CS. Figure 5 shows the FT-IR spectra of native and pretreated CS, and peaks were assigned according to the literature [15]. The peak assignments for both native and pretreated CS samples exhibit similar patterns. The absorption peak at 3319 cm− 1 is related to the stretching of O-H bonds, which are obvious characteristics of cellulose and lignin, respectively [27]. The order of these peak intensities was ASS-CS > AFEX-CS and native CS, suggesting the increase of hydrogen bonds in ASS-CS. The peaks of ASS and AFEX pretreatments disappear significantly in the 1740 cm− 1 band which is attributed to ester carbonyl vibration in feruloyl, p-coumaroyl and acetyl groups in lignin and hemicellulose, indicating that the ester bonds between hemicellulose and lignin were ruptured[23]. Meanwhile, compared with native CS, the peak of ASS and AFEX pretreatment disappearance at 1244 cm− 1 in the FT-IR spectra, which the acetyl group in hemicellulose is removed from CS [28, 29]. Lignin is linked with carbohydrates via acetyl, ferulic acid and p-coumaric acid, etc. The removal of the acetyl group of AFEX and ASS pretreatment samples indicates that the bonds of the LCCs might be cleaved. The characteristic peaks of lignin are found at 1513 cm− 1and 1260 cm− 1, which are assigned to the C-O of the guaiacyl ring and aromatic ring stretch vibrations, respectively [30, 31]. The reduction or shift in the two positions is attributed to either aromatic Page 7/19 ring vibration or C-H deformation vibration of lignin. The peak intensity of ASS-CS is minimal compared with AFEX-CS, which is attributed to the removal or modification of lignin. 3.6. Acid group titration of the AFEX and ASS pretreated substrates Figure 6 shows the conductometric titration curves of AFEX-CS and ASS-CS resulting from the addition of NaOH solution. The conductometric titration curve of AFEX and ASS pretreated substrates begins with a short decreasing portion, a constant portion and a long increasing portion, characterizing the repression of the dissociation of the weak acidic groups. For ASS pretreated substrate, the slightly lower and longer increasing portion is due to the presence of additional free electronics from sulfonates (strongly acidic groups). To show the distinctive between the strong acids and weak acids, a known amount of hydrochloric is added to the suspension before titration. The constant area between the linearly decreasing portion and the linear slowly increasing branch, corresponding to the titration of weak acidic groups, yields a quantitative estimation of equivalents of strong acidic groups in the biomass (0.018 g NaOH/g). After neutralization, the weaker acidic groups begin to dissociate and contribute to the measured conductance. For ASS pretreatment, the change attributed to the proton concentration change is lower and longer in this portion due to the low level of dissociation of the weak acidic groups. When the weak acids are neutralized, the conductivity increases in proportion to the excess of sodium hydroxide added. The conductometric titration curve of ASS-CS is slightly lower and longer, giving the amount of weak acidic equivalent groups in the sulfonates (0.038 NaOH g/g) [32]. 3.7. Zeta potential analysis of different pretreated substrates The zeta potential of ASS-CS and AFEX-CS is measured (Table 2). The result shows that the zeta potential of ASS pretreatment is lower than that of AFEX substrates. The zeta potentials of ASS-CS are − 25.1, -27.2 and − 28.0 mV at 0.025% (w/v), 0.050% (w/v) and 0.100% (w/v) solid loadings, respectively. While zeta potentials of AFEX substrates are − 12.4, -13.9 and − 14.7 mV at the same solid loadings, respectively. In addition, the zeta potential value of AFEX-CS and ASS-CS gradually rises with increasing solid loading. The lower zeta potential (<-20 mV) of ASS-CS indicates stronger electrostatic repulsion between lignin and enzymes. Huang et al reported that lignosulfonate mitigated non-productive binding, due to the sulfoacid group on lignin reducing the negative effects of lignin during enzymatic hydrolysis [33, 34]. Conversely, the relatively higher zeta potential (<-20 mV) of AFEX-CS indicates inadequate electrostatic repulsion to increase free enzymes during the hydrolysis, leaving a large number of enzymes adsorbed on lignin. So electrostatic repulsion is characterized as a key factor in lignin inhibition of enzymatic hydrolysis [35, 36]. Page 8/19 Table 2 Zeta potential of corn stover after ASS and AFEX pretreatment in different substrates loading at pH 4.8 Pretreatment Solid loading(w/v) Zeta potential (mV) ASS 0.025% -25.1 ± 0.3 0.050% -27.2 ± 0.1 0.100% -28.0 ± 0.5 0.025% -12.4 ± 0.7 0.050% -13.9 ± 0.1 0.100% -14.7 ± 0.1 AFEX Conclusions Compared with AFEX pretreatment, ASS pretreatment could effectively improve the digestibility of substrates via sulfonating lignin of lignocellulosic biomass. ASS pretreatment introduces negative charges onto lignin, resulting in electrostatic repulsion against cellulase and reduced non-productive binding, thereby improving enzymatic digestibility. The result shows that glucan conversion from ASS-CS at 4/8 mg protein/g glucan is 43.9%/41.4% higher than those of AFEX-CS. In addition, ASS-CS achieved the same sugar yield as AFEX-CS while requiring half the enzyme loading, demonstrating its superior efficiency. Therefore, ASS pretreatment is proved to be an efficient pretreatment method for lignocellulosic biomass. Abbreviations ASS ammonium sodium sulfite AFEX ammonia fiber expansion FT-IR Fourier transform infrared HPLC high performance liquid chromatography SEM scanning electron microscopy LCCs lignin-carbohydrate complexes. Declarations Page 9/19 Author Contributions Xiangxue Chen performed the experiments, and analyzed the data; Shizhong Yang, Shuiping Ouyang, Xinchuan Yuan, Junling Song and Shuai Ding coordinated and supervised this study; Xiangxue Chen, Yuanyuan Sha and Rui Zhai drafted and revised the manuscript. All authors read and approved the final version. Funding This work was supported by National Natural Science Foundation (Grant No. 22308164) and National Key R&D Program of China (Grant No. 2021YFC2101301). Ethical Approval and Consent to participate Not applicable. Consent for publication Not applicable. Availability of data and materials All data generated or analyzed during this study are included. Competing interests The authors declare that they have no competing interests. Author information School of Environmental and Biological Engineering, Nanjing University of Science and Technology, 200 Xiaolingwei Street, Nanjing 210094, China. *E-mail: ruizhai@njust.edu.cn and shayuanyuan@njust.edu.cn. Mailing address: 200 Xiaolingwei Street, Xuanwu District, Nanjing, Jiangsu Province 210094, China. References 1. Cai, D., Li, P., Luo, Z., Qin, P., Chen, C., Wang, Y., Wang, Z., & Tan, T. (2016). Effect of dilute alkaline pretreatment on the conversion of different parts of corn stalk to fermentable sugars and its application in acetone-butanol-ethanol fermentation. Bioresource Technology, 211, 117–124. 2. Aruwajoye, G. S., Kassim, A., Saha, A. K., & Gueguim Kana, E. B. (2020). Prospects for the improvement of bioethanol and biohydrogen production from mixed starch-based agricultural wastes. Energies, 13(24), 6609. Page 10/19 3. Deralia, P. K., Jensen, A., Felby, C., & Thygesen, L. G. J. B. P. (2021). Chemistry of lignin and hemicellulose structures interacts with hydrothermal pretreatment severity and affects cellulose conversion. Biotechnology Progress, 37(5), e3189. 4. Balan, V., Jin, M., & Ubanwa, B. (2020). in Recent advances in bioconversion of lignocellulose to biofuels and value-added chemicals within the biorefinery concept, pp. 101–132. 5. da Costa Sousa, L., Jin, M., Chundawat, S. P. S., Bokade, V., Tang, X., Azarpira, A., Lu, F., Avci, U., Humpula, J., Uppugundla, N., Gunawan, C., Pattathil, S., Cheh, A. M., Kothari, N., Kumar, R., Ralph, J., Hahn, M. G., Wyman, C. E., Singh, S., Simmons, B. A., Dale, B. E., & Balan, V. (2016). Next-generation ammonia pretreatment enhances cellulosic biofuel production. Energy and Environmental Science, 9(4), 1215–1223. 6. Zhang, Y., Xin, D., Wen, P., Chen, X., Jia, L., Lu, Z., & Zhang, J. (2023). Comparison of alkaline sulfite pretreatment and acid sulfite pretreatment with low chemical loading in saccharification of poplar. Applied Biochemistry and Biotechnology, 195(7), 4414–4428. 7. Wu, J., Chandra, R., Takada, M., Del Rio, P., Kim, K. H., Kim, C. S., Liu, L. Y., Renneckar, S., & Saddler, J. (2020). Alkaline sulfonation and thermomechanical pulping pretreatment of softwood chips and pellets to enhance enzymatic hydrolysis. Bioresource Technology, 315, 123789. 8. Chen, X., Zhai, R., Li, Y., Yuan, X., Liu, Z. H., & Jin, M. (2020). Understanding the structural characteristics of water-soluble phenolic compounds from four pretreatments of corn stover and their inhibitory effects on enzymatic hydrolysis and fermentation. Biotechnology for Biofuels, 13(1), 44. 9. Xiao, L. P., Shi, Z. J., Xu, F., & Sun, R. C. (2013). Hydrothermal treatment and enzymatic hydrolysis of Tamarix ramosissima: evaluation of the process as a conversion method in a biorefinery concept. Bioresource Technology, 135, 73–81. 10. Nirmal, U., Leonardo, S., Shishir, C., Yu, X., Blake, S., Seema, S., Gao, X., Rajeev, K., Charles, W., Bruce, D., & Venkatesh, B. (2014). A comparative study of ethanol production using dilute acid, ionic liquid and AFEX™ pretreated corn stover. Biotechnology for biofuels, 7(72), 1–14. 11. Chen, X. X., Liu, S. M., Zhai, R., Yuan, X. C., Yu, Y., Shen, G. N., Wang, Z., Yu, J. M., & Jin, M. J. (2022). Lime pretreatment of pelleted corn stover boosts ethanol titers and yields without water washing or detoxifying pretreated biomass. Renewable Energy, 192, 396–404. 12. Yuan, Y., Zhai, R., Li, Y., Chen, X., & Jin, M. (2018). Developing fast enzyme recycling strategy through elucidating enzyme adsorption kinetics on alkali and acid pretreated corn stover. Biotechnology for Biofuels, 11(1), 1–12. 13. Chen, X., Zhai, R., Shi, K., Yuan, Y., Dale, B. E., Gao, Z., & Jin, M. (2018). Mixing alkali pretreated and acid pretreated biomass for cellulosic ethanol production featuring reduced chemical use and decreased inhibitory effect. Industrial Crops and Products, 124, 719–725. 14. Yuan, X., Shen, G., Chen, S., Chen, X., Zhang, C., Liu, S., & Jin, M. (2022). Modified simultaneous saccharification and co-fermentation of DLC pretreated corn stover for high-titer cellulosic ethanol production without water washing or detoxifying pretreated biomass. Energy, 247, 123488. Page 11/19 15. Shen, X. J., Wang, B., Pan-li, H., Wen, J. L., & Sun, R. C. (2016). Understanding the structural changes and depolymerization of Eucalyptus lignin under mild conditions in aqueous AlCl3. RSC Advances, 6(51), 45315–45325. 16. Chen, X., Zhai, R., Li, Y., Yuan, X., Liu, Z. H., & Jin, M. (2020). Understanding the structural characteristics of water-soluble phenolic compounds from four pretreatments of corn stover and their inhibitory effects on enzymatic hydrolysis and fermentation. Biotechnology For Biofuels, 13(1), 44. 17. Ferreira, P. F. O., Pereira, A. L. S., Rosa, M. F., & De Santiago-Aguiar, R. S. (2022). Lignin-rich cellulose nanocrystals from coir fiber treated with ionic liquids: Preparation and evaluation as pickering emulsifier. Industrial Crops and Products, 186, 115119. 18. Fourest, E., & Volesky, B. J. E. S. (1996). Contribution of sulfonate groups and alginate to heavy metal biosorption by the dry biomass of Sargassum fluitans. Environmental Science and Technology, 30(1), 277–282. 19. Martinez-Ramirez, A. P., Rincon-Ortiz, S. A., Baldovino-Medrano, V. G., Blanco-Tirado, C., & Combariza, M. Y. (2023). Influence of reaction variables on the surface chemistry of cellulose nanofibers derived from palm oil empty fruit bunches. RSC advances, 13(51), 36117–36129. 20. Tavares, J., Łukasik, R. M., de Paiva, T., & da Silva, F. (2018). Hydrothermal alkaline sulfite pretreatment in the delivery of fermentable sugars from sugarcane bagasse. New Journal of Chemistry, 42(6), 4474–4484. 21. Zhou, H., Gleisner, R., Zhu, J. Y., Tian, Y., & Qiao, Y. (2018). SPORL pretreatment spent liquors enhance the enzymatic hydrolysis of cellulose and ethanol production from glucose. Energy and Fuels, 32(7), 7636–7642. 22. Wang, Z., Zhu, J., Fu, Y., Qin, M., Shao, Z., Jiang, J., & Yang, F. (2013). Lignosulfonate-mediated cellulase adsorption: enhanced enzymatic saccharification of lignocellulose through weakening nonproductive binding to lignin. Biotechnology for Biofuels, 6, 1–10. 23. Yang, M., Wang, J., Hou, X., Wu, J., Fan, X., Jiang, F., Tao, P., Wang, F., Peng, P., Yang, F., & Zhang, J. (2017). Exploring surface characterization and electrostatic property of Hybrid Pennisetum during alkaline sulfite pretreatment for enhanced enzymatic hydrolysability. Bioresource Technology, 244, 1166–1172. 24. Eraghi Kazzaz, A., Hosseinpour Feizi, Z., & Fatehi, P. (2019). Grafting strategies for hydroxy groups of lignin for producing materials. Green Chemistry, 21, 5714–5752. 25. Mansfield, S. D., Mooney, C., & Saddler, J. N. (1999). Substrate and enzyme characteristics that limit cellulose hydrolysis. Biotechnology Progress, 15(15), 804–816. 26. Wang, J., Xu, Y., Meng, X., Pu, Y., Ragauskas, A., & Zhang, J. (2021). Production of xylooligosaccharides from poplar by acetic acid pretreatment and its impact on inhibitory effect of poplar lignin. Bioresource Technology, 323, 124593. 27. Yang, X., Ma, F., Zeng, Y., Yu, H., Xu, C., & Zhang, X. (2010). Structure alteration of lignin in corn stover degraded by white-rot fungus Irpex lacteus CD2. International Biodeterioration and Biodegradation, Page 12/19 64(2), 119–123. 28. Kim, S., Kim, T., & Oh, K. (2018). Deacetylation followed by fractionation of yellow poplar sawdust for the production of toxicity-reduced hemicellulosic sugar for ethanol fermentation. Energies, 11, 404. 29. Deng, L. H., Tang, Y., & Liu, Y. (2014). Detoxification of corncob acid hydrolysate with SAA pretreatment and xylitol production by immobilized Candida Tropicalis. ScientificWorldJournal, 2014, 214632. 30. Abdul, P. M., Jahim, M. J., Harun, S., Markom, M., Lutpi, N. A., Hassan, O., Balan, V., Dale, B. E., & Nor, M., M. T (2016). Effects of changes in chemical and structural characteristic of ammonia fibre expansion (AFEX) pretreated oil palm empty fruit bunch fibre on enzymatic saccharification and fermentability for biohydrogen. Bioresource Technology, 211, 200–208. 31. Yang, M., Wang, J., Hou, X., Wu, J., Fan, X., Jiang, F., Tao, P., Wang, F., Peng, P., Yang, F., & Zhang, J. (2017). Exploring surface characterization and electrostatic property of Hybrid Pennisetum during alkaline sulfite pretreatment for enhanced enzymatic hydrolysability. Bioresource Technology, 244, 1166–1172. 32. Fourest, E., & Volesky, B. J. E. S. (1996). Contribution of sulfonate groups and alginate to heavy metal biosorption by the dry biomass of Sargassum fluitans. Environmental Science and Technology, 30(1), 277–282. 33. Zhang, C., Gleisner, R., Houtman, C. J., Pan, C., & Zhu, C. (2016). j. Sulfite pretreatment to overcome the recalcitrance of lignocelluloses for bioconversion of woody biomass. Biomass Fractionation Technologies for a Lignocellulosic Feedstock Based Biorefiner, pp.499–541. 34. Huang, Y., Sun, S., Huang, C., Yong, Q., Elder, T., & Tu, M. (2017). Stimulation and inhibition of enzymatic hydrolysis by organosolv lignins as determined by zeta potential and hydrophobicity. Biotechnology for biofuels, 10(1), 1–11. 35. Ingle, A. P., Rathod, J., Pandit, R., da Silva, S. S., & Rai, M. (2017). Comparative evaluation of free and immobilized cellulase for enzymatic hydrolysis of lignocellulosic biomass for sustainable bioethanol production. Cellulose, 24(12), 5529–5540. 36. Xue, W., Lei, F., Li, P., & Jiang, J. (2018). Cellulose accessibility and zeta potentials of sugarcane bagasse pretreated by green liquor and ethanol for high hydrolysis efficiency. BioResources 13(1), 1510–1524. Figures Page 13/19 Figure 1 Time course profile of glucose and xylose concentrations during enzymatic hydrolysis of ASS and AFEX pretreated corn stover (a-d) and the conversion of glucan and xylan after enzymatic hydrolysis at different enzyme loadings. (a):4 mg protein/g glucan of CTec 2 enzyme; (b):8 mg protein/g glucan of CTec 2 enzyme; (c):12 mg protein/g glucan of CTec 2 enzyme; (d):16 mg protein/g glucan of CTec 2 enzyme; (e): Conversion. Page 14/19 Figure 2 Schematic representation of adsorption between lignin of ASS and AFEX pretreated corn stover and poplar. Page 15/19 Figure 3 The enzyme protein concentration in liquid streams after enzymatic hydrolysis of ASS and AFEX pretreated corn stover at different enzyme loadings. Page 16/19 Figure 4 SEM images at various magnifications for native and AFEX and ASS pretreated corn stover. (a) native corn stover; (b) AFEX pretreated corn stover; (c) AFEX pretreated corn stover. AFEX: Ammonia fiber expansion, ASS: Ammonia sodium sulfite. Page 17/19 Figure 5 Chemical changes in native and pretreated corn stover were determined by FT-IR. Page 18/19 Figure 6 Potentiometric and conductometric titrations of AFEX and ASS pretreated corn stover. Page 19/19
https://openalex.org/W4389939141	http://journal.asu.ru/wv/article/download/14428/12208	Russian	null	The concept of good in abrahamic religions in the context of social service	Narody i religii Evrazii	2,023	cc-by	7,014	НАРОДЫ И РЕЛИГИИ ЕВРАЗИИ Барнаул Издательство Алтайского государственного университета 2023 ISSN 2542-2332 (Print) ISSN 2686-8040 (Online) 2023 Том 28, № 4 2023 Том 28, № 4 ISSN 2542-2332 (Print) ISSN 2686-8040 (Online) ISSN 2542-2332 (Print) ISSN 2686-8040 (Online) ISSN 2542-2332 (Print) ISSN 2686-8040 (Online) 2023 Vol. 28, № 4 NATIONS AND RELIGIONS OF EURASIA Barnaul Publishing house of Altai State University 2023 ISSN 2542-2332 (Print) ISSN 2686-8040 (Online) 2023 Vol. 28, № 4 НАРОДЫ И РЕЛИГИИ ЕВРАЗИИ Барнаул Издательство Алтайского государственного университета 2023 СОДЕРЖАНИЕ Раздел I АРХЕОЛОГИЯ И ЭТНОКУЛЬТУРНАЯ ИСТОРИЯ Казанцева О. А. Сандиякское городище — памятник раннего Средневековья бассейна Камы .................................................................................................................................7 Захаров С. В. Каменный жезл с человеческой личиной из погребения в курганной группе Майское V .................................................................................................27 Серегин Н. Н., Матренин С. С. Социальная планиграфия некрополя предтюркского времени Чобурак-I (Северный Алтай)........................................................44 Раздел II ЭТНОЛОГИЯ И НАЦИОНАЛЬНАЯ ПОЛИТИКА Бичеев Б. А., Тюмидова М. Е. Ойратский текст «Руководство к практике Прибежища» (психотехника визуализации объектов поклонения).................................59 Кубаев С. Ш. Появление и развитие отопительных систем на территории Узбекистана .....................................................................................................................................73 Ожередов Ю. И., Ярзуткина А. А. «Держатель силы». Мухомор в чукотской культуре ...........................................................................................................................................84 Аюпов Т. М. Генеалогические предания — ценный источник изучения кыргызско-башкирских историко-культурных связей .................................................... 110 Раздел III РЕЛИГИОВЕДЕНИЕ И ГОСУДАРСТВЕННО-КОНФЕССИОНАЛЬНАЯ ПОЛИТИКА Головнев Г. А. Амбивалентное отношение к «внешнему» миру нового религиозного движения «Свидетели Иеговы» ....................................................................122 Набиев Р. А., Ибрагимов М. И. Понятие блага в авраамических религиях в контексте социального служения ........................................................................................134 Слезин А. А. Внедрение новых обрядов в русской деревне 1920‑х гг. как фактор эволюции взаимоотношений крестьянских поколений ...................................................147 ДЛЯ АВТОРОВ...........................................................................................................................163 ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) Цитирование статьи: Набиев Р. А., Ибрагимов М. И. Понятие блага в авраамических религиях в контексте социального служения // Народы и религии Евразии. 2023. Т. 28, № 4. С. 134–146. DOI 10.14258/nreur(2023)4–09. CONTENT NATIONS AND RELIGIONS OF EURASIA 2023 Vol. 28, № 4 Section I ARCHAEOLOGY AND ETNO-CULTURAL HISTORY Kazantseva O. A. Sandiyak hillfort is a monument of the early middle ages of the Kama basin...............................................................................................................................7 Zakharov S. V. Stone rod with a human mask from a burial in the Mayskoye V burial mound group....................................................................................................................................27 Seregin N. N., Matrenin S. S. Social planigraphy of the preturkic period necropolis of Choburak-I (Northern Altai).....................................................................................................44 Section II ETHNOLOGY AND NATIONAL POLICY Bicheev B. A., Tyumidova M. E. The oirat text, “A guide to refuge practice” (psychotechnic technique of visualization of objects of worship).............................................59 Kubaev S. Sh. Emergence and development of heating systems on the territory of Uzbekistan....................................................................................................................................73 Ozheredov Y. I., Yarzutkina A. A. ‘Holder of strength’. Muskrat in Chukotka culture.............84 Ayupov T. M. Genealogical legends are a valuable source for studying the historical and cultural ties of Kyrgyz-Bashkir ........................................................................................... 110 Section III RELIGIOUS STUDIES AND STATE-CONFESSIONAL RELATIONS Golovnev G. A. Ambivalent attitude to ‘the world’ in the organization Jehovah’s Witnesses.........................................................................................................................................122 Nabiev R. A., Ibragimov M. I. The concept of good in abrahamic religions in the context of social service......................................................................................................134 Slezin A. A. Introducing new rituals in the Russian village of the 1920s as a factor in the evolution of the relationship between peasant generations...........................................147 FOR AUTHORS............................................................................................................................163 ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) 134 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. УДК 291.5 DOI 10.14258/nreur(2023)4–09 УДК 291.5 DOI 10.14258/nreur(2023)4–09 УДК 291.5 DOI 10.14258/nreur(2023)4–09 DOI 10.14258/nreur(2023)4–09 Р. А. Набиев, М. И. Ибрагимов Казанский (Приволжский) федеральный университет, Казань (Россия) ПОНЯТИЕ БЛАГА В АВРААМИЧЕСКИХ РЕЛИГИЯХ В КОНТЕКСТЕ СОЦИАЛЬНОГО СЛУЖЕНИЯ В статье рассматривается одно из фундаментальных положений авраамических ре- лигиозных систем, относящихся к морально-этическим вопросам во взаимоотноше- ниях между верующими и реализуемых ими в соответствии с установками сакраль- ных текстов религиях откровения. Представлены особенности интерпретации понятия «благо» в исламе, христианстве и иудаизме. В рамках их социальных доктрин показывается его содержание и социаль- но-общественная значимость через призму социального служения. Рассматриваются священные тексты ислама, христианства и иудаизма. Анализируются различные кон- тексты упоминания блага. Приводятся комментарии богословов, предметом интереса которых является понятие блага в качестве фундаментальной основы концепции со- циального служения в авраамических религиях. Продемонстрированы теологические основы для благотворительной деятельности социальных институций ислама, христианства и иудаизма, раскрыты возможности их межконфессионального сотрудничества на социальной площадке. Изучается про- цесс трансформации понятия блага в контексте социального служения в связи с ци- фровизацией и переходом общества на информационную стадию развития. Выделя- ются предпосылки возникновения различных смыслов и дополнительных контекстов, которые присущи благу как термину в системе философии благотворительности. Ключевые слова. социальное служение, благо, благотворительность, ислам, хри- стианство, иудаизм, религиозная мораль. ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) 135 Nations and religions of Eurasia • 2023 Vol. 28, № 4. P. 134–146. R. A. Nabiev, M. I. Ibragimov Kazan (Volga Region) Federal University, Kazan (Russia) Kazan (Volga Region) Federal University, Kazan (Russia) THE CONCEPT OF GOOD IN THE ABRAHAMIC RELIGIONS IN THE CONTEXT OF SOCIAL SERVICE The article discusses one of the fundamental provisions of the Abrahamic religious systems related to moral and ethical issues in the relationship between believers and implemented by them in accordance with the settings of the sacred texts of the religions of revelation. The features of the interpretation of the concept of “good” in Islam, Christianity, and Judaism are presented. Within the framework of their social doctrines, its content and social significance are shown through the prism of social service.h The sacred texts of Islam, Christianity, and Judaism are considered. Various contexts of mentioning the good are analyzed. The comments of theologians, whose subject of interest is the concept of good as a fundamental basis for the concept of social service in the Abrahamic religions, are given.h The theological foundations for the charitable activities of the social institutions of Islam, Christianity, and Judaism are demonstrated, the possibilities of their interfaith cooperation on the social platform are revealed. The process of transformation of the concept of good in the context of social service in connection with digitalization and the transition of society to the information stage of development is being studied. The prerequisites for the emergence of various meanings and additional contexts that are inherent in good as a term in the philosophy of charity are highlighted. eywords: social service, welfare, charity, Islam, Christianity, Judaism, religious morali Cайт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv For citation: Nabiev R. A., Ibragimov M. I. The concept of good in the abrahamic religions in the context of social service. Nations and religions of Eurasia. 2023. Vol. 28, No 4. P. 134–146 DOI 10.14258/nreur(2023)4–09. For citation: Nabiev R. A., Ibragimov M. I. The concept of good in the abrahamic religions in the context of social service. Nations and religions of Eurasia. 2023. Vol. 28, No 4. P. 134–146 DOI 10.14258/nreur(2023)4–09. Набиев Ринат Ахматгалиевич, доктор исторических наук, профессор, профессор кафедры религиоведения Института социально-философских наук и массовых ком- муникаций Казанского (Приволжского) федерального университета, Казань (Россия). Адрес для контактов: professor.nabiev@yandex.ru. Ибрагимов Марсель Ирекович, аспирант кафедры религиоведения Института социально-философских наук и массовых коммуникаций Казанского (Приволж- ского) федерального университета, преподаватель общенационального благотво- рительного фонда «Ярдам — Помощь», Казань (Россия). Адрес для контактов: marselibragim@yandex.ru. g y Nabiev Rinat Akhmatgalievich, Doctor of Historical Sciences, Professor, Professor of the Department of Religious Studies, Institute of Social Philosophical Science and Mass айт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv 136 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. Communications, Kazan (Volga Region) Federal University, Kazan (Russia). Communications, Kazan (Volga Region) Federal University, Kazan (Russia). Contact address: professor.nabiev@yandex.ru. Ibragimov Marcel Irekovich, post-graduate student of the Department of Religious Studies of the Institute of Social and Philosophical Science and Mass Communications of the Kazan (Volga Region) Federal University, teacher at the National charitable foundation “Yardam — Help”, Kazan (Russia). Contact address: marselibragim@yandex.ru. Введение Понятие блага является одним из центральных в концепции социального служе- ния. Зачастую под благом в авраамических религиях понимается довольно широкий спектр морально-этических категорий, таких как честь, вера, нрав. Экономическое бла- го как понятие также зачастую применяется в общем ряду значений данной категории. В подобных условиях необходимым является выделение блага как важной категории при осмыслении феномена религиозного социального служения. Для выявления наи- более точного и полного определения термина «благо» относительно термина «благо- творительность», которое включает в себя важность оказания взаимопомощи, необ- ходимо рассмотреть благо в контексте социального служения. Авраамические религии являются самыми распространёнными в России. Прини- мая во внимание данный факт, необходимо выделить ряд определений блага в контек- сте социального служения, который присущ священным текстам ислама, христианства и иудаизма. Подобный подход позволит выявить общие черты, присущие значению по- нятия «благо» в подобном контексте для различных авраамических религий, а также структурировать знания и наметить пространство для межрелигиозного диалога. Не- обходимо отметить, что социальное служение, в свою очередь, является методом реа- лизации концепции общественного блага в авраамических религиях. В основе своей в священных текстах благо является морально-этической категорией. Понятие блага в исламе В ходе анализа понятия «благо в исламе» необходимо учитывать, что с точки зрения мусульманской догматики источником блага для человека является бог. Об этом свиде- тельствует следующий аят священного Корана: «Более благого, лучшего удела, чем тот, что предоставляется Господом, не сыскать». Сура аль-Хадж, аят 58 [Коран, 1990: 496]. Важным аспектом для понимания блага в исламе является утверждение о важности стремления к благу в нынешний жизни и после смерти. Социальное служение в исламе позволяет достигать баланса между выделением имущества на нужды бедных и приоб- ретением награды в контексте веры. Приведённый ниже аят подтверждает данный те- зис: «Сделали Мы жертвенное животное (верблюда и верблюдицу) [а также быка и ко- рову, заклание каждого из которых производится от семи человек, в отличие от бара- нов и овец, которые лишь от одного] обрядом, в котором благо для вас [мирское и веч- ное]». Сура аль-Хадж, аят 36 [Коран, 1990: 495]. Для каждого мусульманина крайне важным является задача помощи ближним. За- кят — один из пяти столпов веры. В случае, если человек не имеет экономических ре- сурсов для осуществления благотворительной деятельности, он имеет возможность со- ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) Nations and religions of Eurasia • 2023 Vol. 28, № 4. P. 134–146. 137 вершать благие поступки именно в контексте социального служения, призывая делать добро. Введение Следовательно, даже члены мусульманской уммы, которые с точки зрения боль- шинства определений являются нуждающимися, сами способны совершать благие по- ступки, призывая делать добро и пропагандируя постулаты взаимопомощи. На данный факт указывает следующий аят: «Пусть среди вас будет группа людей, которые будут призывать к благому, повелевать одобряемое и запрещать предосудительное. Именно они окажутся преуспевшими». Сура Али Имран, аят 104 [Коран, 1990: 271]. Благотворительности также посвящено большое количество хадисов (изречений пророка Муххамеда): «Абу Муса сказал: «однажды пророк сказал: «Каждый мусуль- манин должен подавать милостыню». Люди спросили: «О пророк Аллаха, а что делать тому, кто ничего не имеет?» Он ответил: «Он должен зарабатывать своим трудом, при- нося пользу самому себе и подавая милостыню из заработанного». Люди спросили: «А если он окажется не в состоянии подавать милостыню и в этом случае?» Он отве- тил: «Тогда ему следует помочь нуждающемуся, который оказался в беде». Люди сно- ва спросили: «А если он окажется не в состоянии сделать и этого?» Он ответил: «Тогда пусть совершает одобряемое шариатом и воздерживается от всего дурного, и это за- чтётся ему как cадака»; Укба ибн «Амир рассказывал, что слышал, как Посланник Алла- ха сказал: «Каждый человек будет находиться в сени своих пожертвований, пока не за- кончится суд над людьми» [Аз-Зубайди, 2013]. «Абу Хурейра рассказывал, что Послан- ник Аллаха сказал: «Кто просит и выпрашивает у людей богатство ради собственного обогащения, тот просит горящие угли» [Аз-Зубайди, 2013]. «Совершайте намаз и вы- плачивайте закят. Все то благое, что вы предварите для себя, вы найдете у Аллаха. Во- истину, Аллах видит то, что вы совершаете». Сура аль Бакара, аят 110 [Коран, 1990: 152]. Данный аят позволяет убедиться в важности выплаты милостыни в контексте блага. В данном случае речь идёт о том, что Господь как источник блага для всего земного от- кроет для подающего путь к получению блага в жизни и после смерти. А. В. Смирнов считает важнейшей чертой определения блага в исламе, указывая на важность сочетания мирской и вечной жизни. При подобных условиях верующий не отрекается от материального мира с целью конвертировать собственную бедность и аскетичный образ жизни в богатство и блага после смерти [Смирнов, 2008: 155–156]. «А если кто осознанно совершает доброе дело, то это более благое для него самого». Сура аль Бакара, аят 184 [Коран, 1990: 169]. В контексте данных строк Корана особую важность приобретает категория осознанности. Cайт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv Введение «Вы являетесь лучшей из общин, появившейся на благо человечества, повелевая совершать одобряемое, удерживая от предосудительного и веруя в Аллаха». Сура Али Имран, аят 110 [Коран, 1990: 328]. Согласно данному отрывку, миссия мусульманской уммы в рамках человечества заключается в реализации концепции блага для каждого человека, поскольку мусульманин является наместником Бога на земле. Миссией му- сульман как общности является творить добро, поэтому по утверждению имама Аль- Кардави, следует обращать в ислам как можно больше людей. Под совершением бого- угодных деяний понимается совместное совершение благих поступков в масштабе го- сударства или города. Мусульманская община должна стремиться играть важную роль в улучшении жизни окружающих. Т. Э. Кафаров обращает внимание на следующую особенность аятов священно- го Корана, посвящённых благу: ряд формулировок включает в себя определение бла- га как осознанного и осмысленного поступка. Под благом понимается сохранение за- боты о мирском и посмертном. При распределении ресурсов для помощи необходи- мо помнить о том, что в исламе существует чёткое разграничение помощи [Кафаров, 2010: 72–74]. При совершении поступков, направленных на улучшение благосостояния нуждаю- щихся, мусульманин приумножает состояние адресата помощи, тем самым увеличивая своё благосостояние в метафизическом смысле. Однако необходимо помнить, что осо- знанность в данном случае является важнейшей категорией взаимодействия с нуждаю- щимися, осмысленность действий позволит передать дополнительный символический смысл каждому пожертвованию, укрепляя веру мусульманина и выстраивая у людей, не придерживающихся постулатов ислама, позитивный образ данной религии. В. В. Липов в своей работе указывает на важность блага как общественно-значимой категории. Мусульмане, совершая благие поступки, должны выстраивать позитивную коммуникацию с представителями других религий, привлекая их к процессу ислам- ского социального служения. Отсутствие указаний на то, что объектами помощи дол- жны быть только мусульмане, позволяет исламу, наряду с другими авраамическими религиями, расширять понятие хорошего поступка и помощи на представителей всех конфессий. Данное положение дел позволяет расширять деятельность мусульманских благотворительных организаций вне пространства мечети и исламских учебных заве- дений [Липов, 2005: 68–69]. В. В. Липов в своей работе указывает на важность блага как общественно-значимой категории. Мусульмане, совершая благие поступки, должны выстраивать позитивную коммуникацию с представителями других религий, привлекая их к процессу ислам- ского социального служения. Отсутствие указаний на то, что объектами помощи дол- жны быть только мусульмане, позволяет исламу, наряду с другими авраамическими религиями, расширять понятие хорошего поступка и помощи на представителей всех конфессий. Данное положение дел позволяет расширять деятельность мусульманских благотворительных организаций вне пространства мечети и исламских учебных заве- дений [Липов, 2005: 68–69]. Введение Вовлечённость в совершение благих поступков и степень их осознанности является крайне важной при оценке искренно- сти поступка, поскольку лицемерие является одним из самых больших грехов в исла- ме. В приведённом выше аяте, помимо совершения доброго дела, под благом понима- ется осознанность, включённость и вовлечённость. По мнению А. Сальваторе, при выполнении общественно-полезных действий, прак- тикующие верующие руководствуются принципами гуманности и этики, совершая ряд действий в качестве субъектов определённых поведенческих моделей [Salvatore, Dale, 2004: 134]. «Любое добро, которое вы раздаете, должно достаться родителям, близким род- ственникам, сиротам, беднякам, путникам. Что бы вы ни сделали благого, Аллах зна- 138 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. ет об этом». Сура аль Бакара, аят 21 [Коран, 1990: 54]. Легитимность благополучателей в системе исламского социального служения является важным аспектом. Попрошай- ничество и иждивенчество крайне порицаемы с точки зрения исламской философии. Поддержка подобных категорий людей, даже если мусульманин знает об их нечестно- сти, с формальной точки зрения представляется благим поступком. ет об этом». Сура аль Бакара, аят 21 [Коран, 1990: 54]. Легитимность благополучателей в системе исламского социального служения является важным аспектом. Попрошай- ничество и иждивенчество крайне порицаемы с точки зрения исламской философии. Поддержка подобных категорий людей, даже если мусульманин знает об их нечестно- сти, с формальной точки зрения представляется благим поступком. Р. М. Даудов в своей работе указывает на ещё одну важную установку. При оценке благого деяния с точки зрения исламской морали, адресность пожертвования играет одну из ключевых ролей для получения награды от Господа [Даудов, 2014: 4–6]. Р. М. Даудов в своей работе указывает на ещё одну важную установку. При оценке благого деяния с точки зрения исламской морали, адресность пожертвования играет одну из ключевых ролей для получения награды от Господа [Даудов, 2014: 4–6]. «Вы являетесь лучшей из общин, появившейся на благо человечества, повелевая совершать одобряемое, удерживая от предосудительного и веруя в Аллаха». Сура Али Имран, аят 110 [Коран, 1990: 328]. Согласно данному отрывку, миссия мусульманской уммы в рамках человечества заключается в реализации концепции блага для каждого человека, поскольку мусульманин является наместником Бога на земле. Миссией му- сульман как общности является творить добро, поэтому по утверждению имама Аль- Кардави, следует обращать в ислам как можно больше людей. Под совершением бого- угодных деяний понимается совместное совершение благих поступков в масштабе го- сударства или города. Мусульманская община должна стремиться играть важную роль в улучшении жизни окружающих. Cайт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv Введение Сунна пророка, которая является вторым после Корана источником знаний для каж- дого правоверного мусульманина, также, в свою очередь, содержит информацию о раз- личных аспектах блага при пожертвованиях. «Если женщина раздает еду, хранящуюся ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) 139 Nations and religions of Eurasia • 2023 Vol. 28, № 4. P. 134–146. в ее доме, и не расточительствует при этом, то она получит вознаграждение за свое по- жертвование. Ее супруг получит вознаграждение за то, что заработал, и такое же воз- награждение получит слуга, отвечающий за кладовую. При этом ничья награда не ста- нет меньше из‑за того, что вознаграждение получит и другой». (Этот хадис передали аль-Бухари и Муслим). Данный хадис указывает на важность грамотного распределе- ния имущества при осуществлении пожертвований. Если мусульманин будет прене- брегать интересами своей семьи, стремясь за счёт систематического совершения по- даяний заработать благо, то он будет порицаем. Благом является умение не только да- вать милостыню, но и распределять имущество таким образом, чтобы при социальном служении не были ущемлены интересы семьи. Понятие блага в христианстве Понятие блага в христианстве Целью христианского социального учения, включая его социально-этические, соци- ально-педагогические и общественно-политические аспекты, является установление со- циального строя, при котором человек может выполнять волю Бога и вести христиан- ский образ жизни. «Ибо и Сын Человеческий не для того пришел, чтобы Ему служили, но чтобы послужить и отдать душу Свою для искупления многих» [Библия, Мк. 10: 45]. «Вот пост, который Я избрал: разреши оковы неправды, развяжи узы бремени, чтоб освободить угнетенных, и сломать всякое ярмо; Не в том ли он, чтобы ты поделился едой с голодным и дал пристанище бедному скитальцу, видя нагого, одень его, и не от- ворачивайся от родственников. Тогда воссияет твой свет, как заря, и быстро придет твоё исцеление, Твоя праведность пойдёт пред тобою, и слава Господня защитит тебя с тыла. Тогда воззовешь ты — и Господь отзовется, позовешь на помощь — и Он от- ветит „Я здесь”. Если покончишь с ярмом угнетения, прекратишь показывать пальцем и оскорблять, предложишь свою пищу голодному и насытишь страдальца, тогда вос- сияет во тьме твой свет и мрак твой станет как полдень; И Господь всегда будет вести тебя; Он насытит тебя в безводных местах и укрепит твои кости» [Библия, Ис. 58]. Ис- ходя из этой главы книги пророка Исаии можно сделать вывод, что в христианстве так- же важным благодеянием является помощь нуждающимся. В данном фрагменте также не делается акцент на вероисповедании нуждающегося. Р. А. Фатхуллин видит в данном случае определение блага в рамках христианской морали как освобождение от пороков. Забота о ближних и родственниках выделяется особо, данный постулат при определении блага идентичен с повелениями из священ- ного Корана, которые призывают не забывать о родственниках, но вместе с этим по- могать и незнакомым людям [Фатхуллин, 2022: 121–122]. «Он говорит нам, что поскольку это так, то, оказывая им внимание, мы служим Самому Господу — «Говорю вам истину: то, что вы сделали одному из наименьших Моих братьев, вы сделали Мне» [Библия, Мф. 25: 40]. В данном случае речь идёт о том, что благое дело в христианстве, направленное нуждающемуся, является и благим де- лом по отношению к Богу. Д. Холленбач считает важным, что в священном писании упоминается, что любое потраченное на благое дело имущество многократно вернётся человеку после смерти. В данном контексте под благом понимается несоизмеримость награды даже за самый малый добрый поступок [Hollenbach, 2002: 66–67]. 140 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. Л. В. Баева считает, что в Библии зачастую подчёркивается важность искренности при совершении благих деяний. Понятие блага в христианстве Проявляя смирение во время подачи милостыни, веру- ющий должен бороться не только с жадностью, но и с гордыней. Искренность поступ- ка позволяет сделать процесс социального служения очищающим [Баева, 2007: 78–80]. Е. В. Бобырева и О. А. Дмитриева выдвигают гипотезу, согласно которой порицание попрошайничества и чёткое разграничение групп, на которые необходимо направлять помощь, характерно для основных направлений христианства и должно способствовать максимальному вовлечению каждого человека в процесс, поскольку тенденция цифро- визации экономики сделала возможным отправлять пожертвования на механическом уровне, не задумываясь о совершаемом действии [Бобырева, Дмитриева, 2021: 577]. Понятие блага в иудаизме Понятие блага в иудаизме Понятие блага в иудаизме В Торе благотворительности уделено важное внимание. Существует ряд чётко обо- значенных категорий людей, которым необходима помощь в первую очередь. К ним относятся малоимущие, сироты, рабы на выкуп, на которых можно потратить сред- ства, старики и др. Благотворительность в иудаизме охватывает все стороны жизни, от сбора придано- го для невест до помощи в организации похорон. Для благотворителя в иудаизме важ- но оставаться незамеченным. А. В. Коркина делает акцент на некоем намеренном отчуждении еврейских благо- творительных организаций от своих подопечных. Делающие пожертвования зачастую не видят человека или людей, которым помогают. Подобная система призвана исклю- чить из отношений между людьми излишнюю гордость и подобострастие [Коркина, 2016: 47–48]. С точки зрения концепции социального служения, в иудаизме важным является не только процесс помощи человеку, но и конечный результат. Иудейская благотвори- тельность зачастую носит систематически направленный характер на конкретного че- ловека или семью, её цель — помочь накопить определенную сумму для улучшения фи- нансового положения и социального статуса нуждающегося. С точки зрения помога- ющего, благом становится не сам процесс передачи имущества, а поиск для малоиму- щих работы, налаживание их социальных связей и даже психологическая поддержка. Для верующего иудея творение добрых дел должно быть сопряжено с некой систем- ностью, которая позволит вместе с оказанием помощи мотивировать нуждающегося к самостоятельному изменению своего положения в обществе. И. А. Абдулаев обращает внимание на важность для иудейской общины высокой степени вовлечённости в процессе оказания помощи. Данные мероприятия, по мне- нию автора, сплачивают общины внутри иудейского общества [Абдулаева, 2011: 89–90]. р , щ у р уд щ [ ду , ] В иудаизме как благо воспринимается не только богатство или имущество, даро- ванное конкретному человеку, но и как имущество, которым в целом должна распо- ряжаться община, поэтому благотворительные организации в иудаизме называют- ся братствами. Каждый участник должен выделить определённую часть собствен- ного имущества или благосостояния и нести ответственность за то, чтобы его бо- гатство было распределено наиболее эффективно. Такой подход позволяет, с одной стороны, сплачивать общину вокруг социального служения, а с другой — создавать ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) 141 Nations and religions of Eurasia • 2023 Vol. 28, № 4. P. 134–146. определённую категорию людей, для которых благотворительность является посто- янной практикой. определённую категорию людей, для которых благотворительность является посто- янной практикой. Зачастую благо как понятие, относящееся к концепции социального служения, трак- туется с позиции помощи и искренности. При таком подходе упускается аксиологиче- ский аспект взаимопомощи. Е. Н. Cайт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv Понятие блага в иудаизме Исходя из этого человек, убедив- шийся в прозрачности работы благотворительных фондов и инициатив, распростра- няющий информацию о них в масштабе глобальной сети, с точки зрения религиозной 142 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. морали, совершает определённое благо, также становясь агентом социального служе- ния [Дудорова, 2020: 341–343]. Существует несколько иной взгляд на трансформацию понятия блага в условиях развития современных технологий, таких как цифровизация экономических операций. Совершая пожертвования на постоянной основе, верующий пользуется всевозможны- ми интерфейсами для осуществления определённых экономических операций, таких как переводы и оплата. Зачастую в подобных условиях человек отчуждается от процес- са совершения благого дела. Механизация и рутинизация приводят к практически пол- ному отсутствию рефлексии и осмыслении процесса совершения благого поступка, та- ких как помощь ближним или поддержка благотворительных инициатив. При подоб- ном положении дел благим будет возвращение к традиционным практикам социаль- ного служения. Восстановление межличностного характера взаимодействий при соци- альном служении позволит отрефлексировать и оценить уровень искренности каждо- го поступка [Ванюкова, 2015: 27–29]. Поскольку при совершении благих дел в доктрине авраамических религий крайне важна осмысленность, то во многом подобные практики должны способствовать воз- растанию уровня сопереживания и реагирования. Одной из проблем является возмож- ность автоматического списания средств с личного счёта адресанта помощи. Подоб- ная практика фактически полностью отчуждает человека от благополучателя, которо- му он оказывает помощь. При автоматическом списании средств, процесс пожертвова- ния становится абсолютно механическим и лишённым всякой сакральности [Чеснова, 2022: 33]. Осмысление положительных и отрицательных черт современных технологий при со- вершении блага до сих пор занимает важное место в богословском дискурсе авраамиче- ских религий. Процессы глобализации актуализировали на современном этапе пробле- му всеобщего или мирового блага. Некоторые учёные, занимающиеся разработкой по- нятий взаимопомощи и социальной справедливости, считают важным апеллирование категориям блага в масштабе всей планеты. При подобной постановке проблемы зада- чей религиозного благотворительного фонда становится расширение зоны свой дея- тельности за пределы своего города либо страны. Н. И. Тимохина считает, что забота о благе должна охватывать всё человечество. При подобном подходе в идеале образу- ется определённый тип взаимодействия между благотворительными организациями, позволяющий устранить проблемы, которые принято называть общемировыми. В дан- ной интерпретации благо представляется как стремление через социальное служение достичь общественной справедливости не только в масштабах общины, города, стра- ны, но и всего мира [Тимохина, 2019: 241–244]. Схожесть интерпретаций понятия благо, общность духовных ориентиров и мораль- но-этических установок, а также ценностных систем позволяет представителям различ- ных религий налаживать устойчивую долгосрочную коммуникацию в сфере социаль- ного служения. С точки зрения С. Понятие блага в иудаизме Дорф пишет о важности системной поддержки ну- ждающихся для религиозных сообществ с точки зрения влияния на общественное со- знание. В обществах, где религиозные организации оказывают помощь незащищён- ным слоям населения, благодеяние постепенно из религиозной практики постепенно смещается в сферу этики [Dorf, 2004: 200–203]. В своей работе В. Н. Мунгалов так описывает общие черты понятия блага с точки зрения взаимопомощи в авраамических религиях. Для ислама, христианства и иудаиз- ма характерны следующие общие дефиниции при осмыслении блага наряду с поняти- ем социального служения: богатство, отданное в качестве милостыни, вернётся в мно- гократном размере в мирской жизни или в вечности; благо должно распределяться ра- зумно; даже незначительное дело, направленное на улучшение окружающего мира, яв- ляется благим с точки зрения социального служения [Мунгалов, 2014: 146]. Каждая благотворительная инициатива становится частью общей миссии каждой религиозной конфессии по реализации концепции всеобщего блага на земле. Т. В. Би- рюлина считает, что общая направленность морально-этических установок и ориенти- ров позволяет представителям авраамической религии выстраивать позитивную ком- муникацию в сфере благотворительности, создавая предпосылки для межрелигиозно- го диалога [Бирюлина, 2011: 14–15]. Современные тенденции об интерпретации блага в авраамических религ б Современные тенденции об интерпретации блага в авраамических религиях При осмыслении понятия блага как важной части социальной доктрины в любой авраамической религии богословы заостряют своё внимание на развитие современ- ных технологий в сфере массовых коммуникаций. Для ислама, христианства и иуда- изма общим является утверждение о том, что не имеющий имущества человек, пропо- ведующий или рассказывающий о занятии благотворительностью, а также призываю- щий делать добро, совершает благой поступок. Консенсус касательно данного положе- ния позволяет прийти к важности интернета как площадки для формирования опреде- лённого социального капитала путём распространения информации о благе, заключа- ющемся в социальном служении. Человек, не имеющий имущества для помощи окру- жающим, но использующий информационные ресурсы, например, собственную стра- ницу в социальной сети, либо иные возможности информационно-коммуникацион- ных технологий для распространения информации о благотворительных инициати- вах в целом и нуждах конкретных некоммерческих организаций и благотворительных фондов, в частности, тем самым призывает совершать добро [Moussa, 2017: 100–104]. ф д , , р р д р [ , ] Мультимедиацизация и конвергентность позволяют доносить информацию о рели- гиозном активизме и социальном служении, используя различные каналы восприятия. По мнению А. С. Дудоровой, для социального служения представителей различных кон- фессий информационная поддержка крайне важна. Заключение Анализируя категорию блага в контексте социального служения в авраамических религиях, можно отметить ряд общих черт и понятий категорий блага и благодеяний. К ним относятся: помощь ближним, осуществляемая искренне и осознанно; общие по- стулаты, касающиеся наиболее приоритетных групп благополучателей; улучшение жиз- ни окружающих; социальное служение, осуществляемое совместно с единоверцами. Определения блага в различных авраамических религиях в основе своей схожи, од- нако наблюдается ряд разночтений. В исламе и иудаизме Бог является источником бла- га, тогда как в христианстве Бог также наравне с нуждающимися является своего рода объектом помощи. Общие понятия, связанные с благом, воспринимаемые в контексте социального служения, позволяют представителям авраамических религий создавать общие благотворительные инициативы, совместно реализуя таким образом миссию каждой конфессии по установлению на земле всеобщего блага. Понятие блага в иудаизме В. Мельника, тип межрелигиозного диалога в сфере благотворительности имеет название «партнёрский» [Мельник, 2022: 81–82]. р р р На круглом столе, прошедшим в 2021 г., обсуждалась тема социального служения как инструмента установления общественного блага. В данном конкретном случае дис- ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) 143 Nations and religions of Eurasia • 2023 Vol. 28, № 4. P. 134–146. куссия о благе как о центральном понятии в рамках концепции религиозного активиз- ма и социального служения стала своего рода фактором формирования партнёрского межрелигиозного диалога между представителями двух авраамических религий, в дан- ном конкретном случае ислама и христианства. Общие интерпретации категории блага в рамках социального служения обусловли- вает ряд сфер, в которых с наибольшей частотностью происходит межконфессиональ- ное сотрудничество. К таким категориям относятся: строительство и восстановление религиозных сооружений, совместное кормление нуждающихся, оказание помощи пострадавшим от природных катаклизмов. В других сферах межрелигиозный диалог выстраивается скорее из общности традиций и не имеет столь устойчивого характера. Принимая во внимание изложенное, можно сделать вывод, что общее понимание бла- га как базового понятия социального служения является важным фактором при фор- мировании межконфессионального сотрудничества [Kearney, 2018: 40–47]. Cайт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv БИБЛИОГРАФИЧЕСКИЙ СПИСОК № 4 (294). С. 341–343. Кафаров Т. Э. Принципы этики ислама: нравственный ориентир для мусульмани- на // Исламоведение. 2010. № 2. С. 72–74. Кафаров Т. Э. Принципы этики ислама: нравственный ориентир для мусульмани- на // Исламоведение. 2010. № 2. С. 72–74. Коран / перевод И. Ю. Крачковского. М. : Наука, 1990. 831 с. Коркина А. В. Проявление благодати в еврейской библии: аспект заветов // World science. 2016. № 6 (10). С. 47–48. Липов В. В. Религиозные ценности как фактор зависимости от предшествующего развития и формирования социально-экономических моделей // Пространство эко- номики. 2005. № 3. С. 68–69. Мельник С. В. Межрелигиозный диалог: типологизация, методология, формы реа- лизации. М. : ИНИОН РАН, 2022. 398 с. Мунгалов В. Н. Мировоззренческие различия буддизма и авраамических религий // Психология в экономике и управлении. 2014. № 1. С. 144–148. у р р р удд р р Психология в экономике и управлении. 2014. № 1. С. 144–148. Смирнов А. В. «Благо» и «Зло» в исламской традиции и философии (к постановке вопроса) // Этическая мысль. 2008. № 8. С. 155–156. охина Н. И. Особенности коммуникации благотворительных организаций в со- // М й й 2019 №12 (250) С 241 244 Тимохина Н. И. Особенности коммуникации благотворительных организаций в со- циальных сетях // Молодой ученый. 2019. № 12 (250). С. 241–244. О у ц р р ц ых сетях // Молодой ученый. 2019. № 12 (250). С. 241–244. Фатхуллин Р. А. Единый вектор социальной направленности христианства и исла- Фатхуллин Р. А. Единый вектор социальной направленности христианства и исла- ма // Известия Иркутского государственного университета Серия: Политология Ре- Фатхуллин Р. А. Единый вектор социальной направленности христианства и исла- ма // Известия Иркутского государственного университета. Серия: Политология. Ре- лигиоведение. 2022. Т. 40. С. 121–122. ма // Известия Иркутского государственного университета. Серия: Политология. Ре- лигиоведение. 2022. Т. 40. С. 121–122. Чеснова Е. Н. Специфика цифровизации ислама и it-халяль // Исламоведение. 2022. № 3 (53). С. 26–41. Armando Salvatore, Dale F. Eickelman Public islam and the common good. 2004. P. 134 (на англ. яз).h David Hollenbach, S. J. The common good and christian ethics. 2002, рр. 66–67 (на англ. яз). Eliot N. Dorf To Do the Right and the Good, a jewish approach to modern socail ethics. 2004, рр. 200–203 (на англ. яз). Kearney J. From Intra-Religious to Interreligious Dialogue. 2018. P. 40–47 (на англ. яз).hh Moussa M. БИБЛИОГРАФИЧЕСКИЙ СПИСОК Абдулаева И. А. Моральная система ценностей в авраамистических религиях на при- мере традиционных черт иудаизма в Дагестане // Наука и современность. 2011. № 8–3. С. 89–90. Абдулаева И. А. Моральная система ценностей в авраамистических религиях на при- мере традиционных черт иудаизма в Дагестане // Наука и современность. 2011. № 8–3. С. 89–90. Аз-Зубайди, Мухтасар Сахих аль-Бухари. Алматы : ТКәусар-саяхат, 2013. 609 с Аз-Зубайди, Мухтасар Сахих аль-Бухари. Алматы : ТКәусар-саяхат, 2013. 609 с. Баева Л. В. Классические христианские ценности с позиции современной экзистен- циальной антропологии // Государство, религия, церковь в России и за рубежом. 2007. № 3. С. 78–80. Баева Л. В. Классические христианские ценности с позиции современной экзистен- циальной антропологии // Государство, религия, церковь в России и за рубежом. 2007. № 3. С. 78–80. Библия: Книги Священного Писания Ветхого и Нового Завета: в русском переводе с параллельными местами. М. : Библейское общество, 1995. 1376 с. Библия: Книги Священного Писания Ветхого и Нового Завета: в русском переводе с параллельными местами. М. : Библейское общество, 1995. 1376 с. Бирюлина Т. В. Вера Авраама как тема диалога авраамических религий // Исламо- ведение. 2011. № 3. С. 14–15. Бирюлина Т. В. Вера Авраама как тема диалога авраамических религий // Исламо- ведение. 2011. № 3. С. 14–15. Бобырева Е. В., Дмитриева О. А. Сравнительная характеристика ценностных ориен- тиров христианства и буддизма // СИСП. 2021. № 4. С. 568–583. Бобырева Е. В., Дмитриева О. А. Сравнительная характеристика ценностных ориен- тиров христианства и буддизма // СИСП. 2021. № 4. С. 568–583. Ванюкова И. А. Влияние религиозных ценностей на экономику и её развитие // Со- временные научные исследования и инновации. 2015. № 6, ч. 4. С. 27–29. Ванюкова И. А. Влияние религиозных ценностей на экономику и её развитие // Со- временные научные исследования и инновации. 2015. № 6, ч. 4. С. 27–29. 144 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. Верующие разных религий — вместе в деле благотворительности // Институт религии и политики. URL: https://irp. news/verujushhie-raznyh-religij-vmeste-v-dele- blagotvoritelnosti/ (дата обращения: 18.04.2023). Даудов Р. Х. Специфика исламской философии // Гуманитарные, социально-эконо- мические и общественные науки. 2014. № 12–1. С. 4–6. Даудов Р. Х. Специфика исламской философии // Гуманитарные, социально-эконо- мические и общественные науки. 2014. № 12–1. С. 4–6. Дудорова А. С. Отражение феномена киберрелигиозности в практике мировых ре- лигий // Молодой ученый. 2020. № 4 (294). С. 341–343. Дудорова А. С. Отражение феномена киберрелигиозности в практике мировых ре- лигий // Молодой ученый. 2020. БИБЛИОГРАФИЧЕСКИЙ СПИСОК The Use of the Internet by Islamic Social Movements in Collective Action: The Case of Justice and Charity. 2017, рр. 100–104 (на англ. яз). REFERENCES Abdulaeva I. A. Moral'naya sistema cennostej v avraamisticheskih religiyah na primere tradicionnyh chert iudaizma v Dagestane [Moral system of values in Abrahamic religions on the example of traditional features of Judaism in Dagestan]. Nauka i sovremennost' [Science and modernity]. 2011, no. 8–3, pp. 89–90 (in Russian). ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) 145 Nations and religions of Eurasia • 2023 Vol. 28, № 4. P. 134–146. Az-Zubajdi Muhtasar Sahih al' — Buhari. Almaty: TOO “Kәusar-sayahat”, 2013, 609 p. (in Russian). Az-Zubajdi Muhtasar Sahih al' — Buhari. Almaty: TOO “Kәusar-sayahat”, 2013, 609 p. (in Russian). Baeva L. V. Klassicheskie hristianskie cennosti s pozicii sovremennoj ekzistencial'noj antropologii [Classical Christian values from the standpoint of modern existential anthropology]. Gosudarstvo, religiya, cerkov' v Rossii i za rubezhom [State, religion, church in Russia and abroad]. 2007, no. 3, pp. 78–80 (in Russian). Bibliya: Knigi Svyashchennogo Pisaniya Vethogo i Novogo Zaveta: v russkom perevode s parallel'nymi mestami [Bible: Books of the Holy Scriptures of the Old and New Testament: in Russian translation with parallel passages]. Moscow: Biblejskoe obshchestvo, 1995, 1376 p. (in Russian). Biryulina T. V. Vera Avraama kak tema dialoga avraamicheskih religij [Faith of Abraham as a topic of dialogue among Abrahamic religions]. Islamovedenie [Study of Islam]. 2011, no. 3, pp. 14–15 (in Russian). Bobyreva E. V., Dmitrieva O. A. Sravnitel'naya harakteristika cennostnyh orientirov hristianstva i buddizma [comparative characteristics of value guidelines of christianity and buddhism]. SISP [SISP]. 2021, no. 4, pp. 568–583 (in Russian).ii Bobyreva E. V., Dmitrieva O. A. Sravnitel'naya harakteristika cennostnyh orientirov hristianstva i buddizma [comparative characteristics of value guidelines of christianity and buddhism]. SISP [SISP]. 2021, no. 4, pp. 568–583 (in Russian).ii Chesnova E. N. Specifika cifrovizacii islama i it-halyal' [the specificity of the digitalization of islam and it-halal]. Islamovedenie [Study of Islam]. 2022, no. 3 (53), pp. 26–41 (in Russian).iiii Chesnova E. N. Specifika cifrovizacii islama i it-halyal' [the specificity of the digitalization of islam and it-halal]. Islamovedenie [Study of Islam]. 2022, no. 3 (53), pp. 26–41 (in Russian).iiii Daudov R. H. Specifika islamskoj filosofii [Specifics of Islamic Philosophy]. Gumanitarnye, social'no-ekonomicheskie i obshchestvennye nauki [Humanitarian, socio-economic and social sciences]. 2014, no. 12–1, pp. 4–6 (in Russian). Daudov R. H. Specifika islamskoj filosofii [Specifics of Islamic Philosophy]. Gumanitarnye, social'no-ekonomicheskie i obshchestvennye nauki [Humanitarian, socio-economic and social sciences]. 2014, no. 12–1, pp. 4–6 (in Russian). Dudorova, A. S. Cайт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv REFERENCES Otrazhenie fenomena kiberreligioznosti v praktike mirovyh religij [Reflection of the Phenomenon of Cyberreligiosity in the Practice of World Religions]. Molodoj uchenyj [Young scientist]. 2020, no. 4 (294), pp. 341–343 (in Russian). Dudorova, A. S. Otrazhenie fenomena kiberreligioznosti v praktike mirovyh religij [Reflection of the Phenomenon of Cyberreligiosity in the Practice of World Religions]. Molodoj uchenyj [Young scientist]. 2020, no. 4 (294), pp. 341–343 (in Russian). Fathullin R. A. Edinyj vektor social'noj napravlennosti hristianstva i islama [A single vector of the social orientation of Christianity and Islam]. Izvestiya Irkutskogo gosudarstvennogo universiteta. Seriya Politologiya. Religiovedenie [Proceedings of the Irkutsk State University. Political science series. religious studies]. 2022, vol. 40, pp. 121–122. (in Russian). Fathullin R. A. Edinyj vektor social'noj napravlennosti hristianstva i islama [A single vector of the social orientation of Christianity and Islam]. Izvestiya Irkutskogo gosudarstvennogo universiteta. Seriya Politologiya. Religiovedenie [Proceedings of the Irkutsk State University. Political science series. religious studies]. 2022, vol. 40, pp. 121–122. (in Russian). Kafarov T. E. Principy etiki islama: nravstvennyj orientir dlya musul'manina [Principles of Islamic Ethics: A Moral Guideline for a Muslim]. Islamovedenie [Study of Islam]. 2010, no. 2, pp. 72–74 (in Russian). Kafarov T. E. Principy etiki islama: nravstvennyj orientir dlya musul'manina [Principles of Islamic Ethics: A Moral Guideline for a Muslim]. Islamovedenie [Study of Islam]. 2010, no. 2, pp. 72–74 (in Russian). Koran, perevod I. Y. Krachkovskogo [Koran, I. Y. Krachkovsky translation]. Moscow: Nauka, 1990, 831 p. (in Russian). Koran, perevod I. Y. Krachkovskogo [Koran, I. Y. Krachkovsky translation]. Moscow: Nauka, 1990, 831 p. (in Russian). Korkina A. V. Proyavlenie blagodati v evrejskoj biblii: aspekt zavetov [the manifestation of grace in the hebebible: a covenant aspect]. World science [World science]. 2016, no. 6 (10), pp. 47–48 (in Russian). Korkina A. V. Proyavlenie blagodati v evrejskoj biblii: aspekt zavetov [the manifestation of grace in the hebebible: a covenant aspect]. World science [World science]. 2016, no. 6 (10), pp. 47–48 (in Russian). Lipov V. V. Religioznye cennosti kak faktor zavisimosti ot predshestvuyushchego razvitiya i formirovaniya social'no-ekonomicheskih modelej [Religious values as a factor of dependence on the previous development and the formation of socio-economic models]. Prostranstvo ekonomiki [Space of Economics]. 2005, no. 3, pp. 68–69 (in Russian). Mel'nik S. V. Mezhreligioznyj dialog: tipologizaciya, metodologiya, formy realizacii [Interreligious Dialogue: Typology, Methodology, Forms of Implementation]. Moscow: INION RAN, 2022, 398 p. (in Russian). ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online) REFERENCES айт журнала: http://journal.asu.ru/wv • Journal homepage: http://journal.asu.ru/wv 146 Народы и религии Евразии • 2023 Том 28, № 4. C. 134–146. Mungalov V. N. Mirovozzrencheskie razlichiya buddizma i avraamicheskih religij [Worldview differences between Buddhism and Abrahamic religions]. Psihologiya v ekonomike i upravlenii [Psychology in Economics and Management]. 2014, no. 1, pp. 145–148 (in Russian). Smirnov A. V. “Blago” i “Zlo” v islamskoj tradicii i filosofii (k postanovke voprosa) [“Good” and “Evil” in the Islamic tradition and philosophy (to the formulation of the question)]. Eticheskaya mysl' [Ethical thought]. 2008, no. 8, pp. 155–156 (in Russian) Timohina, N. I. Osobennosti kommunikacii blagotvoritel'nyh organizacij v social'nyh setyah [Features of communication of charitable organizations in social networks]. Molodoj uchenyj [Young scientist]. 2019, no. 12 (250), pp. 241–244 (in Russian).hl Timohina, N. I. Osobennosti kommunikacii blagotvoritel'nyh organizacij v social'nyh setyah [Features of communication of charitable organizations in social networks]. Molodoj uchenyj [Young scientist]. 2019, no. 12 (250), pp. 241–244 (in Russian).hl Vanyukova I. A. Vliyanie religioznyh cennostej na ekonomiku i eyo razvitie [The influence of religious values on the economy and its development] Sovremennye nauchnye issledovaniya i innovacii [Modern scientific research and innovation]. 2015, no. 6, part 4, pp. 27–29 (in Russian). Veruyushchie raznyh religij — vmeste v dele blagotvoritel'nosti [believers of different religions — together in the cause of charity] (in Russian). Veruyushchie raznyh religij — vmeste v dele blagotvoritel'nosti [believers of different religions — together in the cause of charity] (in Russian). Armando Salvatore, Dale F. Eickelman Public islam and the common good. 2004, 134 p. (in English). Armando Salvatore, Dale F. Eickelman Public islam and the common good. 2004, 134 p. (in English).h id Hollenbach, S. J. The common good and christian ethics. 2002, pp. 66–67 (in English). David Hollenbach, S. J. The common good and christian ethics. 2002, pp. 66–67 (in English). Eliot N. Dorf To Do the Right and the Good, a jewish approach to modern socail ethics. 2004, pp. 200–203 (in English). Eliot N. Dorf To Do the Right and the Good, a jewish approach to modern socail ethics. 2004, pp. 200–203 (in English). t N. Dorf To Do the Right and the Good, a jewish approach to modern socail ethics. 2004, –203 (in English). Kearney J. From Intra-Religious to Interreligious Dialogue. 2018, pp. 40–47 (in English).hh Kearney J. From Intra-Religious to Interreligious Dialogue. 2018, pp. 40–47 (in English).hh Kearney J. From Intra-Religious to Interreligious Dialogue. 2018, pp. REFERENCES 40–47 (in English).hh Moussa M. The Use of the Internet by Islamic Social Movements in Collective Action: The Case of Justice and Charity. 2017, pp. 100–104 (in English). Moussa M. The Use of the Internet by Islamic Social Movements in Collective Action: The Case of Justice and Charity. 2017, pp. 100–104 (in English). Статья поступила в редакцию: 09.05.2023 Принята к публикации: 12.10.2023 Дата публикации: 12.12.2023 Дата публикации: 12.12.2023 ISSN 2542-2332 (Print) • ISSN 2686-8040 (Online)
https://openalex.org/W3116342345	https://www.researchsquare.com/article/rs-18422/latest.pdf	English	null	Thermal And Chemical Resistance Of Plasma Sprayed Al2o3, Al2o3-Tio2 Coatings	Research Square (Research Square)	2,020	cc-by	4,995	Thermal And Chemical Resistance Of Plasma Sprayed Al2o3, Al2o3-Tio2 Coatings Airingas Šuopys (  airingas.suopys@lei.lt ) Lietuvos energetikos institutas https://orcid.org/0000-0001-5972-3084 Lietuvos energetikos institutas Research Article Keywords: Plasma spray, Al2O3, TiO2, thermal resistance, chemical resistance Posted Date: October 9th, 2020 DOI: https://doi.org/10.21203/rs.3.rs-18422/v2 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Posted Date: October 9th, 2020 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Page 1/13 Page 1/13 Page 1/13 Abstract In this article, the results of thermal and chemical resistance, of plasma sprayed Al2O3 and Al2O3- 13wt.TiO2 coatings on a steel (P265GH) sample are presented. Coatings were formed using air-hydrogen plasma direct current plasma torch at atmospheric pressure. The resistance of coatings was measured by analyzing surface structure, elemental and phase composition of as sprayed coatings and after several heating cycles, imitating application conditions. To make this technology more appealing, by decreasing exploitation cost of standard cast iron grate used in straw pellet broilers, we used cheaper stainless steel with a protective Al2O3 and Al2O3-13wt.% TiO2 coatings. Most of the research done in this field is by using Ar as primary gas, but to decrease the manufacturing cost even further we used Air as primary gas. It was determined that after heat treatment the minor changes in the elemental composition were observed. Meanwhile an additional crystalline phases of metal oxides were detected. Introduction: The need for functional materials is ever increasing, many systems and their individual parts or sections must be durable, resistant to chemical or thermal effects and corrosion. The higher longevity can be reached – the better. One of the most effective and cost-efficient way to prolong the lifespan of certain parts or systems is to apply a protective layer of material on top. Ceramic coatings are quickly becoming a viable alternative to organic and polymer-based coatings for surface protection applications. Most commonly used oxide ceramics today are alumina, zirconia, titania, yttria or their composites coatings. Ceramic coating is a protective film for all kinds of parts and can dramatically increase their useful life and productivity, by achieving longer run times before the repairs or replacement of part is needed. Thermal spray coating is one of the most common ways to apply a protective coating on the material and plasma spray is one of the most widely used techniques for this purpose [1]. Plasma sprayed ceramic coatings possesses lots of excellent properties, such as wear and corrosion resistance, high temperature resistance, therefore is widely used in various industrial equipment [2], and particularly are useful when corrosion and wear resistance is required simultaneously [3]. Alumina coatings are one of the most popular high-performance ceramic coatings being used in the industry today. Such wide applications are related to the high hardness, chemical inertness and low cost of material. To improve coating qualities, such as brittleness and friction coefficient reduction, increase of fracture toughness and corrosion resistance, addition of titanium oxide to alumina powder is widely used [4–8]. It was found that addition of TiO2 reduces the microhardness of alumina coatings [9–12], also titanium oxide increases corrosion resistance, wear rate, elastic modulus and decreases porosity [13]. It was demonstrated that Al2O3-YSZ composite coatings improves the oxidation resistance at high (1050 oC) temperatures [15]. V.C. Misra et al. [14] demonstrated that the tribological properties of Al2O3 coatings prepared using argon-hydrogen plasma strongly depend on process parameters and could be used in dry air and dry nitrogen environment. M. Wang et al. [16] obtained that Al2O3-13wt.% TiO2 sprayed using argon-helium plasma remained stable after thermal shock tests under temperatures below 800 oC. Page 2/13 Page 2/13 Alumina and alumina-based composite coatings are a promising selection to improve the grate of a straw pellet boiler, since manufacturing high quality steel would be extremely expensive. Introduction: The intention was to lower the price of the grate by using stainless P265GH steel instead of cast iron, that is typically being used in straw pellet broilers. According to manufacturing sites (alibaba.com and iron-foundry.com) the cost of cast iron is 1300–1400 $/ton while stainless steel is only 600–800$/ton. This change of the grate material would greatly decrease the cost of manufacturing, not to mention the increased longevity of the grate. The temperature inside the boiler is about 500–600 °C and as other scientist have confirmed, there are almost no phase changes of Al2O3 coatings in the range of 500–700 °C [7,14]. To decrease the manufacturing cost even further, unlike in the work of other authors [4–10], where coatings were formed mostly using Ar as primary and H2 as secondary gas, in this research air was used as primary and H2 as secondary gas. Because of this change it was necessary to verify if the coatings properties would stay the same. Another reason is for using alumina-based coatings, when burning straw potassium, sodium and chlorine compounds formed during combustion react with a furnace surface and destroy it via corrosion reactions. The supreme process of destruction is an etching of the metal due to the chromium reaction with chlorine produced during burning of straws. So, the deposition of the ceramic coatings on the metal surfaces could reduce the production cost and greatly improves the lifetime of the furnace, without the need to replace the grate. Because of the reasons mentioned above, alongside pure alumina, titania was selected as a suitable additive for coatings in our work. In this paper, Al2O3 and Al2O3-13wt.% TiO2 coatings were deposited on steel (P265GH) samples. The aim of this research is to investigate the chemical and thermal resistance of alumina coatings with addition of TiO2. Methodology: 2b) and filled with straw pellets (as shown in Fig. 2a). The cylinder had holes on each side and air was flowing through the bottom one at a constant rate. Then the cylinder was closed and inserted into the heating furnace (model TMH12/38/500–2416) and the temperature was set at 500 °C. The duration of the measurement was 60 minutes after the temperature in the furnace reached 500 oC. Additionally, the temperature inside the cylinder was measured using three thermocouples, placed inside the metal cylinder. The first thermocouple was placed below the samples, while the second and third were placed on the top of Al2O3 and Al2O3-13%wt.TiO2 coatings, respectively. The variation of the temperatures was measured using digital multimeter (Mitutoyo Surftest SJ-210). After the heating was over, the specimens were left in the furnace to cool down for several hours. This sequence completed one cycle of heat treatment and when the five cycles were reached the surface morphology, phase and elemental composition were investigated. In order to imitate the burning process of the straw pellets in the heating furnace the metallic tube was made (Fig. 2). The specimens were heat treated under identical conditions and were as following: the samples were placed into a metal cylinder (Fig. 2b) and filled with straw pellets (as shown in Fig. 2a). The cylinder had holes on each side and air was flowing through the bottom one at a constant rate. Then the cylinder was closed and inserted into the heating furnace (model TMH12/38/500–2416) and the temperature was set at 500 °C. The duration of the measurement was 60 minutes after the temperature in the furnace reached 500 oC. Additionally, the temperature inside the cylinder was measured using three thermocouples, placed inside the metal cylinder. The first thermocouple was placed below the samples, while the second and third were placed on the top of Al2O3 and Al2O3-13%wt.TiO2 coatings, respectively. In order to imitate the burning process of the straw pellets in the heating furnace the metallic tube was made (Fig. 2). The specimens were heat treated under identical conditions and were as following: the samples were placed into a metal cylinder (Fig. 2b) and filled with straw pellets (as shown in Fig. 2a). The cylinder had holes on each side and air was flowing through the bottom one at a constant rate. Methodology: Al2O3 and Al2O3-13wt.TiO2 coatings were sprayed on steel (P265GH) substrate at atmospheric pressure using a direct current plasma torch, developed in Lithuanian Energy Institute [15]. The steel substrates (dimensions of 40 × 10 × 6 mm) were polished and chemically cleaned before the deposition process. The substrates were placed on the water-cooled sample holder. The plasma torch was moving in the x-axis direction back and forth, in order to prevent overheating. Air was used as primary and the powder carrier gas, and hydrogen was used as a secondary gas. The coatings were deposited using the plasma torch parameters, that are listed in the Table 1. Page 3/13 Current 200 A Voltage 198 V Spray distance 70 mm Primary gas flow rate 3.7 g/s Powder carrier gas flow rate 0.75 g/s Secondary gas flow rate 0.06 g/s Deposition duration 60 s Two different compositions of powders were used in this research. Conventional 63–81 µm size (Fig. 1) Al2O3 (MOGUL PC15, purity 99.8) and the same size distribution Al2O3 – 13 wt.% TiO2 (MOGUL PC12) powders were used as feedstock material, which were injected into the reactor nozzle (internal diameter of 7 mm) at 150 mm from the exit. Additionally, the bonding layer was formed from nickel-chromium powders (MOGUL M3, Ni/Cr ratio of 80/20) powders in order to increase the coating adhesion to steel substrate. Before the coating process, all powders were dried. Three different specimens were used in the research: uncoated steel (P265GH), Al2O3 and Al2O3-13%wt.TiO2 coatings. Two different compositions of powders were used in this research. Conventional 63–81 µm size (Fig. 1) Al2O3 (MOGUL PC15, purity 99.8) and the same size distribution Al2O3 – 13 wt.% TiO2 (MOGUL PC12) powders were used as feedstock material, which were injected into the reactor nozzle (internal diameter of 7 mm) at 150 mm from the exit. Additionally, the bonding layer was formed from nickel-chromium powders (MOGUL M3, Ni/Cr ratio of 80/20) powders in order to increase the coating adhesion to steel substrate. Before the coating process, all powders were dried. Three different specimens were used in the research: uncoated steel (P265GH), Al2O3 and Al2O3-13%wt.TiO2 coatings. In order to imitate the burning process of the straw pellets in the heating furnace the metallic tube was made (Fig. 2). The specimens were heat treated under identical conditions and were as following: the samples were placed into a metal cylinder (Fig. Results: Variation of the temperature during the burning of straw pellets is shown in Fig. 3. The temperature of the first thermocouple (below the samples) is quite lower, that is due to the constant air flow from the bottom of the cylinder, and the maximum temperature is considerably lower (450–480 °C), because the burning of straw pellets is happening on top of the samples. The rapid growth of temperature can be observed at 17 min., that is when the straw pellets start to burn, the peak temperature is reached after 25 min. and is between 520–600 °C. Meanwhile, the temperature on the surface was about 100 °C higher than on the bottom when the straw pellets were burning. Once the burning process is over the temperature starts to decrease and stabilizes at 430–470 °C range. The temperature obtained on the top surface of the coatings was about ~ 50 oC higher after burning process was finished. Surface morphology of uncoated steel before the heat treatment and after 20 cycles is shown in Fig. 4. The surface of untreated sample (Fig. 4a) is relatively clean, although some particles can be observed. After 20 heat treatment cycles the chance of the surface is evident. A lot more particles can be observed on the surface, that remained even after ultrasonic cleaning, these are mainly leftover products of burned straw pellets. In addition, the surface seems to be etched, this is the result of ongoing chemical reactions and oxidation during the burning of straw pellets. The SEM images clearly indicated that the surface of P265GH steel was changes and damaged. Surface of sample with the Al2O3 coating is shown in Fig. 5. The surface of untreated sample (Fig. 5a) consists of splats and somewhat unmolten particles, also no defects, such as cracks or delamination, can be seen. Surface of the sample after 20 cycles (Fig. 5b) is similar, but the quantity of small particles is considerably higher. Much like in steel sample, a lot of those small particles are leftover products of burned straw pellets, that were not cleaned completely. Just as in steel sample, the surface is etched and cavities do form, but they are considerably smaller. This is due to the fact, that Al2O3 coating is a lot more resistant to chemical reactions, happening during the burning process of straw pellets, that steel. The surface of Al2O3-13%wt.TiO2 sample is shown in Fig. 6. Methodology: Then the cylinder was closed and inserted into the heating furnace (model TMH12/38/500–2416) and the temperature was set at 500 °C. The duration of the measurement was 60 minutes after the temperature in the furnace reached 500 oC. Additionally, the temperature inside the cylinder was measured using three thermocouples, placed inside the metal cylinder. The first thermocouple was placed below the samples, while the second and third were placed on the top of Al2O3 and Al2O3-13%wt.TiO2 coatings, respectively. The variation of the temperatures was measured using digital multimeter (Mitutoyo Surftest SJ-210). After the heating was over, the specimens were left in the furnace to cool down for several hours. This sequence completed one cycle of heat treatment and when the five cycles were reached the surface morphology, phase and elemental composition were investigated. The specimens were washed using ultrasonic cleaning after each five cycles before the characterization. The surface morphology of the coatings was characterized by scanning electron microscopy (SEM) using a Hitachi S-3400N, elemental composition by energy dispersive X-ray spectroscopy (EDS) using dispersive X-ray spectroscopy (Bruker Quad 5040 spectrometer, AXS Microanalysis GmbH). The measurements were performed from 1.05 mm2 surface area at 4 different points for each sample and the mean values were calculated. Structure of the coatings was analyzed using X-ray diffraction crystallography (XRD) with Bruker D8 instrument whose main parameters were theta-theta configuration, Page 4/13 Page 4/13 CuKα (λ = 0.154059 nm) radiation. The five cycles procedure was repeated, and the testing was finished when 20 cycles of heat treatment were reached. CuKα (λ = 0.154059 nm) radiation. The five cycles procedure was repeated, and the testing was finished when 20 cycles of heat treatment were reached. Results: As in previous sample, untreated coating (Fig. 6a) consists of splats and somewhat unmolten particles, also no cracks or delamination zones can be seen. But unlike in Al2O3 sample the surface after 20 cycles is relatively the same, roughly the same number of small particles and no formed cavities are observed (Fig. 6b). The addition of titania increased coatings resistance to chemical and thermal impact. Elemental composition was determined using energy dispersive X-ray spectroscopy. Samples were tested before treatment, and after 5 and 20 cycles. Each sample was measured in 4 different spots, then mean values were calculated (deviation 1%). The first sample (uncoated steel) consisted mainly of iron (93 at.%), also a small amount of oxygen (4 at.%) and other materials, that came from elemental Page 5/13 Page 5/13 composition of steel, were found. Even after 5 cycles the amount of oxygen dramatically increased to 46%, while iron lowered to 50%, and further increase of cycles had no effect to the amount of oxygen in the sample. This is due to absorbed oxygen during the combustion process, during the burning of straw pellets. Additionally, low traces of silicon, potassium, calcium and sulfur are found, that are leftover products of combustion reactions and were not completely removed with ultrasonic cleaning. Al2O3 sample before treatment consisted of aluminum (31 at.%), oxygen (58 at.%), nickel (6 at.%) and chromium (2 at.%). Also, low amounts of other elements can also be found (carbon, iron, silicon etc.) and are attributed to impurities within the sample. Chromium and nickel originate from NiCr underlayer, that was applied before the coating in order to increase the adhesion of the coating. After 20 cycles aluminum decreased by 4% (to 27 at.%) and oxygen by 1% (to 57 at.%), nickel increased by 2% (to 8 at.%) and chromium remained the same (2 at.%). This happened due to the increase of other elements (iron, carbon, silicon etc.) during burning of straw pellets, and because the combustion products could not be completely removed. Al2O3-13wt.% TiO2 sample consisted of aluminum (22 at.%), oxygen (57 at.%), titanium (4 at.%), nickel (9 at.%) and chromium (3 at.%). Much like in the Al2O3 sample, nickel and chromium are attributed to NiCr underlayer and small traces of other elements (carbon, iron, silicon etc.) are found due to impurities of the sample. Results: However, after 20 cycles the amount of aluminum, titanium and chromium remained the same (22 at.%, 4 at.% and 3 at.% respectively), while oxygen increased by 2% (to 59 at.%) and nickel decreased by 1% (to 8 at.%). Also, just as in Al2O3 coating, number of other elements coming from combustion products increased. The data indicates that addition of titania did increase stability in elemental composition, since even after 20 heat treatment cycles only amount of oxygen altered by more than 1%. Phase composition of the P265GH steel sample is shown in Fig. 7. Before the treatment there is only 2 peaks of Fe at 2θ = 44.7° and 65.1°, but even after 5 cycles several peaks of Fe3O4 appear and become the dominant phase. The further increase of heating cycles continues to increase the intensity of Fe3O4 peaks. These results support the previous statement, that steel sample was heavily damaged during the treatment cycles and protection is necessary under these working conditions. There is noticeably less phase composition change in Al2O3 sample. Firstly, the dominant phase remains α-Al2O3 with peaks when 2θ = 25.6°, 35.7°, 43.5°, 57.7° and 63.1°. Secondly, another phase of alumina also apparent, that is γ-Al2O3 at 2θ = 37.4°, 38.9°, 45.8° and 67.1°. The ratio of α-Al2O3 and γ-Al2O3 most intense peaks before treatment is 1.16 but increases to 2.06 after 20 cycles. This is since α-Al2O3 peak intensity increased, while γ-Al2O3 remained similar. This happens because the temperature in not high enough for phase transition reactions from α-Al2O3 to γ-Al2O3 to occur. The temperature is only enough to initiate transition of amorphous Al2O3 to α-Al2O3. Similar results were found by Dhakar et al [16] where Al2O3 coating were heat treated at 900 °C. Thirdly, besides alumina there are two more peaks that are attributed to iron and nickel at 2θ = 44.3°and 51.6°, their intensities remained the same before and after 20 heat cycles. Page 6/13 Addition of titania in the final sample increased stability of phase composition of alumina coating, therefore even less changes can be observed (Fig. 9). There is a less peaks of α-Al2O3: when 2θ = 35.7°, 43.5° and 63.1°. Also, there are a few γ-Al2O3 peaks at 2θ = 37.4°, 45.8° and 67.1°. The ratio of α-Al2O3 to γ-Al2O3 the most intense peaks of the coating was 1.46. Results: Meanwhile after 20 cycles of treatment the α- Al2O3/γ-Al2O3 ratio increased very slightly up to 1.51, which confirms, that the resistance to heat treatment increased with the addition of titania. Just as in α-Al2O3 sample there are two peaks of at 2θ = 44.3°and 51.6° attributed to the bonding NiCr layer. These results agree with the work of other authors [17], where no significant changes were observed for γ-Al2O3 phase during the annealing of up to 700 °C of plasma sprayed NiCrAl/Al2O3-13wt.%TiO2 coatings. The authors obtained only minor changes in the peak’s intensities of alpha phase. Conclusion: Al2O3, Al2O3-3% TiO2 and Al2O3-13% TiO2 coatings were formed using atmospheric plasma spray technology. Surface morphology was examined using scanning electron microscopy, and the result show that most damage was done to the uncoated steel sample, while the samples with both, alumina and alumina-titania coatings were almost intact. Elemental composition results done by energy dispersive X- ray spectroscopy indicate, that the steel sample was heavily oxidized after 5 heating cycles, since oxygen increased from 1–46%. The change of elemental composition in other samples was insignificant. In Al2O3 coating oxygen decreased from 58–57%, and aluminium decreased from 31–27%. In the Al2O3- 13% TiO2 coating amount of oxygen increased from 57–59%, but aluminium and titanium fraction remained the same. It was determined that after 5 cycles of treatment the uncoated steel substrate is heavily oxidized and Fe2O3 becomes the dominant phase. The X-ray diffraction results swho that even after 20 cycles of heat treatment at 500 °C, the phase composition of Al2O3 coating greatly increased the thermal and chemical resistance of the P256GH substrate and the addition of titania increased its phase stability even further. 1. Wang S, Wang Y, Wang M, Jiang S, Xiao TD, Strutt PR, “Abrasive wear characteristics of plasma sprayed nanostructured alumina/titania coatings.” Wear, 237 (2) 176-185 (2000) 2. Kang JJ, B.S.Xu, Wang CB, Wang HD, “Influence of spraying parameters on the microstructure and properties of plasma sprayed Al2O3/40%TiO2 coating.” Physics Procedia, 50, 269-176 (2013) 3. Zavareh MA, Sarhan, Ahmed Aly Diaa Mohammed, Razak BBA, Basirun WJ, “Plasma thermal spray of ceramic oxide coating on carbon steel with enhanced wear and corrosion resistance for oil and gas applications.” Ceramics International, 40 (9), 14267-14277 (2014) 4. Sun-Hun Y, “Comparative study on wear performance of traditional and nanostructured Al2O3-13 wt.% TiO2 air plasma spray coatings.” Ceramics – Silikaty, 59 (1), 59-63 (2015) References 1. Wang S, Wang Y, Wang M, Jiang S, Xiao TD, Strutt PR, “Abrasive wear characteristics of plasma sprayed nanostructured alumina/titania coatings.” Wear, 237 (2) 176-185 (2000) 2. Kang JJ, B.S.Xu, Wang CB, Wang HD, “Influence of spraying parameters on the microstructure and properties of plasma sprayed Al2O3/40%TiO2 coating.” Physics Procedia, 50, 269-176 (2013) 2. Kang JJ, B.S.Xu, Wang CB, Wang HD, “Influence of spraying parameters on the microstructure and properties of plasma sprayed Al2O3/40%TiO2 coating.” Physics Procedia, 50, 269-176 (2013) 3. Zavareh MA, Sarhan, Ahmed Aly Diaa Mohammed, Razak BBA, Basirun WJ, “Plasma thermal spray of ceramic oxide coating on carbon steel with enhanced wear and corrosion resistance for oil and gas applications.” Ceramics International, 40 (9), 14267-14277 (2014) 3. Zavareh MA, Sarhan, Ahmed Aly Diaa Mohammed, Razak BBA, Basirun WJ, “Plasma thermal spray of ceramic oxide coating on carbon steel with enhanced wear and corrosion resistance for oil and gas applications.” Ceramics International, 40 (9), 14267-14277 (2014) 4. Sun-Hun Y, “Comparative study on wear performance of traditional and nanostructured Al2O3-13 wt.% TiO2 air plasma spray coatings.” Ceramics – Silikaty, 59 (1), 59-63 (2015) 4. Sun-Hun Y, “Comparative study on wear performance of traditional and nanostructured Al2O3-13 wt.% TiO2 air plasma spray coatings.” Ceramics – Silikaty, 59 (1), 59-63 (2015) Page 7/13 Page 7/13 5. Sreekumar Rajesh T, Venkata Rao R, “Experimental investigation and parameter optimization of Al2O3-40% TiO2 atmospheric plasma spray coating on SS316 steel substrate.” Materials Today: Proceedings, 5 (2), 5012-5020 (2018) 6. Li J, Li H, Ke Z, Xue L, Yan Y, “An effective low-temperature strategy for sealing plasma sprayed Al2O3-based coatings.” Journal of the European Ceramic Society, 38 (4), 1871-1877 (2018) 7. Jia S, Zou Y, Xu J, Wang J, Yu L, “Effect of TiO2 content on properties of Al2O3 thermal barrier coatings by plasma spraying.” Transactions of Nonferrous Metals Society of China, 25 (1), 175-183 (2015) 8. Sahab ARM, Saad NH, Kasolang S, Saedon J, “Impact of plasma spray variables parameters on mechanical and wear behaviour of plasma sprayed Al2O3 3%wt TiO2 coating in abrasion and erosion application.” Procedia Engineering, 41, 1689-1695 (2012) 9. Yugeswaran S, Selvarajan V, Vijay M, Ananthapadmanabhan PV, Sreekumar KP, “Influence of critical plasma spraying parameter (CPSP) on plasma sprayed Alumina–Titania composite coatings.” Ceramics International, 36 (1), 141-149 (2010) 9. References Yugeswaran S, Selvarajan V, Vijay M, Ananthapadmanabhan PV, Sreekumar KP, “Influence of critical plasma spraying parameter (CPSP) on plasma sprayed Alumina–Titania composite coatings.” Ceramics International, 36 (1), 141-149 (2010) 10. Dejang N, Watcharapasorn A, Wirojupatump S, Niranatlumpong P, Jiansirisomboon S, “Fabrication and properties of plasma-sprayed Al2O3/TiO2 composite coatings: A role of nano-sized TiO2 addition.” Surface and Coatings Technology, 204 (9), 1651-1657 (2010) 10. Dejang N, Watcharapasorn A, Wirojupatump S, Niranatlumpong P, Jiansirisomboon S, “Fabrication and properties of plasma-sprayed Al2O3/TiO2 composite coatings: A role of nano-sized TiO2 addition.” Surface and Coatings Technology, 204 (9), 1651-1657 (2010) 11. Fervel V, Normand B, Coddet C, “Tribological behavior of plasma sprayed Al2O3-based cermet coatings.” Wear, 230(1), 70-77 (1999) 11. Fervel V, Normand B, Coddet C, “Tribological behavior of plasma sprayed Al2O3-based cermet coatings.” Wear, 230(1), 70-77 (1999) 12. Çelik İ, “Structure and surface properties of Al2O3–TiO2 ceramic coated AZ31 magnesium alloy.” Ceramics International, 42 (12), 13659-13663 (2016) 12. Çelik İ, “Structure and surface properties of Al2O3–TiO2 ceramic coated AZ31 magnesium alloy.” Ceramics International, 42 (12), 13659-13663 (2016) 13. Klyatskina E, Rayón E, Darut G, Salvador MD, Sánchez E, Montavon G, “A study of the influence of TiO2 addition in Al2O3 coatings sprayed by suspension plasma spray.” Surface and Coatings Technology, 278, 25-29, (2015) 13. Klyatskina E, Rayón E, Darut G, Salvador MD, Sánchez E, Montavon G, “A study of the influence of TiO2 addition in Al2O3 coatings sprayed by suspension plasma spray.” Surface and Coatings Technology, 278, 25-29, (2015) 14. Misra V, Chakravarthy Y, Khare N, Singh K, Ghorui S, “Strongly adherent Al2O3 coating on SS 316L: Optimization of plasma spray parameters and investigation of unique wear resistance behaviour under air and nitrogen environment.” Ceramics International, 2019, 14. Misra V, Chakravarthy Y, Khare N, Singh K, Ghorui S, “Strongly adherent Al2O3 coating on SS 316L: Optimization of plasma spray parameters and investigation of unique wear resistance behaviour under air and nitrogen environment.” Ceramics International, 2019, 15. Zhang K, Ning S, Ren C, Luo X, Dong L, Chen H, He Y, “Improvement of oxidation resistance of spatial- network Al2O3/YSZ composite coatings by chemical densification.” Surface & Coatings Technology, 266, 105-112 (2015) 16. Wang M, Zhou Z, Yi Y, Wang Z, “Thermal-shock resistance of plasma sprayed Al2O3-13wt%TiO2 coating evaluated by residual strength method.” Surface & Coatings Technology, 375, 888-893 (2019) 17. References Matthews S, Taliana F, James B, “Heat treatment of plasma sprayed Al2O3 and Al2O3-WO3 coatings between 500 and 1000 °C.” Surface and coatings technology, 212, 109-118 (2012) 18. Valinčius V, Valatkevičius P, Valinčiūtė V, Marcinauskas L, “Electric and thermal characteristics of the linear, sectional DC plasma generator.” Plasma Sources Science and Technology, 13 (2), 199-206 (2004) Page 8/13 Page 8/13 19. Dhakar B, Namdeo A, Chatterejee S, Sabiruddin K, “Heat treatment of plasma sprayed alumina- chromia composite coatings.” Surface Engineering, 34:10, 737-746 (2018) 20. Zhang J, Wang Z, Li M Wang Z, Li T, Wang Z, “Adhesion properties and fracture mechanism of plasma sprayed NiCrAl/Al2O3-13w.%TiO2 coatings by post annealing.” Journal of Adhesion Science and Technology, 29 (4), 256-270 (2015) 19. Dhakar B, Namdeo A, Chatterejee S, Sabiruddin K, “Heat treatment of plasma sprayed alumina- chromia composite coatings.” Surface Engineering, 34:10, 737-746 (2018) 20. Zhang J, Wang Z, Li M Wang Z, Li T, Wang Z, “Adhesion properties and fracture mechanism of 19. Dhakar B, Namdeo A, Chatterejee S, Sabiruddin K, “Heat treatment of plasma sprayed alumina- chromia composite coatings.” Surface Engineering, 34:10, 737-746 (2018) 20. Zhang J, Wang Z, Li M Wang Z, Li T, Wang Z, “Adhesion properties and fracture mechanism of plasma sprayed NiCrAl/Al2O3-13w.%TiO2 coatings by post annealing.” Journal of Adhesion Science and Technology, 29 (4), 256-270 (2015) 20. Zhang J, Wang Z, Li M Wang Z, Li T, Wang Z, “Adhesion properties and fracture mechanism of plasma sprayed NiCrAl/Al2O3-13w.%TiO2 coatings by post annealing.” Journal of Adhesion Scien and Technology, 29 (4), 256-270 (2015) 19. Dhakar B, Namdeo A, Chatterejee S, Sabiruddin K, “Heat treatment of plasma sprayed alumina- chromia composite coatings.” Surface Engineering, 34:10, 737-746 (2018) 20. Zhang J, Wang Z, Li M Wang Z, Li T, Wang Z, “Adhesion properties and fracture mechanism of plasma sprayed NiCrAl/Al2O3-13w.%TiO2 coatings by post annealing.” Journal of Adhesion Science and Technology, 29 (4), 256-270 (2015) Figure 1 Figure 1 Page 9/13 SEM micrograph of Al2O3 powders Page 9/13 s SEM micrograph of Al2O3 powders SEM micrograph of Al2O3 powders Page 9/13 Figure 2 Figure 2 a) samples in tube with grain pellets, b) metal cylinder with the samples, c) experimental setup. a) samples in tube with grain pellets, b) metal cylinder with the samples, c) experimental setup. Figure 3 Figure 3 Temperature graph measured by thermocouples in metal cylinder Temperature graph measured by thermocouples in metal cylinder Temperature graph measured by thermocouples in metal cylinder Page 10/13 Figure 4 SEM micrographs of steel sample, before a) and after b) 20 heating cycles Figure 5 SEM micrographs of Al2O3 sample, before a) and after b) 20 heating cycles Figure 4 SEM micrographs of steel sample, before a) and after b) 20 heating cycles SEM micrographs of steel sample, before a) and after b) 20 heating cycles Figure 5 SEM micrographs of Al2O3 sample, before a) and after b) 20 heating cycles Figure 5 SEM micrographs of Al2O3 sample, before a) and after b) 20 heating cycles Page 11/13 Figure 6 SEM micrographs of Al2O3-13%TiO2 sample, before a) and after b) 20 heating cycles Figure 6 SEM micrographs of Al2O3-13%TiO2 sample, before a) and after b) 20 heating cycles Figure 6 SEM micrographs of Al2O3-13%TiO2 sample, before a) and after b) 20 heating cycles Figure 6 SEM micrographs of Al2O3-13%TiO2 sample, before a) and after b) 20 heating cycles EM micrographs of Al2O3-13%TiO2 sample, before a) and after b) 20 heatin Page 12/13 g p p , ) ) g y X-ray diffraction pattern before and after 5 and 20 treatment cycles of steel sample. Fig 8. Al2O3 coating Fig. 9 Al2O3 – 13%wt.%TiO2 coating Figure 7 X-ray diffraction pattern before and after 5 and 20 treatment cycles of steel sample. Fig 8. Al2O3 coating Fig. 9 Al2O3 – 13%wt.%TiO2 coating Page 13/13
https://openalex.org/W4238240331	https://www.nature.com/articles/s41598-021-92879-5.pdf	English	null	miRNA-1246 in Extracellular Vesicles Secreted from Metastatic Tumor Induces Drug Resistance in Tumor Endothelial Cells	Research Square (Research Square)	2,020	cc-by	12,090	miRNA‑1246 in extracellular vesicles secreted from metastatic tumor induces drug resistance in tumor endothelial cells Chisaho Torii1,2,3,9, Nako Maishi1,2,4,9, Taisuke Kawamoto1,9, Masahiro Morimoto2,4,5, Kosuke Akiyama1, Yusuke Yoshioka6, Takashi Minami7, Takuya Tsumita4, Mohammad Towfik Alam1,2,4, Takahiro Ochiya6, Yasuhiro Hida8 & Kyoko Hida1,2,4* Chisaho Torii1,2,3,9, Nako Maishi1,2,4,9, Taisuke Kawamoto1,9, Masahiro Morimoto2,4,5, Kosuke Akiyama1, Yusuke Yoshioka6, Takashi Minami7, Takuya Tsumita4, Mohammad Towfik Alam1,2,4, Takahiro Ochiya6, Yasuhiro Hida8 & Kyoko Hida1,2,4* Tumor endothelial cells (TECs) reportedly exhibit altered phenotypes. We have demonstrated that TECs acquire drug resistance with the upregulation of P-glycoprotein (P-gp, ABCB1), contrary to traditional assumptions. Furthermore, P-gp expression was higher in TECs of highly metastatic tumors than in those of low metastatic tumors. However, the detailed mechanism of differential P-gp expression in TECs remains unclear. miRNA was identified in highly metastatic tumor extracellular vesicles (EVs) and the roles of miRNA in endothelial cell resistance were analyzed in vitro and in vivo. In the present study, we found that treatment of highly metastatic tumor-conditioned medium induced resistance to 5-fluorouracil (5-FU) with interleukin-6 (IL-6) upregulation in endothelial cells (ECs). Among the soluble factors secreted from highly metastatic tumors, we focused on EVs and determined that miR-1246 was contained at a higher level in highly metastatic tumor EVs than in low metastatic tumor EVs. Furthermore, miR-1246 was transported via the EVs into ECs and induced IL-6 expression. Upregulated IL-6 induced resistance to 5-FU with STAT3 and Akt activation in ECs in an autocrine manner. These results suggested that highly metastatic tumors induce drug resistance in ECs by transporting miR-1246 through EVs. Abbreviations TECs Tumor endothelial cells P-gp P-glycoprotein ABCB1 ATP-binding cassette sub-family B member 1 EVs Extracellular vesicles 5-FU 5-Fluorouracil IL-6 Interleukin-6 ECs Endothelial cells VEGF Vascular endothelial growth factor NECs Normal endothelial cells miRNA Micro RNA HMVECs Human dermis microvascular endothelial cells MEM Minimum essential medium FBS Fetal bovine serum 1Department of Vascular Biology, Hokkaido University Graduate School of Dental Medicine, Sapporo 060‑8586, Japan. 2Vascular Biology, Frontier Research Unit, Institute for Genetic Medicine, Hokkaido University, Sapporo 060‑0815, Japan. 3Department of Oral and Maxillofacial Surgery, Hokkaido University Graduate School of Dental Medicine, Sapporo 060‑8586, Japan. 4Department of Vascular Biology and Molecular Pathology, Hokkaido University Graduate School of Dental Medicine, Sapporo 060‑8586, Japan. 5Department of Oral Diagnosis and Medicine, Hokkaido University Graduate School of Dental Medicine, Sapporo 060‑8586, Japan. 6Institute of Medical Science, Tokyo Medical University, Tokyo 160‑0023, Japan. 7Division of Molecular and Vascular Biology, Institute of Resource Development and Analysis, Kumamoto University, Kumamoto 860‑0811, Japan. www.nature.com/scientificreports www.nature.com/scientificreports 8Dep of Cardiovascular and Thoracic Surgery, Hokkaido University Faculty of Medicine, Sapporo 060‑8638, Japa authors contributed equally: Chisaho Torii, Nako Maishi and Taisuke Kawamoto. email: khida@den.hoku ts \| (2021) 11:13502 \| https://doi.org/10.1038/s41598-021-92879-5 miRNA‑1246 in extracellular vesicles secreted from metastatic tumor induces drug resistance in tumor endothelial cells Chisaho Torii1,2,3,9, Nako Maishi1,2,4,9, Taisuke Kawamoto1,9, Masahiro Morimoto2,4,5, Kosuke Akiyama1, Yusuke Yoshioka6, Takashi Minami7, Takuya Tsumita4, Mohammad Towfik Alam1,2,4, Takahiro Ochiya6, Yasuhiro Hida8 & Kyoko Hida1,2,4 Tumor endothelial cells (TECs) reportedly exhibit altered phenotypes. We have demonstrated t TECs acquire drug resistance with the upregulation of P-glycoprotein (P-gp, ABCB1), contrary t traditional assumptions. Furthermore, P-gp expression was higher in TECs of highly metastatic tumors than in those of low metastatic tumors. However, the detailed mechanism of differenti expression in TECs remains unclear. miRNA was identified in highly metastatic tumor extracellu vesicles (EVs) and the roles of miRNA in endothelial cell resistance were analyzed in vitro and in In the present study, we found that treatment of highly metastatic tumor-conditioned medium induced resistance to 5-fluorouracil (5-FU) with interleukin-6 (IL-6) upregulation in endothelial (ECs). Among the soluble factors secreted from highly metastatic tumors, we focused on EVs a determined that miR-1246 was contained at a higher level in highly metastatic tumor EVs than metastatic tumor EVs. Furthermore, miR-1246 was transported via the EVs into ECs and induce expression. Upregulated IL-6 induced resistance to 5-FU with STAT3 and Akt activation in ECs i autocrine manner. These results suggested that highly metastatic tumors induce drug resistanc ECs by transporting miR-1246 through EVs. Abbreviations TECs Tumor endothelial cells P-gp P-glycoprotein ABCB1 ATP-binding cassette sub-family B member 1 EVs Extracellular vesicles 5-FU 5-Fluorouracil IL-6 Interleukin-6 ECs Endothelial cells VEGF Vascular endothelial growth factor NECs Normal endothelial cells miRNA Micro RNA HMVECs Human dermis microvascular endothelial cells MEM Minimum essential medium FBS Fetal bovine serum OPEN 1Department of Vascular Biology, Hokkaido University Graduate School of Dental Medicine, Sapporo 0 Japan. 2Vascular Biology, Frontier Research Unit, Institute for Genetic Medicine, Hokkaido U Sapporo 060‑0815, Japan. 3Department of Oral and Maxillofacial Surgery, Hokkaido University Graduate S Dental Medicine, Sapporo 060‑8586, Japan. 4Department of Vascular Biology and Molecular Pathology, H University Graduate School of Dental Medicine, Sapporo 060‑8586, Japan. 5Department of Oral Diagn Medicine, Hokkaido University Graduate School of Dental Medicine, Sapporo 060‑8586, Japan. 6Ins Medical Science, Tokyo Medical University, Tokyo 160‑0023, Japan. 7Division of Molecular and Vascular Institute of Resource Development and Analysis, Kumamoto University, Kumamoto 860‑0811, Japan. miRNA‑1246 in extracellular vesicles secreted from metastatic tumor induces drug resistance in tumor endothelial cells 8Department of Cardiovascular and Thoracic Surgery, Hokkaido University Faculty of Medicine, Sapporo 060‑8638, Japan. 9These authors contributed equally: Chisaho Torii, Nako Maishi and Taisuke Kawamoto. email: khida@den.hokudai.ac.jp \| https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 www.nature.com/scientificreports/ FITC Fluorescein isothiocyanate GAPDH Glyceraldehyde-3-phosphate dehydrogenase AR Androgen receptor UNC5B Unc-5 netrin receptor B CTSC Cathepsin C CM Conditioned medium PBS Phosphate buffered saline BCA Bicinchoninic acid NC Negative control RIPA Radio immunoprecipitation assay HRP Horseradish peroxidase IPA Ingenuity pathway analysis MDR-1 Multidrug resistance 1 Tumor blood vessels contribute to tumor growth by providing nutrition and oxygen1 and play important roles in tumor progression and metastasis. Anti-angiogenic therapy has been proposed as a novel anti-cancer strategy more than decades ago2. Current anti-angiogenic therapies mostly target vascular endothelial growth factor (VEGF) and its receptors, which have prolonged cancer patients’ overall survival. On another note, recent clini- cal studies have shown the limited results of these drugs. These agents sometimes cause adverse effects, such as lethal hemoptysis3,4 and intestinal perforation5–7, because the VEGF pathway is also important for normal endothelial cells (NECs)8, and drug resistance has been reported. Traditionally, it is considered that the advantage of targeting endothelial cells (ECs) rather than tumor cells is that unlike tumor cells, ECs are genetically stable and do not develop drug resistance9,10. However, there have been reports showing that tumor endothelial cells (TECs) themselves may acquire resistance to these drugs11. Indeed, cytogenetically abnormal TECs were also detected in malignant tumors, such as lymphoma12. We also previously reported that TECs show a degree of cytogenetic abnormalities, such as aneuploidy or abnormal centrosomes, in mouse tumors13 and human renal carcinomas14. These abnormalities were not caused by tumor cell contamination13 but may be affected by the tumor microenvironment. These cytogenetic abnormalities are suggestive of the genetic instability of TECs, predicting the possibility that TECs may acquire drug resistance. p g p y y q g Several TECs are resistant to certain drugs15,16. For example, renal carcinoma ECs are resistant to vincristine15, whereas hepatocellular carcinoma ECs are resistant to 5-fluorouracil (5-FU)17. We recently demonstrated that TECs are resistant to paclitaxel with ABCB1 upregulation, which is a drug transporter. We revealed that VEGF secreted from tumor cells induced P-gp expression in ECs18,19. It was also demonstrated that TEC phenotypes, such as drug resistance, chromosomal abnormality, or stemness, are different depending on tumor malignancy20, suggesting that tumor-secreting factors may induce resistance in ECs18. Naito et al. miRNA‑1246 in extracellular vesicles secreted from metastatic tumor induces drug resistance in tumor endothelial cells reported that there are TECs that express P-gp and cause resistance to anti-angiogenic drugs21. Taken together, TECs clearly acquire drug resistance; however, the detailed mechanism of acquiring drug resistance in TECs has remained unknown.f Cancer cells have been known to secrete extracellular vesicles (EVs), including exosomes, and these EVs affect tumor progression and metastasis22–24. EV is a general term for exosomes or microvesicles that vary from 50 to 1000 nm in diameter. These vesicles contain protein, mRNA, or micro RNA (miRNA) inside, and EVs play a role in intercellular communication25. We have reported that tumor-derived EVs enhanced proangiogenic phe- notypes in ECs26. Regarding the effect of EVs on ECs, tumor-derived EVs reportedly enhanced brain metastasis by altering the brain–blood barrier27. Hsu et al. reported that EVs that were secreted from hypoxic lung cancer cells increased angiogenesis and permeability in ECs28. However, the mechanisms of TEC abnormality related to tumor EVs still lack elucidation. Recently, miRNAs have emerged as key regulators of multiple physiologic and pathologic cellular responses, and multiple reports have implicated miRNA in cancer pathogenesis. Additionally, miRNAs represent a class of small (20–25 nucleotides), non-coding RNAs that generally inhibit several target messenger RNAs by repress- ing the translation of deducing mRNA stability29. Furthermore, miRNA is exported to extracellular spaces by EVs or by forming complexes with protein or lipids30,31, and then, extracellular miRNAs contribute to cell–cell communication. However, miRNA function in tumor-derived EVs on ECs in primary tumors remains unclear. In the present study, we identified that miR-1246 was more contained in highly metastatic tumor-derived EVs than in lowly metastatic tumor EVs, and the role of this EV miR-1246 in acquiring drug resistance was elucidated. y g p p p communication. However, miRNA function in tumor-derived EVs on ECs in primary tumors remains unclear. In the present study, we identified that miR-1246 was more contained in highly metastatic tumor-derived EVs than in lowly metastatic tumor EVs, and the role of this EV miR-1246 in acquiring drug resistance was elucidated. p y In the present study, we identified that miR-1246 was more contained in highly metastatic tumor-derived than in lowly metastatic tumor EVs, and the role of this EV miR-1246 in acquiring drug resistance was elucid www.nature.com/scientificreports/ Sigma, St Louis, MO, USA). Human cervical cancer cell HeLa, human oral squamous carcinoma cell HSC-3, human renal clear cell carcinoma cell OS-RC-2, and human colon carcinoma cell HCT119 were purchased from RIKEN Cell Bank (Tsukuba, Japan). Neonatal human foreskin fibroblast BJ was purchased from ATCC. The Hela, HSC-3, HCT119, and BJ cells were cultured in Dulbecco’s modified eagle medium (Sigma) supplemented with 10% FBS. The OS-RC-2 cells were cultured in RPMI1640 medium (Sigma–Aldrich) with 10% FBS. Normal human epidermal melanocytes were purchased from Kurabo Industries Ltd. (Osaka, Japan) and cultured in melanocyte growth medium (PromoCell, Germany).hth y g ( y) The TECs were isolated from tumors that were subcutaneously xenografted with A375 or A375SM. The NECs were isolated from the dermis of tumor-free nude mice, as previously described13. Furthermore, ECs were isolated using a magnetic cell sorter device (Miltenyi Biotec, Bergisch Gladbach, Germany) and a flow cytometer (FACS Aria II; BD Biosciences, San Jose, CA, USA) using an anti-CD31 antibody after removing leukocytes with an anti-CD45 antibody. The CD31-positive cells were sorted and seeded on 1.5% gelatin-coated culture plates with EGM-2MV containing 15% FBS. After EC isolation, diphtheria toxin (500 ng/mL; Calbiochem, San Diego, CA, USA) was added to the TEC subcultures to eliminate the remaining human tumor cells and NEC subcultures for technical consistency. The subcultured ECs were sorted through the second round of purification using fluorescein isothiocyanate (FITC)-conjugated Bandeiraea simplicifolia lectin–isolectin B4 (Vector Laboratories, Burlingame, CA, USA). The cells were cultured at 37 ℃ in a humidified atmosphere containing 5% CO2. RT‑PCR and real‑time PCR. Total RNA was extracted from each EC type using the ReliaPrep RNA Cell Miniprep System (Promega, Madison, WI, USA). First-strand cDNA was synthesized using ReverTra-Plus (Toyobo Co., Osaka, Japan). Real-time RT-PCR was conducted using the KAPA SYBR Fast qPCR Kit (Nippon Genetics, Tokyo, Japan). The cycling conditions were done according to the manufacturer’s instructions, and the CFX Manager (Bio-Rad, Hercules, CA, USA) was used for analysis. The expression levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) levels and were analyzed using the 2−∆∆Ct. www.nature.com/scientificreports/ The prim- ers used were as follows: mouse GAPDH: forward, 5′-TCTGACGTGCCGCCTGGAG-3′; reverse, 5′-TCGCAG GAGACAACCTGGTC-3′, human GAPDH: forward, 5′-ACAGTCAGCCGCATCTTCTT-3′; reverse, 5′-GCC CAATACGACCAAATCC-3′, mouse interleukin-6 (IL-6): forward, 5′-AGCTGGAGTCACAGAAGGAGTGGC -3′; reverse, 5′-GGCATAACGCACTAGGTTTGCCGAG-3′, human IL-6: forward, 5′-TACCCCCAGGAGAAG ATTCC-3′; reverse, 5′-TTTTCTGCCAGTGCCTCTTT-3′, human androgen receptor (AR): forward, 5′-GAC TGCCAGGGACCATGTTTTG-3′; reverse, 5′-GCGCACAGGTACTTCTGTTTCC-3′, human unc-5 netrin receptor B (UNC5B): forward, 5′-ACAAGGCAGAAAGTACCCTCCC-3′; reverse, 5′-GTTCAGGGTCTCCTC ATCCAGG-3′, human cathepsin C (CTSC): forward, 5′-AACTGGCCATGAACAGACGTTGGGG-3′; reverse, 5′-AGCTGCCTTGGAGGTAGGTCACCAG-3′. Tumor‑conditioned medium (CM) preparation. The tumor CM was collected from the culture super- natant of the tumor cells (A375SM or A375 cells) as previously described18. Briefly, the tumor cells (1 × 106 cells) were cultured in 10% FBS MEM for 24 h. The tumor CM was obtained from culture supernatant after passing through a 0.22 μm syringe filter (Millipore, Billerica, MA, USA). The control CM was collected from the culture supernatant from HMVECs in 10% FBS MEM in the same way as tumor CM. The HMVECs were treated with tumor CM or control CM in several periods in each experiment. When the cells were treated for 5 days, the medium was changed to freshly collected CM at the third day. The IL-6 level in each CM was measured using an IL-6 ELISA kit (R&D systems, Q6000B, Minneapolis, MN). DNA microarray. HMVECs were treated with tumor CM or control CM for 1, 8, and 24 h. Total RNA was extracted from each treated cell as described in the previous text. DNA microarray was conducted according to the manufacturer’s instructions (Thermo Fisher, Affymetrix) as previously described33. Data were analyzed according to the MIAME rule. The NCBI accession number for the array data reported in this paper is GSE49426. Preparation and characterization of EVs from tumor cell culture supernatant. A375 and A375SM cells (3 × 106 cells) were cultured on a 15 cm-cell culture dish for 48 h in 20 mL of MEM supplemented with 10% EV-depleted FBS. EV-depleted FBS was prepared by ultracentrifugation at 110,000×g using Beckman 45Ti rotor for 70 min at 4℃. HMVECs were cultured for 48 h in EBM2 supplemented with 5% EV-depleted FBS. The CM was centrifuged at 2000×g for 10 min at 4 ℃, and the supernatant was filtered through a 0.22 μm filter unit (Thermo Scientific Nalgene) to thoroughly remove the cellular debris. Methods Mi S Mice. Six week-old female nude mice (BALB/c AJcl-nu/nu, Clea, Japan) were housed under specific-path- ogen-free conditions. All procedures for animal care and experimentation adhered to institutional guidelines with the approval of the Hokkaido University Animal Committee and were carried out in compliance with the ARRIVE guidelines. Cell lines and culture conditions. Human dermis microvascular endothelial cells (HMVECs) were obtained from Lonza (Lonza, Basel, Switzerland) and cultured in EC growth medium for microvascular cells (EGM-2MV) (Lonza). A375 cells were obtained from American Type Culture Collection (ATCC) (Manas- sas, VA). A375SM cells (super metastatic human melanoma cells) were kindly provided by Dr. Isaiah J. Fidler, M.D. (Anderson Cancer Center, Houston, TX)32. A375 and A375SM cells were cultured in minimum essential medium (MEM) (Gibco, Grand Island, NY) supplemented with 10% heat-inactivated fetal bovine serum (FBS, https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ www.nature.com/scientificreports/ Then, to prepare EVs, the superna- tant was ultracentrifuged (110,000×g, 70 min at 4 ℃) using Beckman 45Ti rotor, the pellet was washed twice with phosphate buffered saline (PBS) by ultracentrifugation at 110,000×g using Beckman 45Ti rotor for 70 min at 4 ℃, and EVs were eluted in 200 μL of PBS. The Micro bicinchoninic acid (BCA) Protein Assay Kit (Thermo Sci- entific, MA, USA) was used to measure the protein concentration. The isolated EVs were characterized through nanoparticle tracking analysis (Nanosight: Quantum Design, Japan) and western blot procedure with anti-CD9 (1:500, Santa Cruz, #59140), anti-CD63 (1:500, Biolegend, #353013), anti-Cytochrome C (1:500, BD Pharmin- gen, #556433) antibodies, and anti-Mouse Anti-Hsp70 (1:1000, BD Pharmingen, #610607). Scientific Reports \| (2021) 11:13502 \| https://doi.org/10.1038/s41598-021-92879-5 www.nature.com/scientificreports/ Nanoparticle tracking analysis (NTA). All NTA measurements were performed using the NS300 unit (Malvern) equipped with a 488 nm laser and the camera sCOMOS. All samples were diluted to provide a con- centration of 1 × 109 particles/mL counted using NTA. For each run about 1 mL of the sample was injected into the chamber with a sterile 1-mL syringe controlled via an external syringe pump (syringe pump speed/AU:30). NTA 3.4 software was used for all analysis, using standard settings. In the analysis the mode size (main peak), mean size and its standard derivation values were obtained. All counts were performed in replicates of 5 for each sample, collecting 30 s videos. Observation of isolated EVs. The observation of EVs was based on the method reported by Yamashita et al.35. EVs were fixed with 4% paraformaldehyde (PFA) for 30 min and then placed on a grid (PVF-C15 STEM Cu150 Grid, Okenshoji, Tokyo, Japan) for 20 min. The grids were treated with 1% glutaraldehyde for 2 min and washed 8 times with dH2O. They were stained with 1% uranyl acetate for 10 min. EVs were visualized with a transmission electron microscope (JEM-1400, JEOL, Tokyo, Japan). miRNA array. For CM-derived miRNA analysis, the CM collected by culturing normal cells (BJs, melano- cytes, and HMVECs) and tumor cells (A375, A375SM, HSC-3, OS-RC-2, and Hela) with serum-free medium (1 × 105 cells, 1 mL/9 cm2) for 48 h were used. The CM was centrifuged at 200×g for 5 min, and then, the super- natant was further centrifuged at 2000×g for 15 min at 4 ℃, and after centrifugation, 200 μL of the supernatant was collected. www.nature.com/scientificreports/ The total RNA extraction method for miRNA analysis using the miRNeasy mini kit (QIAGEN, Netherlands) is briefly shown in the succeeding text. Then, 200 μL of CM was mixed with 1 mL of QIAZOL (QIAGEN, Netherlands) and allowed to stand for 3 min. The mixed solution was vigorously mixed with 140 μL of chloroform and allowed to stand for 10 min. The mixed solution was centrifuged at 15,000×g for 15 min, and 600 μL of the supernatant was collected. Then, 900 μL of ethanol was added to the supernatant, and the miRNA was purified by the same method as mRNA. The final volume of the total RNA solution was adjusted to 30 μL and used for array analysis. EV-derived RNA extraction was carried out in the same method as CM after adjust- ing the EVs to 50 μg/200 μL with PBS. For comprehensive analysis of CM and EV-derived miRNA, 3-D Gene miRNA Microarray Platform was used (Toray, Japan). The miRNA array data were normalized using the global normalization method and used for subsequent analysis. Analysis of miR‑1246 level in EVs and cells in vitro. To analyze the miR-1246 level in EVs, the EVs were collected from the CM of NECs (HMVEC) and tumor cells (A375 and A375SM cells) as described in the previous text. QIAZOL was mixed with the EV solution, and cel-miR-39 (Hokkaido System Science, Japan) was added (final concentration: 1 nM). Afterward, the total RNA was extracted as described in the previous text. The cells were seeded on six-well plates at 1 × 105 cells/well to analyze the miR-1246 level. Then, 500 µL of QIA- ZOL was added to the cells, and the total RNA was extracted as described in the previous text. To analyze the miR-1246 level, real-time RT-PCR was conducted using TaqMan MicroRNA Assays (Applied Biosystems, CA, USA) and a Universal PCR Master Mix II (Applied Biosystems), according to the manufacturer’s instructions. The cycling conditions were done according to the manufacturer’s instructions, and the CFX Manager program (Bio-Rad) was used for analysis. The primers and probes were defined as miR-1246 (Assay ID: CSN1EFS). Cel- miR-39 (Assay ID: 000200; UCACCGGGUGUAAAUCAGCUUG) level was used as the external control and RNU6B (Assay ID; 001093; CGCAAGGATGACACGCAAATTCGTGAAGCGTTCCATATTTTT) was used as internal control. Analysis of miR‑1246 level in EVs from human blood samples. Human melanoma patients’ serum amples were purchased from BizComJapan, Inc. The cell death was measured by FACS Aria II (BD) and calculated as relative cell ratio using three different experiments data. The HMVECs were treated with recombinant IL-6 (10 ng/mL) or anti-IL-6 receptor antibody (1 μg/mL) for 5 days and then treated with 5-FU (10 μM). Measurement of dead cells. After 24 h of miR-1246 transfection, the HMVECs were treated with 5-FU (10 μM). The cells were stained with FITC-annexin V and propidium iodide using an Annexin V-FLUOS stain- ing kit (Roche, Basel, Switzerland). The cell death was measured by FACS Aria II (BD) and calculated as relative cell ratio using three different experiments data. The HMVECs were treated with recombinant IL-6 (10 ng/mL) or anti-IL-6 receptor antibody (1 μg/mL) for 5 days and then treated with 5-FU (10 μM). Cell survival under 5‑FU treatment. Each cell was seeded on 96-well plate at 5 × 103 cells/well in EGM- 2MV. After 6 h, 5-FU was added to the medium. Cell survival was measured through MTS [3-(4,5-dimethylthyl- thiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] assay (Promega) at 72 h. siRNA transfection. UNC5B or AR siRNA was transfected using lipofectamine transfection reagent (Inv- itrogen, Tokyo, Japan), according to the manufacturer’s instructions. UNC5B siRNA (UNC5B si#1: 5′-UAC GUGUUCACGGGCGAGUCCUAUU-3′, 5′-AAUAGGACUCGCCCGUGAACACGUA-3′, UNC5B si#2: 5′-UCCACAGAGCUCACCUGCAAGAUCU-3′, 5′-AGAUCUUGCAGGUGAGCUCUGUGGA-3′, UNC5B si#3: 5′-GAGGGCCAGAUAUUCCAGCUGCAUA-3′, 5′-UAUGCAGCUGGAAUAUCUGGCCCUC-3′) (HSS137260, 137261, 137262, stealth, Thermo Fisher Scientific, USA) and AR siRNA (AR si#1: 5′-GACUCC UUUGCAGCCUUGCUCUCUA-3′, 5′-UAGAGAGCAAGGCUGCAAAGGAGUC-3′, AR si#2: 5′-GAUGAA GCUUCUGGGUGUCACUAUG-3′, 5′-CAUAGUGACACCCAGAAGCUUCAUC-3′, AR si#3: 5′-CCGGAA GCUGAAGAAACUUGGUAAU-3′, 5′-AUUACCAAGUUUCUUCAGCUUCCGG-3′) (HSS100619, 179972, 179973, stealth, Thermo Fisher Scientific, USA). Non-targeting siRNA (Invitrogen) was used as control. Western blotting. The cells and EVs were lysed using radio immunoprecipitation assay (RIPA) buffer (Cell Signaling Technology). The total protein concentration was determined using a BCA Protein Assay kit (Pierce, Rockford, IL, USA). Equal amounts of whole protein extracts (10 μg for EVs and 20 μg for others) were sepa- rated on SDS-PAGE gels (10% for STAT3 and Akt and 12% for IL-6 analyses). Western blotting was conducted according to standard methods using antibodies specific for STAT3 (Cell Signaling Technology, 12640S, Beverly, MA), pSTAT3 (Cell Signaling Technology, 9145S), Akt (Cell Signaling Technology, 4685S), pAkt (Cell Signaling Technology, 4060S), IL-6 (Abcam, ab6672), β-actin (Cell Signaling Technology, 4970L), and a horseradish per- oxidase (HRP)-conjugated secondary antibody as previously described20. The signals were detected using ECL Western Blotting Detection Reagent (GE Healthcare, Little Chalfont, UK) and LAS-4000 Mini image analyzer (FUJIFILM, Tokyo, Japan). Statistical analysis. Analysis of miR‑1246 level in EVs from human blood samples. Human melanoma patients’ serum amples were purchased from BizComJapan, Inc. The cells were harvested 48 h after transfection, and the Rennila luciferase activity was measured and normalized to the firefly luciferase activity using the Dual-Glo Luciferase Assay System (Pro- mega). Relative luminescence was obtained by normalizing the values against the transfection of the NC mimic group. All assays were repeated at least three times, and the representative results are shown. system (Promega) and ligated into the Not I site of the 3′ UTR of the Rennila luciferase gene in the psiCHECK™-2 plasmid (Promega). The following primer sequences were used (shown 5′ to 3′): For vector 1, forward: CACTGT GTTTGCTAGTGCCC, reverse: CATCAGGTGTGATCTGGAAC. For vector 2, forward: CCCACCTGTCTC TTAGCCTG, reverse: CAAATCTGGCCTGTCACCTC. The HEK293 cells were cultured overnight in 96-well tissue culture plates at densities of 1 × 104 cells/well, respectively, and each construct was co-transfected with hsa-miR-1246 mimic or negative control (NC) mimic using the DharmaFECT Duo Transfection Reagent (Dhar- macon, Horizon Discovery). The cells were harvested 48 h after transfection, and the Rennila luciferase activity was measured and normalized to the firefly luciferase activity using the Dual-Glo Luciferase Assay System (Pro- mega). Relative luminescence was obtained by normalizing the values against the transfection of the NC mimic group. All assays were repeated at least three times, and the representative results are shown. system (Promega) and ligated into the Not I site of the 3′ UTR of the Rennila luciferase gene in the psiCHECK™-2 plasmid (Promega). The following primer sequences were used (shown 5′ to 3′): For vector 1, forward: CACTGT GTTTGCTAGTGCCC, reverse: CATCAGGTGTGATCTGGAAC. For vector 2, forward: CCCACCTGTCTC TTAGCCTG, reverse: CAAATCTGGCCTGTCACCTC. The HEK293 cells were cultured overnight in 96-well tissue culture plates at densities of 1 × 104 cells/well, respectively, and each construct was co-transfected with hsa-miR-1246 mimic or negative control (NC) mimic using the DharmaFECT Duo Transfection Reagent (Dhar- macon, Horizon Discovery). The cells were harvested 48 h after transfection, and the Rennila luciferase activity was measured and normalized to the firefly luciferase activity using the Dual-Glo Luciferase Assay System (Pro- mega). Relative luminescence was obtained by normalizing the values against the transfection of the NC mimic group. All assays were repeated at least three times, and the representative results are shown. Measurement of dead cells. After 24 h of miR-1246 transfection, the HMVECs were treated with 5-FU (10 μM). The cells were stained with FITC-annexin V and propidium iodide using an Annexin V-FLUOS stain- ing kit (Roche, Basel, Switzerland). Analysis of miR‑1246 level in EVs from human blood samples. Human melanoma patients’ serum amples were purchased from BizComJapan, Inc. All data are expressed as the mean ± standard deviation of triplicate independent experiments. Statistical significance was determined using unpaired Student’s t test between two groups or one- way ANOVA followed by Tukey’s test. p values less than 0.05 were considered statistically significant. Ethics approval and consent to participate. The investigation was conducted in accordance with the ethical standards, the Declaration of Helsinki, and national and international guidelines. All methods have been approved by the Institutional Ethics Committee of Hokkaido University (Sapporo, Hokkaido, Japan), and writ- ten informed consent was obtained from each patient before surgery. All procedures for animal care and experi- mentation adhered to institutional guidelines and were approved by the local animal research authorities. Analysis of miR‑1246 level in EVs from human blood samples. Human melanoma patients’ serum amples were purchased from BizComJapan, Inc. p p p 400 μL of serum was ultracentrifuged (210,000×g, 40 min at 4 ℃) using Beckman SW55Ti rotor, the pellet was washed twice with PBS, and EVs were eluted in 100 μL of PBS. miRNA analysis methods were as described in the previous text, and hsa-miR-451 (Assay ID: 001105; AAACCGUUACCAUUACUGAGUUU) was used as an internal control. Uptake of EVs‑derived miR‑1246. The HMVECs were cultured for 24 h with EV-free EBM2 (serum concentration: 5%) and pretreated with the endocytosis inhibitor dynasore (20 µM) for 2 h. The EVs were added at a concentration of 50 μg/mL, and the miR-1246 level was analyzed through qPCR after 12 h. miR‑1246 transfection. The HMVECs were seeded on six-well plates at 1 × 105 cells/well. After 24 h, miRIDIAN microRNA Human hsa-miR-1246-Mimic (CN-001040, Dharmacon, GE Dharmacon, Lafayette, CO, USA) and miRIDIAN microRNA Mimic Negative Control (CN-001000-01-05) were transfected into the HMVECs using Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA, USA) to a final concentration of 50 nM, respectively. After 6 h, the medium was replaced and used for subsequent analysis. 3′ UTR reporter assay. A 328-base pair fragment (vector 1) and a 264-base pair fragment (vector 2) from the 3′ UTR of AR, which contained the predicted target sequences of miR-1246 (located at positions 4714 to 4719 (vector 1), 5870 to 5874 (vector 2), and 5903 to 5908 (vector 2) of the AR 3′ UTR), was cloned through PCR from the total RNA isolated from HEK293 cells. A 3′ polyadenylate overhang was added to the PCR products after 15 min of regular Taq polymerase treatment at 72 ℃. The PCR products were cloned into a pGEM®-T Easy Vector https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ system (Promega) and ligated into the Not I site of the 3′ UTR of the Rennila luciferase gene in the psiCHECK™-2 plasmid (Promega). The following primer sequences were used (shown 5′ to 3′): For vector 1, forward: CACTGT GTTTGCTAGTGCCC, reverse: CATCAGGTGTGATCTGGAAC. For vector 2, forward: CCCACCTGTCTC TTAGCCTG, reverse: CAAATCTGGCCTGTCACCTC. The HEK293 cells were cultured overnight in 96-well tissue culture plates at densities of 1 × 104 cells/well, respectively, and each construct was co-transfected with hsa-miR-1246 mimic or negative control (NC) mimic using the DharmaFECT Duo Transfection Reagent (Dhar- macon, Horizon Discovery). Results (E) Dead cells were analyzed through flow cytometry using PI under 5-FU treatment condition for 72 h after treatment with recombinant IL-6 and/ or IL-6R neutralizing antibody (NA) in HMVECs for 5 days. Data are presented as mean ± SD, n = 3. (F) After stimulation of HMVECs using recombinant IL-6 at the indicated doses, the HMVECs were lysed, and the levels of phosphorylated Akt (pAkt) and total Akt (Akt) were determined through western blotting. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments (P < 0.05 versus 5 and 10 ng/mL of rIL-6; one-way ANOVA followed by post hoc Tukey’s test). (G) IL-6 levels in in each CM were analyzed through ELISA. Control CM was prepared from HMVECs (P < 0.01 vs. A375CM and control CM, one-way ANOVA followed by post hoc Tukey’s test; data are presented as mean ± SD, n = 3). induced cell survival in HMVECs compared to control, which suggested that the tumor CM from A375SM induced resistance to 5-FU in HMVECs (Fig. 1D). Anti-IL-6 receptor antibody treatment canceled 5-FU resist- ance induced by IL-6, showing that IL-6 is involved in 5-FU resistance in HMVECs (Fig. 1E). Furthermore, Akt phospholylation level was further increased by 5 ng/mL of IL-6 treatment in HMVEC, although Akt is constitu- tively activated, because of endothelial cell culture medium containing the several growth factors, such as VEGF (Fig. 1F). It was suggested that IL-6 is also one of additional activator for EC survival. In contrast, IL-6 was not detected in both A375 and A375SM CMs, suggesting that the tumor cells did not secrete IL-6, unlike HMVECs (Control CM) (Fig. 1G). These results suggested that IL-6 was derived from HMVECs after stimulation by tumor secretion factors, but not from tumor cells. Tumor secretion factors induce drug resistance through IL-6 upregu- lation in HMVECs. Tumor EVs induce IL‑6 mRNA expression. We focused on EVs as factors that may be contained in tumor CM and induce IL-6 upregulation in ECs. EVs from A375 and A375SM cells were collected and charac- terized. The size of the isolated EVs was approximately 140 nm in diameter, which was similar to the reported EV size38 (Fig. S1A,C). The particle numbers of A375SM EV were more compared to A375 EV (Fig. S1B). Results IL‑6 upregulation by tumor CM induced drug resistance in ECs in an autocrine manner. We have reported that A375SM-derived tumor CM induced resistance to paclitaxel by upregulating MDR-1/ABCB1 in human NEC:HMVEC18. Furthermore, we found that TECs isolated from A375SM tumors (A375SM-TECs) were resistant to 5-FU17, which is not a substrate of ABCB1, suggesting the presence of a different mechanism apart from ABCB1. Since IL-6 is a known molecule that is related to drug resistance36,37 and was upregulated in A375SM-TECs, we focused on IL-6. The A375SM-TECs showed a significantly higher level of IL-6 than A375 tumor-derived TECs (A375-TECs) or mouse normal dermis-derived ECs (NECs) (Fig. 1A,B). We assumed that A375SM cell-derived factors induced IL-6 upregulation in A375SM-TECs in tumor microenvironment. Then, we treated HMVECs with A375SM-derived CM to check IL-6 expression level. Treatment of tumor CM from A375SM increased IL-6 mRNA expression level in HMVECs (Fig. 1C). The tumor CM from A375SM https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ Figure 1. IL-6 upregulation by tumor-conditioned medium induces drug resistance in ECs via Akt activation. (A) IL-6 expression was evaluated in mouse normal dermis-derived ECs (NECs), TECs derived from A375 and A375SM tumors (A375-TEC and A375SM-TEC), respectively, through real-time PCR (P < 0.01 versus NEC and A375-TEC; one-way ANOVA followed by post hoc Tukey’s test; data are presented as mean ± SD, n = 4, real- time RT-PCR runs). (B) The levels of IL-6 in mouse normal dermis-derived ECs (NECs), TECs derived from A375 and A375SM tumors (A375-TEC and A375SM-TEC) were determined through western blotting. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of western blot from three independent experiments. (C) IL-6 expression was evaluated in HMVECs treated with conditioned medium (CM) from HMVECs (Control CM) and CM from A375SM (Tumor CM) through real-time PCR after treatment of each CM for 24 h (P < 0.01 vs. Control CM, two-sided Student’s t-test; data are presented as mean ± SD, n = 4 real-time RT-PCR runs). (D) Survival of HMVECs was analyzed through MTS assay under 5-FU treatment condition for 72 h after culture in each CM for 5 days. Control CM and Tumor CM were prepared from HMVECs and A375SM, respectively (P < 0.01 vs. Control CM, two-sided Student’s t-test; data are presented as mean ± SD, n = 3). Results Control, two-sided Student’s t-test; data are mean ± SD, n = 4 real-time RT-PCR runs). (B) IL-6 expression in HMVECs was evaluated through real-time PCR after treatment of each EV for 24 h (P < 0.01 vs. A375-EVs, two-sided Student’s t-test; data are presented as mean ± SD, n = 4 real-time RT-PCR runs). Figure 2. Tumor EVs induces IL-6 mRNA expression. (A) IL-6 expression in HMVECs was evaluated through real-time PCR after treatment of HMVECs-derived EVs (Control) or A375SM-derived EVs (A375SM-EVs), respectively, for 24 h (P < 0.01 vs. Control, two-sided Student’s t-test; data are mean ± SD, n = 4 real-time RT-PCR runs). (B) IL-6 expression in HMVECs was evaluated through real-time PCR after treatment of each EV for 24 h (P < 0.01 vs. A375-EVs, two-sided Student’s t-test; data are presented as mean ± SD, n = 4 real-time RT-PCR runs). Figure 2. Tumor EVs induces IL-6 mRNA expression. (A) IL-6 expression in HMVECs was evaluated through real-time PCR after treatment of HMVECs-derived EVs (Control) or A375SM-derived EVs (A375SM-EVs), respectively, for 24 h (P < 0.01 vs. Control, two-sided Student’s t-test; data are mean ± SD, n = 4 real-time RT-PCR runs). (B) IL-6 expression in HMVECs was evaluated through real-time PCR after treatment of each EV for 24 h (P < 0.01 vs. A375-EVs, two-sided Student’s t-test; data are presented as mean ± SD, n = 4 real-time RT-PCR runs). AR inhibition and induction of drug resistance in HMVECs by miR‑1246. To address the role of miR-1246 in drug resistance, we attempted to identify the target molecule of miR-1246. First, 4585 genes that were downregulated more than 50% in miR-1246 transfected HMVECs, compared with those in control miR- transfected HMVECs, were selected through DNA microarray analysis (Step 1), as shown in Fig. 4A. Second, among these genes, 176 predicted genes were picked up as miR-1246 targets by three different databases for miRNA target prediction: target scan (32 genes), miR DB (46 genes), and Diana (139 genes) (Step 2). Finally, three genes that were reported to downregulate IL-6 by Ingenuity Pathway Analysis (IPA) were selected as can- didate target genes of miR-1246 (Fig. 4A): Androgen receptor (AR), unc-5 netrin receptor B (UNC5B), and Cathepsin C (CTSC). As observed, the mRNA expression levels of AR and UNC5B decreased by miR-1246 transfection in HMVECs; however, CTSC mRNA expression level did not (Fig. 4B). Results Addi- tionally, EV markers, such as HSP70, CD63 and CD9, were detected in the extracted protein from each EV, whereas cytochrome C, a cytoplasm protein, was not detected (Fig. S1D), while tumor cell lysate western blot- ting showed the expression of all of them. The vesicular structures of the EVs isolated from A375 and A375SM were observed by transmission electron microscopy (Fig. S1E). IL-6 mRNA was upregulated in HMVECs when they were treated with A375SM-EVs, not with HMVECs-derived EVs (Control) (Fig. 2A). Furthermore, the EVs isolated from A375 were added to the HMVEC culture. A375SM-EV treatment induced IL-6 mRNA at a higher level than did A375-EV treatment (Fig. 2B). These data suggested that the factors in A375SM-EVs induced IL-6 mRNA upregulation in HMVECs. miR‑1246 in EVs from A375SM enhanced IL‑6 mRNA expression level in HMVECs. Next, miRNA expression levels were determined to identify the factor in tumor CM or in EVs that induced IL-6 in HMVECs. Here, miRNA was isolated from tumor CM, which was collected from several tumor cell lines (A375, A375SM, HSC3, OS-RC-2, and HeLa) and normal cells (BJs, melanocytes. and HMVECs), and compared with each other through miRNA array. Furthermore, miRNA was also compared between A375SM-EV and A375-EV through miRNA array (Table S1), since we have found that A375SM-TEC showed a higher level of IL-6 than did A375-TEC, as shown in Fig. 1A. Furthermore, miR-1246 was picked up because its level was higher in the CM from tumor cells than in that from normal cells and also in A375SM-EV than in A375-EV (Fig. 3A). Indeed, it was confirmed that miR-1246 was higher in A375SM-EV than in A375-EV or in HMVEC-EV, as determined through PCR (Fig. 3B). The endocytosis inhibitor dynasore canceled miR-1246 induction through A375SM-EV treatment in HMVECs, suggesting that miR-1246 is transported into ECs by A375SM-EVs (Fig. 3C). Further- more, transfection of miR-1246 in HMVECs increased IL-6 mRNA expression levels (Fig. 3D) consistently with the results of EV treatment, as shown in Fig. 2A. Furthermore, miR-1246 transfection caused resistance to 5-FU in HMVECs (Fig. 3E). Scientific Reports \| (2021) 11:13502 \| https://doi.org/10.1038/s41598-021-92879-5 www.nature.com/scientificreports/ scientificreports/ Scientific Reports \| (2021) 11:13502 \| https://doi.org/10.1038/s41598-021-92879-5 www.nature.com/scientificreports/ tificreports/ Figure 2. Tumor EVs induces IL-6 mRNA expression. (A) IL-6 expression in HMVECs was evaluated through real-time PCR after treatment of HMVECs-derived EVs (Control) or A375SM-derived EVs (A375SM-EVs), respectively, for 24 h (P < 0.01 vs. Results This suggested that AR and UNC5B are targets of miR-1246 but not CTSC. Similarly, A375SM-EVs treatment caused downregulated AR and UNC5B mRNA expressions (Fig. 4C). IL-6 mRNA expression in HMVECs was analyzed after AR or UNC5B inhibition to examine the contribution of these genes on IL-6 expression. IL-6 expression level was increased by the knockdown of AR but not of UNC5B, suggesting that AR suppresses IL-6 expression in HMVECs (Fig. 4D). Furthermore, 3′ UTR reporter assay was conducted to address whether miR-1246 directly binds to the AR gene. The transfection of a vector containing AR 3′UTR decreased luciferase activity. This result showed that miR-1246 binds to the 3′UTR of the AR gene, suggesting that AR is an miR-1246 target (Fig. 4E, Supplementary Fig. S2). Furthermore, we analyzed the effect of AR knockdown on drug resistance to 5-FU in HMVECs by MTS assay. Cell survival of HMVECs under 5-FU treatment increased by AR knockdown, which suggests that AR knock- down caused resistance to 5-FU in HMVECs (Fig. 4F). miR‑1246 caused drug resistance via activation of STAT3 and Akt. It is well known that IL-6 activates STAT3, a substance known to be related to drug resistance39. Tumor CM induced STAT3 activation, whereas its activation was canceled by S3I-201, a STAT3 inhibitor (Fig. 5A). It was confirmed that tumor EVs (A375SM-EVs) (Fig. 5B) and miR-1246 transfection (Fig. 5C) activated STAT3 in HMVECs. STAT3 phospho- rylation levels were not elevated in UNC5B knockdown in HMVECs, but they were increased 1.69 times in AR knockdown (AR si#2) endothelial cells compared to control si (Fig. 5D, Supplementary Fig. S3). Akt activation has been reported to be involved in cell survival or drug resistance38. Akt phosphorylation levels was 5 times higher in tumor EV treated in HMVEC compared to control (Fig. 5E). In addition, miR-1246 transfection dou- bled the phosphorylation level of Akt (Fig. 5F). miR‑1246 levels in EVs were higher in melanoma patients than in healthy volunteers. Finally, EVs were isolated from sera in melanoma patients and healthy volunteers, and miR-1246 levels in EVs were determined. miR-1246 levels were significantly higher in melanoma patients’ EVs than those in healthy volun- teers’ EVs (Fig. 6A). In the present study, we revealed three findings. First, miR-1246 was contained in highly metastatic tumor EVs. Second, miR-1246 in tumor EVs caused IL-6 induction through AR downregulation in ECs. Third, miR-1246 in tumor EVs activated STAT3 and Akt, causing drug resistance in ECs (Fig. 6B). Discussion I h In the present study, we revealed three findings. First, miR-1246 was contained in highly metastatic tumor EVs. Second, miR-1246 in tumor EVs caused IL-6 induction through AR downregulation in ECs. Third, miR-1246 in tumor EVs activated STAT3 and Akt, causing drug resistance in ECs (Fig. 6B). Scientific Reports \| (2021) 11:13502 \| https://doi.org/10.1038/s41598-021-92879-5 www.nature.com/scientificreports/ Figure 3. miR-1246 in A375SM EVs enhances IL6 mRNA expression in HMVECs. (A) miRNA array analys in conditioned medium (CM) from tumor and normal cells (left) and in extracellular vesicles (EVs) from A375SM and A375 (right). Each miRNA was plotted. The black rhombus shows miR-1246. miR-1246 was high in both CM from tumor cells and EVs from A375SM. (B) miR-1246 expression in each EV was evaluate through real-time PCR (P < 0.01 vs. HMVEC-EVs and A375-EVs, one-way ANOVA; data are presented as mean ± SD, n = 4 real-time RT-PCR runs). (C) miR-1246 expression in HMVECs was evaluated through real- time PCR with or without A375SM-EVs and dynasore (P < 0.05, one-way ANOVA followed by post hoc Tuk test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (D) IL-6 expression was evaluated throug real-time PCR after transfection of miR-1246 in HMVECs. Control miRNA-transfected HMVECs was shown control (P < 0.01 vs. control, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-P runs). (E) Dead cells were analyzed through flow cytometry using PI under 5-FU treatment condition for 72 after transfection of miR-1246 in HMVECs. Control miRNA-transfected HMVECs was shown as Control. D are presented as mean ± SD, n = 3. Figure 3. miR-1246 in A375SM EVs enhances IL6 mRNA expression in HMVECs. (A) miRNA array analysis in conditioned medium (CM) from tumor and normal cells (left) and in extracellular vesicles (EVs) from A375SM and A375 (right). Each miRNA was plotted. The black rhombus shows miR-1246. miR-1246 was high in both CM from tumor cells and EVs from A375SM. (B) miR-1246 expression in each EV was evaluated through real-time PCR (P < 0.01 vs. HMVEC-EVs and A375-EVs, one-way ANOVA; data are presented as mean ± SD, n = 4 real-time RT-PCR runs). (C) miR-1246 expression in HMVECs was evaluated through real- time PCR with or without A375SM-EVs and dynasore (P < 0.05, one-way ANOVA followed by post hoc Tukey’s test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). Discussion I h The major mechanism involved in developing resistance is tumor cell phenotypic change. For example, tumor cells express other angiogenic factors in response to VEGF inhibition40. However, we have suggested that another mechanism may be responsible for the abnormalities in TECs. We previously compared the characteristics of TECs with that of NECs and found that TECs contain several abnormalities, such as cytogenetic abnormalities13,14 and drug resistance with multidrug resistance 1 (MDR1) upregulation, which is a well-known stem marker and an ABC transporter18,41. p Furthermore, we previously demonstrated that TECs are heterogeneous depending on tumor malignancy. Highly metastatic tumor-derived TECs (A375SM-TECs) showed different characteristics from low metastatic tumor-derived TECs (A375-TECs). For example, A375SM-TECs showed higher expression levels of MDR1 than did A375-TECs and showed a more resistant profile to paclitaxel, an anti-cancer drug that is a substrate of ABCB120. We have found that tumor-secreted VEGF-A induced MDR1 mRNA upregulation in NECs through Y box-binding protein 1 activation18. These findings suggested that tumor-derived factors can cause EC resist- ance, and these alterations may be dependent on tumor malignancy. Additionally, TECs showed resistance to 5-FU, which is not a substrate of ABCB117, suggesting that TECs possess other mechanisms of drug resistance besides ABCB1. We focused on IL-6, because IL-6 was upregulated by tumor CM treatment, and it is reportedly involved in drug resistance in tumor cells36,37,42. Transcriptional factor C/EBP was activated through the IL-6 autocrine loop mechanism, causing MDR-1 upregulation and inducing resistance to doxorubicin, vincristine, and taxol in IL-6-expressing mammary carcinoma cells36. Furthermore, the IL-6 autocrine loop upregulated expressions of anti-apoptotic genes, such as Bcl-2, Bcl-xL, and XIAP, and drug resistance genes, such as MDR-1 and GSTpi, accompanied by the activation of the Ras/MEK/ERK and PI3K/Akt pathways in uterine cancer cells37. We previously reported that TECs expressed higher levels of IL-6 than did NECs43, and Akt was more activated in TECs than in NECs43. In the present study, the IL-6 autocrine loop induced 5-FU resistance by activating STAT3 and Akt in HMVECs. It was suggested that tumor CM-induced resistance in HMVECs was caused by the IL-6 autocrine loop and not by tumor derived-IL-6. p y EVs are the generic term for exosomes or microvesicles, which contain biomolecules, such as protein, mRNA, and miRNA, playing roles in intercellular communications25. Discussion I h (D) IL-6 expression was evaluated through real-time PCR after transfection of miR-1246 in HMVECs. Control miRNA-transfected HMVECs was shown as control (P < 0.01 vs. control, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (E) Dead cells were analyzed through flow cytometry using PI under 5-FU treatment condition for 72 h after transfection of miR-1246 in HMVECs. Control miRNA-transfected HMVECs was shown as Control. Data are presented as mean ± SD, n = 3. https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ Figure 4. miR-1246 targets androgen receptor (AR) and induces drug resistance in HMVECs. (A) Schematic diagram of steps to narrow miR-1246 targets using DNA microarray and IPA. Of note that the three genes AR, UNC5B and CTSC were selected at the Step 3. (B) Each mRNA expression was evaluated through real-time PCR after transfection of miR-1246 in HMVECs. Control miRNA-transfected HMVECs was shown as Control (P < 0.01 vs. control, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (C) AR and UNC5B expressions were evaluated through real-time PCR after treatment of A375SM-derived EVs (EVs) cells in HMVECs. HMVEC-EV treated HMVEC was shown as Control (P < 0.01 vs. control, two- sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (D) IL-6 expression was evaluated through real-time PCR after transfection of control, AR or UNC5B siRNA (si) in HMVECs (P < 0.01 vs. control si, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs. N.S. not significant). (E) Luciferase reporter assays with AR 3′ UTR. HEK293 cells were co-transfected with miR- 1246 mimic or NC mimic and vector 1, vector 2, or control vector (without AR 3′ UTR). After 48 h, luciferase activities were measured (n = 6 each group). (F) Survival of HMVECs was analyzed through MTS assay under 5-FU treatment condition for 72 h after transfection of Control or AR siRNA (P < 0.01 vs. Control si, two-sided Student’s t-test; data are presented as mean ± SD, n = 3). ◂ It has been considered that anti-angiogenic therapy has advantages, because ECs are genetically stable and do not develop drug resistance, unlike tumor cells10. However, resistance to this type of therapy has been reported. Discussion I h PI3K/Akt signaling plays an important role in the acquisition of tumors cells’ resistance to anti-cancer drugs58. miR-1246 have other target molecules in ECs, and AR was found to be one. In cancer biology, there are several reports about AR53. It has been reported that AR expression was negative in triple-negative breast carcinoma, and it was suggested that the loss of AR may be correlated with poor prognosis54. Additionally, decreased AR expression induced stemness phenotypes in prostate cancer cells through STAT3 activation55. Although the role of AR in ECs is still unclear, previous reports may support our findings that AR downregulation induced drug resistance through STAT3 activation and this activation contributed to cell survival56. It has been reported that the feedback activation of STAT3 was a side effect of chemotherapy57. PI3K/Akt signaling plays an important role in the acquisition of tumors cells’ resistance to anti-cancer drugs58. Although the detailed molecular mechanisms by which highly metastatic tumor-secreted EVs contained higher levels of miR-1246 still remain unclear, our results demonstrated first evidence for tumor-derived EVs causing drug resistance in ECs. Furthermore, in this study we could not completely deny the miRNA1246 in the external contaminants because we did not serum free culture medium to collect tumor EVs. We tried to collect EV after culture with FBS-free medium, but it was difficult to obtain enough amounts of EVs. And we considered that it is important to analyze the communication between actively growing tumor cells and endothelial cells, which resemble tumor microenvironment. That is why we have cultured tumor EVs with 10% FBS (EV-depleted)- containing medium. However, we could detect miR1246 in A375SM-EVs even after RNase treatment, although the level is a half of nontreated EVs. But miR1246 level is 10 times higher in A375SM-EVs than in A375-EVs as described in Fig. 3B, so the biological effect of A375SM-EVs comparing with A375-EVs, which we show in this study, is due to miR1246, at least partially. y p y Recently, resistance to anti-angiogenic therapy has been reported, and we have demonstrated that ABC trans- porter upregulation is one of the mechanisms involved in acquiring drug resistance18,19,41. However, our results suggest that TECs can acquire resistance to other drugs, such as anti-angiogenic drugs (anti-VEGF drugs) and not just paclitaxel, by increasing IL-6 and Akt or STAT3 activation. Particularly, miR-1246-expressing cancers may produce such results. Discussion I h It has been known that tumor cells take advantage of their EV secretion for tumor growth and progression in an autocrine manner22–24, and EV secretion was increased by microenvironmental factors, such as hypoxia, low nutrition, and inflammation44–46. Furthermore, EVs affect parenchymal cells in a paracrine manner. There are several reports about the effects of tumor EVs on blood vessels. We have reported that tumor-derived EVs were taken into ECs through endocytosis, result- ing in pro-angiogenic phenotypes in ECs26. A recent study reported that breast cancer cell EVs enhanced brain metastasis by breaking the blood–brain barrier27. However, whether tumor EVs affect TECs in the primary site has remained unclear. It is known that miRNAs are small non-coding RNAs that reduce protein expression through destabilization and/or translational suppression of target RNA (mRNA) molecules29. Thus, miRNAs are key regulators of mul- tiple physiologic and pathologic cellular responses. Multiple reports have implicated that miRNA are involved in cancer progression. Furthermore, miRNAs are one of the important biomolecules contained in EVs and exported to extracellular spaces by EVs, playing important roles in intercellular communication31. In the present study, we focused on miR-1246 that was contained in highly metastatic tumor-derived EVs. Recent studies related to the function of miR-1246 in cancer cells have been reported47–52. For example, miR-1246 inhibits CCNG2 expres- sion, causing drug resistance and stemness in pancreatic cancer cells50. Furthermore, miR-1246 in EVs secreted from colon cancer cells were imported into ECs and induced angiogenesis by downregulating promyelocytic leukemia expression, resulting in Smad 1/5/8 activation47. However, the function of miR-1246 remains unla- beled. We found that CCNG2 expression was not downregulated by miR-1246 in HMVECs (data not shown), although CCNG2 was reported to be a target of miR-1246 and is involved in drug resistance. It was suggested that Scientific Reports \| (2021) 11:13502 \| https://doi.org/10.1038/s41598-021-92879-5 e.com/scientificreports/ www.nature.com/scientificreports/ https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ Figure 5. miR-1246 causes drug resistance in HMVECs via STAT3 and Akt activation. (A) HMVECs were pre- incubated with S3I-201, a STAT3 phosphorylation inhibitor. After stimulation of control CM from HMVECs or tumor CM from A375SM for 30 min, HMVECs were lysed, and the levels of phosphorylated STAT3 (pSTAT3) and STAT3 were determined through Western blotting. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments. Discussion I h Our results showed that EVs in the blood of melanoma patients contained more miR- 1246 than those of healthy volunteers’ EVs. It would be worth addressing the correlation between EV-miR1246 and prognosis, especially treatment outcomes in the future, although the present study did not investigate the correlation between EV-miR1246 and metastasis or prognosis. Discussion I h (B) After stimulation of A375SM tumor EVs (EVs), the HMVECs were lysed, and the levels of pSTAT3 and STAT3 were determined through Western blotting. PBS was used as Control. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments. (C) Control miRNA (Control) or miR-1246 transfected HMVECs were lysed, and the levels of pSTAT3 and STAT3 were determined through western blotting. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments. (D) Control, UNC5B or AR siRNA-transfected HMVECs were lysed, and the levels of pSTAT3 and STAT3 were determined through Western blotting. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments. (E) After stimulation of A375SM tumor EVs, the HMVECs were lysed and the levels of phosphorylated Akt (pAkt) and Akt were determined through western blotting. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments. (F) Control miRNA (Control) or miR-1246-transfected HMVECs were lysed, and the levels of pAkt and Akt were determined through western blotting. β-actin was used as an internal control. The value shows the average of relative band intensities, which are taken from densitometric analysis of Western blot from three independent experiments (P < 0.05 vs. Control, two-sided Student’s t-test; data are presented as mean ± SD, n = 3). ◂ miR-1246 have other target molecules in ECs, and AR was found to be one. In cancer biology, there are several reports about AR53. It has been reported that AR expression was negative in triple-negative breast carcinoma, and it was suggested that the loss of AR may be correlated with poor prognosis54. Additionally, decreased AR expression induced stemness phenotypes in prostate cancer cells through STAT3 activation55. Although the role of AR in ECs is still unclear, previous reports may support our findings that AR downregulation induced drug resistance through STAT3 activation and this activation contributed to cell survival56. It has been reported that the feedback activation of STAT3 was a side effect of chemotherapy57. Conclusion miR-1246 in tumor EVs caused IL-6 induction through AR downregulation in ECs. In addition, miR-1246 in tumor EVs activated STAT3 and Akt, causing drug resistance in ECs. It was suggested that tumor makes TEC resistant via transporting miR-1246 by EV and keeps blood supply and pathway to metastasize to distant organ. https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ .com/scientificreports/ Data availability Figure 6. miR-1246 levels in EVs in plasma, and a schematic diagram of the study. (A) miR-1246 levels in EVs in healthy volunteers (n = 20) and melanoma patients (n = 42) were determined through real-time PCR (P < 0.01 vs. healthy volunteers, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (B) A schematic diagram of the study. miR-1246 in tumor EVs was taken into endothelial cells, and AR expression was inhibited, which induced IL-6 secretion and caused drug resistance in an autocrine manner through STAT3 and Akt activation. Figure 6. miR-1246 levels in EVs in plasma, and a schematic diagram of the study. (A) miR-1246 levels in EVs in healthy volunteers (n = 20) and melanoma patients (n = 42) were determined through real-time PCR (P < 0.01 vs. healthy volunteers, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (B) A schematic diagram of the study. miR-1246 in tumor EVs was taken into endothelial cells, and AR expression was inhibited, which induced IL-6 secretion and caused drug resistance in an autocrine manner through STAT3 and Akt activation. Figure 6. miR-1246 levels in EVs in plasma, and a schematic diagram of the study. (A) miR-1246 levels in EVs in healthy volunteers (n = 20) and melanoma patients (n = 42) were determined through real-time PCR (P < 0.01 vs. healthy volunteers, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (B) A schematic diagram of the study. miR-1246 in tumor EVs was taken into endothelial cells, and AR Figure 6. miR-1246 levels in EVs in plasma, and a schematic diagram of the study. (A) miR-1246 levels in EVs in healthy volunteers (n = 20) and melanoma patients (n = 42) were determined through real-time PCR (*P < 0.01 vs. healthy volunteers, two-sided Student’s t-test; data are presented as mean ± SD, n = 3 real-time RT-PCR runs). (B) A schematic diagram of the study. Data availability Th d d d y The datasets used and analysed during current study are available from the corresponding author on reasonable request. Received: 27 August 2020; Accepted: 16 June 2021 Conclusion miR-1246 in tumor EVs was taken into endothelial cells, and AR expression was inhibited, which induced IL-6 secretion and caused drug resistance in an autocrine manner through STAT3 and Akt activation. References 1. Folkman, J. Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat. Med. 1, 27–31 (1995). , J g g , , , ( ) 2. Folkman, J. Angiogenesis: An organizing principle for drug discovery? Nat. Rev. Drug Discov. 6, 273–286 (2007). 3 J h D H t l R d i d h II t i l i b i b l b l ti d lit l ith b 3. Johnson, D. H. et al. Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non-small-cell lung cancer. J. Clin. Oncol. 22, 2184–2191 (2004). https://doi.org/10.1038/s41598-021-92879-5 www.nature.com/scientificreports/ www.nature.com/scientificreports/ 4. Keedy, V. L. & Sandler, A. B. Inhibition of angiogenesis in the treatment of non-small cell lung cancer. Cancer Sci. 98, 1825–1830 (2007). 4. Keedy, V. L. & Sandler, A. B. Inhibition of angiogenesis in the treatment of non-small cell lung cancer. Cancer Sci. 98, 1825–1830 (2007). 5 P Rib M t l A ti i i th li it li t i f t t i d l l i i d di t t ( ) 5. Paez-Ribes, M. et al. Antiangiogenic therapy elicits malignant progression of tumors to increased local invasion and distant metastasis. Cancer Cell 15, 220–231 (2009).t ( ) 6. Ebos, J. M. L. et al. Accelerated metastasis after short-term treatment with a potent inhibitor of tumor angiogenesis. Cancer Cel 15, 232–239 (2009).fl 7. Jain, R. K., Duda, D. G., Clark, J. W. & Loeffler, J. S. Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin. Pract. Oncol. 3, 24–40 (2006). 8. Rini, B. I. Vascular endothelial growth factor-targeted therapy in renal cell carcinoma: Current status and future directions. Clin Cancer Res. 13, 1098–1106 (2007). 9. Auerbach, R., Akhtar, N., Lewis, R. L. & Shinners, B. L. Angiogenesis assays: Problems and pitfalls. Cancer Metast. Rev. 19, 167–172 (2000). ( ) 0. Kerbel, R. S. Clinical trials of antiangiogenic drugs: Opportunities, problems, and assessment of initial results. J. Clin. Oncol. 19 45S-51S (2001). 11. Zhang, G. et al. Repression of human hepatocellular carcinoma growth by regulating Met/EGFR/VEGFR-Akt/NF-κB pathways with theanine and its derivative, (R)-2-(6,8-Dibromo-2-oxo-2H-chromene-3-carboxamido)-5-(ethylamino)-5-oxopentanoic ethyl ester (DTBrC). J. Agric. Food Chem. 64, 7002–7013 (2016). 12. Streubel, B. et al. Lymphoma-specific genetic aberrations in microvascular endothelial cells in B-cell lymphomas. N. Engl. J. Med. 351, 250–259 (2004). d l d d h l l ll h b l ( ) ( ) 13. Hida, K. et al. Tumor-associated endothelial cells with cytogenetic abnormalities. Cancer Res. 64, 8249–8255 (2004). 14. Akino, T. et al. Cytogenetic abnormalities of tumor-associated endothelial cells in human malignant tumors. Am. J. Pathol. 175, 2657–2667 (2009). 5. Bussolati, B., Deambrosis, I., Russo, S., Deregibus, M. C. & Camussi, G. Altered angiogenesis and survival in human tumor-derived endothelial cells. FASEB J. https://doi.org/10.1096/fj.02-0557fje,1-3 (2003). p g fj f 6. Xiong, Y. Q. et al. Human hepatocellular carcinoma tumor-derived endothelial cells manifest increased angiogenesis capability and drug resistance compared with normal endothelial cells. Clin. Cancer Res. 15, 4838–4846 (2009). 7. www.nature.com/scientificreports/ S., Calin, S., Yang, Y., You, M. J. & Calin, G. A. Cell-to-cell miRNA transfer: From body homeostasis to therapy. Pharmacol. Rev. 32, 326–348 (2012). 31. Redis, R. S., Calin, S., Yang, Y., You, M. J. & Calin, G. A. Cell-to-cell miRNA transfer: From body homeostasis to therapy. Pharmacol. Rev. 32, 326–348 (2012). 31. Redis, R. S., Calin, S., Yang, Y., You, M. J. & Calin, G. A. Cell-to-cell miRNA transfer: From body homeostasis to therapy. Pharmacol. h 1. Redis, R. S., Calin, S., Yang, Y., You, M. J. & Calin, G. A. Cell-to-cell miRNA transfer: From body homeostasis to therapy. Pharmacol Ther. 136, 169–174 (2012). h 32. Kozlowski, J. M., Hart, I. R., Fidler, I. J. & Hanna, N. A human melanoma line heterogeneous with respect to metastatic capacity in athymic nude mice. J. Natl. Cancer Inst. 72, 913–917 (1984). y 3. Suehiro, J. et al. Genome-wide approaches reveal functional vascular endothelial growth factor (VEGF)-inducible nuclear factor of activated T cells (NFAT) c1 binding to angiogenesis-related genes in the endothelium. J. Biol. Chem. 289, 29044–29059 (2014) 33. Suehiro, J. et al. Genome-wide approaches reveal functional vascular endothelial growth factor (VEGF)-inducible nuclear factor of activated T cells (NFAT) c1 binding to angiogenesis-related genes in the endothelium. J. Biol. Chem. 289, 29044–29059 (2014). 34. Enabling Technologies Consortium et al. EV-TRACK: Transparent reporting and centralizing knowledge in extracellular vesicle research. Nat. Methods 14, 228–232 (2017).f of activated T cells (NFAT) c1 binding to angiogenesis related genes in the endothelium. J. Biol. Chem. 289, 29044 29059 (2014). 34. Enabling Technologies Consortium et al. EV-TRACK: Transparent reporting and centralizing knowledge in extracellular vesicle research. Nat. Methods 14, 228–232 (2017).f 35. Yamashita, T., Takahashi, Y., Nishikawa, M. & Takakura, Y. Effect of exosome isolation methods on physicochemical properties of exosomes and clearance of exosomes from the blood circulation. Eur. J. Pharm. Biopharm. 98, 1–8 (2016). l d f l k l d b ll 6. Conze, D. et al. Autocrine production of interleukin 6 causes multidrug resistance in breast cancer cells. Cancer Res. 61, 8851–8858 (2001). 37. Wang, Y. et al. Autocrine production of interleukin-6 confers cisplatin and paclitaxel resistance in ovarian cancer cells. Cancer Lett. 295, 110–123 (2010).i 38. Andaloussi, S. E. L., Mager, I., Breakefield, X. O. & Wood, M. J. Extracellular vesicles: Biology and emerging therapeutic opportuni- ties. Nat. Rev. Drug Discov. 12, 347–357 (2013). g 9. Ara, T. et al. https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| www.nature.com/scientificreports/ Hida, K. et al. Tumor endothelial cells with high aldehyde dehydrogenase activity show drug resistance. Cancer Sci. 108, 2195–2203 (2017). 18. Akiyama, K. et al. Tumor endothelial cells acquire drug resistance by MDR1 up-regulation via VEGF signaling in tumor micro- environment. Am. J. Pathol. 180, 1283–1293 (2012). 9. Kikuchi, H. et al. Chemotherapy-induced IL-8 upregulates MDR1/ABCB1 in tumor blood vessels and results in unfavorable outcome. Cancer Res. https://doi.org/10.1158/0008-5472.CAN-19-3791 (2020). et al. Heterogeneity of tumor endothelial cells. Am. J. Pathol. 180 l d h l l d l ll b 20. Ohga, N. et al. Heterogeneity of tumor endothelial cells. Am. J. Pathol. 180, 1294–1307 (2012). . Ohga, N. et al. Heterogeneity of tumor endothelial cells. Am. J. P 1. Naito, H. et al. Endothelial side population cells contribute to tumor angiogenesis and antiangiogenic drug resistance. Can. Res 76, 3200–3210 (2016).h 22. Yang, C. & Robbins, P. D. The roles of tumor-derived exosomes in cancer pathogenesis. Clin. Dev. Immunol. 2011, 842849 (2011). h 3. Bobrie, A. et al. Rab27a supports exosome-dependent and -independent mechanisms that modify the tumor microenvironmen and can promote tumor progression. Cancer Res. 72, 4920–4930 (2012). g 4. Peinado, H. et al. Melanoma exosomes educate bone marrow progenitor cells toward a pro-metastatic phenotype through MET Nat. Med. 18, 883–891 (2012). 5. Valadi, H. et al. Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells Nat. Cell Biol. 9, 654–659 (2007). 6. Kawamoto, T. et al. Tumor-derived microvesicles induce proangiogenic phenotype in endothelial cells via endocytosis. PLoS ONE 7, e34045–e34049 (2012). 7. Tominaga, N. et al. Brain metastatic cancer cells release microRNA-181c-containing extracellular vesicles capable of destructing blood-brain barrier. Nat. Commun. 6, 6716 (2015). 8. Hsu, Y. L. et al. Hypoxic lung cancer-secreted exosomal miR-23a increased angiogenesis and vascular permeability by targeting prolyl hydroxylase and tight junction protein ZO-1. Oncogene 36, 4929–4942 (2017). p y y y g j p g 29. Bartel, D. P. MicroRNAs: Target recognition and regulatory functions. Cell 136, 215–233 (2009). Z Zh l l l f b k d d 0. Zen, K. & Zhang, C. Y. Circulating microRNAs: A novel class of biomarkers to diagnose and monitor human cancers. Med. Res Rev. 32, 326–348 (2012). 30. Zen, K. & Zhang, C. Y. Circulating microRNAs: A novel class of biomarkers to diagnose and monitor human cancers. Med. Res. Rev. 32, 326–348 (2012). 31. Redis, R. Acknowledgements g We would like to express our appreciation to Dr. I. J. Fidler for providing the super-metastatic human malignant melanoma cell line (A375-SM). We also thank Drs. T. Hojo, M. Yanagiya, D. Annan, H. Kikuchi and Ms. Y. Suzuki for their technical support. g We would like to express our appreciation to Dr. I. J. Fidler for providing the super-metastatic human malignant melanoma cell line (A375-SM). We also thank Drs. T. Hojo, M. Yanagiya, D. Annan, H. Kikuchi and Ms. Y. Suzuki for their technical support. www.nature.com/scientificreports/ Prognostic significance of activated Akt expression in pancreatic ductal adenocarcinoma. Clin. Cancer Res. 10, 2846–2850 (2004). Author contributions C.T. performed and analyzed experiments and wrote the manuscript. N.M. helped conceive the study and con- tributed to writing the manuscript. T.K., M.M., K.A., Y.Y., T.T. and M.T.A. contributed to the completion of various experiments. T.M. performed DNA microarray and provided valuable advice. T.O. and Y.H. provided valuable advice. K.H. conceived, designed and supervised the study and wrote the manuscript. www.nature.com/scientificreports/ www.nature.com/scientificreports/ 47. Yamada, N. et al. Colorectal cancer cell-derived microvesicles containing microRNA-1246 promote angiogenesis by activating Smad 1/5/8 signaling elicited by PML down-regulation in endothelial cells. Biochem. Biophys. Acta 1839, 1256–1272 (2014).i g g y g p y 8. Takeshita, N. et al. Serum microRNA expression profile: miR-1246 as a novel diagnostic and prognostic biomarker for oesophagea squamous cell carcinoma. Br. J. Cancer 108, 644–652 (2013). i squamous cell carcinoma. Br. J. Cancer 108, 644–652 (2013). q J ( ) 49. Chen, J. et al. MiR-1246 promotes SiHa cervical cancer cell proliferation, invasion, and migration through suppression of its target gene thrombospondin 2. Arch. Gynecol. Obstet. 290, 725–732 (2014). g p y 0. Hasegawa, S. et al. MicroRNA-1246 expression associated with CCNG2-mediated chemoresistance and stemness in pancreatic cancer. Br. J. Cancer 111, 1572–1580 (2014). 50. Hasegawa, S. et al. MicroRNA 1246 expression associated with CCNG2 mediated chemoresistance and stemness in pancreatic cancer. Br. J. Cancer 111, 1572–1580 (2014). 51. Sun, Z. et al. MicroRNA-1246 enhances migration and invasion through CADM1 in hepatocellular carcinoma. BMC Cancer 14, cancer. Br. J. Cancer 111, 1572–1580 (2014). 51. Sun, Z. et al. MicroRNA-1246 enhances migration and invasion through CADM1 in hepatocellular carcinoma. BMC Cancer 14, 616 (2014) 1. Sun, Z. et al. MicroRNA-1246 enhances migration and invasion through CADM1 in hepatocellular carcinoma. BMC Cancer 14 616 (2014).h 52. Huang, W., Li, H. & Luo, R. The microRNA-1246 promotes metastasis in non-small cell lung cancer by targeting cytoplasmic polyadenylation element-binding protein 4. Diagn. Pathol. 10, 127 (2015). p y y g p g 3. Leach, D. A. & Buchanan, G. Stromal androgen receptor in prostate cancer development and progression. Cancers (Basel) 9, 10 (2017).h ( ) 54. Thike, A. A. et al. Loss of androgen receptor expression predicts early recurrence in triple-negative and basal-like breast cancer. Mod. Pathol. 27, 352–360 (2014). , ( ) 55. Schroeder, A. et al. Loss of androgen receptor expression promotes a stem-like cell phenotype in prostate cancer through STAT3 signaling. Cancer Res. 74, 1227–1237 (2014). g g ( ) 56. Wake, M. S. & Watson, C. J. STAT3 the oncogene—Still eluding therapy? FEBS J. 282, 2600–2611 (2015). 57. Zhao, C. et al. Feedback activation of STAT3 as a cancer drug-resistance mechanism. Trends Pharmacol. Sci. 37, 47–61 (20 Y S l P fi f d Ak d l d Cl C R 58. Yamamoto, S. et al. www.nature.com/scientificreports/ Critical role of STAT3 in IL-6-mediated drug resistance in human neuroblastoma. Can. Res. 73, 3852–3864 (2013). C O kl G h b f f G l 39. Ara, T. et al. Critical role of STAT3 in IL 6 mediated drug resistance in human neuroblastoma. Can. Res. 73, 3852 3864 (2013). 40. Casanovas, O., Hicklin, D. J., Bergers, G. & Hanahan, D. Drug resistance by evasion of antiangiogenic targeting of VEGF signaling in late-stage pancreatic islet tumors. Cancer Cell 8, 299–309 (2005).f g p 1. Akiyama, K. et al. Inhibition of multidrug transporter in tumor endothelial cells enhances antiangiogenic effects of low-dose metronomic paclitaxel. Am. J. Pathol. 185, 572–580 (2015). metronomic paclitaxel. Am. J. Pathol. 185, 572–580 (2015). p 42. Gilbert, L. A. & Hemann, M. T. DNA damage-mediated induction of a chemoresistant niche. Cell 143, 355–366 (2010).i 42. Gilbert, L. A. & Hemann, M. T. DNA damage-mediated indui 43. Ohmura-Kakutani, H. et al. Identification of tumor endothelial cells with high aldehyde dehydrogenase activity and a h angiogenic phenotype. PLoS ONE 9, e113910–e113917 (2014). 44. King, H. W., Michael, M. Z. & Gleadle, J. M. Hypoxic enhancement of exosome release by breast cancer cells. BMC Cancer 12, 421 (2012). 45. Garcia, N. A., Ontoria-Oviedo, I., Gonzalez-King, H., Diez-Juan, A. & Sepulveda, P. Glucose starvation in cardiomyocytes enhances exosome secretion and promotes angiogenesis in endothelial cells. PLoS ONE 10, e0138849 (2015).l p g g 46. Kucharzewska, P. et al. Exosomes reflect the hypoxic status of glioma cells and mediate hypoxia-dependent activation of vas cells during tumor development. Proc. Natl. Acad. Sci. U.S.A. 110, 7312–7317 (2013). https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \| Fundingh g This research was funded by JSPS Grants-in-Aid for Scientific Research Innovative Areas on integrated analysis and regulation of cellular diversity (to K. Hida, JP18H05092), JSPS Grants-in-Aid for Scientific Research (to N. Maishi, JP18K09715; Y. Hida, JP18H02891; and K. Hida, JP18H02996), and Grants from Japan Agency for Medical Research and Development (to N. Maishi, JP18ck0106198h0003 and K. Hida, JP19ck0106406h0002). p g y Medical Research and Development (to N. Maishi, JP18ck0106198h0003 and K. Hida, JP19ck0106406h0002). Competing interests The authors declare no competing interests. Additional information Supplementary Information The online version contains supplementary material available at https://doi.org/ 10.1038/s41598-021-92879-5. Correspondence and requests for materials should be addressed to K.H. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2021 © The Author(s) 2021 Competing interests h p g The authors declare no competing interests. Additional information Supplementary Information The online version contains supplementary material available at https://doi.org/ 10.1038/s41598-021-92879-5. Correspondence and requests for materials should be addressed to K.H. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and nstitutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2021 https://doi.org/10.1038/s41598-021-92879-5 Scientific Reports \| (2021) 11:13502 \|
https://openalex.org/W2430728306	https://escholarship.org/content/qt4bg3q8kq/qt4bg3q8kq.pdf?t=rxv28i	English	null	Naïve CD8+ T cell derived tumor-specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment	Scientific reports	2,016	cc-by	9,007	Title Title Naïve CD8+ T cell derived tumor-specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment Permalink https://escholarship.org/uc/item/4bg3q8kq UC Davis UC Davis Previously Published Works Title Naïve CD8+ T cell derived tumor-specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment ISSN 2045-2322 Authors Nguyen, Hong Hanh Kim, Therasa Song, Sang Yun et al. Publication Date 2016-06-01 DOI 10.1038/srep28208 Peer reviewed ISSN 2045-2322 Authors Nguyen, Hong Hanh Kim, Therasa Song, Sang Yun et al. Publication Date 2016-06-01 DOI 10.1038/srep28208 Peer reviewed Permalink https://escholarship.org/uc/item/4bg3q8kq Journal Scientific Reports, 6(1) Naïve CD8+ T cell derived tumor- specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment received: 18 March 2016 accepted: 23 May 2016 Published: 16 June 2016 Hong Hanh Nguyen1,, Therasa Kim2,, Sang Yun Song3, Somang Park1, Hyang Hee Cho1, Sung-Hoon Jung4, Jae-Sook Ahn4, Hyeoung-Joon Kim4, Je-Jung Lee1,4, Hee-Ok Kim5, Jae-Ho Cho5 & Deok-Hwan Yang1,4 Despite of the potential implications for cancer immunotherapy, conventional approaches using in vitro expanded CD8+ T cells have suboptimal outcomes, mostly due to loss of functionality from cellular exhaustion. We therefore investigated the phenotypic and functional differences among in vitro activated CD8+ T cells of three different sources, namely naïve (NTeff), memory (MTeff) and tumor- infiltrating lymphocytes (TILeff) from human and mice, to better understand mechanisms behind potent effector functions and potential for overcoming current limitations. In line with the greater proliferation activity and longer telomere lengths of NTeff populations, cells of naïve origin exhibited significantly less amounts of T cell exhaustion markers than those of MTeff and TILeff, and moreover, acquired distinct expression patterns of memory-promoting transcription factors, T-bet and Eomes, induced in a rapid and sustainable manner. NTeff cells appeared to have lower expression of Foxp1 and were refractory to apoptosis upon TGF-β conditioning, implying better survival potential and resistance to tumor- induced immune suppression. Of CD8+ T cell pools activated to tumor-specific CTLs, naïve cell generated effectors possessed the most potent cytotoxic activity, validating implications for use in rational design of adoptive immunotherapy. Adoptive immunotherapy, or the infusion of ex vivo activated and expanded tumor-specific CD8+ T cells into cancer patients, is a strategy involving removal of CD8+ T cells from the tumor environment and provision of stimulatory conditions necessary for their optimal activation, in attempts to overcome poor T-cell responsiveness to tumors. Adoptive T-cell transfer therapy was first attempted in the late 1980s to early 1990s, following the iden- tification of the first tumor associated antigens and isolation of tumor reactive CD8+ T-cell clones from cancer patients. A sufficient number of activated CD8+ effector T cells were obtained and subsequently transferred intra- venously into patients, mediating tumor elimination1. However, current articles have reported that immunother- apy employing the use of CD8+ cytotoxic T lymphocytes (CTLs) is limited by chronic activation and functional impairment of effector cells induced by immunosuppressive factors2–4. Powered by the California Digital Library University of California Powered by the California Digital Library University of California eScholarship.org www.nature.com/scientificreports www.nature.com/scientificreports received: 18 March 2016 accepted: 23 May 2016 Published: 16 June 2016 Naïve CD8+ T cell derived tumor- specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment Investigation of these cells has revealed a so called ‘exhaustion profile’ that includes cell dysfunction, loss of effector function, and progressive increase in the amount and diversity of check point inhibitors such as programmed cell death protein 1 (PD1), cytotoxic T lymphocyte antigen 4 (CTLA4), lymphocyte activation gene 3 protein (LAG3), and killer cell lectin-like receptor 1Research Center for Cancer Immunotherapy, Chonnam National University Hwasun Hospital, Hwasun, Jeollanam-do, Republic of Korea. 2Department of Medical Sciences, Chonnam National University Hwasun Hospital, Hwasun, Jeollanam-do, Republic of Korea. 3Department of Thoracic and Cardiovascular Surgery, Chonnam National University Hwasun Hospital, Hwasun, Jeollanam-do, Republic of Korea. 4Department of Hematology- Oncology, Chonnam National University Hwasun Hospital, Hwasun, Jeollanam-do, Republic of Korea. 5Academy of Immunology & Microbiology (AIM), Institute for Basic Science (IBS), Pohang, Republic of Korea. These authors contributed equally to this work. Correspondence and requests for materials should be addressed to D.H.Y. (email: drydh1685@hotmail.com) Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 1 www.nature.com/scientificreports/ G1 (KLRG1)2–4. It has also been shown that CTL function is altered by transforming growth factor-β (TGF-β), a lymphocyte inhibitor frequently overexpressed in the tumor mircroenvironment (TME) of multiple tumors5,6. Stephen et al.7 have characterized the mechanism of T cell unresponsiveness in TME driven by Forkhead box protein P1 (Foxp1) upregulation in response to TGF-β, by which antigen-primed CD8+ T cells are prevented from proliferating and promoting cytotoxic function.h p g p g y The pool of lymphocytes used for adoptive immunotherapy can be derived from any of the CD8+ subsets, including tumor-infiltrating lymphocytes (TILs), naive T cells (TN), and antigen-experienced memory T cells (TM), which can be divided into central memory (TCM) and effector memory (TEM) subsets that differ in pheno- type, homing, and function8. Memory CD8+ T-cell subsets are studied more often than naïve cells, because of their tumor specificity and greater ability to rapidly proliferate when exposed to previously encountered antigens. Moreover, Berger et al.9 showed that effector cells derived from TCM rather than TEM possess increased potential to survive and establish immunologic memory after infusion in macaques. These findings have been consistent in mice memory subsets10. However, a more recent study has suggested that TN subsets may serve as a suitable can- didate for immunotherapy because of their low level of ‘exhaustion’ in effector cells, sustained replicative potential correlative to telomere length, and minimal effector differentiation translating to greater efficacy after infusion11. Results S f Successful expansion of human naïve, memory, and TIL CD8+ T cells in vitro. CD8+ T cell subsets could be defined by the surface markers CD62L, CD45RO, CCR7, and CD45RA to naïve (CCR7+CD45RA+ or CD62L+CD45RO−) and memory (CD45RA− or CD45RO+) cells. Tumor-infiltrating CD8+ T cells were CD44+, mostly with loss of homing receptors CD62L and CCR7 expression, and composed of two main sub-populations CD45RA− and CD45RA+, suggesting a relatively differentiated status. After MACS isolation and stimulation by CD3/CD28 Dynabeads on anti-CD2 coated plates, successful expansion was demonstrated in all three groups by either CFSE peak division of NTeff and MTeff subsets, or enlarged populations of CD44+CD8+ TILeff cells (Fig. 1A). CFSE staining of TILeff proliferation was observed but peak division was indistinct due to the small number of initial TILs (data not shown). Activation kinetics differed for cell subsets where exponential expansion and accumulation was observed starting at day 3 for NTeff and later at day 5 for MTeff and TILeff cells, and thus further analyses of all effector cells were performed on day 5 post-stimulation.f f Phenotypic characteristics of effector cell populations were assessed by surface expression of activation mark- ers CD62L, CD25, CD44, and OX40. Effector cells from all progenitors (naïve, memory, and TILs) exhibited an effector phenotype with significant up-regulation of activation markers in comparison to their progenitors, and were considered activated as CD62L−CD25+CD44+OX40+ populations (Fig. 1B). Despite variability in the per- centage of effector cells produced among donors, a significantly higher percentage of OX40 expressing cells was observed for NTeff in comparison to MTeff (p < 0.05) or TILeff (p < 0.005) cells (Fig. 1C). To further compare the proliferative potential among cell populations, relative telomere length (RTL), which correlate with replicative capacity, of NTeff, MTeff, and TILeff cells were investigated against CCRF-CEM control cell line. Telomere length was greatest in NTeff, shorter in MTeff, and shortest in TILeff cell subsets (Fig. 1D). This result was consistent with previously published reports of longer telomeres in human naive T cells, leading to increased proliferation of NTeff cells in comparison to MTeff cells after in vitro stimulation9. Exhaustion phenotypes differ among in vitro generated human and murine effector cells NTeff, MTeff, and TILeff. Inhibitory receptors on T cell surfaces such as PD-1, CTLA-4, and KLRG-1, have been shown to facilitate T cell exhaustion by interaction with ligands on antigen presenting cells or tumor cells3–5. Naïve CD8+ T cell derived tumor- specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment Although self or tumor-reactive naïve CD8+ T cells have the capacity to respond to tumor antigens and differ- entiate into cytotoxic effector cells, some major limitations are the relatively low number of self/tumor-specific lymphocytes that result from negative selection during thymic maturation and inadequate magnitude of stimula- tion from antigens in the tumor environment12. Studies have shown that CD8+ T cells can be expanded by T cell growth factor interleukin-2 (IL-2) in the absence of T-cell receptor (TCR) signals, and in this situation, IL-2 can induce a unique signaling pathway that is associated with strong lymphocyte-specific protein tyrosine kinase/ JAK3-dependent activation of the PI3K/AKT pathway13,14. Such signaling pathways not only induce rapid prolif- eration of IL-2 activated naïve CD8+ T cells but also upregulate expression of eomesodermin (Eomes), signifying the differentiation fate of primary naïve effectors into long-lived memory cells. Several current studies performed in murine models have demonstrated the role of T-box transcription factors, T-bet and Eomes, in regulation of both effector and memory functions where high expression of T-bet promotes effector function and Eomes is linked to formation of the long-term memory population15–17. g y p p Considering some of the current limitations in adoptive T-cell transfer, we sought to evaluate the phenotypic and functional differences of effectors derived from naive, memory, and tumor-infiltrating CD8+ T cells (NTeff, MTeff, and TILeff, respectively), and to further investigate the possibility of generating optimal tumor-reactive CTLs with high-dose IL-2 activation for application in clinical trials. We report that exhaustion phenotypes and expression kinetics of T-box transcription factors for NTeff cell populations show possible advantages over other subgroups for use in cell transfer therapy. Our results also suggest that effector cells of naïve origin are relatively resistant to immunosuppressive effects of TGF-β with a trend in lower Foxp1 expression. High-dose IL-2 prim- ing of naïve T cells successfully expands tumor-specific effector cells that exert favorable cytotoxic function, and serves as a potential means for overcoming some of the current barriers in immunotherapy. Results S f We therefore compared the expression of these inhibitory receptors on the three effector cell subtypes. All effector cells showed significant increase of exhaustion phenotypes during proliferation, but NTeff cells showed signifi- cantly less expression of PD-1 and CTLA4 compared to MTeff and TILeff subsets (Fig. 2A). Check point inhibitors showed varying levels of expression dependent on the time elapsed from activation with peak expression on days 4–5 for PD-1, days 5–7 for CTLA-4, and days 4–7 for KLRG-1 (data not shown). To investigate functional Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 2 www.nature.com/scientificreports/ Figure 1. Characteristics of human naïve, memory, and TIL effector CD8+ T cells. (A) Isolation of CD8+ T cell subsets and proliferation assays. MACS isolation and FACS Aria analyses displayed in vitro expansion of naïve (CCR7+CD45RA+), memory (CD45RA−), and TIL (CD44+) subpopulations, 5 days post TCR stimulation. CFSE assay showed positive proliferation results after 3 to 5 days of culture. (B) Activation markers of effector cell subsets. Effector naïve and memory CD8+ cells (NTeff and MTeff, respectively) are characterized by CD62L- CD25+CD44+OX40+ expression. Re-stimulated CD8+ TIL (TILeff) populations also showed similar phenotypes. (C) OX40+ cell percentage of effector cell subsets. NTeff showed significantly higher values compared to MTeff and TILeff (P < 0.05 vs TILeff, *P < 0.005 vs MTeff). (D) Telomere length comparison showed the longest telomeres in NTeff and shortest in TILeff cells (P < 0.05). RTLs of effector cells were measured against CCRF-CEM control cells (1:1 ratio, total 5.0 × 105 cells) with a DAKO Telomere PNA Kit. Data are representative of three to four independent experiments and are presented as mean ± SD. Figure 1. Characteristics of human naïve, memory, and TIL effector CD8+ T cells. (A) Isolation of CD8+ T cell subsets and proliferation assays. MACS isolation and FACS Aria analyses displayed in vitro expansion of naïve (CCR7+CD45RA+), memory (CD45RA−), and TIL (CD44+) subpopulations, 5 days post TCR stimulation. CFSE assay showed positive proliferation results after 3 to 5 days of culture. (B) Activation markers of effector cell subsets. Effector naïve and memory CD8+ cells (NTeff and MTeff, respectively) are characterized by CD62L- CD25+CD44+OX40+ expression. Re-stimulated CD8+ TIL (TILeff) populations also showed similar phenotypes. (C) OX40+ cell percentage of effector cell subsets. NTeff showed significantly higher values compared to MTeff and TILeff (P < 0.05 vs TILeff, P < 0.005 vs MTeff). Results S f (D) Telomere length comparison showed the longest telomeres in NTeff and shortest in TILeff cells (P < 0.05). RTLs of effector cells were measured against CCRF-CEM control cells (1:1 ratio, total 5.0 × 105 cells) with a DAKO Telomere PNA Kit. Data are representative of three to four independent experiments and are presented as mean ± SD. Figure 2. Exhaustion phenotypes of effector CD8+ T cell subpopulations in human and mice. Effector cells (OX40+) derived from naïve cells (NTeff) expressed lower levels of exhaustion markers PD1, CTLA4, KLRG1, and LAG3 compared to effectors from memory cells or TILs (MTeff or TILeff) (A) in human (P < 0.05 for PD-1, P < 0.005 for CTLA4) and (B) in mice at 5 days post-stimulation with anti-CD3/CD28 and anti-CD2. MFI levels are the average result of three experiments and are presented as mean ± SD. (C) In vitro functional analysis of human effector cells by cytokine production. An equal number of progenitors (2 × 105 cells per subtype) activated to OX40+ effectors were characterized by flow cytometry for direct comparison among subtypes. NTeff cells demonstrated a greater increase in perforin+ granzyme B+ populations compared to MTeff and TILeff subsets at two time points (day 3 and 5 post-stimulation). Data are representative of three independent experiments. Figure 2. Exhaustion phenotypes of effector CD8+ T cell subpopulations in human and mice. Effector cells (OX40+) derived from naïve cells (NTeff) expressed lower levels of exhaustion markers PD1, CTLA4, KLRG1, and LAG3 compared to effectors from memory cells or TILs (MTeff or TILeff) (A) in human (P < 0.05 for PD-1, *P < 0.005 for CTLA4) and (B) in mice at 5 days post-stimulation with anti-CD3/CD28 and anti-CD2. MFI levels are the average result of three experiments and are presented as mean ± SD. (C) In vitro functional analysis of human effector cells by cytokine production. An equal number of progenitors (2 × 105 cells per subtype) activated to OX40+ effectors were characterized by flow cytometry for direct comparison among subtypes. NTeff cells demonstrated a greater increase in perforin+ granzyme B+ populations compared to MTeff and TILeff subsets at two time points (day 3 and 5 post-stimulation). Data are representative of three independent experiments. Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 3 www.nature.com/scientificreports/ Figure 3. Expression of T-box transcriptional factors and characteristics of human effector CD8+ T cells under TGF-β conditioning. Results S f Increase in Foxp1 MFI levels pre and post TGF-β conditioning was minimal for naïve effector cells and NTeff showed the least Foxp1 protein expression, although not statistically significant. MFI levels and protein density ratios are the average result of three experiments and shown as mean ± SD. (E) Annexin V assay by flow cytometry. Quadrant plots show the percent distribution of early and late apoptotic cells (first quadrant) with or without 48-hour exposure to TGF-β. Naïve cell derived effectors (NTeff) showed minimal increase in apoptotic cell populations relative to MTeff and TILeff subgroups. All experiments were repeated three times and are presented by one representative figure. i 48-hour exposure to TGF-β. Naïve cell derived effectors (NTeff) showed minimal increase in apoptotic cell populations relative to MTeff and TILeff subgroups. All experiments were repeated three times and are presented by one representative figure. relevance of exhaustion phenotypes, we then evaluated the secretory function of cytotoxic cytokines such as granzyme B, perforin, and IFN- ɤ from different human effectors. During days 3 to 5 post-stimulation, the secre- tion of granzyme B and perforin gradually increased in all effector cells derived from naïve, memory, and TIL populations (Fig. 2C). Among fully activated day 5 effector cells, NTeff possessed the highest expression level of perforin and granzyme B, and similar levels of IFN-ɤ (data not shown) compared to MTeff and TILeff cells. In the murine model, effectors derived from CD8+ naïve and memory phenotype (MP) T cells, and TILs collected from OVAp-expressing EL4-EG7 tumors were successfully proliferated with CD44+OX40+ phenotypes. Expression levels of inhibitory receptors PD-1, KLRG1, and LAG3 in murine CD8+ effectors showed similar patterns to those of human cells (Fig. 2B). Expression kinetics of transcription factors T-bet and Eomes differ among CD8+ T cell effector populations. For further characterization of the CD8+ T cell subpopulations, the expression levels of two T-box transcription factors, T-bet and Eomes, were examined using intracellular staining for FACS analysis and confirmed with Western blot. Although naïve CD8+ T cells expressed lower levels of T-bet and Eomes in com- parison to memory cells before stimulation, activated effectors NTeff, MTeff, and TILeff showed relatively similar MFI levels of T-bet and Eomes, demonstrating the capacity of naïve or memory cells to induce similar effector populations (Fig. 3A). Results S f (A) Intracellular staining of T-bet and Eomes expressed on TCR stimulated effectors NTeff, MTeff, and TILeff. MFI levels are shown for day 0 (initial peak) and day 5 (right side peak) post- stimulation. For TILs, day 0 peaks are lost due to insufficient cell amount pre-expansion. (B) Western blot analysis of naïve and memory T cells, and TILs on days 0, 3, 5 and 8 post TCR activation. Kinetics of T-bet and Eomes expression in differentiating CD8+ T cells suggest rapid and sustained stimulation for NTeff populations. Effector groups from day 5 post-stimulation are treated with TGF-β (10 ng/mL) for 48 hours and checked for Foxp1 expression (day 7 post-stimulation) by (C) intracellular FACS staining and (D) Western blot. Increase in Foxp1 MFI levels pre and post TGF-β conditioning was minimal for naïve effector cells and NTeff showed the least Foxp1 protein expression, although not statistically significant. MFI levels and protein density ratios are the average result of three experiments and shown as mean ± SD. (E) Annexin V assay by flow cytometry. Quadrant plots show the percent distribution of early and late apoptotic cells (first quadrant) with or without 48-hour exposure to TGF-β. Naïve cell derived effectors (NTeff) showed minimal increase in apoptotic cell populations relative to MTeff and TILeff subgroups. All experiments were repeated three times and are presented by one representative figure. Figure 3. Expression of T-box transcriptional factors and characteristics of human effector CD8+ T cell Figure 3. Expression of T-box transcriptional factors and characteristics of human effector CD8+ T cells under TGF-β conditioning. (A) Intracellular staining of T-bet and Eomes expressed on TCR stimulated effectors NTeff, MTeff, and TILeff. MFI levels are shown for day 0 (initial peak) and day 5 (right side peak) post- stimulation. For TILs, day 0 peaks are lost due to insufficient cell amount pre-expansion. (B) Western blot analysis of naïve and memory T cells, and TILs on days 0, 3, 5 and 8 post TCR activation. Kinetics of T-bet and Eomes expression in differentiating CD8+ T cells suggest rapid and sustained stimulation for NTeff populations. Effector groups from day 5 post-stimulation are treated with TGF-β (10 ng/mL) for 48 hours and checked for Foxp1 expression (day 7 post-stimulation) by (C) intracellular FACS staining and (D) Western blot. Results S f After conditioning with TGF-β, flow cytometry detection of apoptosis with Annexin V displayed limited induction of apoptotic cell populations for NTeff cells relative to MTeff and TILeff subgroups (Fig. 3E). Taken together, these results suggest that CD8+ effector subgroups differ in Foxp1 expression and apoptotic response to TGF-β exposure, and that NTeff cells maintain resistance to inhibitory factors induced from the tumor microenvironment. Cytotoxic activity against targeted tumor cells is superior for effector cells of naïve T cell origin in both human and mouse models. To investigate the cytotoxic function of tumor-specific human CTLs, antigen (U266 MM cell line) loaded autologous mature dendritic cells (mDCs) were co-cultured for 7 days with IL-2 primed naïve and memory CD8+ T cells and the generated CTLs were assessed via IFN-ɤ ELISPOT assay and LDH Cytotoxicity Detection Kit. A total of five cell groups were harvested for cytotoxic analyses including CD8+ (Naïve D14, Naïve D7, Naïve D0, and Memory) and CD3+ CTLs. Briefly, Naïve D14 and Naïve D7 cells refer to CD8+ naïve cells co-cultured with high-dose IL-2 (1 μg/mL) for 14 and 7 days respectively, whereas Naïve D0 and Memory cells are co-cultured with conventional dose IL-2 (10 ng/mL) and are of naïve and memory T cell origin, respectively (refer to Methods). IFN-ɤ ELISPOT was measured in the presence and absence of MHC class I specific monoclonal antibody (W6/32, 20 µg/mL) with CTL alone as the control. High dose IL-2 primed Naïve D14 effector cells showed the highest secretion of IFN-ɤ against target U266 cells when co-cultured for one day at a 10:1 ratio of effector (E) to target (T) cells (Fig. 4A.). Although CD3+ cells also secreted higher levels of IFN-ɤ relative to Naïve D7, Naïve D0, and Memory CD8+ cells, they also displayed non-specific secretion as observed in the control CTL alone group. Direct cytotoxicity testing against U266 target cells using the LDH cytotoxic assay revealed enhanced CTL activity for Naïve D14 cells that was significantly greater than other effector CTLs when E:T ratios were at least 10:1 (Fig. 4B.). In the murine model, in vivo CTL activity against inoculated OVAp-expressing EL4-EG7 lymphoma cells in B6 mice was examined for NTeff, MTeff, and TILeff CD8+ effector subsets obtained from OT-1 Thy1.1 mice. Results S f However, considering the lower MFI for pre-stimulated naïve cells, this cell population displays a greater increase in T-bet and Eomes expression levels post-stimulation compared to memory T cells. Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 4 www.nature.com/scientificreports/ To clarify and confirm the expression profile of these T-box transcription factors in the effector subpopula- tions, Western blot analysis was performed using whole cell lysates of naïve and memory T cells, and TILs on days 0, 3, 5 and 8 post TCR activation. Our experiments revealed clear differences in the kinetics of T-bet and Eomes expression during CD8+ T cell activation (Fig. 3B). Effectors derived from naïve cells increased in T-bet expression appreciably starting on day 3, reaching maximum on day 8, whereas memory cell derived effectors gained the highest expression of T-bet earlier on day 3 and were quickly down-regulated on days 5 through 8. TILeff cells demonstrated poor expression of T-bet throughout stimulation time. While naïve and memory cells showed increasing Eomes expression with stimulation, TILs exhibited decreasing intensity during day 5 through 8 post-activation. Such kinetics suggest that NTeff cells express higher levels of both T-bet and Eomes, and expres- sion is sustained for longer periods during activation, in comparison to other effector groups MTeff and TILeff. Collectively, naïve T cell derived effectors appear to differentiate relatively rapidly, proliferate actively, and have potential to be converted into long-lived memory cells. Foxp1 expression and apoptosis studies under TGF-β conditioning show favorable results for CD8+ NTeff cells. To determine the immunosuppressive resistance potential of CD8+ NTeff cells in the tumor microenvironment, we examined expression of Foxp1 among effector cells derived from naïve, memory, and TIL progenitors in the presence of TGF-β. After 5 days of proliferation, effector cells were exposed to TGF-β (10 ng/mL) for 48 hours and Foxp1 expression (day 7 post-stimulation) was evaluated by intracellular FACS staining. MTeff and TILeff populations showed increased Foxp1 expression with TGF-β conditioning, whereas NTeff cells main- tained their lower expression levels in three independent experiments (Fig. 3C). To confirm the increased expres- sion of Foxp1 in CD8+ effector cells in the tumor microenvironment, Western blot analysis was performed for effector T cell subsets under TGF-β exposure using non-stimulated naïve cells as the control. As expected, NTeff populations expressed the lowed amount of Foxp1 protein expression relative to other effector groups, consistent with intracellular staining results (Fig. 3D). Results S f Tumor growth was negative for at least 35 days after tumor inoculation in mice injected with NTeff cells, whereas MTeff, and TILeff injected mice showed positive tumor growth within the same time frame (Fig. 4C). Tumor growth became positive in one NTeff injected mouse at 42 days post inoculation (data not shown). Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 Discussion Specific cytotoxicity lysis percentages against U266 MM target cells were highest for Naïve D14 effectors compared to other groups when E:T ratios were at least 10:1 (p < 0.05). (C) In vivo CTL activity in the murine model. Following injection of NTeff, MTeff, and TILeff CD8+ T cells, OVAp-expressing EL4-EG7 cells (8 × 105 cells/200 µ L) were inoculated subcutaneously in B6 recipient mice in a 1:2 ratio of target tumor to effector cells. All B6 mice injected with TILeff had positive mass (1 cm3) formation after 28 days, whereas tumor growth was suppressed in NTeff harboring B6 mice for longer than 35 days. All data are representative of three independent experiments and shown as mean ± SD. Figure 4. Tumor-specific CTL function of in vitro human and in vivo murine CD8+ T cells. (A) In vitro comparison of tumor-specific CTL function of IL-2 primed human CD8+ T cells against U266 MM cells by IFN-ɤ ELISPOT assay. CTL alone represents the control group and U266 refers to cytotoxic activity of effectors against U266 target cells (E:T ratio of 10:1) in the absence of MHC class I restriction, expressed as the number of IFN-ɤ spots observed. Naïve T cells activated under high-dose IL-2 for 14 days (Naïve D14) displayed highest activity against U266 target cells. (B) LDH cytotoxicity assay of IL-2 primed human CD8+ T cells. Specific cytotoxicity lysis percentages against U266 MM target cells were highest for Naïve D14 effectors compared to other groups when E:T ratios were at least 10:1 (*p < 0.05). (C) In vivo CTL activity in the murine model. Following injection of NTeff, MTeff, and TILeff CD8+ T cells, OVAp-expressing EL4-EG7 cells (8 × 105 cells/200 µ L) were inoculated subcutaneously in B6 recipient mice in a 1:2 ratio of target tumor to effector cells. All B6 mice injected with TILeff had positive mass (1 cm3) formation after 28 days, whereas tumor growth was suppressed in NTeff harboring B6 mice for longer than 35 days. All data are representative of three independent experiments and shown as mean ± SD. effector and memory CD8+ T cell development, short- and long-term effects of T-bet and Eomes have been relatively well identified in the murine model, but are still limited to the context of infection and disease in humans16,24–27. Discussion Current adoptive immunotherapy involves mainly the use of CD3-collected T cells or autologous genetically engineered TILs, with increasing prospect for memory CD8+ cells18,19. Autologous CD8+ subpopulations hold the advantage of avoiding contamination of immunosuppressive regulatory T cells and decrease in terminal differ- entiation with repeated clonal proliferation in cancer patients. In order to demonstrate the superior background of tumor-specific CTLs derived from naïve CD8+ populations we focused on clarifying the different exhaustion characteristics, including expression check point inhibitors and transcription factors, and divergence of cytotoxic function among different effector groups NTeff, MTeff, and TILeff. We found that naïve CD8+ effectors have rela- tively favorable resistance patterns to immunosuppression and strong CTL potency against tumor targets. This may be explained by the “fresh” nature of these cell groups that display lower levels of inhibitory immune check- point receptors such as PD-1, CTLA-4, and KLRG1, leading to a greater activation and replicative potential20. Our results were consistent with reports by Hinrichs et al.21 indicating effectors derived from naïve CD8+ T cells may be the preferable candidate for T-cell adoptive immunotherapy based on evaluation of KLRG1, CD57, CD27 expression and telomere length. Such markers represent tendency of MTeff or TILeff cells to quickly enter into a T cell “exhaustion” phase with elevated levels of inhibitory receptors following long-term exposure to tumor antigen and ultimately acquire a non-functional state4,22,23. Another point of consideration is the expression of two T-box transcriptional factors, T bet and Eomes, as well as their effect on CTL differentiation. Among various transcription factors that function in pairs to regulate Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 5 www.nature.com/scientificreports/ tificreports/ Figure 4. Tumor-specific CTL function of in vitro human and in vivo murine CD8+ T cells. (A) In vitro comparison of tumor-specific CTL function of IL-2 primed human CD8+ T cells against U266 MM cells by IFN-ɤ ELISPOT assay. CTL alone represents the control group and U266 refers to cytotoxic activity of effectors against U266 target cells (E:T ratio of 10:1) in the absence of MHC class I restriction, expressed as the number of IFN-ɤ spots observed. Naïve T cells activated under high-dose IL-2 for 14 days (Naïve D14) displayed highest activity against U266 target cells. (B) LDH cytotoxicity assay of IL-2 primed human CD8+ T cells. Discussion Previous studies have emphasized on the role of high T-bet expression, which is associated with long-term resilience, low expression of inhibitory receptors, and protection from CD8+ T-cell exhaustion27,28, as well as Eomes induced inhibition of cell death leading to generation or survival of memory cells17,29. In our experiments, data showed that NTeff cells gradually increase and maintain high expression levels of both T-bet and Eomes throughout activation, whereas other effectors quickly down-regulated these factors. It has been suggested that high expression of T-bet correlates with a granzyme B+perforin+ phenotype where high perforin secretion was only detected in T-bethiEomeshi/lo cells15. Stable expression of Eomes levels in NTeff and MTeff subsets show their ability to become memory-precursor effector cells (MPECs), whereas TILeff cells enter a terminal differenti- ation phase via down regulation of Eomes. Collectively, our data indicates that NTeff populations co-express sus- tained, high levels of T-beta and Eomes, and a granzyme B+perforin+ phenotype representative of full cytotoxic function. To clarify the relationship between patterns of Foxp1 expression and TGF-β, we activated T cells by in vitro TCR stimulation and evaluated Foxp1 levels of these effectors before and after TGF-β exposure with intention to mimic the immunosuppressive TME. As shown in a recent article7, overexpression of Foxp1 is thought to be mediated by TGF-β signaling through Smad protein interaction, which in turn mediates c-Myc and c-Jun tran- scriptional repression, and is expected to be upregulated after homing to the TME. However, in this study, only whole mouse CD8+ cells were investigated and did not address subsets of human CD8+ T cells. As inferred from our results, CD8+ NTeff are less influenced by tumor-derived inhibitory factor TGF-β, i.e., effectors derived from naïve cells are less affected by tumor suppression mechanisms driven by Foxp1 in response to TGF-β. Importantly, 48 hours post treatment of TGF-β, a smaller percentage of NTeff cells entered apoptosis compared to MTeff and TILeff subsets, which further offers evidence for the superior features of naïve cell derived effectors in resistance to the TGF-β induced apoptosis effect. Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 Methods Ethical approval of studies and informed consent. All experimental protocols were approved by the institutional ethical committee at ChonnamNational University Hwasun Hospital (CNUHH2014-146) and meth- ods were carried out in accordance with the approved guidelines. Human peripheral blood and tumor masses were donated from healthy donors or lung cancer patients after obtaining written informed consent from all sub- jects. Animal experiments were performed with ethical approval from the Chonnam National University Animal Research Committee (HCRL15001-2). Isolation of human CD8+ T cell subsets and tumor-infiltrating lymphocytes (TILs). Peripheral blood mononuclear cells (PBMCs) were obtained from healthy donors by density gradient centrifugation with Lymphoprep (AXIS-SHIED Rodelokka, Oslo, Norway). CD3+ and CD8+ lymphocytes were isolated by a magnetic-activated cell sorter (MACS) using CD3 Microbeads and a CD8 T cell isolation kit (MiltenyiBiotec), respectively. The cells were then labeled with anti-CD8, anti-CCR7, and anti-CD45RA (Ebiosciences) and sorted into naïve and memory CD8+ T cell subsets on a BD fluorescence activated cell sorting (FACS) Aria sorter (BD Biosciences). For extraction of TILs, tissue from lung tumor samples were minced and digested with collagenase (2.5 mg/ml collagenase I) at 37 °C for one hour. Cell suspension was then twice filtered through 100-μm and 40-μm cell strainers (BD Biosciences) to obtain single cells. TILs were isolated from single cells utilizing density gradient centrifugation with Lymphoprep and further purified using CD8 MicroBeads (MiltenyiBiotec). Generation of human CD8+ effector subgroups and proliferation assay. Isolated naïve, memory, and TIL subpopulations were mixed in a 1:1 ratio with Human T-Activator CD3/CD28 Dynabeads (Gibco, Life Technologies) and cultured on CD2-coated plates (CD2-Biotin functional grade, MiltenyiBiotec). Carboxyl fluo- rescein succinimidyl ester (CFSE) studies were performed by labeling cells with 2.5 μM CFSE (Life Technologies) before stimulation. Three and 5 days post-stimulation, CFSE division of cells was determined by flow cytometry and analyzed using FlowJo Version 10.0.7 software. After 5 days of stimulation, activation of effector cells was determined by FACS analysis based on CD62L, CD25, CD44, and OX40 expression. Telomere lengths. Relative telomere lengths (RTL) of effector cells were determined by comparing test cells to a control cell line (human T-cell lymphoblast-like cell line CCRF-CEM, Sigma-Aldrich) using a DAKO Telomere PNA Kit (DAKO). Test cells (NTeff, MTeff, and TILeff) and control cells (cell line 1301) were washed in PBS and mixed in a 1:1 ratio. www.nature.com/scientificreports/ generation, and secondary expansion31. In addition, high-dose IL-2 may strongly induce the proliferation as well as stimulation of naïve CD8+ T cells both in vitro and in vivo32. In our research, high-dose IL-2 (1 μg/mL) stimulated naïve CD8+ T cells, which were then co-cultured with autologous DCs, could successfully generate strong CTLs within a short duration. By evaluation of ELISPOT and LDH assay, Naïve D14 and Naïve D7 both generated CTLs with a high number of IFN-ɤ spots and better (Naïve D14) or similar (Naïve D7) cytotoxic effects compared to CD3+ CTLs. As mentioned, high concentrations of IL-2 (0.1–1 μg/mL) are necessary for stimula- tion of CD8+ naïve T cells through the antigen-independent manner13. Although our antigen-loaded DCs could induce antigen-specific CTLs from CD8+ naive T cells, higher than conventional dose IL-2 was required not only for proliferation of the limited number of antigen-specific naïve CD8+ T cells but also for generation of potent CTLs. During cytokine stimulation, high-dose IL-2 plays an important role in both pre-activation (during 7 days co-culture with naïve T cells) and co-activation (during 7 days co-culture with IL-2 primed T cells and DCs). In the co-activation period, high-dose IL-2 could potentially be used with other ɤc cytokines (IL-7, IL-15) for greater generation of CTLs. However, the optimal exposure times of high-dose IL-2 to tumor-specific naïve CD8+ T cells remains to be determined.f Taken together, our results provide a relatively complete framework for collecting effector cells derived from naïve origin and present a practical method to effectively generate strongly functional NTeff populations with the use of high-dose IL-2 (1 μg/mL). Nowadays, human T lymphocytes, which can be genetically engineered prior to adoptive transfer, can express virtually any target gene, by application of techniques involving genes encoding T-cell receptors (TCRs) or chimeric antigen receptors (CARs) to provide the desired T-cell specificity11. CAR-redirected T lymphocytes (CAR-T cells) are mostly limited to memory and effector T cells33–35 due to potent cytotoxic function. However, the appropriate substrate of effector T cells for clinical use is still controversial and the development of antigen escape variants may represent the limitation of monoclonal specificity. www.nature.com/scientificreports/ To obtain an adequate amount of effector cells with exceeding cytotoxic function, we suggest that naïve cells should be consid- ered as a candidate for CAR-T therapy.f py In summary, primary effector cells derived from naïve CD8+ T cells showed the highest potential for adoptive cell transfer therapy, with lower expression of inhibitory surface markers, higher secretion of cytotoxic cytokines, and improved resistance to TGF-β induced suppression in the TME, in comparison to secondary effectors derived from memory or TIL CD8+ cells. High-dose IL-2 priming amplifies the number of tumor-specific naïve CD8+ T cells without clonal exhaustion and results in the generation of potent tumor-reactive CTLs with capacity to overcome tumor-derived immune suppression. Discussion These findings provide indication for the use of TGF-β inhibitors29, NTeff engineered to express a dominant-negative mutant of the TGF-β receptor30, and Foxp1-deficient NTeff, with objec- tives to enhance the immunosuppressive resistance of lymphocytes in the TME in various clinical applications.h pp y p y pp The essential role of high-dose IL-2 in cancer has now been widely discussed, especially because IL-2 is suspected to optimize all stages of CD8+ T cell response, including primary expansion, contraction, memory Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 6 www.nature.com/scientificreports/ Methods For Western analysis, whole-cell protein lysates were obtained from effector CD8+ T cells at 4 time points (0, 3, 5, and 8 days post-stimulation) with lysis buffer Proprep (INTRON Biotechnology) by suspending 106 cells in 10 μl buffer and incubating on ice for 30 minutes in the presence of Halt Protease & Phosphatase inhibitors cocktail (Thermo Scientific). Cell lysates were resolved by 10% Bis-Tris SDS PAGE Gel, transferred onto PVDF membrane (Merck), blocked with 5% skim milk in TBST buffer containing 0.1% Tween- 20, and probed with the following mAbs: T-bet (4B10, Santa Cruz), Eomes (Y-20, Santa Cruz), Foxp1 (polyclonal, Abcam), and β-actin (C4, Santa Cruz). Quantification of detected protein was performed with an Intelligent Dark Box unit (LAS-3000; Fujifilm) and normalized for loading with the amount of β-actin (1:5000) detected in each lane. Generation of human tumor-specific CTLs. Mature dendritic cells (mDCs) were pulsed with apoptotic body (irradiated U266 cell line) at a 2:1 ratio (2 × 105 DC: 105 U266) from day 6 to day 8 and then presented to CD8+ T cells at a 1:10 ratio (2 × 105 DC: 2 × 106 CD8+ T cell). Naïve and memory T cells were isolated from the same donor and cultured in different conditions. To visualize cytokine-activated proliferation and determine the optimal co-culturing start time with antigen-loaded mDCs post high-dose IL-2 activation, activated naïve CD8+ T cells were labeled with CFSE and analyzed at various time points (Supplementary Fig. S1). Initiation of peak expansion at 7 days post high-dose IL-2 lead to the decision to start co-culture of cells with mDCs after 7 days of high-dose IL-2 pretreatment. Naïve cells were activated with high-dose IL-2 (1 µg/mL) in two conditions; 14 day exposure to high-dose IL-2 (Naïve D14; 7 days pretreatment and during 7 days of mDC co-culture) and 7 day exposure to high-dose IL-2 (Naïve D7; 7 days pretreatment only followed by 7 days of mDC co-culture without high-dose IL-2). For control, we cultured naïve and memory cells under equal conditions; 7 day mDC co-culture with IL-2 (10 ng/mL) and IL-7 (10 ng/mL), labeled Naïve D0 and Memory, respectively. After mDC co-culture, 4 groups of CD8+ CTLs (Naïve D14, Naïve D7, Naïve D0, and Memory) and CD3+ CTLs (cultured as previously described) were harvested for cytotoxic analysis. Cytotoxic functional evaluation of human tumor-specific CTLs. Methods A total of 5 × 105 cells were resuspended in 300 μl of hybridization solution contain- ing 70% formamide either without a probe (unstained control) or with a fluorescein-conjugated telomere PNA probe. The cells were heated for 10 min at 82 °C for DNA denaturation and hybridization was performed over- night at room temperature. Subsequently, cells were washed and resuspended in 0.5 ml of DAKO DNA staining solution and incubated at 2–8 °C for 2 hours. Samples were then analyzed by flow cytometry using logarithmic scale FL1-H for probe fluorescence and linear scale FL3-H for DNA staining. RTLs of the sample cells were cal- culated per manufacturer’s instructions and reported as mean ± standard deviation (SD) for triplicate samples. Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 7 www.nature.com/scientificreports/ Surface and intracellular staining using FACS. For surface staining, cells were labeled with mAbs against various targets including CD8, CD62L, CCR7, CD45RA, CD45RO, CD25, CD44, OX40, PD-1, CTLA-4, and KLRG-1 (Ebiosciences), incubated at 4 °C for 15 minutes, then washed twice with PBS 1% FBS (FACS buffer), and finally fixed in PBS containing 1% paraformaldehyde (Fix buffer). For T-bet (4B10) and Eomes (Dan11mag), and Foxp1 intracellular staining, cells were first labeled with surface markers CD8, CD62L, OX40 (Ebiosciences) for T-bet and Eomes, or CD25 (BD Biosciences), CD8, CD27, and CD44 (Ebiosciences) for Foxp1 detection, respectively, and then fixed and permeabilized with Foxp3/Transcription Factor Fixation/Permeabilization Concentrate and Diluent (Ebiosciences) in a 96-round-well plate. Intracellular labeling for T-bet and Eomes (Ebiosciences), and Foxp1 (LifeSpan Technologies) was performed according to manufacturer’s instruction. For evaluation of Perforin, Granzyme B, and IFN-ɤ cytokine secretion, effector cells were incubated with brefeldin A for 4 hours at 37 °C to disrupt Golgi-mediated transport and accumulate cytokines. Cells were then surface stained, permeabilized with BD Cytofix/Cytoperm™ Fixation/Permeabilization Solution Kit (BD Biosciences), and intracellularly stained as recommended. Flow cytometry data was analyzed using FlowJo Version 10.0.7 software. Apoptosis under TGF-β conditioning. For evaluation of apoptosis, individual effector CD8+ T cell populations were exposed under TGF-β (10 ng/mL) 48 hrs and stained with Annexin V FITC (FITC Annexin V Apoptosis Detection Kit I). Briefly, 1 × 106 cells were washed twice with cold PBS and stained with 5 μL of Annexin-V-FITC in binding buffer for 15 minutes at room temperature. Apoptotic cells were evaluated by flow cytometry according to the manufacturer’s protocol (BD Pharmigen). Western blot protocol. Methods The U266 cell line was chosen as target cells for our cytotoxic assays based on its expression of E-Cadherin and PD-L1 (Supplementary Fig. S2), which are ligands of KLRG-1 and PD-1, respectively. ELISPOT assay (BD Biosience) was performed to quantify antigen-specific IFN-ɤ releasing effector T cells. 2 × 105 of the effector T cells were co-cultured with 2 × 104 U266 target cells in a 96-well nitrocellulose flat-bottom plate for 24 hours at 37 °C. IFN-ɤ ELISPOT was measured in the presence or absence of MHC class I specific monoclonal antibody (W6/32) with CTLs alone as the control, and spots were counted with ImmunoSpot Reader (Cellular Technology Ltd, Ohio). Data is presented as the mean number of spots ± SD of IFN- ɤ secreting cells per well of triplicate samples. p g p p p Evaluation of CD8+ and CD3+ CTL functional activity was performed using the Cytotoxicity Detection Kit LDH (Roche Applied Science, Basel, Switzerland) according to the manufacturer’s protocol36,37. Cytotoxicity of CTLs was calculated according to the following formula: % cell lysis = (experimental – effector spontaneous – low control) x 100/(high control – low control), where “experimental” corresponds to the experimental signal value, “effector spontaneous” to the spontaneous background signal value of the effector cells alone, “low control” to the spontaneous background signal value of target tumor cells alone, and “high control” to the maximum sig- nal value of target cells in medium containing 1% Trixton X-100. Cytotoxicity assays were performed for varying effector cell (E) to target cell (T) ratios, and specific cytotoxicity lysis percentages are reported as the mean ± SD of triplicate samples. T cell purification, in vitro stimulation, and in vivo CTL assay of the murine model. C57BL/6 (B6) mice were purchased from Orient Bio (Iksan, South Korea) and OT-1.Thy1.1 TCR transgenic mice on a B6 back- ground were obtained from the Institute for Basic Science (IBS; Pohang, South Korea). All mice were used in experiments at 6 to 12 weeks of age according to protocol. Pooled lymph nodes from OT-1 transgenic mice were stained with fluorochrome-conjugated antibodies to CD4, CD8, CD25, CD44 and CD62L and sorted to obtain 8 Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 www.nature.com/scientificreports/ CD44−CD62L+ naïve or CD44+ memory-phenotype (MP) CD8+ T cells using FACS. CD8+ TILs were obtained from B6 mice that were inoculated subcutaneously with OVAp-expressing EL4-EG7 tumor cell lines. References Smith, C. et al. Endogenous antigen presentation impacts on T-box transcription factor expression and functional maturation o CD8+ T cells. Blood 120, 3237–3245 (2012).hf 6. Thaventhiran, J. E., Fearon, D. T. & Gattinoni, L. Transcriptional regulation of effector and memory CD8+ T cell fates. Curren Opinion in Immunology 24, 321–328 (2014). 7. Banerjee, A.et al. Cutting edge: the transcription factor eomesodermin enables CD8+ T cells to compete for the memory cell niche J. Immunol. 185, 4988–4992 (2010). 8. Restifo, N. P., Dudley, M. E. & Rosenberg, S. A. Adoptive immunotherapy for cancer: harnessing the T cell response. Nat Rev Immunol. 12, 269–81 (2012). 9. Busch, D. H., Fräßle, S. P., Sommermeyer, D., Buchholz, V. R. & Riddell, S. R. Role of memory T cell subsets for adoptive immunotherapy. Semin Immunol. 16, S1044–5323 (2016). . Wherry, E. J. et al. Molecular signature of CD8+ T cell exhaustion 0. Wherry, E. J. et al. Molecular signature of CD8+ T cell exhaustion during chronic viral infection. Immunity 27, 670–684 (2007). 1. Hinrichs, C. S. et al. Adoptively transferred effector cells derived from naïve rather than central memory CD8+ T cells mediat superior antitumor immunity. Proc Natl Acad Sci. USA 106, 17469–17474 (2009).h y g g y 21. Hinrichs, C. S. et al. Adoptively transferred effector cells derived from naïve rather than central memory CD8+ T cells mediate superior antitumor immunity. Proc Natl Acad Sci. USA 106, 17469–17474 (2009). , p yf y superior antitumor immunity. Proc Natl Acad Sci. USA 106, 17469–17474 (2009).h The blockade of immune checkpoints in cancer immunotherapy. N h 3. Kim, P. S. & Ahmed, R. Features of responding T cells in cancer and chronic infection. Curr. Opin. Immunol. 22, 223–230 (2010). k d l f d l d fi l d ff h hl f l fi h 23. Kim, P. S. & Ahmed, R. Features of responding T cells in cancer and chronic infection. Curr. Opin. Immunol. 22, 223 230 (2010). 24. Makedonas, G. et al. Perforin and IL-2 upregulation define qualitative differences among highly functional virus-specific human CD8 T cells. PLoS Pathog. 6, e1000798 (2010).if CD8 T cells. PLoS Pathog. 6, e1000798 (2010)i 5. Hertoghs, K. M. et al. Molecular profiling of cytomegalovirus-induced human CD8+ T-cell differentiation. J Clin Invest. 120 4077–4090 (2010).i 6. Hersperger, A. R. et al. Increased HIV-specific CD8 T-cell cytotoxic potential in HIV elite controllers is associated with T-be expression. Methods For in vitro stimulation, sorted naïve, MP, and TIL CD8+ T cells were cultured on plate-bound anti-CD3 and anti-CD28 with IL-2 (10 ng/mL) and IL-15 (10 ng/mL). Effector cell surface staining procedure for PD-1, KLRG-1, and LAG3 labeling was identical to human cell methods previously described. In vivo CTL functional assay was performed using OVAp-expressing EL4-EG7 tumor cell lines as target cells. EL4-EG7 tumor cells (0.8 × 105 cells/200 µL) were inoculated subcutaneously in recipient B6 mice and 8 hours later, NTeff, MTeff, and TILeff were injected intra- venously in a 2:1 ratio of effector to target cells. A tumor mass growth of greater than 1 cm3 was considered as positive tumor formation. Statistical analysis. The two-tailed student’s t test was used to determine differences in the mean. P values of less than 0.05 were considered statistically significant. Statistical analysis. The two-tailed student’s t test was used to determine differences in the mean. P values of less than 0.05 were considered statistically significant. References 1. Perret, R. & Ronchese, F. Memory T cells in cancer immunotherapy: which CD8+ T-cell population provides the best protection against tumours? Tissue Antigens 72, 187–194 (2008). g g 2. Wherry, E. J. T cell exhaustion. Nature Immunology 12, 492–499 (2011). 2. Wherry, E. J. T cell exhaustion. Nature Immunology 12, 492–499 (2011). y gy 3. Baistch, L. et al. Extended co-expression of inhibitory receptors by human CD8 T-cells depending on differentiation antigen specificity and anatomical localization. Plos One 7, 1–10 (2012). i y 4. Joel, C. et al. T cell anergy, exhaustion, senescence, and stemness in the tumor microenvironment. Current Opinion in Immunology 25, 1–8 (2012). 5. Caja, F. & Vannucci, L. TGFβ: A player on multiple fronts in the tumor microenvironment. J Immunotoxicol. 12, 300–7 (2015) l ll b k Th l f ll h 6. Flavell, R. A., Sanjabi, S., Wrzesinski, S. H. & Licona-Limón, P. The polarization of immune cells in the tumour environment by TGFbeta. Nat Rev Immunol. 10, 554–67 (2010). 7. Stephen, T. L. et al. Transforming growth factor β-mediated suppression of antitumor T cells requires foxp1 transcription facto expression. Immunity 41, 427–39 (2014).i 8. Rosenberg, S. A. et al. Use of tumor-infiltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma. A preliminary report. N Engl J Med. 310, 1676–1680 (1988).f 9. Berger, C. et al. Adoptive transfer of effector CD8+ T cells derived from central memory cells establishes persistent T cell memory in primates. J. Clin. Invest. 118, 294–305 (2008).f p 0. Klebanoff, C. A. et al. Central memory self/tumor-reactive CD8+ T cells confer superior antitumor immunity compared with effector memory T cells. Proc Natl Acad Sci. USA 102, 9571–9576 (2005).f f 11. Christian, S. H. et al. Human effector CD8+ T cells derived from naïve rather than memory subsets possess superior traits for adoptive immunotherapy. Blood 20, 808–814 (2011).h y 12. Baitsch, L., Fuertes-Marraco, S. A., Legat, A., Meyer, C. & Speiser, D. E. The three main stumbling blocks for anticancer T cells. Trends Immunol. 33, 364–72 (2012). 13. Cho, J. H. et al. Unique features of naïve CD8+ T cell activation by IL-2. J. Immunology 191, 5559–5573 (2013). G l T ll l l d l b l d b ll l C ll 14. Voisinne, G. et al. T cells integrate local and global cues to discriminate between structurally similar antigens. Cell Rep. 26, 1208–19 (2015) 5. References Blood 117, 3799–3808 (2011).l p ( ) 27. Lim, D. G., Bieganowska Bourcier, K., Freeman, G. J. &Hafler, D. A. Examination of CD8+ T cell function in humans using MHC class I tetramers: similar cytotoxicity butvariable proliferation and cytokine production among different clonal CD8+ T cells specific to a single viral epitope. J. Immunol. 165, 6214–6220 (2000).f g p p 28. Intlekofer, A. M.et al. Effector and memory CD8+ T cell fate coupled by T-bet and eomesodermin. Nat. Immunol. 6, 1236–1244 (2005). 29. Connolly, E. C., Freimuth, J. & Akhurst, R. J. Complexities of TGF-β targeted cancer therapy. Int J Bio Sci. 8, 964–978(2012). l f f ll 30. Zitvogel, L. & Kroemer, G. Targeting foxp1 for reinstating anticancer immunosurveillance. Immunity previews, Immunity 41, 245–247 (2014).hf 31. Boyman, O., Cho, J. H. & Sprent, J. The role of interleukin‑2 in memory CD8 cell differentiation. Adv. Exp. Med. Biol. 684, 28–41 (2010). 2. Cho, J. H. et al. An intense form of homeostatic proliferation of naive CD8+ cells driven by IL‑2. J. Exp. Med. 204, 1787–1801 (2007) 33. Pule, M. A. et al. Virus-specific T cells engineered to coexpress tumor-specific receptors: persistence and antitumor activity in individuals with neuroblastoma. Nat. Med. 14, 1264–1270 (2008).i 4. Savoldo, B. et al. CD28 costimulation improves expansion and persistence of chimeric antigen receptor-modifid T cells in lymphoma patients. J. Clin. Invest. 121, 1822–1826 (2011).f 35. Kalos, M. et al. T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci. Transl. Med. 3, 95ra73 (2011). Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 9 www.nature.com/scientificreports/ 36. Neri, S., Mariani, E., Meneghetti, A., Cattini, L.&Facchini, A. Calcein-acetyoxymethyl cytotoxicity assay: standardization of a method allowing additional analyses on recovered effector cells and supernatants.Clinical and Diagnostic Laboratory Immunology 8, 1131–1135 (2001).ii , ( ) 37. Bae, J. et al. A novel immunogenic CS1-specific peptide inducing antigen-specific cytotoxic T lymphocytes targeting multiple myeloma. British Journal of Haematology 157, 687–701 (2012). ( ) 37. Bae, J. et al. A novel immunogenic CS1-specific peptide inducing antigen-specific cytotoxic T lymphocytes targeting multiple myeloma. British Journal of Haematology 157, 687–701 (2012). Acknowledgementsh g This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2013R1A1A1061513), by Chonnam National University Hwasun Hospital Institute for Biomedical Sciences (HCRI 14030-21), and by Chonnam National University Research Institute of Medical Sciences (2013-CURIMS-DR007). g This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2013R1A1A1061513), by Chonnam National University Hwasun Hospital Institute for Biomedical Sciences (HCRI 14030-21), and by Chonnam National University Research Institute of Medical Sciences (2013-CURIMS-DR007). Author Contributions D.-H.Y. and J.-H.C. designed the study; H.H.N., S.P. and H.H.C. performed the experiments; D.-H.Y., H.H.N. and T.K. analyzed the data; H.H.N. and T.K. prepared the manuscript; S.Y.S., J.-S.H., J.-S.A., H.-J.K., H.-O.K. and J.-J.L. critically reviewed the manuscript. D.-H.Y. and J.-H.C. designed the study; H.H.N., S.P. and H.H.C. performed the experiments; D.-H.Y., H.H.N. and T.K. analyzed the data; H.H.N. and T.K. prepared the manuscript; S.Y.S., J.-S.H., J.-S.A., H.-J.K., H.-O.K. and J.-J.L. critically reviewed the manuscript. Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 Additional Information Supplementary information accompanies this paper at http://www.nature.com/srep Supplementary information accompanies this paper at http://www.nature.com/srep Competing financial interests: The authors declare no competing financial interests. How to cite this article: Nguyen, H. H. et al. Naïve CD8+ T cell derived tumor-specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment. Sci. Rep. 6, 28208; doi: 10.1038/srep28208 (2016). How to cite this article: Nguyen, H. H. et al. Naïve CD8+ T cell derived tumor-specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment. Sci. Rep. 6, 28208; doi: 10.1038/srep28208 (2016). How to cite this article: Nguyen, H. H. et al. Naïve CD8+ T cell derived tumor-specific cytotoxic effectors as a potential remedy for overcoming TGF-β immunosuppression in the tumor microenvironment. Sci. Rep. 6, 28208; doi: 10.1038/srep28208 (2016). This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ Scientific Reports \| 6:28208 \| DOI: 10.1038/srep28208 10
https://openalex.org/W3103281328	https://europepmc.org/articles/pmc6689026?pdf=render	English	null	Tunable giant magnetoresistance in a single-molecule junction	Nature communications	2,019	cc-by	6,668	ARTICLE 9-11587-x Results Topographic image of FePc. After adsorption on Au(111), the FePc molecule appears in STM images as a cross with a central protrusion (Fig. 2a). The FePc molecule under study adsorbs at the bridge site of Au(111) (Supplementary Fig. 1). Our DFT calculation shows that the total spin magnetic moment of Fe is 2.05 μB and, in the ground state, it orients close to the plane of the molecule (easy-plane magnetic anisotropy). g y Here, we measured the transport through a single iron phthalocyanine (FePc) molecule using a scanning tunneling microscope (STM), which provides an ideal experimental plat- form to study the single-molecule transport in an atomically well- deﬁned environment5,9,11–13. Our single-molecule device consists of an FePc molecule attached to a Au(111) surface and an STM tip (Fig. 1). The two metal electrodes (Au and tip) serve as the source and drain. Normally, an electrostatic gate is used to shift the chemical potential of the molecular level12,14. In our experi- ment, we use an external magnetic ﬁeld as a handle to control the spin states and the conductance of the single-molecule junction. Temperature and magnetic-ﬁeld dependence of dI/dV spectra. To resolve the molecular spin states, we measured differential conductance (dI/dV) spectra at the Fe center of the molecule. At zero magnetic ﬁeld, the dI/dV spectrum exhibits a sharp dip around EF (Fig. 2a). The dip intensity decreases rapidly with increasing temperature and disappears completely above 8 K (Fig. 2b). The sharp dip at EF is attributed to the Kondo effect since the dip conductance decreases logarithmically with decreasing temperature (Supplementary Fig. 2a), which is a characteristic feature of the Kondo effect27. The FePc molecule belongs to a class of planar metal–phthalocyanine molecules, which have an indispensable role in spintronic applications due to the wide range of tunability of the spin-bearing centers1,15–20. The interaction of the magnetic mole- cules with a metal substrate can lead to a collective quantum behavior, such as the Kondo effect in which the spin moment is screened by the coherent spin–ﬂip process of the conduction elec- trons, giving rise to a Kondo resonance at the Fermi level (EF)10,21. The line shape of Kondo resonance in the conductance spectra is sensitive to the local electronic and magnetic environment, such as conformational changes22, inter-molecular interaction11, charge donation23, axial coordination24,25, and ﬁlling of d- orbitals16,23,26. Figure 2c shows the evolution of the dI/dV spectra at different magnetic ﬁeld. ARTICLE ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x S ingle molecular magnets in contact with metal electrodes are exciting playgrounds to study spin-dependent transport through a single molecule1–5, to examine fundamental magnetic interactions6 and quantum many-body phenomena3,4, and they are essential building blocks for spintronics devices1,7,8. S ingle molecular magnets in contact with metal electrodes are exciting playgrounds to study spin-dependent transport through a single molecule1–5, to examine fundamental magnetic interactions6 and quantum many-body phenomena3,4, and they are essential building blocks for spintronics devices1,7,8. The molecular spin states can be controlled mechanically3, elec- trically4, or by magnetic ﬁeld4,9,10, providing a useful handle to tune the ﬂow of current. For example, using a magnetic ﬁeld, spin-dependent transport phenomena can be tuned at the single- molecule level, such as the negative differential resistance9 or the highly correlated Kondo effect10. in the tip–FePc–Au junction and demonstrate that the electrons travel through the FePc by two possible d orbitals: dz2 and dπ (dπ represents dxz or dyz), and that the single-electron passage through the FePc can be controlled by the external magnetic ﬁeld (Fig. 1). Density functional theory (DFT) calculations reveal that this unique tunability originates from the reorientation of the magnetic moment on the Fe atom. The molecular spin states can be controlled mechanically3, elec- trically4, or by magnetic ﬁeld4,9,10, providing a useful handle to tune the ﬂow of current. For example, using a magnetic ﬁeld, spin-dependent transport phenomena can be tuned at the single- molecule level, such as the negative differential resistance9 or the highly correlated Kondo effect10. Results The sharp Kondo dip at EF becomes shallower but its width remains almost unchanged with increasing magnetic ﬁeld for B < 4 T. At B = 4 T the Kondo dip is completely suppressed. When the magnetic ﬁeld is above 4 T, a peak emerges at EF and becomes more pronounced and broader with increasing magnetic ﬁeld. We use the line shape and spatial distribution of the Kondo resonance of FePc as indicators to monitor the electron pathway Spatial distribution of the Kondo resonance. The dip-to-peak transition of the Kondo resonance indicates the change of the B weak B strong V V I I Au(111) Au(111) Tip Tip a b dxz dz² dxz dz² Fig. 1 A tunable single-molecule device in the STM junction. a and b Schematics of the electron transport process through an FePc molecule adsorbed on a Au(111) surface at different magnetic ﬁelds. The Au substrate and STM tip are the two terminals of the single-molecule device. The magnetic ﬁeld can be viewed as a gate to control the molecular spin states. The arrays of green lines indicate the magnetic ﬁeld lines. During the electron transport through the FePc, the tunneling electron has two possible passages corresponding to two molecular orbitals (dz2 or dxz/dyz). At weak magnetic ﬁeld as in a, the spin direction of Fe is in plane and currents ﬂow by the dxz/dyz orbital. At strong magnetic ﬁeld as in b, Fe spin is aligned to the magnetic ﬁeld and electrons tunnel preferentially through the dz2 orbital b a Fig. 1 A tunable single-molecule device in the STM junction. a and b Schematics of the electron transport process through an FePc molecule adsorbed on a Au(111) surface at different magnetic ﬁelds. The Au substrate and STM tip are the two terminals of the single-molecule device. The magnetic ﬁeld can be viewed as a gate to control the molecular spin states. The arrays of green lines indicate the magnetic ﬁeld lines. During the electron transport through the FePc, the tunneling electron has two possible passages corresponding to two molecular orbitals (dz2 or dxz/dyz). At weak magnetic ﬁeld as in a, the spin direction of Fe is in plane and currents ﬂow by the dxz/dyz orbital. Tunable giant magnetoresistance in a single- molecule junction Kai Yang1,5, Hui Chen1,5, Thomas Pope 2,5, Yibin Hu3,5, Liwei Liu1, Dongfei Wang 1, Lei Tao1, Wende Xiao 1, Xiangmin Fei1, Yu-Yang Zhang 1, Hong-Gang Luo4, Shixuan Du1, Tao Xiang1, Werner A. Hofer1,2 & Hong-Jun Gao 1 Kai Yang1,5, Hui Chen1,5, Thomas Pope 2,5, Yibin Hu3,5, Liwei Liu1, Dongfei Wang 1, Lei Tao1, Wende Xiao 1, Xiangmin Fei1, Yu-Yang Zhang 1, Hong-Gang Luo4, Shixuan Du1, Tao Xiang1, Werner A. Hofer1,2 & Hong Jun Gao 1 Controlling electronic transport through a single-molecule junction is crucial for molecular electronics or spintronics. In magnetic molecular devices, the spin degree-of-freedom can be used to this end since the magnetic properties of the magnetic ion centers fundamentally impact the transport through the molecules. Here we demonstrate that the electron pathway in a single-molecule device can be selected between two molecular orbitals by varying a magnetic ﬁeld, giving rise to a tunable anisotropic magnetoresistance up to 93%. The unique tunability of the electron pathways is due to the magnetic reorientation of the transition metal center, resulting in a re-hybridization of molecular orbitals. We obtain the tunneling electron pathways by Kondo effect, which manifests either as a peak or a dip line shape. The energy changes of these spin-reorientations are remarkably low and less than one millielectronvolt. The large tunable anisotropic magnetoresistance could be used to control electronic trans- port in molecular spintronics. 1 Institute of Physics & University of Chinese Academy of Sciences, Chinese Academy of Sciences, 100190 Beijing, China. 2 School of Natural and Environmental Sciences, Newcastle University, Newcastle NE1 7RU, UK. 3 State Key Laboratory of Infrared Physics, Shanghai Institute of Technical Physics, Chinese Academy of Sciences, 200083 Shanghai, China. 4 School of Physical Science and Technology, Lanzhou University, 730000 Lanzhou, China. 6These authors contributed equally: Kai Yang, Hui Chen, Thomas Pope, Yibin Hu. Correspondence and requests for materials should be addressed to W.A.H. (email: Werner.Hofer@newcastle.ac.uk) or to H.-J.G. (email: hjgao@iphy.ac.cn) 1 NATURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunicatio TURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x a dI/dV spectra taken on the Fe center at 0.4 K and zero magnetic ﬁeld, showing a Kondo dip superimposed on a broad feature near EF (setpoint: I = 0.2 nA, Vb = −0.1 V). Inset: STM image of an FePc on Au(111) (3.2 nm × 3.2 nm, I = 10 pA, Vb = −0.2 V). b Evolution of the dI/dV spectra of FePc at increasing temperatures in the absence of magnetic ﬁeld (setpoint: I = 0.3 nA, Vb = −60 mV). Successive spectra are offset for clarity. The dip at EF vanishes above 8 K, making two broad resonances visible. The broad two-peak feature was originally attributed to a Kondo feature in ref. 15. However, in this work we note the similarity between the spectra and the density of states for the underlying gold bulk (Supplementary Fig. 7), suggesting that this is in fact the source of the broader feature. c Evolution of the dI/dV spectra of FePc with increasing magnetic ﬁeld at 0.4 K (setpoint: I = 0.3 nA, Vb = −60 mV), showing a dip-to-peak transition. a.u., arbitrary units molecular spin states with magnetic ﬁeld. To gain a better understanding of the spin states, we studied the spatial distribu- tion of the Kondo resonance over the FePc molecule by per- forming spectroscopic mapping near EF under varying magnetic ﬁelds (Fig. 3 and Supplementary Fig. 3). At zero or low ﬁelds, the Kondo resonance shows a non-circular spatial distribution (Fig. 3a). This distribution is not due to drifting during mea- surement, as conﬁrmed by dI/dV spectra taken along different axes over the FePc molecule (Fig. 3b, c). It shows that the Kondo resonance (dip) decays slower along one axis of the feature (Fig. 3c) than the other (Fig. 3b). In comparison, at high magnetic ﬁelds (above 4 T), the Kondo resonance (peak) is more localized on the molecular center, with a radial symmetry (Fig. 3d–f). At intermediate ﬁeld strengths of 3–6 T, the Kondo resonance is largely invisible (Supplementary Fig. 3c, d). different d orbital character in the ground states. A level crossing within the low-energy spin multiplet with quenched orbital angular momentum cannot explain the change of the spatial distribution of the Kondo resonance with increasing magnetic ﬁeld, since this scenario would only result in a tiny change of the orbital composition in the wave function. Reorientation of the Fe magnetic moment. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x Sample bias (mV) Kondo FePc [112] 1 nm [110] –200 –100 0 100 200 dI/dV (a.u.) a Sample bias (mV) –20 –10 0 10 20 Sample bias (mV) –20 –10 0 10 20 Sample bias (mV) Kondo FePc [112] 1 nm [110] –200 –100 0 100 200 dI/dV (a.u.) dI/dV (a.u.) dI/dV (a.u.) a b c 0.4 K 4.2 K 5.3 K 6.3 K 7.8 K 8.9 K 10.9 K 0 T 2 T 3 T 4 T 5 T 6 T 9 T 11 T At 0 tesla At 0.4 K Fig. 2 dI/dV spectra under variation of temperature and magnetic ﬁeld. a dI/dV spectra taken on the Fe center at 0.4 K and zero magnetic ﬁeld, showing a Kondo dip superimposed on a broad feature near EF (setpoint: I = 0.2 nA, Vb = −0.1 V). Inset: STM image of an FePc on Au(111) (3.2 nm × 3.2 nm, I = 10 pA, Vb = −0.2 V). b Evolution of the dI/dV spectra of FePc at increasing temperatures in the absence of magnetic ﬁeld (setpoint: I = 0.3 nA, Vb = −60 mV). Successive spectra are offset for clarity. The dip at EF vanishes above 8 K, making two broad resonances visible. The broad two-peak feature was originally attributed to a Kondo feature in ref. 15. However, in this work we note the similarity between the spectra and the density of states for the underlying gold bulk (Supplementary Fig. 7), suggesting that this is in fact the source of the broader feature. c Evolution of the dI/dV spectra of FePc with increasing magnetic ﬁeld at 0.4 K (setpoint: I = 0.3 nA, Vb = −60 mV), showing a dip-to-peak transition. a.u., arbitrary units Sample bias (mV) –20 –10 0 10 20 dI/dV (a.u.) b 0.4 K 4.2 K 5.3 K 6.3 K 7.8 K 8.9 K 10.9 K At 0 tesla Sample bias (mV) –20 –10 0 10 20 dI/dV (a.u.) c 0 T 2 T 3 T 4 T 5 T 6 T 9 T 11 T At 0.4 K Sample bias (mV) –20 –10 0 10 20 dI/dV (a.u.) c 0 T 2 T 3 T 4 T 5 T 6 T 9 T 11 T At 0.4 K b a c Sample bias (mV) Sample bias (mV) Sample bias (mV) Fig. 2 dI/dV spectra under variation of temperature and magnetic ﬁeld. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x To conﬁrm this, we performed noncollinear DFT calculations for both the in-plane conﬁguration (Fe magnetic moment in the plane of the mole- cule), and the out-of-plane conﬁguration (Fe magnetic moment aligned to the magnetic ﬁeld, perpendicular to the molecular plane). Our DFT calculations show that at zero magnetic ﬁeld the Fe magnetic moment is in the plane of the molecule. The reorientation of the magnetic moment at strong ﬁeld is possible because the Zeeman energy due to the external magnetic ﬁeld (~4 T) is comparable with the in-plane magnetic anisotropy of the Fe spin (~1 meV from DFT calculations). Similar energy scale of this magnetic anisotropy energy was obtained by other simulations30. Figure 4a, b show the calculated band- decomposed partial density of states (PDOS) of the d orbitals for both magnetic conﬁgurations. By comparing spatial dis- tribution of the molecular orbitals closest to EF (Fig. 4c, d) with the experimental spatial distribution of the Kondo resonance, we ﬁnd that the non-circular depression at low ﬁeld is due to the dπ orbital (Fig. 4c), while the circular protrusion at high ﬁeld is due to the dz2 orbital (Fig. 4d). We also ﬁnd that the relative population of the two orbitals changes as the magnetic moment orientates onto the direction of the magnetic ﬁeld. The re-hybridization of the molecular orbitals of the magnetic ground state thus gives rise to the change of spatial distribution of the zero-bias conductance (Kondo resonance) over the FePc molecule as shown in Fig. 3. Note that the reorientation of the magnetic moment doesn’t mean that the Zeeman energy induces direct orbital transitions, since only the low-energy density distribution of the d orbitals near EF changes. g y y g The spatial distribution of the Kondo resonance is determined by the electron orbital responsible for the resonance, that is, the local distribution of electron charge of the responsible orbital11,28. The two different spatial distributions (non-circular or circular) suggest two distinct Kondo screening channels originating from two d orbitals at different magnetic ﬁelds. The change of spatial distributions of Kondo resonance can be understood by considering the different spatial symmetries of the dπ and dz2 orbitals (Fig. 1). Results At strong magnetic ﬁeld as in b, Fe spin is aligned to the magnetic ﬁeld and electrons tunnel preferentially through the dz2 orbital NATURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunications 2 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x NATURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunications NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x The Kondo resonance (peak) displays a nearly circular spatial distribution in d Sample bias (mV) –20 –10 0 10 20 Sample bias (mV) –20 –10 0 0 1 2 3 4 5 6 10 20 Distance from the center (Å) 0 1 2 3 4 5 6 Distance from the center (Å) 5 Å 5 Å 0 T 0 T b c 5 Å 0 T a Sample bias (mV) –20 –10 0 10 20 0 1 2 3 4 5 6 Distance from the center (Å) 5 Å 0 T c b c Distance from the center (Å) Sample bias (mV) LO HI –20 –10 0 0 1 2 3 4 5 6 10 20 Distance from the center (Å) Sample bias (mV) –20 –10 0 10 20 0 1 2 3 4 5 6 Distance from the center (Å) 5 Å 5 Å 9 T 9 T e f HI Sample bias (mV) –20 –10 0 10 20 0 1 2 3 4 5 6 Distance from the center (Å) 5 Å 9 T f Sample bias (mV) LO –20 –10 0 0 1 2 3 4 5 6 10 20 Distance from the center (Å) 5 Å 9 T e LO HI 5 Å 9 T d d LO HI Fig. 3 Spatial distribution of the Kondo resonance at different magnetic ﬁelds. a dI/dV mapping taken around the Fermi energy at 0 T (setpoint: I = 0.3 nA, Vb = −40 mV. Image size: 2.2 nm × 2.2 nm). b and c dI/dV spectra taken along different directions on the FePc molecule (as indicated by the arrows). The spectra start from the Fe center and end at about 6 Å away. The black dashed line roughly indicates the position where the Kondo resonance is too weak to see in the dI/dV spectra. Inset: STM image (2.2 nm × 2.2 nm) of the FePc. d–f, same as a–c except taken at B = 9 T. The Kondo resonance (peak) displays a nearly circular spatial distribution in d The re-hybridization in the ground state of FePc can also be directly visualized in the change of the line shapes of the Kondo resonance, in addition to its spatial distribution. The dip-to-peak transition of the Kondo resonance with external magnetic ﬁeld (Fig. 2c) can be understood by considering the different spatial symmetries of the d orbitals. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x The line shapes of the Kondo resonance are determined by the Fano interference during the electron tunneling process31. Electrons originating from the tip either tunnel through the discrete Kondo resonance or the continuum states of the conduction electrons. The Fano asymmetry factor q is proportional to the ratio of the probabilities of the two paths. Since different d orbitals couple differently to the tip due to their alignments and shapes, this will result in different Fano interference with the tunneling into the Au state as the magnetic ﬁeld increases. At high magnetic ﬁeld, the magnetic ground state is mainly contributed by the dz2 orbital (Fig. 4d), which couples more to the tip states due to its favorable spatial distribution (Fig. 1b). Hence at high ﬁeld the Kondo resonance appears as a peak (large q-factor). Similarly, the weaker coupling between the dπ and the tip states results in the Kondo dip at low magnetic ﬁelds (Fig. 1a). magnetic moment is reoriented. In the in-plane conﬁguration, the magnetic moment is nearly perpendicular to the magnetic ﬁeld and thus the product is small. At high ﬁeld, the reorientation allows for a much larger product and, so, noticeable Zeeman splitting (Supplementary Fig. 5). Tunneling anisotropic magnetoresistance (TAMR) of FePc. The magnetic reorientation of the Fe center by varying the magnetic ﬁeld also leads to the TAMR effect. The TAMR effect describes the dependence of the magnetoresistance on the magnetization orientation, and has been found in metal ﬁlm33,34, single adatoms29,35, and molecular tunneling junc- tions36–39. In addition, the TAMR effect with non-magnetic electrodes has been shown in mechanical break junctions40,41. Here, the magnetic FePc molecule is attached directly to a gold metal electrode in our tunnel junction and there is no addi- tional magnetic layer. g y The TAMR can be deﬁned as TAMR(M) = (dI(0)/dV–dI(M)/ dV)/(dI(0)/dV), where dI(M)/dV denotes the differential conductance of FePc at the magnetic reorientation with an out-of-plane component M39. Increasing the magnetic ﬁeld corresponds to a change of the magnetization angle from 0° to 90°. We show the TAMR of FePc as a function of bias voltage in Fig. 4e. In the Kondo resonance region (between ±5 mV), the TAMR can be enhanced up to 93%, much larger than the TAMR values reported in other molecular-based TAMR devices36–39. At intermediate magnetic ﬁeld (Fig. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x Sample bias (mV) LO HI LO HI –20 –10 0 0 1 2 3 4 5 6 10 20 Distance from the center (Å) Sample bias (mV) –20 –10 0 10 20 Sample bias (mV) –20 –10 0 10 20 Sample bias (mV) –20 –10 0 0 1 2 3 4 5 6 10 20 Distance from the center (Å) 0 1 2 3 4 5 6 Distance from the center (Å) 0 1 2 3 4 5 6 Distance from the center (Å) 5 Å 5 Å 5 Å 5 Å 5 Å 5 Å 0 T 0 T 0 T 9 T 9 T 9 T a b c d e f Fig. 3 Spatial distribution of the Kondo resonance at different magnetic ﬁelds. a dI/dV mapping taken around the Fermi energy at 0 T (setpoint: I = 0.3 nA, Vb = −40 mV. Image size: 2.2 nm × 2.2 nm). b and c dI/dV spectra taken along different directions on the FePc molecule (as indicated by the arrows). The spectra start from the Fe center and end at about 6 Å away. The black dashed line roughly indicates the position where the Kondo resonance is too weak to see in the dI/dV spectra. Inset: STM image (2.2 nm × 2.2 nm) of the FePc. d–f, same as a–c except taken at B = 9 T. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x The hybridization of dπ orbital with the neighboring atoms inside the FePc molecule results in a more extended Kondo resonance distribution at low ﬁeld; while the dz2 orbital, due to its different symmetry, hybridizes less with the molecular orbitals, giving rise to the more localized spatial distribution of Kondo resonance at high magnetic ﬁeld. The change of orbital characteristics results from the reorientation of the Fe magnetic moment induced by the applied magnetic ﬁeld. This reorientation changes the orbital character through spin–orbit interaction29. The applied magnetic ﬁeld ﬂips the magnetic moment of Fe from the in-plane to the out-of-plane direction when the Zeeman energy is larger than the magnetic anisotropy energy. Thus, due to the spin–orbit interaction, the reorientation of the magnetic moment changes the mixing of 3 3 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x –6 –4 –2 0 2 4 6 –100 –80 –60 –40 –20 0 20 0 2 4 6 8 10 –100 –80 –60 –40 –20 0 TAMR (%) TAMR (%) –100 –50 0 50 100 150 200 PDOS (a.u.) E−EF (meV) b B = 11 T PDOS (a.u.) B = 0 T a c dz2 d dxy d Sample bias (mV) Magnetic field B (T) e f Fig. 4 Controlling the tunneling anisotropic magnetoresistance (TAMR). a and b Calculated lm-decomposed partial density of states (PDOS) of the dxy, dπ, and dz2 bands of the Fe atom with a magnetization vector in the plane of the molecule a and in the direction of the applied magnetic ﬁeld, perpendicular to the plane of the molecule b. The insets show the direction of the magnetic moment schematically. The energy is given with respect to EF. Black lines represent the energy of the molecular orbital closest to EF. c and d Band-decomposed charge density plots for the molecular orbitals close to EF, showing the similar spatial distribution of the Kondo resonance as in Fig. 3a, d. We note that the orbital density on the C atoms in c is unlikely to have a signiﬁcant overlap with the substrate and thus will not contribute signiﬁcantly to the signals in the dI/dV maps. e TAMR values at different bias voltages calculated from the dI/dV spectra taken on the Fe center at B = 0 and 11 T. f TAMR values at different magnetic ﬁeld calculated from the dI/dV spectra taken on the Fe center at the bias voltage of −1.06 mV. a.u. arbitrary units –100 –50 0 50 100 150 200 PDOS (a.u.) E−EF (meV) b B = 11 T PDOS (a.u.) B = 0 T a c dz2 d dxy d a c PDOS (a.u.) b d d 0 2 4 6 8 10 –100 –80 –60 –40 –20 0 TAMR (%) Magnetic field B (T) f –6 –4 –2 0 2 4 6 –100 –80 –60 –40 –20 0 20 TAMR (%) Sample bias (mV) e e TAMR (%) Fig. 4 Controlling the tunneling anisotropic magnetoresistance (TAMR). NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x a and b Calculated lm-decomposed partial density of states (PDOS) of the dxy, dπ, and dz2 bands of the Fe atom with a magnetization vector in the plane of the molecule a and in the direction of the applied magnetic ﬁeld, perpendicular to the plane of the molecule b. The insets show the direction of the magnetic moment schematically. The energy is given with respect to EF. Black lines represent the energy of the molecular orbital closest to EF. c and d Band-decomposed charge density plots for the molecular orbitals close to EF, showing the similar spatial distribution of the Kondo resonance as in Fig. 3a, d. We note that the orbital density on the C atoms in c is unlikely to have a signiﬁcant overlap with the substrate and thus will not contribute signiﬁcantly to the signals in the dI/dV maps. e TAMR values at different bias voltages calculated from the dI/dV spectra taken on the Fe center at B = 0 and 11 T. f TAMR values at different magnetic ﬁeld calculated from the dI/dV spectra taken on the Fe center at the bias voltage of −1.06 mV. a.u. arbitrary units In conclusion, we have demonstrated that the pathway of single electrons through the orbitals of a magnetic molecule can be tuned at a very low energy by varying the external magnetic ﬁeld. This unique tunability originates from the reorientation of the magnetic moment on the metal center, which alters the electron distribution in the d orbitals. While at zero ﬁeld the electron density scattered through the Kondo resonance will be exclusively at the dπ orbital, it will pass through the dz2 orbital at high ﬁeld, and a varying part of the density will pass through both orbitals in the intermediate regime. Note that a local control of the spin reorientation could in principle be achieved by using the mag- netic ﬁeld from a spin-polarized STM tip42. Our work shows that the multi-orbital nature and the spin–orbit coupling can be employed to control the single-electron process in a single- molecule device at sub-meV energies. Discussion T d To demonstrate the generality of tuning the tunneling pathway with magnetic ﬁelds, we also studied the Kondo effect of an Fe- porphyrin derivative on Au(111). The Kondo resonance there shows similar dip-to-peak transitions with magnetic ﬁelds (Sup- plementary Fig. 6d). We expect the behavior of the magnetic moment to be replicated in the Fe-porphyrin derivative. The dI/ dV mapping of the Kondo resonance exhibits a similar transition from an extended distribution to a more concentrated distribu- tion (Supplementary Fig. 6e–h). We can treat the electron tunneling through the FePc molecule as single-electron transport. The time lag from one electron to the next through the molecule is in the range of nanoseconds (cor- responding to a tunneling current of nanoamperes), but other processes at the atomic scale occur at a much faster time scale. The fastest processes, electronic relaxations, typically occur within femtoseconds. The dynamics of the atomic cores, manifested by phonons and vibrations, are within the picosecond range, with relaxation times such that they typically do not exceed a few hundred picoseconds. The transport phenomena in the STM junction are thus too slow for vibrations or electronic relaxations to inﬂuence electron transport. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x 4f), the TAMR effect (with respect to the in-plane magnetization at zero ﬁeld) increases monotonically with the magnetic ﬁeld near zero bias. We ﬁt the Kondo resonances at different magnetic ﬁeld with a Fano function (Supplementary Fig. 4)21. The ﬁtting shows that the Fano asymmetry factor q drops suddenly from positive to negative values across the transition magnetic ﬁeld (4 T), corresponding to the dip-to-peak spectral transition. The line width of the Kondo resonance increases with magnetic ﬁeld above 6 T due to Zeeman splitting27,32. This sudden change in the spectral feature—going from a conﬁguration in which it is unaffected by Zeeman splitting into a conﬁguration in which Zeeman splitting is measured—supports the argument that the NATURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunications 4 4 ARTICLE References Efﬁciency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comp. Mater. Sci. 6, 15–50 (1996). 12. Martínez-Blanco, J. et al. Gating a single-molecule transistor with individual atoms. Nat. Phys. 11, 640–644 (2015). 45. Kresse, G. & Furthmüller, J. Efﬁcient iterative schemes for ab initio total- energy calculations using a plane-wave basis set. Phys. Rev. B 54, 11169–11186 (1996). 13. Esat, T., Friedrich, N., Tautz, F. S. & Temirov, R. A standing molecule as a single-electron ﬁeld emitter. Nature 558, 573–576 (2018). g 14. Fernández-Torrente, I. et al. Gating the charge state of si 14. Fernández-Torrente, I. et al. Gating the charge state of single molecules by local electric ﬁelds. Phys. Rev. Lett. 108, 036801 (2012). 46. Hobbs, D., Kresse, G. & Hafner, J. Fully unconstrained noncollinear magnetism within the projector augmented-wave method. Phys. Rev. B 62, 11556–11570 (2000). 15. Gao, L. et al. Site-speciﬁc Kondo effect at ambient temperatures in iron-based molecules. Phys. Rev. Lett. 99, 106402 (2007). 47. Blöchl, P. E. Projector augmented-wave method. Phys. Rev. B 50, 17953–17979 (1994). 16. Stepanow, S. et al. Mixed-valence behavior and strong correlation effects of metal phthalocyanines adsorbed on metals. Phys. Rev. B 83, 220401 (2011). 48. Kresse, G. & Joubert, D. From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 59, 1758–1775 (1999). 17. Tsukahara, N. et al. Evolution of Kondo resonance from a single impurity molecule to the two-dimensional lattice. Phys. Rev. Lett. 106, 187201 (2011). 18. Ballav, N. et al. Emergence of on-surface magnetochemistry. J. Phys. Chem. Lett. 4, 2303–2311 (2013). 49. Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865–3868 (1996). 19. Auwarter, W., Ecija, D., Klappenberger, F. & Barth, J. V. Porphyrins at interfaces. Nat. Chem. 7, 105–120 (2015). 50. Grimme., S. Semiempirical GGA-type density functional constructed with a long-range dispersion correction. J. Comput. Chem. 27, 1787–1799 (2006). 20. Pope, T., Du, S., Gao, H.-J. & Hofer, W. A. Electronic effects and fundamental physics studied in molecular interfaces. Chem. Commun. 54, 5508–5517 (2018). 51. Liechtenstein, A. I., Anisimov, V. I. & Zaanen, J. Density-functional theory and strong interactions: orbital ordering in Mott–Hubbard insulators. Phys. Rev. B 52, R5467–R5470 (1995). 21. Madhavan, V. et al. Tunneling into a single magnetic atom: spectroscopic evidence of the Kondo resonance. Science 280, 567–569 (1998). 52. Dudarev, S. L. et al. References 36. Grünewald, M., Homonnay, N., Kleinlein, J. & Schmidt, G. Voltage-controlled oxide barriers in organic/hybrid spin valves based on tunneling anisotropic magnetoresistance. Phys. Rev. B 90, 205208 (2014). 1. Bogani, L. & Wernsdorfer, W. Molecular spintronics using single-molecule magnets. Nat. Mater. 7, 179–186 (2008). 1. Bogani, L. & Wernsdorfer, W. Molecular spintronics using single-molecule magnets. Nat. Mater. 7, 179–186 (2008). 37. Barraud, C. et al. Unidirectional spin-dependent molecule-ferromagnet hybridized states anisotropy in cobalt phthalocyanine based magnetic tunnel junctions. Phys. Rev. Lett. 114, 206603 (2015). 2. Liang, W. et al. Kondo resonance in a single-molecule transistor. Nature 417, 725–729 (2002). 3. Parks, J. J. et al. Mechanical control of spin states in spin-1 molecules and the underscreened Kondo effect. Science 328, 1370–1373 (2010). j y 38. Kamiya, T., Miyahara, C. & Tada, H. Large tunneling anisotropic magnetoresistance in La0.7Sr0.3MnO3/pentacene/Cu structures prepared on SrTiO3 (110) substrates. Appl. Phys. Lett. 110, 032401 (2017). 4. Roch, N. et al. Quantum phase transition in a single-molecule quantum dot. Nature 453, 633–637 (2008). pp y 39. Wang, K. et al. Effect of orbital hybridization on spin-polarized tunneling across Co/C60 interfaces. ACS Appl. Mater. Interfaces 8, 28349–28356 (2016). 5. Schmaus, S. et al. Giant magnetoresistance through a single molecule. Nat. Nanotechnol. 6, 185–189 (2011). 6. Hirjibehedin, C. F., Lutz, C. P. & Heinrich, A. J. Spin coupling in engineered atomic structures. Science 312, 1021–1024 (2006). 40. Hayakawa, R. et al. Large magnetoresistance in single-radical molecular junctions. Nano Lett. 16, 4960–4967 (2016). 7. Sanvito, S. Molecular spintronics. Chem. Soc. Rev. 40, 3336–3355 (2011 8. Aradhya, S. V. & Venkataraman, L. Single-molecule junctions beyond electronic transport. Nat. Nanotechnol. 8, 399–410 (2013). 41. Xie, Z. et al. Large magnetoresistance at room temperature in organic molecular tunnel junctions with nonmagnetic electrodes. ACS Nano 10, 8571–8577 (2016). 9. Warner, B. et al. Tunable magnetoresistance in an asymmetrically coupled single-molecule junction. Nat. Nanotechnol. 10, 259–263 (2015). 42. Yan, S. et al. Control of quantum magnets by atomic exchange bias. Nat. Nanotechnol. 10, 40–45 (2015). 10. Scott, G. D. & Natelson, D. Kondo resonances in molecular devices. ACS Nano 4, 3560–3579 (2010). 43. Kresse, G. & Hafner, J. Ab initio molecular dynamics for liquid metals. Phys. Rev. B 47, 558–561 (1993). 11. Mugarza, A. et al. Spin coupling and relaxation inside molecule–metal contacts. Nat. Commun. 2, 490 (2011). 44. Kresse, G. & Furthmüller, J. ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x 24. Stróżecka, A., Soriano, M., Pascual, J. I. & Palacios, J. J. Reversible change of the spin state in a manganese phthalocyanine by coordination of CO molecule. Phys. Rev. Lett. 109, 147202 (2012). technique with a sinusoidal modulation (0.09 mV rms–0.3 mV rms) at a frequency of 973 Hz. The electrochemically etched tungsten tips were calibrated against the Au(111) surface state before spectroscopic measurements to ensure no tip-related features on the recorded dI/dV spectra. y 25. Liu, L. W. et al. Reversible single spin control of individual magnetic molecule by hydrogen atom adsorption. Sci. Rep. 3, 1210 (2013). 26. Kügel, J. et al. Relevance of hybridization and ﬁlling of 3d orbitals for the Kondo effect in transition metal phthalocyanines. Nano Lett. 14, 3895–3902 (2014). Ab-initio calculations. All non-collinear spin-polarized DFT calculations were performed with the Vienna ab-initio Simulation Package (VASP)43–46 and the projector augmented wave (PAW) method47,48. The exchange-correlation potential was described by the Perdew–Burke–Ernzerhof functional49 extended to incor- porate a Van der Waals correction50. We perform a geometry minimization on the system, until the residual forces were smaller than 0.02 eV Ang−1. To improve the description of band structure and magnetic properties of the adsorbed FePc molecule, a Hubbard-like + U correction is adopted51,52. The Hubbard-like on-site Coulomb and exchange parameters for the Fe 3d orbitals are, respectively, chosen to be U = 2 eV and J = 1 eV. These values have been used in previous studies on the same system53–55. In calculating the density of states, we employed Gaussian smearing with a width of 5 meV. 27. Cronenwett, S. M., Oosterkamp, T. H. & Kouwenhoven, L. P. A tunable Kondo effect in quantum dots. Science 281, 540–544 (1998). 28. Pacchioni, G. E. et al. Two-orbital Kondo screening in a self-assembled metal–organic complex. ACS Nano 11, 2675–2681 (2017). g p 29. Néel, N. et al. Tunneling anisotropic magnetoresistance at the single-atom limit. Phys. Rev. Lett. 110, 037202 (2013). 30. Hu, J. & Wu, R. Control of the magnetism and magnetic anisotropy of a single-molecule magnet with an electric ﬁeld. Phys. Rev. Lett. 110, 097202 (2013). 31. Markus, T., Andreas, J. H. & Wolf-Dieter, S. Spectroscopic manifestations of the Kondo effect on single adatoms. J. Phys.: Condens. Matter 21, 053001 (2009). Received: 5 February 2019 Accepted: 9 July 2019 Received: 5 February 2019 Accepted: 9 July 2019 Received: 5 February 2019 Accepted: 9 July 2019 34. Hervé, M. et al. Stabilizing spin spirals and isolated skyrmions at low magnetic ﬁeld exploiting vanishing magnetic anisotropy. Nat. Commun. 9, 1015 (2018). 35. Serrate, D. et al. Imaging and manipulating the spin direction of individual atoms. Nat. Nanotechnol. 5, 350–353 (2010). Methods E i t Experiment. The atomically ﬂat Au(111) surface was prepared by repeated cycles of sputtering with argon ions and annealing at 800 K. Commercial FePc molecules (Sigma-Aldrich, 97% purity) were sublimated from a Knudsen-type evaporator after thermal puriﬁcation, while the Au(111) substrate was held at room tem- perature. STM images were acquired in the constant-current mode and all given voltages are referred to the sample. dI/dV spectra were measured using a lock-in TURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunications 5 Data availability The data that support the ﬁndings of this study are available from the corresponding authors upon reasonable request. The data that support the ﬁndings of this study are available from the corresponding authors upon reasonable request. 32. Goldhaber-Gordon, D. et al. Kondo effect in a single-electron transistor. Nature 391, 156–159 (1998). 33. Bode, M. et al. Magnetization-direction-dependent local electronic structure probed by scanning tunneling spectroscopy. Phys. Rev. Lett. 89, 237205 (2002). Reprints and permission information is available online at http://npg.nature.com/ reprintsandpermissions/ 55. Wang, Y., Li, X., Zheng, X. & Yang, J. Spin switch in iron phthalocyanine on Au(111) surface by hydrogen adsorption. J. Phys. Chem. 147, 134701 (2017). Peer review information: Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available. Acknowledgements Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Work at IOP was supported by the Strategic Priority Research Program of Chinese Academy of Sciences (XDB30000000), the International Partnership Program of Chinese Academy of Sciences (112111KYSB20160061), National Key Research and Development Projects of China (2016YFA0202300, 2017YFA0302900), and the National Natural Sci- ence Foundation of China (61888102). W.A.H. acknowledges support for the North East Centre for Energy Materials, EPSRC Grant EP/R021503/1. This research made use of the Rocket High Performance Computing service at Newcastle University. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/. Author contributions K.Y., H.C., L.W.L., D.F.W., W.D.X., and X.M.F. performed STM measurements. T.P., Y.H., L.T., Y.-Y.Z., H.-G.L., S.X.D., T.X., and W.A.H. carried out theoretical calculations. K.Y., H.C., T.P., Y.H., W.D.X., S.X.D., W.A.H., and H.-J.G. analyzed the data and wrote the manuscript. H.-J.G. designed and coordinated the project. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-019-11587-x Competing interests: The authors declare no competing interests. 54. Wang, Y., Zheng, X. & Yang, J. Environment-modulated Kondo phenomena in FePc/Au(111) adsorption systems. Phys. Rev. B 93, 125114 (2016). Reprints and permission information is available online at http://npg.nature.com/ reprintsandpermissions/ References Electron-energy-loss spectra and the structural stability of nickel oxide: an LSDA+U study. Phys. Rev. B 57, 1505–1509 (1998). 22. Komeda, T. et al. Observation and electric current control of a local spin in a single-molecule magnet. Nat. Commun. 2, 217 (2011). 23. Krull, C., Robles, R., Mugarza, A. & Gambardella, P. Site- and orbital- dependent charge donation and spin manipulation in electron-doped metal phthalocyanines. Nat. Mater. 12, 337–343 (2013). 53. Gottfried, J. M. Surface chemistry of porphyrins and phthalocyanines. Surf. Sci. Rep. 70, 259–379 (2015). 6 NATURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunicatio ARTICLE Additional information Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 019-11587-x. Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467- 019-11587-x. © The Author(s) 2019 7 NATURE COMMUNICATIONS \| (2019) 10:3599 \| https://doi.org/10.1038/s41467-019-11587-x \| www.nature.com/naturecommunications
https://openalex.org/W2346230755	https://air.unimi.it/bitstream/2434/629695/2/SMS%20Local%20Publics.pdf	English	null	Political Participation of Local Publics in the Unemployment Field: A Comparison of Lyon and Turin	Social movement studies	2,016	public-domain	10,030	1 In Italy the right gained control of the government in 2001. The ephemeral return of the left to power between 2006 and 2008 had no major consequence, particularly in terms of unemployment policy. In Political Participation of Local Publics in the Unemployment Field: A Comparison of Lyon and Turin Matteo Bassoli* & Manlio Cinalli** Matteo Bassoli* & Manlio Cinalli** Matteo Bassoli* & Manlio Cinalli** * School of law, eCampus Online University, Novedrate (CO), Italy. *CEVIPOF, SciencesPo, Paris, France School of law, eCampus Online University, Novedrate (CO), Italy. **CEVIPOF, SciencesPo, Paris, France Abstract: This article focuses on political participation of local publics in the unemployment field, examining networks of collective actors in Lyon and Turin. Our main question is: Is the participation of local publics fostered under conditions of more developed governance that increases bottom-up access (formal or informal) to elites and institutions in the policy domain? Drawing upon the most recent developments in literatures on social movement theory, governance, and network analysis, this article discusses the main variations in terms of political participation of local publics in Lyon and Turin. It then enquires into the main explanatory factors accounting for these variations, thus showing that the openness of governance does influence the level of political participation of local publics. The main argument is that in an open context participation is low, while in a closed (or underdeveloped) context local publics participate more, with differential access to decision-making according to their resources. Keywords: Local governance, Social networks, Local publics, Unemployment, Multi-organisational field 1. Introduction Unemployment has emerged in the last decades as one of most compelling issues in Europe. Eurostat data show increasing unemployment rates, particularly for the younger cohorts of populations (Eurostat 2011). The European Commission has provided funds to implement relevant measures through Structural Funds, promoting policy interventions that are flexible enough to allow for distinct national answers and participation of the largest possible number of private and public stakeholders. For example, in the aftermath of economic recession of late 2000s, both the European Commission and the Council of Europe have focused on the inclusion of the unemployed through interventions such as the ‘Agenda 2020’, providing a comprehensive policy effort that is cross-sectorial and highly articulated. Within Europe, however, unemployment has been more significant in some places than in others. In what follows, we compare two cities, Lyon and Turin, which are located in two European countries with traditionally high rates of unemployment. Whether referring to ‘precarious’ workers, the ‘excluded’, or to more colourful expressions such as the Italian ‘cococo’, unemployment and precariousness have stood out in France and Italy as the objects of political intervention throughout the 2000s. In addition, both countries share a number of common political dynamics, such as the predominance of right-wing politics at a general level,1 a slow yet consistent attempt 1 to increase the flexibility of labour market access, and a focus on redistribution of job opportunities in public rhetoric. This article uses a network approach to deal with the interplay of organisational attributes, on the one hand, and multi-level opportunity structures on the other hand. The main aim is to account for variations of political participation across a large range of organisations that engage in the unemployment field. In particular, our argument promotes the use of the relational approach already used in social movement scholarship (Diani 1992 and 1995; Gould 1995; Diani and McAdam 2003; Diani and Bison 2004), but with a different emphasis derived from literature on governance (Cinalli 2007a; Christopoulos 2008; Christopoulos and Quaglia 2009; Feiock and Scholz 2010; Sohn and Giffinger 2015). Instead of using networks to understand the nature of social movements, the governance literature has used relational data to examine the different levels of actors’ involvement in the political sphere. Scholars have also focused on informal groups, associations, cooperatives and unions (Piselli and Ramella 2008), providing the possibility of testing different theories of political participation (resource-based or social capital related). France the 2000s opened with the “cohabitation” between a right-wing President and a left-wing Prime Minister. But 2002 marked the definitive defeat of the left, as it lost both control of government and its majority in Parliament. 1. Introduction That is, attention needs to be focused on the key role of the political environment in which participation of local publics in Lyon and Turin takes place, following those scholars who argue that the consideration of joblessness as a simple grievance is insufficient to explain political developments and the debate in the unemployment field (Baglioni et al. 2008; Baxandall 2001 and 2002). Our comparison looks at the role of ‘political opportunities’ (Tilly 1978, Kitschelt 1986, Kriesi et al. 1995 Tarrow 1998) and ‘welfare regimes’ (Giugni et al. 2009, Cinalli and Giugni 2010) to account for variations of political participation in the unemployment multi-organisational field. In particular, the specific focus on local publics in Lyon and Turin draws upon a number of studies that, since the seminal work of Eisinger (1973), have emphasised the explanatory role of the sub-national contextual environment, while taking into consideration the complex dynamics linking the public to the policy domain on the one hand (Broadbent 1998) and local to national opportunities on the other (Cinalli and El Hariri 2011; Della Porta and Andretta 2002). Our consideration of the more unstable aspects of the political context (Gamson and Meyer 1996) also aims to link the analysis of ‘opportunities’ to that of ‘governance’. For the latter, we can thus move from a simplistic approach based on distinct sectors (public, private for profit and non-profit) to a more nuanced version based on the concept of (local) governance (Bassoli and Polizzi 2011, Grote 2008, John 2001, Mayntz 2003). We thus follow studies that have dealt with the important role of the local level in policymaking (Harlow and Rawlings 2006, Jordan 2001), focusing on private-public partnership (Osborne 2007), multi-sectorial partnership (Sullivan and Skelcher 2002), and participatory governance arrangements (Bassoli 2010 and 2012; Grote and Gbikpi 2002), such as participatory budgeting. Amongst our dimensions of analysis, we consider how much genuine access private actors have to local policy making, above all in the crucial phase of decision making, but also during discussion and implementation phases (Polizzi 2011). The existence of political opportunities rooted in local governance cannot be considered the only determinant of participation. Following the long-standing tradition of Resource Mobilization Theory (Obershall 1973, McCarthy and Zald 1977), many studies have shown the importance of budgets, organisational assets, and increasing resource dependence on broader civil society (Giugni and Passy 2001). 1. Introduction Our approach thus offers the possibility to engage in full with the analysis of a whole ‘multi-organisational field’ (Curtis and Zurcher 1973), allowing for a systematic assessment of political participation within formal and informal decision-making processes. Hence, ‘local publics’, which are understood as made up of all organisations that are active in the local public domain, can be placed at the core of a number of questions. What is the relationship between different political contexts and variations in participation of local publics in the unemployment field? What is the space for direct intervention of local publics in governance processes related to unemployment across the two cities of Lyon and Turin? And beyond the role of political opportunities, what is the role of endogenous attributes in predicting the political participation of various actors? Are variations in participation dependent upon activities, for example, the ‘policy’ or the ‘services’ orientation that different actors may privilege respectively? And what about the specific role of social capital? The following section clarifies our theoretical framework in terms of main dependent and independent variables. Afterwards, the article moves on to present relevant contextual information enabling the comparison between Lyon and Turin (section 3). We then proceed in section 4 to identify and measure political participation as a dependent variable. Section 5 examines the relationship between participation and a range of independent variables at the level of collective actors (resources and orientations) and networks in the unemployment field (governance opportunities and relational structures). Finally, section 6 sums up the empirical findings and indicates directions for future research. 2. Political Context, Governance and Endogenous Characteristics Unemployment has stood out as a crucial issue in Europe, leading to demands for ‘reforms’ of the labour market, social policy, and unemployment regulations. A brief 2 consideration of these reforms in France and Italy confirms a common trend, with the promotion of a secondary labour market (Jessoula et al. 2010) consisting of mini-jobs and short-term contracts, and the elaboration of restrictive measures to pressure the long-term unemployed into accepting available jobs. The organisations that are active in the unemployment field have thus faced a number of common pressures. Yet it should be emphasised that labour market mechanisms, social protection systems, and social services have remained specific to each distinct nation-state (Cinalli and Giugni 2010). 1. Introduction On the other hand, we examine the roles of a number of additional variables that may account for variations of participation. methodological tools for examining the multi-level and multi-organisational field of unemployment.2 Drawing on previous research that has put the study of relational structures at the core of unemployment politics (Cinalli 2010; Cinalli and Fuglister 2008), our networks are primarily made of collective actors (nodes) connected by three main directed relationships (ties): exchange of information, cooperation in projects, and overlapping members. The field is thus defined in terms of a set of (potentially) reciprocal interactions that represent the possibility of accessing specific resources and additionally behavioural constraints or loyalties that constrain organisations (Cinalli 2007b). For the purpose of the following analysis we also consider administrative level as a node within a ‘two-mode network’. Each collective actor may be active at (i.e. connected to) the district, local, national, regional or EU level. We generated several two-mode networks depending on the nature of the relationships represented by network ties and report network densities below. The construction of matrices of connections among collective actors, and between actors and administrative levels also allows graphical representation as in Figures 4.1, 4.2 and 5.1. To deepen our analysis further we also consider the different attributes that our collective actors have (e.g. size or budget), which allows the investigation of correlations between attributes (i.e. variables) and network outcomes through a procedure detailed in section 5 below. On the one hand, therefore, the analysis below assesses the extent to which political participation of local publics in the two cities translates into specific network patterns between local publics and policy actors. These networks are also linked with the concept of social capital (Lin 2001), which is most often understood to increase the organisation's chance of being politically active, given the its number of connections possessed. On the other hand, we examine the roles of a number of additional variables that may account for variations of participation. Overall, relevant questions for the comparison of local publics in Lyon and Turin are given a relational twist. What are the channels being forged in the two cities across the public and the policy domain at the local level and at the other levels beyond that? Does unemployment stand out as an issue that divides certain types of actors from others? 3 For an overview of broader debate about network analysis and governance see Christopoulos 2008, Dowding 1995, 2001, Marsh and Smith 2001, Marshall and Staehel 2015, and Raab 2001. 2 For a treatment of network analysis see Knoke and Kuklinsky 1982, Scott 2000, and Wasserman and Faust 1994. 1. Introduction Or, alternatively, do actors of different types forge extensive ties, reducing the traditional distance between them? Different governance dynamics can be disentangled in the multi-level and multi-organisational unemployment field, shedding light on the variety of actors in the public domain and their variable access to the policy domain across different levels (Feiock and Scholz 2010). In doing this, the article further advances the network conceptualisation of governance (Cinalli 2007a; Rhodes 1996 and 1997), responding to the pressing need to use networks as an effective instrument of empirical analysis, rather than as a simple heuristic device (Christopoluos 2008).3 2 For a treatment of network analysis see Knoke and Kuklinsky 1982, Scott 2000, and Wasserman and Faust 1994. 3 For an overview of broader debate about network analysis and governance see Christopoulos 2008, Dowding 1995, 2001, Marsh and Smith 2001, Marshall and Staehel 2015, and Raab 2001. 1. Introduction Accordingly, this article deals with a number of endogenous characteristics that may have an impact on variations in participation of local publics in Lyon and Turin. In particular, we focus on organisational structuring of local publics, how long different organisations have been active in the unemployment field, their degree of formalisation, their ownership of relevant assets such as an office space, as well as the main orientations of actors, identified in their prevailing engagement in policies or, alternatively, services (Lelieveldt et al. 2007, Torpe and Ferrer-Fons 2007). In what follows network analysis is used to provide key conceptual and 3 Deleted: 2 – Deleted: methodological tools for examining the multi-level and multi-organisational field of unemployment.2 Drawing on previous research that has put the study of relational structures at the core of unemployment politics (Cinalli 2010; Cinalli and Fuglister 2008), our networks are primarily made of collective actors (nodes) connected by three main directed relationships (ties): exchange of information, cooperation in projects, and overlapping members. The field is thus defined in terms of a set of (potentially) reciprocal interactions that represent the possibility of accessing specific resources and additionally behavioural constraints or loyalties that constrain organisations (Cinalli 2007b). For the purpose of the following analysis we also consider administrative level as a node within a ‘two-mode network’. Each collective actor may be active at (i.e. connected to) the district, local, national, regional or EU level. We generated several two-mode networks depending on the nature of the relationships represented by network ties and report network densities below. The construction of matrices of connections among collective actors, and between actors and administrative levels also allows graphical representation as in Figures 4.1, 4.2 and 5.1. To deepen our analysis further we also consider the different attributes that our collective actors have (e.g. size or budget), which allows the investigation of correlations between attributes (i.e. variables) and network outcomes through a procedure detailed in section 5 below. On the one hand, therefore, the analysis below assesses the extent to which political participation of local publics in the two cities translates into specific network patterns between local publics and policy actors. These networks are also linked with the concept of social capital (Lin 2001), which is most often understood to increase the organisation's chance of being politically active, given the its number of connections possessed. Comparing the Unemployment Field in Lyon and Turin Having experienced a similar situation of unemployment grievance over the past few years, with peaks in their regions at 9% (INSEE 2011; ISTAT 2011), Lyon and Turin are comparable on the basis of additional key indicators such as economic size, economic structure, and type of unemployment (level of de-industrialisation). At the 4 same time, Lyon and Turin are two major non-capital cities in countries that have gone through a genuine policy of decentralization. In France this process has been sanctioned by the Constitutional Law of March 28, 2003, acknowledging explicitly the role of sub-national authorities. As regards Italy, the Constitutional Law 3/2001 has provided further policy responsibilities to local authorities. The cities also share a number of common political characteristics. Electoral systems in both cities are predominantly majoritarian. In line with the tradition of French Presidentialism, Lyon is characterised by the dominant role of the mayor, but Turin has equally seen the gradual strengthening of the local executive following the law 81/1993 (Fabbrini, 2001). The leftist governments in Lyon and Turin are in opposition to the rightist French and Italian national governments respectively. And both cities lack effective direct democracy tools.4 However, the two cities have been facing contrasting policy developments, allowing for the analysis of different answers that policy makers have elaborated at the national and sub-national levels. The welfare literature shows crucial divergences between the French and the Italian model, in spite of their common roots in ‘corporatism’ (Esping- Andersen 1990). In Italy, a key role has been handed over to the family as a crucial agent of social protection (Ferrera 1996), while France has preferred to intervene extensively over welfare in an attempt to reach a virtuous balance between corporatist and universalist prescriptions. Although both countries have engaged in a number of reforms that have targeted the labour market to relax employment protections (OECD 2010), specific policies dealing with unemployment and social inclusion have been developed more extensively in France than in Italy. In particular, the two countries diverge when looking at various indicators of state intervention in terms of unemployment regulations – for example, the pre-requisites for obtaining social provisions, the level and extension of benefit coverage, the role of employment agencies, and the scope of sanctions (Giugni et al. 2009; Cinalli and Giugni 2010). 4 Local statute in Turin allows only consultative referenda, though they can be proposed by citizens. In Lyon ‘decisional referenda’ have been instituted in 2003, yet, they need to be initiated by local authorities. Comparing the Unemployment Field in Lyon and Turin In this context, Italy exhibits a less inclusive unemployment legislation, in which a high share of the workforce has low social protection (Jessoula et al. 2010). The French unemployment protection system can rely not only on the corporatist legacies such as the régime d’assurance (financed and managed by the trades unions and entrepreneurial associations), but also on the universalist régime de solidarité (financed and managed by the state). By contrast, unemployment provisions and social protection are very limited in Italy, especially for the recent wave of precarious youth, who could join the labour force only as ‘atypical’ workers. What is more, evidence at the local level shows that the municipality of Turin is less active in unemployment policies than its counterpart in Lyon, which in fact supports an increasing number of projects for job creation. The 'right to experiment', whereby local authorities can diverge from national arrangements, provides the Lyon municipality with further space of intervention in terms of social and economic policy. Differences are also noticeable in terms of public information and support services. Employment offices have been set up in Turin (the Centri per l'impiego), but public information and services are still limited. By contrast, the Maison de l’Emploi et de la Formation in Lyon provides information for the unemployed, helps 5 employers with recruitment, and gathers a large volume of economic and social data. The local level is also important by virtue of the large volume of actors that it involves in service delivery or policy implementation. Hence, the comparison of Lyon and Turin allows for dealing with key explanatory variables at the local level relating to various organisational resources and the networks created by actors. No doubt, the two cities differ in terms of inclusion of local publics within the policy domain. Lyon has inclusive governance dynamics that are supposed to promote a wider access of local publics to the policy domain. For example, the action of the Maison de l’Emploi et de la Formation has been shaped on the objectives of the 2005 loi de programmation pour la cohésion sociale with the main aim of ensuring a stronger partnership of actors in the public sector alongside representatives of civil society. By contrast, Turin stands out for its restrictive stance in terms of governance arrangements owing to the marginal participation of public domain actors in the policy process. 5 For example vocational training is coordinated by the Piedmont Region, and the centri per l'impiego are managed by the Province. Comparing the Unemployment Field in Lyon and Turin While relevant competences are often located beyond the scope of the municipality,5 the vast externalisation to private actors at the local level has brought about a clear shift of responsibility from public to private actors in the delivery of social services (Ranci 1999; Ranci and Montagnini 2009). Two actors play a pivotal role in the Italian labour market: the agenzie interinali (recruitment agencies) and the sistema cooperativo (cooperative system). The latter has been regulated by the national law so as to promote employability of young people and of the long-term unemployed. As regards the agenzie interinali, they are private enterprises that take care of ‘temping’, dealing with different administrative procedures related to fixed- term contracts or jobs on demand. 7 Therefore we move far from the SNA approach to social movements (Diani and Bison 2004), considering SMO only those informal groups that are part of a social movement, although at some point actors could be in a “social movement process” (ibid. 283). While performing the fieldwork there was not a clear social movement related to either unemployment or precariousness. 8 “In the last 2 years, has your organisation been called to participate in decision-making processes in any of the following ways? [Interviewer: Only when interviewee mentions one, ask] Did your organisation finally participate?” 9 Since there are four possible roles in the policy-making process, two matrices were created: a 19 by 4 (Lyon) and a 26 by 4 (Turin). 6 The political parties are not considered here because they could not be formally invited to policy- making, while, at the same time, they had a privileged and informal access to the political arena. Further information on the sampling strategies and local networks can be found in Bassoli and Cinalli (2014). 4. Local Publics in Lyon and Turin O h i b d 56 i i Without presenting the organisational ecology in detail (see Bassoli and Cinalli 2014), it is important to stress that the two cities differ with respect to the organisations interviewed and the results of the mapping process. The organisational fields are different not only in terms of numbers (19 in Lyon and 26 in Turin) but also in terms of degree of heterogeneity. While Turin stands out as having a highly heterogeneous field, which is made of antagonistic actors (such as social movements and spontaneous committees), pivotal actors for social stability (such as trade unions), and strategic partners to deliver public policies (such as cooperatives and training schools), Lyon is characterised by a much higher homogeneity and a remarkably large number of formal associations, with no presence of spontaneous groups of any kind. 4. Local Publics in Lyon and Turin O h i b d 56 i i 8 “In the last 2 years, has your organisation been called to participate in decision-making processes in any of the following ways? [Interviewer: Only when interviewee mentions one, ask] Did your organisation finally participate?” 9 Since there are four possible roles in the policy-making process, two matrices were created: a 19 by 4 (Lyon) and a 26 by 4 (Turin). Deleted: Baglioni and Giugni Deleted: 36 Deleted: 50 Deleted: (Figure 4.1 about here) Deleted: informal access Deleted: Baglioni and Giugni Deleted: 36 Deleted: 50 Deleted: (Figure 4.1 about here) Deleted: Baglioni and Giugni Deleted: 36 Deleted: 50 Deleted: (Figure 4.1 about here) (SMOs), religious organisations (ROs), trade unions (TUs) and political parties (PPs, beyond the scope of this article).6 Actors were defined by their legal status (CSOs, PPs and TUs), by their close connection with a religious institution (ROs, either formally dependent or not), and by the absence of a clear legal status coupled with an engagement in contentious politics (SMO).7 Other actors such as cooperatives and foundations were also included. Without presenting the organisational ecology in detail (see Bassoli and Cinalli 2014), it is important to stress that the two cities differ with respect to the organisations interviewed and the results of the mapping process. The organisational fields are different not only in terms of numbers (19 in Lyon and 26 in Turin) but also in terms of degree of heterogeneity. While Turin stands out as having a highly heterogeneous field, which is made of antagonistic actors (such as social movements and spontaneous committees), pivotal actors for social stability (such as trade unions), and strategic partners to deliver public policies (such as cooperatives and training schools), Lyon is characterised by a much higher homogeneity and a remarkably large number of formal associations, with no presence of spontaneous groups of any kind. (SMOs), religious organisations (ROs), trade unions (TUs) and political parties (PPs, beyond the scope of this article).6 Actors were defined by their legal status (CSOs, PPs and TUs), by their close connection with a religious institution (ROs, either formally dependent or not), and by the absence of a clear legal status coupled with an engagement in contentious politics (SMO).7 Other actors such as cooperatives and foundations were also included. 9 Since there are four possible roles in the policy-making process, two matrices were created: a 19 by 4 (Lyon) and a 26 by 4 (Turin). (Table 4.1 about here) Deleted: (Figure 4.1 about here) Deleted: (Figure 4.1 about here) Deleted: informal access 4. Local Publics in Lyon and Turin O h i b d 56 i i 4. Local Publics in Lyon and Turin Our research is based on 56 interviews with various civil society organisations, trade unions and unemployed groups that engage in the unemployment field in Lyon and Turin. The interview schedule was specifically designed to include not only qualitative in-depth questions, but also sets of standardised questions inquiring into political activities, inter-organisational networks spanning the public and the policy domain, as well as information about actors’ organisational structure and resources. Selection of actors started with the systematic examination of official lists of organisations active in Lyon and Turin. Since this effort could only target certain types of actors, the selection was completed through examination of practitioners’ publications, access to online sources (blogs of precarious groups, alternative information portals, etc.) and, in particular, through indications coming from actors themselves. At the same time, a wide range of secondary sources, such as organisations’ publications, press articles, and official documents from political and institutional authorities, was used to deepen the analysis of further contextual dimensions. 4.1 The organisational ecology g gy A first look at types of actors that are present in the unemployment field shows that Lyon and Turin have very distinct local publics. According to the sample criteria, the study included civil society organisations (CSOs), social movement organisations 6 (SMOs), religious organisations (ROs), trade unions (TUs) and political parties (PPs, beyond the scope of this article).6 Actors were defined by their legal status (CSOs, PPs and TUs), by their close connection with a religious institution (ROs, either formally dependent or not), and by the absence of a clear legal status coupled with an engagement in contentious politics (SMO).7 Other actors such as cooperatives and foundations were also included. Without presenting the organisational ecology in detail (see Bassoli and Cinalli 2014), it is important to stress that the two cities differ with respect to the organisations interviewed and the results of the mapping process. The organisational fields are different not only in terms of numbers (19 in Lyon and 26 in Turin) but also in terms of degree of heterogeneity. 4. Local Publics in Lyon and Turin O h i b d 56 i i While Turin stands out as having a highly heterogeneous field, which is made of antagonistic actors (such as social movements and spontaneous committees), pivotal actors for social stability (such as trade unions), and strategic partners to deliver public policies (such as cooperatives and training schools), Lyon is characterised by a much higher homogeneity and a remarkably large number of formal associations, with no presence of spontaneous groups of any kind. (Table 4.1 about here) 4.2 The political involvement of local publics Our findings also show major differences in terms of political involvement of local publics in the two cities. The ‘two-mode approach’ to relational data (David et al. 1941; Everett 2005) can be deployed to evaluate whether civil society actors have been called to participate in decision-making processes as well as the extent to which they have in fact participated.8 As already stated, data are directed: for example, they assess whether actors have participated in public decision making, with ties being coded between two sets of nodes – actors and specific administrative levels– and moving from the first to the second. The first finding is that all actors who were called to participate in the local policy process eventually did participate. In particular, four possible roles as participants can be identified: “permanent member of the district or neighbourhood council”, “permanent member of a municipal council on specific issues (social services, women, education, etc.)”, “occasional member in a municipal committee to solve a specific problem” and “member of a municipal consultation committee or group for a specific policy or issue”.9 Actors could also be invited to participate at other levels. This question, however, was open-ended, and it was later 6 The political parties are not considered here because they could not be formally invited to policy- making, while, at the same time, they had a privileged and informal access to the political arena. Further information on the sampling strategies and local networks can be found in Bassoli and Cinalli (2014). 7 Therefore we move far from the SNA approach to social movements (Diani and Bison 2004), considering SMO only those informal groups that are part of a social movement, although at some point actors could be in a “social movement process” (ibid. 283). While performing the fieldwork there was not a clear social movement related to either unemployment or precariousness. 4.2 The political involvement of local publics 4.2 The political involvement of local publics Our findings also show major differences in terms of political involvement of local publics in the two cities. The ‘two-mode approach’ to relational data (David et al. 1941; Everett 2005) can be deployed to evaluate whether civil society actors have been called to participate in decision-making processes as well as the extent to which they have in fact participated.8 As already stated, data are directed: for example, they assess whether actors have participated in public decision making, with ties being coded between two sets of nodes – actors and specific administrative levels– and moving from the first to the second. The first finding is that all actors who were called to participate in the local policy process eventually did participate. In particular, four possible roles as participants can be identified: “permanent member of the district or neighbourhood council”, “permanent member of a municipal council on specific issues (social services, women, education, etc.)”, “occasional member in a municipal committee to solve a specific problem” and “member of a municipal consultation committee or group for a specific policy or issue”.9 Actors could also be invited to participate at other levels. This question, however, was open-ended, and it was later 6 The political parties are not considered here because they could not be formally invited to policy- making, while, at the same time, they had a privileged and informal access to the political arena. Further information on the sampling strategies and local networks can be found in Bassoli and Cinalli (2014) Deleted: informal access 7 recoded according to five different levels: other local, provincial, regional, national and European.10 A synthetic representation of the four main political networks can be offered through measures of ‘density’ (Wasserman and Faust 1994: 181), which gives the percentage of existing ties out of possible ties. Accordingly, data clearly suggest key differences between Lyon and Turin. As regards political participation at any level, actors in the public domain participate more in Turin (0.22) than in Lyon (0.18). The same holds true for the local level. Yet, while in Turin at the figure is higher for the local level (0.25), for Lyon it is lower (0.13). Thus, the gap between the two cities at the local level is almost three times higher than the gap measured by the general involvement. 10 In this case, we created two other matrices of 19 by 5 (Lyon) and 26 by 5 (Turin). 11 The distinction between policy oriented organisations and service oriented organisations is crucial in the third sector. The former target the political system – they try to influence it, or they directly take part in the policy process – whereas the latter rather focus on their clients. These categories are not easily respected empirically, as organizations can do both type of actions simultaneously although usually one of these activities prevails over the other. 12 The measure of centrality is given by the total amount of ties considering the number of nodes, therefore the higher level of centrality in Turin is mirrored by an increased amount of ties (per node) in the figure. 4.2 The political involvement of local publics Overall, the case of Lyon shows that a low degree of political participation, especially at the local level, may well follow a more developed process of (formalised) governance. This result is somewhat counter-intuitive since an inclusive political process is expected to encourage political participation overall, in spite of the decreasing recourse to more radical forms of political mobilisation. By contrast, data for Turin, which is characterised by a more restrictive participatory context, show that local publics participate extensively (the density is much higher than Lyon), especially at the local level. Deleted: 2 Figure 4.1 represents the network of participation in formal decision making at any administrative level. The presence of a tie between two nodes (actor ‘A’ and level ‘X’) means that that actor (A) participates at the decision-making process at that level (X). Acronyms indicate the legal form: ‘A’ stands for associations, ‘S’ for informal groups/social movements, ‘T’ for trade unions and ‘C’ for cooperatives. The size of each node increases with the the length of time since the actor first entered the unemployment field. The shape of nodes provides another key piece of information (Lelieveldt et al. 2007): squares are given to actors involved in service provision and circles to other actors (mixed or policy-oriented).11 The number of ties in Lyon is lower than in Turin12; yet, in both cities half of the actors are not active, while the other half are active at more than one level. Another important distinction can be seen is the different levels involved: while in Lyon all levels are accessed with the exception of the EU level, in Turin the national level is also precluded. Deleted: 2 Deleted: 2 Deleted: so the Deleted: 3 Deleted: 3 10 In this case, we created two other matrices of 19 by 5 (Lyon) and 26 by 5 (Turin). 13 Organisations with a budget lower than 10 thousand euro are represented by a circle (as well as missing), between 10 and 200 thousand are square, triangle above 200 thousand. 14 The minimum articulation is given by organisations with only an assembly (score 1), while the higher level is given by those organisations with a board, a leader / president, a chairperson, a secretary, a spokesperson, a treasurer, a general assembly, and committees (score 8). 15 In the permutation test the standard assumptions on independence and random sampling are not required. 16 The routine measures the correlation between the dependent variable (degree centrality – that is, the number of ties each actor has) and the independent variable considered. 17 The first two categories have been considered both mutually exclusive and not, the third is an independent category. 18 This variable has been measured using network data, considering information received by the local organisation by given political actors (or levels). (Fig. 4.1 about here) In addition, Figure 4.1 shows that the local level remains the preferred locus for political participation of local publics in the two cities and so the second step of analysis thus concerns the formal decision-making mechanism set up at the local level (Figure 4.2). In order to put forward some of the resource variables used in the research, in Figure 4.2 nodes have a different shape according to economic In addition, Figure 4.1 shows that the local level remains the preferred locus for political participation of local publics in the two cities and so the second step of analysis thus concerns the formal decision-making mechanism set up at the local level (Figure 4.2). In order to put forward some of the resource variables used in the research, in Figure 4.2 nodes have a different shape according to economic 8 Deleted: 3 resources13 while size changes proportionally to level of organisational complexity of the collective actor, referred to here as internal articulation14. In line with the density figures, the level of participation sharply decreases moving to the local level in Lyon, while it increases in Turin. Clearly both local networks are characterised by a high percentage of isolates (63% and 58%). Once again, only a rather small number of actors in both cities participate; yet active nodes in Turin have forged more extensive ties than in Lyon. (Fig. 4.2 about here) (Fig. 4.2 about here) Deleted: 3 5. Explaining the political participation of local publics So how can we account for these differences between Lyon and Turin? Does the fact that the unemployment field of Turin is formed by spontaneous groups and weak associations account for the differences in participation? Or should we explain this outcome with reference to the development of participatory structure? In what follows, a number of single variables are taken into account, checking them independently. In this way the analysis does not explain the participation of single actors using individual variables as predictive factors; rather, it appraises the extent to which similar actors (grouped according to a given variable) behave similarly. This can be easily done by the ANOVA routine in UCINET (Borgatti et al. 2002). 19 “Regardless of whether you ever get invitations to participate in decision-making processes, does your organisation receive official information concerning the decision-making processes on questions relevant for your organisation from any of these authorities?” This question produced (again) two matrices of 19 by 5 (for Lyon) and 26 by 5 (for Turin). (Tab. 5.1 about here) Generally speaking, participation in Lyon seems randomly distributed. As regards participation at all levels, only one variable is not evenly shared among the actors (the diffusion of advocacy services), while for local participation a single variable seems relevant (the distribution of part-time workers). Hence, data suggest that the political context in Lyon may well overcome individual limitations of resources. By contrast, scarce political opportunities and thinner processes of governance in the case of Turin allow for participation of only some specific actors, namely, the older, more formally structured, richer organisations with the capacity for running public-programmes, as well as those that devote themselves the most to service-provision for internal members. Further differences emerge when we consider the openness of the political system more specifically. A crude indicator of openness is represented by the different magnitude of flow of information received by organisations. Data were gathered on whether the organisation had been receiving information from any specific level of government (European, national, regional, local or district)19. In this case, the density of the French information network is much higher compared to the Italian network, that is, 0.495 against 0.231. This means that, out of all possible connections, 49.5% of them have been forged in the unemployment field in Lyon, but only 23.1% have been forged in Turin. Since some levels of government beyond the local may be crucial in one city but irrelevant in the other (for example, the national level that is so quintessential under the long-standing tradition of French Republicanism), the local flow of information has also been singled out. In this case, figures are much more similar for the two cities, with 0.342 for Lyon and 0.365 for Turin. Nonetheless this similarity is only apparent, since the network patterns are quite different, as can be grasped from Figure 5.1. The networks have been graphed so as to group the actors according to the number of ties that they have forged, and according to the direction of these ties. (Fig. 4.1 about here) The routine undertakes an analysis of variance using a permutation to assess if the independent variable is evenly distributed among the nodes or if it is (significantly) correlated with the dependent variable.15 That is, the significant variables do not explain participation according to a ‘causal’ direction, yet they do show whether the selected variables are unevenly distributed amongst the actors:16 those who participate have either more of a given variable, or less. While one cannot determine the causal direction, it is quite logical to assume the impact of resources on the level of participation rather than vice-versa. The test has been run on a number of variables, many of which are not evenly distributed. These variables include different kinds of resources (budget, recruited staff, office ownership), different dimensions of formalisation (level of internal articulation, being registered), different aims (policy-oriented service-oriented actors, actors running public programmes),17 as well as different levels of activity (only local, local and national, only national) and of information inflow (as a proxy of visibility by the public authorities).18 The significant variables, listed in Table 5.1, reinforce the 9 9 general idea about the open governance structure of Lyon vis-à-vis Turin. Indeed the role of resources is more important in Turin than in Lyon. Out of all possible variables, quite a few resources are more abundant across the actors that participate in the Italian city: the degree of institutionalisation, the possibility of having full-time workers, the capacity for running public-programmes. Only one variable is crucial in both contexts, namely the provision of services (with service providers participating to a greater extent in both cities). Focussing on local participation, the overall picture does not change. While resources affect the French case (Table 5.1) only via the employment of part-time workers, the Italian case still shows the higher salience of resources: length of activity (the older the actor the more it participates), internal articulation (the more articulated the actor, the more it participates), finances (the richer the actor the more it participates), and the diversification of service provision (the actors which diversify the most are also those which participate the most). (Fig. 5.1 about here) 10 As expected, the national level plays a major role in Lyon (with the district being almost absent), while the Turin network contains more isolates than that for Lyon. Yet some specifications are necessary given the different organisational ecology of the two cities. Considering the legal form of organisations in Turin, a strong correlation can be found with network centrality. Thus the fact of being a social movement organisation deeply hinders the possibility of having equal access to the policy domain in terms of 'receiving information' (Table 5.2). Measures have been assessed through two main variables. The first one is the ‘legal form’, distinguishing between trades unions, formal organisations and ‘other’ (that is social movements and cooperatives). The second variable focuses on ‘degree of institutionalisation’, ranking organisations alongside the distinctions between SMOs (1), associations (2), cooperatives (3), and trades unions (4). While the first variable works well when accounting for flows of information at the local level (R square 0.43 at the level of 99%), the second variable performs equally well in terms of the overall picture (R square 0.42 at the level of 99%). Crucially, the SMOs are isolated, alongside with some trade unions, while cooperatives, associations and the largest trades unions take better advantage of the communication channels in Turin. 20 The measure was the out-degree centrality, that is, the number of perceived partners which a given organisation has. 21 The measure was the in-degree centrality, that is, a measure of popularity, since it refers to the number of times that a given organisation has been mentioned as a partner by other actors. 6. Conclusions In this article we have conducted a comparison of the multi-level and multi- organisational field of unemployment in Lyon and Turin in order to analyse, first, variations of participation of local publics, and second, the main explanatory factors accounting for cross-city variations. The focus on unemployment has allowed for analysis of a field that has become increasingly contentious throughout the 2000s, and which may thus be exemplary for understanding the variable impact of different policy answers to potentially contentious issues of the future. In particular, our research has shed light on the complex dynamics that operate across the public domain (where local publics mobilise) and the policy domain (where elites and institutions shape their interventions), with the aim of contributing to the development of a set of theoretical approaches that deal with political participation within the two distinct scholarly traditions of governance and political opportunities. Our main point is that even the most articulated sets of inclusive arrangements fostering the participation of local publics need to be appraised in terms of actual outcomes across the public and policy domains. Following the nexus between context and behaviour that is at the core of the scholarship of opportunities, more theorisation is needed in the governance field about the consequences of variable policy processes for political participation. We have thus argued that a number of crucial questions need to be asked about the complex interactive dynamics that enable stakeholders to engage across the public and the policy domains. Accordingly, amongst the main findings, we have shown that a low level of participation may well fit with more extensive channels of governance (just as it happens in Lyon), whilst higher levels of participation may well follow situations of underdeveloped governance (as in the case of Turin). That is, the allegedly virtuous relationship between governance and participation is in our research far from being proved, warning us that a more thorough examination of the normative underpinnings of the same concept of governance is necessary. Furthermore, our analysis has revealed that provision of services, overall resources, and time devoted to the specific issue of unemployment (amongst the other possible issues competing for members’ attention) stand out as predictors of higher levels of participation in Turin, whereas groups with fewer resources find a better place within the more inclusive structure of Lyon’s governance regime. quintessential expectations of the social capital literature. (Table 5.3 about here) (Table 5.2 about here) As regards the social capital created by the overall relational structure in which local publics are embedded, the ANOVA procedure can be useful to correlate the degree centrality in different networks. This undertakes the variance analysis comparing the distribution of the centrality in two different networks. The procedure has been used to assess the correlation between centrality in the participation network and three other variables: societal importance (level of centrality in the local field of unemployment, as perceived by the organisation20 or by others21), centrality in information networks, and the political involvement in networks created by wider civil society partnerships beyond the unemployment field. Findings show that in both cities participation at the local level co-varies with participation at the national level, yet with some crucial differences. In Lyon the impact is much stronger and works only in one direction: the local participation 'moves' the participation at all levels (R-square 0.577). By contrast, in Turin the two types of participation move together but to a lesser extent (R-square equal 0,268 and 0.299 respectively). At the same time, information stands out as a key factor in Turin. This finding fits with the idea that the political context is negative for civil society actors of the Italian city. Its local public is cut off, while the rare channels are selective and biased towards 'well-known' actors so that information is provided only to the 'usual suspects'. Clearly, information strongly co-varies with both participation and local participation (at a 99% level of confidence). By contrast, the ‘social centrality is irrelevant, since scores in both cities cannot explain political participation. This is again a counter-intuitive finding, which hardly matches the 11 quintessential expectations of the social capital literature. References Baglioni, S., Della Porta D. & Graziano, P. (2008) The contentious politics of unemployment: the Italian case in comparative perspective, European Journal of Political Research, 47 (6), pp. 827-851. Deleted: Baglioni, S. & Giugni, M. (2014) Civil Society Organization, Unemployment, and Precarity in Europe. Organizational Activities and Networks, Houndmills, Basingstoke, Hampshire, England; New York: Palgrave Macmillan.¶ Bassoli, M. (2010) Local Governance Arrangements and Democratic Outcomes, Governance, An International Journal of Policy, Administration, and Institutions, 23(3), pp. 485-508. ( ) pp Bassoli, M., (2012) Participatory Budgeting in Italy: An Analysis of (Almost Democratic) Participatory Governance Arrangements. International Journal of Urban and Regional Research, 36(6), pp.1183–1203. g ( ) pp Bassoli, M & Cinalli, M. (2014) Networks within the Multi-Organizational Field of Unemployment: A Tale of Seven Cities in Europe. In S. Baglioni & M. Giugni (eds.) Bassoli, M & Cinalli, M. (2014) Networks within the Multi-Organizational Field of Unemployment: A Tale of Seven Cities in Europe. In S. Baglioni & M. Giugni (eds.) Civil Society Organizations Unemployment and Precarity in Europe Houndmills Bassoli, M & Cinalli, M. (2014) Networks within the Multi-Organizational Field of Unemployment: A Tale of Seven Cities in Europe. In S. Baglioni & M. Giugni (eds.) Basingstoke, Hampshire, England; New York: Palgrave Macmillan, pp. 109–135. Bassoli, M. & Polizzi, E. (2011) (eds.) La governance dei territori. Partecipazione e rappresentanza della società civile nelle politiche locali, Milano: Angeli. Borgatti, S.P. (2002) NetDraw: Graph Visualization Software. Harvard: Analytic Technologies Borgatti, S.P., Everett, M.G. & Freeman, L.C. (2002) Ucinet for Windows: Software for Social Network Analysis. Harvard, MA: Analytic Technologies. f y y g Broadbent, J. (1998), Environmental Politics in Japan Networks of Power and Protest, Cambride, Cambridge University Press. g y Cinalli, M. (2010) “National Multi-Organisational Fields and Unemployment in Europe: A Comparison of Britain and France”, in Royall F. and D. Chabanet (eds.), Mobilising Against Marginalisation in Europe, Cambridge Scholars Publishing: Cambridge. - (2007a) “Between horizontal bridging and vertical governance: Pro-beneficiary movements in New Labour Britain”, in Pordue D. (ed.), Civil societies and social movements: potentials and problems, London: Routledge, pp. 88-108. - (2007b), The Impact of Relational Structures upon Collective Action: A Comparison of Unemployment and Asylum in New Labour Britain, CEVIPOF Working Paper, 23. Cinalli, M. (2010) “National Multi-Organisational Fields and Unemployment in Europe: A Comparison of Britain and France”, in Royall F. and D. Chabanet (eds.), Mobilising Against Marginalisation in Europe, Cambridge Scholars Publishing: Cambridge. 6. Conclusions We have thus shown that more attention has to be focused on specific interactions taking place between the exogenous characteristics of political contexts within which local publics are situated on the one hand, and other endogenous explanations in terms of resources and main orientations on the other. Findings indicate that orientations may indeed play a crucial role, especially in more amorphous and locally-rooted contexts such as Turin. In the Italian city, more established groups define the field. Yet, this happens not just because these groups have an official status, as in Lyon (although legal status grants access to information that may increase the chance of being politically involved), but also because they provide specific services. As regards resources, our analyses have dealt with financial capabilities, internal articulation, as well as diversification in the 12 scope of service provision. We have thus found that the impact of the political context needs to be appraised in terms of its interaction with the availability of certain resources, for example when governance is variably (un)structured. Most crucially, our findings are in line with earliest arguments in the literature stressing the ‘tyranny of structurelessness’ (Freeman 1973), or less normatively, the fact that resources have a greater impact upon participation in the absence of formal structures. This is a further reason to continue to inquire into the vices and virtues of governance. 13 13 Deleted: Baglioni, S. & Giugni, M. (2014) Civil Society Organization, Unemployment, and Precarity in Europe. Organizational Activities and Networks, Houndmills, Basingstoke, Hampshire, England; New York: Palgrave Macmillan.¶ References - (2007a) “Between horizontal bridging and vertical governance: Pro-beneficiary movements in New Labour Britain”, in Pordue D. (ed.), Civil societies and social movements: potentials and problems, London: Routledge, pp. 88-108. - (2007b), The Impact of Relational Structures upon Collective Action: A Comparison of Unemployment and Asylum in New Labour Britain, CEVIPOF Working Paper, 23. g p Cinalli, M. & El Hariri, A. (2011), “Contentious Opportunities in the Field of Immigration in France and Italy”, in Carmel E., Cerami A. and T. Papadopoulos (eds.), Migration and Welfare in the 'New' Europe. Social Protection and the Challenges of Integration, Policy Press: Bristol. Cinalli, M. & Giugni, M. (2010), “Mapping the Contentious Politics of Unemployment in Europe”, in Giugni M. (ed), The Contentious Politics of Unemployment in Europe: Welfare States and Political Opportunities, Palgrave: London. Cinalli, M. & Füglister, K. (2008), Networks and Political Contention over Unemployment: A Comparison of Britain, Germany and Switzerland, Mobilization : An International Journal, 13(3), pp. 259-76. ( ) pp Christopoulos, D. (2008) The Governance of Networks: Heuristic or Formal Analysis?, Political Studies, 54/2. Christopoulos, D. & Quaglia, L. (2009) Influence and Brokerage: Network Constraints in EU Banking Regulation, Journal of Public Policy, 29/2, pp. 179- 200. Curtis, R.L. & Zurcher L.A. (1973) Stable resources of protest movements: The Curtis, R.L. & Zurcher L.A. (1973) Stable resources of protest movements: The 14 multi-organizational field, Social Forces 52, pp. 53-61. multi-organizational field, Social Forces 52, pp. 53-61. Della Porta, D. & Andretta, M. (2002) Changing forms of environmentalism in Italy: the protest campaign on the high speed railway system, Mobilization : An International Journal, 7(1), pp. 59-77. Diani, M. (1992) “Analysing social movement networks” in Diani M. and R. Eyerman (eds.), Studying Collective Action. London: Sage. Diani, M. (1992) “Analysing social movement networks” in Diani M. and R. Eyerman (eds.), Studying Collective Action. London: Sage. - (1995) Green Networks: A Structural Analysis of the Italian Environmental Movement. Edinburgh: Edinburgh University Press. g g y Diani, M. & Bison I. (2004) Organizations, coalitions, and movements, Theory and Society, 33, pp. 281–309. Diani, M. & McAdam, D. (2003) Social Movements and Networks. Relational Approaches to Collective Actions. Oxford: Oxford University Press. Dowding, K. (1995) Model or Metaphor? A Critical Review of the Policy Network Approach, Political Studies, 45 (1), pp. 136–58. Dowding, K. (1995) Model or Metaphor? A Critical Review of the Policy Network Approach, Political Studies, 45 (1), pp. 136–58. References - (2001) There Must be End to Confusion: Policy Networks, Intellectual Fatigue, and the Need for Political Science Methods Courses in British Universities, Political Studies, 49 (1), pp. 89–105. Eisinger, P. K. (1973) The Conditions of Protest Behaviour in American Cities, American Political Science Review, 67, pp.11-28. Esping- Andersen, G. (1990) The three worlds of welfare capitalism, Princeton: Princeton University Press. Fabbrini, S. (2001) Features and Implications of Semi-Parliamentarism: The Direct Election of Italian Mayors, South European Society and Politics, 6(2), pp. 47 – 70. Feiock, R. & Scholz, J. (2010) (eds.) Self-Organizing Governance: Collaborative Mechanisms to Mitigate Institutional Collective Action Dilemmas. New York: Cambridge University Press. Ferrera, M. (1996) The ‘Southern Model’ of Welfare in Social Europe, Journal of European Social Policy, 6(1), pp. 17-37. Freeman, J. (1973) “The Tyranny of Structurelessness.” Berkeley Journal of Sociology, 17, pp.151–64. Gamson, W.A. & Meyer, D. (1996) “The Framing of Political Opportunity”, in McAdam D., McCarthy J.D. and M.N. Zald (eds.), Opportunities, Mobilizing Structures, and Framing: Comparative Applications of Contemporary Movement Theory, CUP: Cambridge, pp. 275-90. Glendinning, C., Powell, M. & Rummery, K. (eds.) (2002) Partnerships, New Labour and the governance of welfare, Bristol: The Policy Press. Giugni, M., Berclaz, M. & and Füglister, K. (2009) “Welfare States, Labour Markets, and the Political Opportunities for Collective Action in the Field of Unemployment: A Theoretical Framework.” in Giugni M. (ed.), The Politics of Unemployment in Europe: State and Civil Society Responses, Aldershot: Ashgate, pp. 133-149. Giugni, M. & Passy, F. (2001) (eds.) Political Altruism? Solidarity Movements in International Perspective. Lanham, Md.: Rowman & Littlefield Publishers. Gould, R.V. (1995) Insurgent Identities: Class, Community and Protest in Paris from 1848 to the Commune. Chicago: University of Chicago Press. Grote, J. (2008) “Local Governance and Organized Civil Society: Concepts and Applications.” Paper presented at the CINEFOGO conference “New Forms of Local Governance and Civil Society,” Trento, June 20–21. y, , Grote, J.& Gbikpi, B. (2002) Participatory Governance, Opladen: Leske & Budrich. J.& Gbikpi, B. (2002) Participatory Governance, Opladen: 15 Harlow, C. & Rawling, R. (2006) Promoting Accountability in Multi-Level Governance: A Network Approach. European Governance Papers (EuroGov), C-06-02, http://www.connex-network.organizzazioni/eurogov/pdf/egp-connex- C-06-02.pdf. INSEE (2011) Lyon - Evolution et structure de la population, http://www.statistiques- locales.insee.fr/FICHES/DL/DEP/69/COM/DL_COM69123.pdf ISTAT (2011) tasso di disoccupazione – livello provinciale, http://dati.istat.it/ index.aspx?queryid=298 Jessoula M., Graziano, P. & Madama, I. References (2010) ‘Selective Flexicurity’ in Segmented Labour Markets: The Case of Italian ‘Mid-Siders, Journal of Social Policy, 39(4), pp. 561–583. John, P. (2001) Local Governance in Western Europe, London : Sage. Jordan, A. (2001) The European Union: an evolving system of multi-level governance … or government?, Policy & Politics, 29.2, pp.193-208. Kingdon, J. (1995), Agendas, alternatives, and public policies (2d ed.), Boston: Little, Brown. Kitschelt, H. (1986) Political Opportunity Structures and Political Protest: Anti- Nuclear Movements in Four Democracies, British Journal of Political Science 16, pp. 57-85. Kriesi, H., Koopmans, R., Duyvendak, J.W. & Giugni, M. (1995) New Social Movements in Western Europe. A Comparative Analysis. Minneapolis, MN: University of Minnesota Press. Knoke, D. & Kuklinsky, J. (1982) Network Analysis. London: Sage. Lelieveldt, H., Astudillo, J. & Stevenson, L. (2007) “The Spectrum of associational activities: from self-help to lobbying”, in W. Maloney and S. Rossteutcher (eds.), Social Capital and Associations in European Democracies, London: Routledge, pp. 81-95. g pp Lin, N. (2001) Social Capital: A Theory of Social Structure and Action. New York: Cambridge University Press. Mayntz, R. (2003) “From Government to Governance: Political Steering in Modern Societies” Paper for the Summer Academy on IPP, Wuerzburg, September, 7– 11. Marsh, D. & Smith, M. J. (2001) There is more than one way to do political science: on different ways to study policy networks, Political Studies, 49, pp. 528 – 41. Marshall, D.J. & Staeheli, L. (2015). Mapping civil society with social network analysis: Methodological possibilities and limitations. Geoforum, 61, pp.56–66. McCarthy, J.D. & Zald, M.N. (1977) Resource Mobilization and Social Movements: A Partial Theory, American Journal of Sociology 82, pp. 1212-41. Obershall, A. (1973) Social Conflict and Social Movements. Englewood Cliffs: Prentice-Hall. OECD (2010) Online OECD Employment database, webpage www.oecd.org/document/34/0,3343,en_2649_33927_40917154_1_1_1_1,00.ht ml (last access 10/10/10). Osborne, S. (2007) Theory and Practice in International Perspective, London: Routledge. Piselli, F. & Ramella, F. (2008) Patti sociali per lo sviluppo locale, Roma: Donzelli. Polizzi, E. (2011) “Conclusioni” in Bassoli M. and E. Polizzi (eds.), La governance dei territori. Partecipazione e rappresentanza della società civile nelle politiche locali, Milano: Angeli. 16 Raab, C. (2001) Understanding policy networks: a comment on Marsh and Smith, Political Studies, 49, pp.551–6. Ranci, C. (1999) Oltre il welfare state, Bologna, Il Mulino Ranci, C. & Montagnini, E. (2009), “The impact of commodification of social care on the role and identity of third sector in Italy” in Zimmer A. and A. p Matteo Bassoli, eCampus Online University - School of Law, Via Isimbardi 10 , Novedrate (CO) 22060, Italy. References Evers (eds.) Third Sector Organizations Facing Turbulent Environments: Sports, Culture and Social Services in five European countries, Frankfurt: Nomos. Rhodes, R.A.W. (1996) The new governance: governing without government. Political Studies, 54, pp. 652-667. Rhodes, R.A.W. (1996) The new governance: governing without government. Political Studies, 54, pp. 652-667. - (1997) Understanding Governance: Policy Networks, Governance, Reflexivity and Accountability, Buckingham: Open University Press. pp - (1997) Understanding Governance: Policy Networks, Governance, Reflexivity and Accountability, Buckingham: Open University Press. Scott, J. (2000) Social Network Analysis. London: Sage. Sohn, C. & Giffinger, R. (2015) A Policy Network Approach to Cross-Border Metropolitan Governance: The Cases of Vienna and Bratislava. European Planning Studies, 23(6), pp.1187–1208. g ( ) pp Sullivan, H. & Skelcher, C. (2002) Working Across Boundaries: Collaboration in Public Services, Basingstoke: Macmillan-Palgrave. Tarrow, S. (1998) Power in Movement. Cambridge: Cambridge University Press. Tilly, C. (1978) From Mobilization to Revolution. Reading, Mass.: Addison-Wesley. Torpe, L., & Ferrer-Fons, M. (2007) “The internal structure of associations”, in W. Maloney and S. Rossteutcher (eds.), Social Capital and Associations in European Democracies, London: Routledge, pp. 81-95 Wasserman, S. & Faust, K. (1994) Social Network Analysis: Method and Applications. Cambridge: Cambridge University Press. 17 17
https://openalex.org/W3045050403	https://link.springer.com/content/pdf/10.1007/s00701-020-04494-4.pdf	English	null	Ergonomics and musculoskeletal disorders in neurosurgery: a systematic review	Acta neurochirurgica	2,020	cc-by	5,830	Abstract Meling torsteinrmeling@gmail.com 1 Department of Clinical Neurosciences, Division of Neurosurgery, Geneva University Hospitals, Geneva, Switzerland 2 Department of Neurosurgery, Western General Hospital, Edinburgh, UK 3 Faculty of Medicine, University of Geneva, Geneva, Switzerland Abbreviations WMSDs Work-related musculoskeletal disorders WHO World Health Organization CNS Congress of Neurosurgical Surgeons ETV Endoscopic third ventriculostomy OR Operating room CTS Carpal tunnel syndrome HMD Head-mounted display This article is part of the Topical Collection on Neurosurgery general * Torstein R. Meling torsteinrmeling@gmail.com 1 Department of Clinical Neurosciences, Division of Neurosurgery, Geneva University Hospitals, Geneva, Switzerland 2 Department of Neurosurgery, Western General Hospital, Edinburgh, UK 3 Faculty of Medicine, University of Geneva, Geneva, Switzerland https://doi.org/10.1007/s00701-020-04494-4 Acta Neurochirurgica (2020) 162:2213–2220 https://doi.org/10.1007/s00701-020-04494-4 Acta Neurochirurgica (2020) 162:2213–2220 REVIEW ARTICLE - NEUROSURGERY GENERAL Ergonomics and musculoskeletal disorders in neurosurgery: a systematic review Alexandre Lavé1 & Renato Gondar1 & Andreas K. Demetriades2 & Torstein R. Meling1,3 Received: 6 June 2020 /Accepted: 14 July 2020 # The Author(s) 2020 / Published online: 23 July 2020 Abstract Background Work-related musculoskeletal disorders (WMSDs) are a growing and probably undervalued concern for neurosur- geons and spine surgeons, as they can impact their quality of life and career length. This systematic review aims to ascertain this association and to search for preventive measures. Methods We conducted a PRISMA-P-based review on ergonomics and WMSDs in neurosurgery over the last 15 years. Twelve original articles were included, of which 6 focused on spine surgery ergonomics, 5 cranio-facial surgery (mainly endoscopic), and one on both domains. Results We found a huge methodological and content diversity among studies with 5 surveys, 3 cross-sectional studies, 2 retrospective cohorts, and 2 technical notes. Spine surgeons have sustained neck flexion and neglect their posture during surgery. In a survey, low back pain was found in 62% of surgeons, 31% of them with a diagnosed lumbar disc herniation, and 23% of surgery rate. Pain in the neck (59%), shoulder (49%), finger (31%), and wrist (25%) are more frequent than in the general l ti C l t l d h d li l ti hi ith i i l ti h f i ti Results We found a huge methodological and content diversity among studies with 5 surveys, 3 cross-sectional studies, 2 retrospective cohorts, and 2 technical notes. Spine surgeons have sustained neck flexion and neglect their posture during surgery. In a survey, low back pain was found in 62% of surgeons, 31% of them with a diagnosed lumbar disc herniation, and 23% of surgery rate. Pain in the neck (59%), shoulder (49%), finger (31%), and wrist (25%) are more frequent than in the general population. Carpal tunnel syndrome showed a linear relationship with increasing cumulative hours of spine surgery practice. Among cranial procedures, endoscopy was also significantly related to shoulder pain while pineal region surgery received some attempts to optimize ergonomics. Conclusions Ergonomics in neurosurgery remains underreported and lack attention from surgeons and authorities. Improvements shall target postural ergonomics, equipment design, weekly schedule adaptation, and exercise. Keywords Surgery . Systematic review . Ergonomics . Neurosurgical practice . Spine surgery . Neuro-endoscopy . Musculoskeletal disorders Abbreviations WMSDs Work-related musculoskeletal disorders WHO World Health Organization CNS Congress of Neurosurgical Surgeons ETV Endoscopic third ventriculostomy OR Operating room CTS Carpal tunnel syndrome HMD Head-mounted display This article is part of the Topical Collection on Neurosurgery general * Torstein R. Data collection The following parameters were extracted and registered for each article: (1) study ID; (2) study characteristics (author, year, country, prospective, or retrospective study); (3) patient demographics; (4) sample size; (5) ergonomics assessment; (6) type of surgery performed; (7) outcome measurements; (8) follow-up (FU) time. If necessary, the consensus was reached by both authors through discussions with the senior author (TRM). Data from each study reporting prevalence estimates were extracted. We conducted a systematic review and meta-analysis of published studies on this topic. Our goals were to determine (1) the prevalence of WMSDs in neurosurgery and spine sur- gery, (2) their burden on quality of life and ability to work, and (3) the possible preventive interventions to minimize their consequences. Risk of bias and quality of study lead aprons, and high surgical loads pose risks for repetitive strain injury, carpal tunnel syndrome, and wrist or hand dys- function. Other examples of preventable disorders affecting muscles, tendons, and nerves include tendinitis, degenerative spine diseases, thoracic outlet syndrome, and tension neck syndrome. The accepted articles were independently graded by two au- thors (AL and RG) according to the Newcastle–Ottawa Quality Assessment Scale [30] for quality assessment of non-randomized studies. The level of evidence for each study was evaluated using the Oxford Centre for Evidence-Based Medicine guidelines. WMSDs may lead to time off work and could nega- tively impact the quality of a surgeon’s performance [14]. An increasing number of studies have drawn at- tention to the importance of postural ergonomics, aiming to reduce musculoskeletal fatigue among neuro- surgeons [1]. Search strategy and study selection Results for continuous variables are reported as mean ± stan- dard deviation (SD) or range. For articles that did not report mean and SD, we estimated the mean and SD according to the methodology described by Hozo et al. [15]. Categorical vari- ables are presented as median and quartiles or by absolute and relative frequencies. According to the PRISMA-P guidelines (Preferred Reporting Items for Systematic review and Meta-Analysis Protocols checklist) [26], the authors systematically reviewed the pub- lished literature on ergonomics and musculoskeletal disorders in neurosurgery. The literature search was performed on MEDLINE (US National Library of Medicine), Embase Library, and Google Scholar, including papers from 2006 to 2020. The following Medical Subject Headings (MeSH) terms were used: “ergonomics”[MeSH Terms] OR “ergonomics”[All Fields]) AND (“neurosurgical procedures”[MeSH Terms] OR (“neurosurgical”[All Fields] AND “procedures”[All Fields]) OR “neurosurgical procedures”[All Fields] OR “neurosurgery”[All Fields] OR “neurosurgery”[MeSH Terms]”. Introduction Neurosurgeons and spine surgeons are exposed to work- related musculoskeletal disorders (WMSDs), which can neg- atively impact their quality of life and career length. According to the World Health Organization definition, WMSDs include all health problems of the locomotor appa- ratus (the skeleton plus muscles, tendons, cartilage, ligaments, and nerves) and all relevant forms of ill health, ranging from mild or transitory disorders to irreversible and disabling inju- ries [31]. Spine neurosurgeons in particular are exposed to WMSDs by adopting sustained non-neutral positions, with prolonged neck flexion and coronal misalignment as they operate in a standing position, frequently leaning over the operating table [19]. A significant prevalence of neck and back pain has been reported among spine surgeons [2]. The use of vibrating pow- er tools, Kerrison punches with repetitive hand movements, 2214 Acta Neurochir (2020) 162:2213–2220 Results In total, 18 titles were retained for full-paper screening. Six papers were excluded from the qualitative analysis as these were systematic reviews or not written in English. Thus, twelve studies were included in the final analysis (Fig. 1). Six articles [1, 2, 12, 17, 19, 20] focused on spine surgery ergonomics and related issues; five others discussed cranio- facial surgery [5, 9, 22, 23, 28], mainly endoscopic, and one survey reported on both spinal and cranial surgery [14]. No limits were defined regarding the year of publication, language, or publication status, and no other additional filters were applied after running the search in the abovementioned databases (Fig. 1). It was not possible to aggregate demographic results as the variability of study designs and purposes was significant. Among the spine-oriented studies, three were surveys, three cross-sectional studies (randomized or not), and one was a technical note. As for the cranio-facial-oriented studies, these comprised two surveys, two retrospective descriptive cohorts, and one technical note. Among these last ones, only one in- cluded microscopic approaches, while the rest focused on en- doscopic sinus and skull base surgery. Due to the diversity of parameters measured and questions reported, no quantitative analysis was performed. The results are shown in Table 1. After removing duplicates, the titles and abstracts from 428 articles were screened. Two authors (AL and RG) independently screened titles and abstracts of all identi- fied articles, and full-text copies of all relevant articles were acquired. The following inclusion criteria were used: (1) peer- reviewed research articles, prospective, or retrospective, on ergonomics in the field of neurosurgery and spine surgery and (2) studies written in the English language. 2215 Acta Neurochir (2020) 162:2213–2220 Fig. 1 PRISMA flowchart Fig. 1 PRISMA flowchart Discussion Like other groups working in medical health, our review shows that surgeons are significantly impacted by musculo- skeletal disorders. According to the statistics of the Health and Safety Executive, the UK government’s agency responsible for the encouragement, regulation, and enforcement of work- place health, there were 60,000 work-related cases of the mus- culoskeletal disorder among people working in human health and social work in Great Britain in 2019 [16]. This rate is significantly higher than for workers across all other industrial branches. Surgeons are not spared from musculoskeletal dis- orders and are even considered at high risk of developing WMSDs [3, 10]. The current review allowed us to summarize the state-of- affairs in a growing field that has been neglected for many years; to our knowledge, this is the first of its kind in neurosurgery. Firstly, ergonomics in neurosurgery remains underreported and lack attention from surgeons, hospital administrations, surgical material designers, and health in- surance companies. Secondly, the challenges in ergonom- ics are quite different between spine surgery and cranio- facial surgery as the volume of pathology, as well as the instruments and positions used, are substantially different. Thirdly, the impact of such problems on surgical perfor- mance and surgeons’ longevity and quality of life is largely unknown. Lastly, any formal attention given to such as- pects during the formative years of neurosurgical training seems non-existent. Among spine surgeons, sustained neck flexion is seen as a major risk factor for musculoskeletal disorders [14, 19, 20]. Such posture is frequently necessary during procedures, par- ticularly during cervical spine approaches. An additional con- tributing factor may be that surgeons tend to neglect their Acta Neurochir (2020) 162:2213–2220 2216 Table 1 List of the clinical data available from the 12 articles included in the review Source Year of publication Design Specialty Geographic location Study participant Principal outcomes Gadjradj et al. [14] 2020 Survey Cranio-facial and spine The Netherlands N = 417 - 73.6% experienced WMSDs. - Spine surgery as first cause of WMSDs. - 11.3% of the respondents had to take time off work. Auerbach et al. [2] 2011 Survey Spine USA N = 561 - 38% of respondents experienced neck pain/strain/spasm. - 31% of respondents reported lumbar disc herniation/radiculopathy. - 28% cases of cervical disc herniation/radiculopathy. - 24% of rotator cuff disease among the respondents. - 20% reported varicose veins or peripheral edema - 18% experienced lateral epicondylitis. Forst et al. Discussion [9] 2017 Technical note Cranio-facial UK N = 350 - Semi-sitting position, with the addition of having the patient’s head flexed and turned to the side of the surgeon allowed a more com- fortable and intuitive placement of the operator’s arms in relation to the nares. Ramakrishnan et al. [23] 2017 Survey Cranio-facial UK - - Discomfort after endoscopic procedure in the standing position was worse in the legs and low back, whereas, in the sitting position, it was predominant in the upper back and arms. Ito et al. [17] 2015 Technical note Spine Japan N = 14 - Simulations performed using the support significantly reduced the musculoskeletal loading on the ventral side of the left foot by 70%. - The device generated circa 10% of the musculoskeletal load on the right hand. posture during surgery as the procedure itself requires full concentration. Auerbach et al. [2] surveyed the type and prevalence of musculoskeletal disorders among 561 surgeon members of the Scoliosis Research Society; the majority of complaints were low back pain (62%), neck pain (59%), and shoulder pain (49%). A significant number of surgeons suffered also from leg pain with radiculopathy (31%), finger pain (31%), neck pain with radiculopathy (28%), elbow pain (28%), and wrist pain (25%) [2]. The prevalence of neck pain in the gen- eral population is approximately 20% [25]. Another statistic from this study is particularly insightful: 4.6% of the respon- dents had undergone surgical intervention for cervical disc disease. This is in major contrast to the prevalence of 0.35% of cervical radiculopathy observed in the general population [13] and this difference was attributed by some authors [2] to the position with the neck flexed during extended time pe- riods. Importantly, 31% of the surgeons in this survey reported lumbar disc herniation and symptomatic back pain with radiculopathy, with a surgery rate of 23% [2]. Here, again, there is a gap between the prevalence of symptomatic lumbar herniated disc in the general population (up to 5% [29]) and the prevalence observed among the 561 surgeons surveyed. Prolonged operative times while standing, awkward spine po- sition and movement, or lifting heavy patients have been im- plicated to explain such a difference. Similar findings were reported by Gadjradj et al. in a sur- vey among members of the Congress of Neurological Surgeons [14]. Discussion [12] 2006 Survey Spine USA N = 371 - 107 reported cases (28.8%) of CTS. - Average latency from beginning of residency to the development of CTS: 17.1 years. - Neurosurgeons as a subgroup showed a significant increase in risk of CTS (adjusted OR = 2.03). - Kerrison rongeur was the greatest occupational risk to the upper extremities (adjusted OR = 2.72). - 30.8% of the surgeons with CTS reported that the chronic pain interferes with their work, and 23.1% modify the way they perform surgery because of CTS. Park et al. [19] 2012 Randomized crossover Spine South Korea N = 12 - Lumbar lordosis, cervical lordosis, and occipital angle closer to natural standing values when discectomy was performed with a loupe. - Whole spine angles closer to natural standing position with table height at midpoint between the umbilicus and sternum. Park et al. [20] 2013 Randomized crossover Spine South Korea N = 18 - Whole spine angles closer to natural standing position closer to natural standing values when discectomy was performed with a microscope. - When using a microscope, lumbar lordosis, thoracic kyphosis, and cervical lordosis showed no differences according to table heights above the umbilicus. Van Lindert et al. [28] 2004 Retrospective Cranio-facial The Netherlands N = 269 - HMD decreased visual strain. - Working position was considered to be more comfortable when wearing the HMD. - Eye-hand coordination was improved thanks to HMD. Choque-Velasquez et al. [5] 2018 Retrospective Cranio-facial Finland/Italy N = 56 - Steeper sitting position with the upper torso and the head of the patient bent forward and downward allows a more ergonomic working position for the neurosurgeon allowing his/her arms to rest over the patient’s shoulders during the procedure. Albayrak et al. [1] 2007 Cross-sectional Spine The Netherlands N = 7 - EMG results showed an average reduction of 44% for the erector spinae muscle, 20% for the semitendinosus muscle, and 74% for the gastrocnemius muscle. - Support considered as comfortable, safe, and simple to use Acta Neurochir (2020) 162:2213–2220 2217 Source Year of publication Design Specialty Geographic location Study participant Principal outcomes Ramakrishnan et al. [22] 2019 Survey Cranio-facial USA N = 250 - Operating in the standing position resulted in physical discomfort mostly of the trunk and lower body. - Operating in the sitting position shifted the muscular strain to the upper back, neck, and shoulders. Ekanayake et al. Discussion Neck, shoulder, and back complaints were the most prevalent in the context of three procedures: lumbar discectomy, ventriculo-peritoneal shunting, and endoscopic third ventriculostomy (ETV). Neck complaints were the most prevalent in respondents performing lumbar discectomy. Again, surgical procedures for degenerative spinal diseases were most frequently mentioned as preceding factors before such complaints. Univariate analysis showed that tenure of less than 15 years and an operating room (OR) furnished er- gonomically were associated with fewer WMSDs. Multivariate analysis, however, only revealed that having a tenure of less than 15 years was protective. Finally, height, sex, age, dominant hand, caseload, the average duration of the procedure, hours per week in the OR, and having the OR furnished ergonomically were all non-significant factors. A high rate of carpal tunnel syndrome (CTS) among spine surgeons was reported by Forst et al. [12]. The survey includ- ed 371 respondents including 274 spine specialists, where 16.4% were neurosurgeons and 57.4% were orthopedic sur- geons. When compared with a group of non-surgeons, the odds of developing CTS were at least twice as high for spine surgeons as for non-surgical medical practitioners, with 29% of the surgeons reporting CTS [12]. The risk of developing CTS showed a linear relationship with increasing cumulative hours of surgery. A particularly relevant observation for 2218 Acta Neurochir (2020) 162:2213–2220 neurosurgical practice is that the use of the Kerrison rongeur appeared to be a major risk factor for the development of CTS. Indeed, the neurosurgeon subgroup showed a significant in- crease in the risk of developing CTS, with an adjusted odds ratio of 2 [12]. Other significant predictors of CTS among all medical personnel were obesity and a length of professional practice of more than 5 years. Auerbach et al. [2] reported a lower rate of CTS of approximately 9%. Some discrepancies in the rate of CTS observed between Forst et al. [12] and Auerbach et al. [2] could be explained by the potential re- sponse bias in the survey of Forst et al. [12], with a low response rate of 16.8%. comparable with a natural standing position in terms of values of lumbar lordosis, thoracic kyphosis, cervical lordosis, and occipital angle. Moreover, all parameters were also close to the natural standing position when a microscope was used during the simulation [20] if an optimal height of the operating table (midpoint between the umbilicus and sternum) is pro- vided. Discussion As recently pointed out by Demetriades et al., micro- surgery with a microscope is ergonomically more efficient than with loupes and headgear, at least in neurosurgery [8]. In order to improve ergonomics in neurosurgery, especially during spinal procedures that can be long and tiring, engineers at the Delft University of Technology developed ergonomic body support that supports surgeons during both open and min- imally invasive procedures [1]. The aim was to reduce a sur- geon’s muscle activity in the lower back and extremities while keeping the surgical field unobstructed. EMG results from sev- en independent participating surgeons showed that the proto- type was effective in reducing the activity of the lower back and leg muscles during open surgery. This prototype, despite being described by some of the surgeons as restrictive to their move- ments, was considered safe in use [1]. Similarly, Ito et al. [17] proposed a body support device to reduce musculoskeletal loading for surgeons performing neurosurgical microsurgical procedures. This ergonomic device, tested on 14 neurosur- geons, efficiently improved the stability and smoothness of the surgeon’s motion while reducing musculoskeletal loading during procedures performed in the standing position [17]. According to Gadjradj et al. [14], ETV is associated with a high rate of muscular complaints. In the literature, endoscopic and laparoscopic procedures have also been associated with high rates of WMSDs [11, 22–24, 27]. Neurosurgeons are not performing laparoscopic procedures, but endoscopic surgery in a standing position is frequent in neurosurgical practice. In particular, upper limb complaints are frequently reported [2, 14]. Forty-nine percent of surgeons in the survey by Auerbach et al. [2] suffered from shoulder pain, compared with a lower percentage at 24.7% in Gadjradj et al. [14]. Interestingly, 100% of the surgeons surveyed in the latter reported shoulder pain while performing ETVs. In an online survey among members of the European Rhinologic Society [24], 80% of the 250 responders reported musculoskeletal problems. Neck and back were the commonest sites of symptoms, in approximately 60% of cases. A positive correlation was found between musculoskel- etal symptoms and the standing position. Ramakrishnan et al. [23] compared the sitting and standing positions in a study on endoscopic sinus surgery and showed that the standing posi- tion led to increased discomfort in the lower extremities and back compared with the sitting position, which caused more discomfort in the upper extremities. Discussion Hand, neck, and eye dis- comfort were reported in both conditions. Rather than reduc- ing muscle strain, the authors pointed out that the positional switch only transferred the tension to other muscle groups. Thus, there is still some controversy as to whether a sitting or standing position offers the most optimal ergonomic posi- tion during endoscopic sinus surgery. Ramakrishnan et al. [22] recommended some simple mea- sures to decrease WMSDs among surgeons performing skull base or endoscopic procedures. Appropriate monitor placement at 80–120 cm in front of the surgeon, as well as a correct adjust- ment of the table to keep the hand in line with the elbow at approximately 10 cm, with the arms slightly abducted and inter- nally rotated could efficiently decrease the physical discomfort. As described earlier, the sitting position did not show a signifi- cant advantage in reducing muscular strain. The authors also cautioned surgeons not to look directly through an endoscope lens, to limit the time wearing headlight or loupes, and to limit wrist flexion, deviation, and rotation to less than 15 degrees [22]. Surgery of the pineal region is considered as a long and tiring procedure. In this context, some authors [5] proposed a variant of the classic position (semi-sitting or sitting position) to improve ergonomics, while reducing feared complications like air embolism [21]. Thus, they proposed a “praying posi- tion,” a position in which the patient is sitting steeper, with the head bent forward and downward and tilted about 30° [5]. Ergonomic modifications in patient positioning were also pro- posed by some authors [9] during endoscopic pituitary surger- ies to improve operating by setting up conditions more com- fortable for the neurosurgeon [9]. As described above, lengthy hyperflexion of the cervical spine is associated with musculoskeletal disorders. To deter- mine the optimal sagittal balance of the spine during neuro- surgical procedures, Park et al. [19, 20] compared whole spine angles of 12 spine surgeons simulating surgeries on spine models. Three different methods to visualize the surgical field (naked eye, loupe, and out of loupe) were used and three different operating table heights were suggested. The authors showed that appropriate operating table height was the first effective step to improve ergonomics in the OR. Discussion Utilization of loupes should be coupled with a table height situated at the midpoint between the umbilicus and sternum and is Lastly, head-mounted display (HMD) utilization during neuro-endoscopic procedures (mainly ETV and transsphenoidal pituitary surgery) was assessed in the Acta Neurochir (2020) 162:2213–2220 2219 retrospective study of van Lindert et al. [28]. This seemed to improve the ergonomics of neuro-endoscopic and other endoscope-controlled procedures by allowing the surgeon to have a neutral head position, regardless of the surgeon’s posi- tion relative to the surgical field. The authors showed that visual strain and ocular fatigue was decreased [28]. factors; and (vi) maintenance of good practice. Continuous self-improvement and lifelong learning are becoming essential components of a competent and successful surgeon. Limitations The small amount of literature available on a specific topic is a natural limitation. The risk of selection bias was minimized by conducting a two-author data collection, as per PRISMA guide- lines. Due to the diversity of parameters measured and questions reported, no quantitative analysis was performed. Lastly, there was not enough evidence synthesized in order to propose recom- mendations, other than further research in this area. Future directions Trainees and young neurosurgeons must be aware of the sig- nificant risk of suffering from WMSDs and learn good practice early, as well as develop an awareness of the risk they are exposed to (for instance, the long-term use of the Kerrison rongeurs or loupes). Exercise also seems to reduce the risk of developing WMSDs in other standing and manual professions [7, 18, 32]. For example, some companies have set up work- place fitness programs, consisting of stretching and warm-up exercises for over-taxed joints. In the literature, exercise seems to have a moderate, but the definite impact, on WMSDs [31]. Challenges in ergonomics are quite different between the spine and cranio-facial surgery. Globally, fewer musculoskel- etal disorders among cranio-facial surgeries specialist are re- ported, and most of the complaints are related to upper limb disorders [14, 23, 28]. Thus, ameliorations of ergonomics in cranio-facial surgeries are dependent upon correct patient po- sitioning, utilization of HMDs, and correct positioning of the endoscope screen. Interestingly, neither of the sitting or stand- ing position seems to show any advantage to reduce global muscular strain. In spine surgeries, improvements in postural ergonomic can be achieved by ideal positioning of the OR table [19, 20], the ideal positioning of the patient [5], the use of a microscope during spine surgeries [19, 20], and the use of body support to decrease surgeon’s muscle strain [1, 17]. Trainees and young neurosurgeons must be aware of the sig- nificant risk of suffering from WMSDs and learn good practice early, as well as develop an awareness of the risk they are exposed to (for instance, the long-term use of the Kerrison rongeurs or loupes). Exercise also seems to reduce the risk of developing WMSDs in other standing and manual professions [7, 18, 32]. For example, some companies have set up work- place fitness programs, consisting of stretching and warm-up exercises for over-taxed joints. In the literature, exercise seems to have a moderate, but the definite impact, on WMSDs [31]. Authors’ contributions Data collection: A.L., R.G. Review of the literature: A.L., R.G. Manuscript drafting: A.L., T.R.M., A.K.D., R.G. Critical revision: all authors Compliance with ethical standards This article does not contain any studies with human participants or an- imals performed by any of the authors Conclusions Ergonomics for neurosurgeons has seemingly been neglected for many years and remains underreported. Surgeons, hospital ad- ministrations, surgical material designers, and health insurance schemes have a role to play in taking action to protect surgeons from this potential burden and occupational health hazard. Challenges in ergonomics are quite different between the spine and cranio-facial surgery. Globally, fewer musculoskel- etal disorders among cranio-facial surgeries specialist are re- ported, and most of the complaints are related to upper limb disorders [14, 23, 28]. Thus, ameliorations of ergonomics in cranio-facial surgeries are dependent upon correct patient po- sitioning, utilization of HMDs, and correct positioning of the endoscope screen. Interestingly, neither of the sitting or stand- ing position seems to show any advantage to reduce global muscular strain. In spine surgeries, improvements in postural ergonomic can be achieved by ideal positioning of the OR table [19, 20], the ideal positioning of the patient [5], the use of a microscope during spine surgeries [19, 20], and the use of body support to decrease surgeon’s muscle strain [1, 17]. The impact of WMSDs in surgical performance and sur- geon’s longevity and quality of life is largely unknown. A systematic approach into the field of ergonomics in neu- rosurgery will require a sustained effort with multi- disciplinary input aiming at eventual long-lasting benefits for both neuro and spine surgeons, trainees, and patients. Authors’ contributions Data collection: A.L., R.G. Review of the literature: A.L., R.G. Manuscript drafting: A.L., T.R.M., A.K.D., R.G. Critical revision: all authors Neurosurgeons may have to take time off work because of these WMSDs, and their surgical skills could also be altered by these. Specific costs generated from these forced vacations are not known, but overall, the National Research Council estimated that workers’ compensation costs associated with lost workdays due to musculoskeletal disorders range from $13 to $20 billion annually [6]. However, indirect costs must be added to the cost of compensation claims, leading to esti- mates as high as $45 to $54 billion annually for musculoskel- etal disorders reported as work related [6]. Funding information Open Access funding provided by University of Geneva. References 17. Ito K, Horiuchi T, Seguchi T, Hongo K (2015) Usefulness of a device for body support during operations performed while stand- ing. Neurosurg Rev 38:731–737. https://doi.org/10.1007/s10143- 015-0633-3 1. Albayrak A, van Veelen MA, Prins JF, Snijders CJ, de Ridder H, Kazemier G (2007) A newly designed ergonomic body support for surgeons. Surg Endosc 21:1835–1840. https://doi.org/10.1007/ s00464-007-9249-1 18. Kelly D, Shorthouse F, Roffi V, Tack C (2018) Exercise therapy and work-related musculoskeletal disorders in sedentary workers. Occup Med 68:262–272. https://doi.org/10.1093/occmed/kqy054 2. Auerbach JD, Weidner ZD, Milby AH, Diab M, Lonner BS (2011) Musculoskeletal disorders among spine surgeons: results of a survey of the Scoliosis Research Society membership. Spine (Phila Pa 1976) 36: E1715–E1721. https://doi.org/10.1097/BRS.0b013e31821cd140 19. Park JY, Kim KH, Kuh SU, Chin DK, Kim KS, Cho YE (2012) Spine surgeon’s kinematics during discectomy according to operating table height and the methods to visualize the surgical field. Eur Spine J 21: 2704–2712. https://doi.org/10.1007/s00586-012-2425-6 3. Berguer R (1999) Surgery and ergonomics. Arch Surg 134:1011– 1016. https://doi.org/10.1001/archsurg.134.9.1011 20. Park JY, Kim KH, Kuh SU, Chin DK, Kim KS, Cho YE (2014) Spine surgeon’s kinematics during discectomy, part II: operating table height and visualization methods, including microscope. Eur Spine J 23:1067–1076. https://doi.org/10.1007/s00586-013-3125-6 4. Cambridge Dictionary Ergonomics. (2020). https://dictionary. cambridge.org/fr/dictionnaire/anglais/ergonomics. Accessed 03. 05.2020 21. Porter JM, Pidgeon C, Cunningham AJ (1999) The sitting position in neurosurgery: a critical appraisal. Br J Anaesth 82:117–128. https://doi.org/10.1093/bja/82.1.117 5. Choque-Velasquez J, Colasanti R, Resendiz-Nieves JC, Gonzales- Echevarria KE, Raj R, Jahromi BR, Goehre F, Lindroos AC, Hernesniemi J (2018) Praying sitting position for pineal region surgery: an efficient variant of a classic position in neurosurgery. World Neurosurg 113:e604–e611. https://doi.org/10.1016/j.wneu. 2018.02.107 22. Ramakrishnan VR (2019) Ergonomics in endoscopic sinus surgery. Curr Opin Otolaryngol Head Neck Surg 27:25–28. https://doi.org/ 10.1097/MOO.0000000000000504 23. Ramakrishnan VR, Milam BM (2017) Ergonomic analysis of the surgical position in functional endoscopic sinus surgery. Int Forum Allergy Rhinol 7:570–575. https://doi.org/10.1002/alr.21911 6. Council NR, Medicine Io (2001) Musculoskeletal disorders and the workplace: low back and upper extremities. The National Academies Press, Washington, DC. doi:https://doi.org/10.17226/ 10032 24. Rimmer J, Amin M, Fokkens WJ, Lund VJ (2016) Endoscopic sinus surgery and musculoskeletal symptoms. Rhinology 54:105– 110. https://doi.org/10.4193/Rhin15.217 7. da Costa BR, Vieira ER (2008) Stretching to reduce work-related musculoskeletal disorders: a systematic review. J Rehabil Med 40: 321–328. https://doi.org/10.2340/16501977-0204 25. Rubin DI (2007) Epidemiology and risk factors for spine pain. Neurol Clin 25:353–371. https://doi.org/10.1016/j.ncl.2007.01.004 8. References Demetriades AK, Meling TR, Ringel FA, Schaller K (2020) Postural ergonomics and micro-neurosurgery: microscope has an edge over loupes. J Am Coll Surg. https://doi.org/10.1016/j. jamcollsurg.2020.04.026 26. Shamseer L, Moher D, Clarke M, Ghersi D, Liberati A, Petticrew M, Shekelle P, Stewart LA, Group P-P (2015) Preferred reporting items for systematic review and meta-analysis protocols (PRISMA- P) 2015: elaboration and explanation. BMJ 350:g7647. https://doi. org/10.1136/bmj.g7647 9. Ekanayake J, Baudracco I, Quereshi A, Vercauteren T, Dorward NL (2018) The conversational position in endoscopic pituitary sur- gery. Br J Neurosurg 32:44–46. https://doi.org/10.1080/02688697. 2017.1406058 27. Stomberg MW, Tronstad SE, Hedberg K, Bengtsson J, Jonsson P, Johansen L, Lindvall B (2010) Work-related musculoskeletal dis- orders when performing laparoscopic surgery. Surg Laparosc Endosc Percutan Tech 20:49–53. https://doi.org/10.1097/SLE. 0b013e3181cded54 10. Epstein S, Sparer EH, Tran BN, Ruan QZ, Dennerlein JT, Singhal D, Lee BT (2018) Prevalence of work-related musculoskeletal dis- orders among surgeons and interventionalists: a systematic review and meta-analysis. JAMA Surg 153:e174947. https://doi.org/10. 1001/jamasurg.2017.4947 28. van Lindert EJ, Grotenhuis JA, Beems T (2004) The use of a head- mounted display for visualization in neuroendoscopy. Comput Aided Surg 9:251–256. https://doi.org/10.3109/10929080500165476 j g 11. Esposito C, El Ghoneimi A, Yamataka A, Rothenberg S, Bailez M, Ferro M, Gamba P, Castagnetti M, Mattioli G, Delagausie P, Antoniou D, Montupet P, Marte A, Saxena A, Bertozzi M, Philippe P, Varlet F, Lardy H, Caldamone A, Settimi A, Pelizzo G, Becmeur F, Escolino M, De Pascale T, Najmaldin A, Schier F (2013) Work-related upper limb musculoskeletal disorders in pae- diatric laparoscopic surgery A multicenter survey. J Pediatr Surg 48:1750–1756. https://doi.org/10.1016/j.jpedsurg.2013.01.054 29. Vialle LR, Vialle EN, Suárez Henao JE, Giraldo G (2010) Lumbar disc herniation. Rev Bras Ortop (English Edition) 45:17–22. https:// doi.org/10.1016/s2255-4971(15)30211-1 30. Wells GA, Shea B, O’Connell D, Peterson J, Welch V, Losos M, Tugwell P (2019) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www. ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed 27. 05.2020 12. Forst L, Friedman L, Shapiro D (2006) Carpal tunnel syndrome in spine surgeons: a pilot study. Arch Environ Occup Health 61:259– 262. https://doi.org/10.3200/AEOH.61.6.259-262 31. Work-Related Musculoskeletal Disorders & Ergonomics. (2020) National Center for Chronic Disease Prevention and Health Promotion. https://www.cdc.gov/workplacehealthpromotion/ health-strategies/musculoskeletal-disorders/index.html. Accessed 07.05.2020 p g 13. Frymoyer JW (1992) Lumbar disk disease: epidemiology. Instr Course Lect 41:217–223 14. Gadjradj PS, Ogenio K, Voigt I, Harhangi BS (2020) Ergonomics and related physical symptoms among neurosurgeons. World Neurosurg 134:e432–e441. https://doi.org/10.1016/j.wneu.2019.10.093 32. Conflict of interest The authors declare that they have no conflict of interest. Conflict of interest The authors declare that they have no conflict of interest. Conflict of interest The authors declare that they have no conflict of interest. As ergonomics is defined as the scientific study of people and their working conditions, especially done in order to im- prove effectiveness [4], it is a field that deserves investment, both in knowledge and in practice. It is evident that we still need to define optimal ergonomics for our specialty, paying due attention to (i) postural ergonomics; (ii) instrument design and technology; (iii) the environment around the surgeon (e.g., op- erating table, chairs, lights, monitors, navigation equipment); (iv) composition of operating lists (length of procedure, posture during each, weekly schedule); (v) exercise to avoid risk Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adap- tation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, pro- vide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 2220 Acta Neurochir (2020) 162:2213–2220 Publisher’s note Springer Nature remains neutral with regard to jurisdic- tional claims in published maps and institutional affiliations. References Yao Y, Zhao S, An Z, Wang S, Li H, Lu L, Yao S (2019) The associations of work style and physical exercise with the risk of work-related musculoskeletal disorders in nurses. Int J Occup Med Environ Health 32:15–24. https://doi.org/10.13075/ijomeh. 1896.01331 15. Hozo SP, Djulbegovic B, Hozo I (2005) Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 5:13. https://doi.org/10.1186/1471-2288-5-13 16. Human health and social work activities statistics in Great Britain 2019. (2019). https://www.hse.gov.uk/statistics/industry/health. pdf. Accessed 27.05.2020 Publisher’s note Springer Nature remains neutral with regard to jurisdic- tional claims in published maps and institutional affiliations.
https://openalex.org/W4214857548	https://refubium.fu-berlin.de/bitstream/fub188/34442/1/Di%20Filippo_Manuelli_2021_464737.published.pdf	English	null	Hic sunt leones. Iconographic analysis and computational modelling for the study of the Iron Age free-standing lions of the Elbistan plain (south-eastern Anatolia)	Asia Anteriore antica	2,022	cc-by	17,211	Asia Anteriore Antica Journal of Ancient Near Eastern Cultures Asia Anteriore Antica Journal of Ancient Near Eastern Cultures 1 A detailed synthesis is reported by Hawkins (2000: 329). Asia Anteriore Antica. Journal of Ancient Near Eastern Cultures 3: 43-72, 2021 ISSN 2611-8912 (online) \| DOI: 10.36253/asiana-1204 Citation: Francesco Di Filippo, Feder- ico Manuelli (2021) Hic sunt leones. Icono- graphic analysis and computational modelling for the study of the Iron Age free-standing lions of the Elbistan plain (south-eastern Anatolia). Asia Anteriore Antica. Journal of Ancient Near East- ern Cultures 3: 43-72. doi: 10.36253/ asiana-1204 Francesco Di Filippo1, Federico Manuelli2 1 Institute for Studies on the Mediterranean (CNR – ISMed), Rome 2 Institute of Heritage Science (CNR - ISPC), Rome and Institut für Altorientalistik (FU Berlin), Berlin francesco.difilippo@cnr.it; federico.manuelli@cnr.it Francesco Di Filippo1, Federico Manuelli2 1 Institute for Studies on the Mediterranean (CNR – ISMed), Rome 2 Institute of Heritage Science (CNR - ISPC), Rome and Institut für Altorientalistik (FU Berlin), Berlin francesco.difilippo@cnr.it; federico.manuelli@cnr.it Copyright: © 2021 Francesco Di Filippo, Federico Manuelli. This is an open access, peer-reviewed article published by Firenze University Press (http:// www.fupress.com/asiana) and distrib- uted under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, pro- vided the original author and source are credited. Abstract. Lying between the central Anatolian plateau and the Euphrates region, the Elbistan plain represents an ideal environment for inspecting forms of cultural interconnection. During the Iron Age, this territory was marked by the presence of notable inscribed monuments, the study of which allowed scholars to establish relationships with the most significant Neo-Hittite dynasties. This region is also characterized by the presence of sets of anepigraphic portal lions, positioned seem- ingly at random in the open landscape and with no apparent relationship with coeval archaeological remains, which have never been concretely integrated into the historical picture. In this contribution, the iconographic and stylistic analysis of these sculptures will allow us to situate them in their chronological and historical framework. A computational spatial model is further used to evaluate the meaning of their positioning as markers of a visual networking system that may have repre- sented the most significant thoroughfares to and from the Elbistan plain. Data Availability Statement: All rel- evant data are within the paper and its Supporting Information files. Competing Interests: The Author(s) declare(s) no conflict of interest. Competing Interests: The Author(s) declare(s) no conflict of interest. Keywords. Elbistan plain, Iron Age, Euphrates region, Free-standing lions, Syro- Anatolian art, Settlement pattern, Semi-automated landform classifica- tion, Least Cost Paths. Firenze University Press www.fupress.com/asiana Firenze University Press www.fupress.com/asiana 1. INTRODUCTION: BACKGROUND AND GOALS Since the mid-19th century AD, German and British explorers travelling across the Taurus regions reported the presence of two free-standing stone lions located approximately 15 km south of Darende, in the middle of the soft foothills at the northern border of the Elbistan plain in central-eastern Anatolia.1 Because of their presence the spot has always been referred to 44 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli by local inhabitants as ῾Arslantaş᾽ (lion stone) (Fig. 1). Not much attention has been typically dedicated to these sculptures and the lions have generally been mentioned as indicators of the presence of a gateway, drawing on com- parisons with similar stone guardians known from the Assyrian palaces (Sterret 1888: 299). Only later was the attention of scholars drawn to the fact that these standing in situ sculptures seem not to be associated with any building or even any proper site (Hawkins 2000: 329), representing an isolated piece of archaeological evidence in the open countryside (Özgüç, Özgüç 1949: 63-64). Approximately 35 km southeast of Arslantaş, at the eastern border of the Elbistan plain, a second pair of standing lions has been found in the proximity of the village of Sevdiliköy. They were moved in 1961 to the local museum of Kahramanmaraş (Eralp 1995). Nowadays, only one of the two sculptures is exhibited, while the second is kept in the museum storehouse. Their original location was on top of a rock outcrop which probably also repre- sents the quarry from which the lions had been realized (Orthmann 1971: 533). Indeed, according to the locals, the sculptures had been found here lying on their sides. Once again at the beginning of the 1960s, a single lion sculpture was discovered in the village of Hunu/ Arıtaş, at the western edge of the Elbistan plain (Kökten 1960: 43; Dumankaya, Topaloğlu 2017: 291). Despite the fact that the village was built on top of a mound (the so-called Arıtaş Höyük), it is still debatable whether the sculpture was originally located on the site. The fact that the lion was left unfinished and/or reused allows us to assume that it was only later transported and employed at the site. In any case, the sculpture was moved to the Kahramanmaraş museum, where it is still currently exhibited. hi The standing lions of the Elbistan plain have never been the focus of any specific analysis. 1. INTRODUCTION: BACKGROUND AND GOALS Travelers and schol- ars have long argued over whether the lions from Arslantaş could have been given one or more inscriptions (Ram- say, Hogarth 1893: 92-96; Charles 1911: 31-35; Meriggi 1975: 316; Hawkins 2000: 329). Nowadays, there are still Fig. 1: The lions of Arlantaş in 1881 or 1882. Credit: John Henry Haynes archive, courtesy of Special Collections, Fine Arts Library, Harvard University. n 1881 or 1882. Credit: John Henry Haynes archive, courtesy of Special Collections, Fine Arts Library, Fig. 1: The lions of Arlantaş in 1881 or 1882. Credit: John Henry Haynes archive, courtesy of Special Collections, Fine Arts Library, Harvard University. 45 Hic sunt leones doubts about the location and even the existence of any potential inscription(s) and the continuous exposure of the lions to the austere winter conditions of the Anatolian plateau will certainly not improve the situation. Very few attempts have also been made to set the three groups of sculptures within their chronological and cultural context through iconographic and stylistic comparisons (Özgüç, Özgüç 1949: 64; Orthmann 1971: 118; Eralp 1995: 118-119). Moreover, the sculptures have hardly ever been considered as a coherent whole or analyzed in order to understand their possible meaning and their relationships and positioning within the surrounding terri- tory (Harmanşah 2011: 77, Fig. 3).i y ş g However, their main characteristics stand out at first glance. First of all, they are unique in the whole context of Syro-Anatolian art, considering that they are the only lions carved on all their surfaces and arranged to be seen for a full four-side view.2 Second, their location is anything but random, since they are all positioned in strategic areas marking the existence of possible passages and accesses from and to their territory as well as a special rela- tionship with the surrounding landscape. g When plotted on a map along with the rest of the archaeological evidence of the Elbistan region, the portal lions of Arslantaş, Sevdiliköy, and Hunu immediately evoke an impressive significance. Indeed, they all border the outer fringes of the inhabited plain, distanced from any other settlement of the region (Fig. 2).hh The following pages present the lions from the Elbistan plain. Their main iconographic and stylistic charac- teristics will be described and situated within the corpus of the Syro-Anatolian art. 2 The term ῾Syro-Anatolian᾽ is used here to identify the region that geographically includes south-eastern Turkey and north-eastern Syria. For the use of different terminologies in accordance with historical, geographical or ethnic issues see Gilibert (2011: 1-6). 3 We follow here the numeration proposed by Hawkins (2000: 329) who identified lion A with the one on the left entering the poten- tial ῾gate᾽ and lion B on the right. The sculptures are made of limestone and are placed at approximately four meters distance to each other; lion A is 2.05 m in height and 2.55 m in length, lion B is 1.95 m in height and 2.55 m in length (Özgüç, Özgüç 1949: 64). 1. INTRODUCTION: BACKGROUND AND GOALS Their positioning will be then evaluated within the surrounding territory through computational spatial models and their symbolic, political, and historical meanings will be further discussed. Fig. 2: Map of the study area with cited toponyms. Base map by Stamen Design, CC-BY-3.0. Fig. 2: Map of the study area with cited toponyms. Base map by Stamen Design, CC-BY-3.0. 2 The term ῾Syro-Anatolian᾽ is used here to identify the region that geographically includes south-eastern Turkey and north-eastern Syria. For the use of different terminologies in accordance with historical, geographical or ethnic issues see Gilibert (2011: 1-6). 2 The term ῾Syro-Anatolian᾽ is used here to identify the region that geographically includes south-eastern Turkey and north-eastern Syria. For the use of different terminologies in accordance with historical, geographical or ethnic issues see Gilibert (2011: 1-6). 46 Francesco Di Filippo, Federico Manuelli 2. ICONOGRAPHY AND STYLE 2.1. Description The most renowned case of this fascinating group of sculptures is represented by the couple of still in situ por- tal lions from Arslantaş (Fig. 1).3 The lions are free-standing and only the stone blocks underneath the bodies and between the legs were not sculpted away. This resulted in an almost fully three-dimensional shape where the body details are carved for a four-side view and the shoulders and hind legs are seen sideways (Fig. 3). This provides the sculptures with a peculiar natural stance, as is also stressed by their upper outline that steeply follows the curve of the animals’ spines. The lions are relatively slender with slightly rounded edges and smoothed surfaces. Shoulders and hind legs are separated from the torso by well-defined and soft curves. The limbs are elon- gated and slim, but they look rather stiff and immobile, despite the front legs being slightly advanced. The heads are three-dimensionally figured all around and conceived for a multi-side view. However the unnatural position of the heads should be noted – completely retracted and embedded into the shoulders – as well as their cubic, unusu- ally long and wide proportions. The rendering of the animals’ details is unfortunately not always easily comprehensible. This is especially evi- dent on lion B, where the details of the muzzle are almost completely washed away. The lions’ mouths are open wide with a hanging tongue just slightly visible on lion A (Fig. 4). The lips are round and large while the fangs, probably originally four sharp pointed shapes, are now only discernible by means of two conical bulges. The upper parts of the muzzles are deeply damaged, so as not to allow for a precise reconstruction of their details. This gener- ates a bizarre perception of the mouths, as though they were wider than the originals. The outline of a broad nose Fig. 3: The lions of Arlantaş: lion A (top) and lion B (bottom). Drawings by R. Zaher based on: Özgüç, Özgüç 1949 and Bilgin 2021. Fig. 3: The lions of Arlantaş: lion A (top) and lion B (bottom). Drawings by R. Zaher based on: Özgüç, Özgüç 1949 and Bilgin 2021. Fig. 3: The lions of Arlantaş: lion A (top) and lion B (bottom). Drawings by R. Zaher based on: Özgüç, Özgüç 1949 and Bilgin 2021 47 Hic sunt leones is in any case visible especially on lion A. 2.1. Description The eyes of the animals are also wide with the upper orbital parts round- ed and notably protruding. Despite the absence of a neck due to their posture, the muscles that surround the animals’ heads are rendered by a soft band that is still partially visible only on the external side of lion A that merges with the above-mentioned protruding eye socket. The cheek musculature is also very prominent and is especially emphasized on the external side of lion A. The ear shape is in contrast almost totally indiscernible. They give the impression of being large, retreated, and triangular (Fig. 5). i g ( g ) Despite the fact that they are characterized by rounded and well-defined shoulders, the forelegs are out of pro- portion: extremely wide on the upper part and thinner in the lower one. Moreover, the joint between the two seg- ments is characterized by a very unnatural angle that recalls a protruding spur, as is especially noticeable on the back of the external foreleg of lion A. On the same lion, the upper sides of both forepaws are also visible, and are characterized by four elongated and stylized claws (Fig. 6). Unfortunately, their ends on both the front and the lat- eral sides cannot be reconstructed because they are hidden by the soil of the field. y yi A thick mane entirely covers the external sides of the upper parts of both lions, approximately down to the belly line and reaching almost the limit of the hind leg (Fig. 5). On lion A, the mane recalls a pattern characterized by irregular lozenges or leaf-shapes that overlap both the fore and the back shoulders. This is less readable on lion B, where the limestone encrustations and a series of irregular grooves on the surface have been weathered, giving the impression of the presence of some rude circular patterns. This might also be a consequence of the long period spent by the lion lying on this side and touching the earth. h The bodies are slim, and the abdomens are rendered with an arch shape that is extremely stretched and thin on lion A but heavier and more solid on lion B. As far as the hind legs are concerned, they are also well Fig. 4: Arlantaş in 2011, the lions’ head: lion A (left) and lion B (right). Credit: Bilgin 2021. Fig. 4 The exhibited lion is made of andesite and is 2.04 m in height (Eralp 1995: 115). 2.1. Description 4: Arlantaş in 2011, the lions’ head: lion A (left) and lion B (right). Credit: Bilgin 2021. 48 Francesco Di Filippo, Federico Manuelli Fig. 5: Arlantaş in 2014, the lions’ external foreside: lion A (left) and lion B (right). Credit: Wikimedia Commons, https://com- mons.wikimedia.org/wiki/Category:Aslanta%C5%9F_(Darende), CC-BY-SA-3.0. Fig. 5: Arlantaş in 2014, the lions’ external foreside: lion A (left) and lion B (right). Credit: Wikimedia Commons, https://com- mons.wikimedia.org/wiki/Category:Aslanta%C5%9F_(Darende), CC-BY-SA-3.0. Fig. 6: Arlantaş in 1947, lion A forepaws. Adapted from: Özgüç, Özgüç 1949: 16, res. 18-19. Fig. 6: Arlantaş in 1947, lion A forepaws. Adapted from: Özgüç, Özgüç 1949: 16, res. 18-19. and softly modelled, as is especially evident in the rendering of the gentle and rounded curves of the shoulders. However, as for the forelegs, little attention has been paid to their proportions, namely that the lower parts are unnaturally thinner than the upper ones. Only the left paw of lion A is visible. Once again, this is done by ren- dering the four claws with thin and stylized traits, while the paw side is once again not visible. A thick tail is observable on the upper parts of the backs of both figures, disappearing between the legs and turning sideways approximately at the point where the limbs became thinner. Indeed, the faint trace of a carved tail is visible on the external side of lion A, where it sharply runs diagonally downwards with a final thickened upturned curl, maybe symbolizing the hair tuft. and softly modelled, as is especially evident in the rendering of the gentle and rounded curves of the shoulders. However, as for the forelegs, little attention has been paid to their proportions, namely that the lower parts are unnaturally thinner than the upper ones. Only the left paw of lion A is visible. Once again, this is done by ren- dering the four claws with thin and stylized traits, while the paw side is once again not visible. A thick tail is observable on the upper parts of the backs of both figures, disappearing between the legs and turning sideways approximately at the point where the limbs became thinner. Indeed, the faint trace of a carved tail is visible on the external side of lion A, where it sharply runs diagonally downwards with a final thickened upturned curl, maybe symbolizing the hair tuft. 49 Hic sunt leones Hic sunt leones The main traits that characterize the lions from Arslantaş also recur on the Sevdiliköy one.4 The above-men- tioned imbalance of rounded natural surfaces and out-of-proportion anatomical details represents the basic char- acteristic of this sculpture as well (Fig. 7). Moreover, the lion is thought to be a free-standing monument with its almost entirely three-dimensional shape made for a four-side view (Fig. 8). Considering its general iconographic and stylistic aspects as well as dimensions, the lion from Sevdiliköy is essentially quite identical to those from Arslantaş. The figure is slender but characterized by an unnatural triangular and static shape of the body. The head is retracted into the shoulders and shows cubic and unbalanced proportions. The shoulder muscles are once again well-defined but still characterized by rigidity and flatness. il Most of the details of the heads of the Arslantaş lions are also visible here, such as the big and protruding eyes, the wide and deep open mouth with faintly visible hanging tongue, the emphasized cheek muscles, the squared and broad nose, as well as the band that surrounds the head. In any case, it should be noted that the sculpture is better preserved than those from Arslantaş, allowing for a better analysis of some anatomic details (Fig. 9). Fig. 7: The lion of Sevdiliköy in 2011. Credit: Bilgin 2021. Fig. 7: The lion of Sevdiliköy in 2011. Credit: Bilgin 2021. Fig. 7: The lion of Sevdiliköy in 2011. Credit: Bilgin 2021. Fig. 7: The lion of Sevdiliköy in 2011. Credit: Bilgin 2021. 50 Francesco Di Filippo, Federico Manuelli Fig. 8: The lion of Sevdiliköy. Drawings by R. Zahler based on: Eralp 1995 and Bilgin 2021. Fig. 8: The lion of Sevdiliköy. Drawings by R. Zahler based on: Eralp 1995 and Bilgin 2021. Fig. 8: The lion of Sevdiliköy. Drawings by R. Zahler based on: Eralp 1995 and Bilgin 2021. Fig. 8: The lion of Sevdiliköy. Drawings by R. Zahler based on: Eralp 1995 and Bilgin 2021. Fig. 9: The lion of Sevdiliköy in the 1990s, the foreside. Adapted from: Eralp 1995: Lev. E6, F1, courtesy of Gül Eralp Kania. Fig. 9: The lion of Sevdiliköy in the 1990s, the foreside. Adapted from: Eralp 1995: Lev. E6, F1, courtesy of Gül Eralp Kania. The muscles around the mouth are softly defined by a thin curved surface, while the muzzle is high, squared and pronounced. Hic sunt leones A thick tail starts from the back of the animal, disappears between its legs and appears then again on its right side running horizontally and finally turning up with a curl (Fig. 10). The ears rise up from the band that surrounds the head in the shape of two large patches. The eyes are oval-shaped and well-defined. Uncommonly, the mouth does not show any trace of fangs, but a series of fractures at its far ends let us assume that they should have been there originally. In contrast to what can be observed at Arslantaş, no trace of mane or fur is visible on the lion surface. Moreover, the sculpture entirely stands on a substructure that slightly protrudes outside the limits of the animal shape. The forelegs are rounded and straight with the right limb slightly advanced, providing a certain idea of move- ment. In any case the left shoulder is definitively out of proportion, occupying a great part of the body space and being itself as long as the rest of the leg. Moreover, the joint between the upper and the lower parts of the limb is 51 Hic sunt leones Fig. 10: The lion of Sevdiliköy, the lion’ right side in 2014 (left) and back side in the 1990s (right). Left, credit: Wikimedia Com- mons, https://commons.wikimedia.org/wiki/File:Kahramanmaras_Museum_L%C3%B6we_Sevdilli.jpg, CC-BY-SA-3.0. Right, adapted from: Eralp 1995: Lev E5, F3, courtesy of Gül Eralp Kania. Fig. 10: The lion of Sevdiliköy, the lion’ right side in 2014 (left) and back side in the 1990s (right). Left, credit: Wikimedia Com- mons, https://commons.wikimedia.org/wiki/File:Kahramanmaras_Museum_L%C3%B6we_Sevdilli.jpg, CC-BY-SA-3.0. Right, adapted from: Eralp 1995: Lev E5, F3, courtesy of Gül Eralp Kania. once again very unnatural and sharp. Although barely detectable, the paws of the forelegs are rendered by four stylized claws visible on their upper side only. Even more unbalanced are the hind legs. First of all, the right back shoulder is longer compared to the left one, capturing almost all the space designated to the entire leg. As a result, the lower right hind leg is completely projected forwards, giving the impression that the animal is crouched on this side. The lower left leg is better proportioned, but its paw is extremely elongated, probably in an attempt to com- pensate for the position of the right limb. At the extremity of the left paw the faint traces of four elongated claws are again visible. 5 The lion is made of basalt and is 1.30 m in height and 1.60 in length (Kökten 1960: 43). Hic sunt leones Interestingly, the points where the upper and the lower hind legs join are, in both cases, naturalis- tically rendered by means of a small and round protrusion. hf The lion from Hunu shows instead an important set of differences compared to Arslantaş and Sevdiliköy (Fig. 11).5 This is especially evident in its smaller size, as well as its rendering and design (Fig. 12). In-depth observations are difficult because the lion almost doubtlessly is both reworked and unfinished. Three different carving stages are indeed visible on the sculpture. The right backside is only roughly hewn and the front and front-right is carved with finishing details, while the entire left side is smoothed and outlined (Fig. 13). it Despite the fact that these activities are easily recognizable on the stone, the establishment of their temporal order is difficult. However, the occurrence of the different carving steps on the sculpture can hardly be explained other than as an unfinished carving process of reusing the stone block. It gives the impression that its front fin- ished side represents either the earliest or the latest carving activity. The hammering and smoothing traces on the two sides mark instead an even later reuse that was probably never finished. With this is mind it is clear that estab- lishing either its original or final shape is virtually impossible. However, it is also very interesting to note that there are more than a few correspondences with Arslantaş and Sevdiliköy in some of the iconographic and stylistic details still visible on the sculpture. First of all, the upper curved outline of the stone block reproduces once again the spine and the backside of the animal. It allows the 52 52 Francesco Di Filippo, Federico Manuelli Fig. 11: The lion of Hunu in 2015. Credit: Bilgin 2021. Fig. 11: The lion of Hunu in 2015. Credit: Bilgin 2021. Fig. 12: The lion of Hunu. Drawings by R. Zahler based on: Kökten 1960 and Bilgin 2021. Fig. 12: The lion of Hunu. Drawings by R. Zahler based on: Kökten 1960 and Bilgin 2021. Fig. 12: The lion of Hunu. Drawings by R. Zahler based on: Kökten 1960 and Bilgin 2021. 53 Hic sunt leones Fig. 13: The lion of Hunu in 2015, the lion’ left side. Credit: Bilgin 2021. Fig. 13: The lion of Hunu in 2015, the lion’ left side. Credit: Bilgin 2021. Hic sunt leones assumption that this lion as well was not thought to be a structural element but rather a free-standing sculpture. Moreover, the general posture of the lion, with its retracted and unnatural position of the head as well as its round- ed, plastically sculpted but at the same time cubic and stiff character, are very close to the other lions. Despite the fact that the front face is almost completely lost, probably erased together with the left side of the lion, interesting observations can be drawn from its right side. The head is surrounded by a curved band on which the traces of the depressions left by a small triangular retreated ear are still visible. The eye is only slightly perceptible. It was certainly wide, with its upper orbital part protruding. The mouth was also wide and open, as well as characterized by a soft line that defines its lateral muscles. Because of all these characteristics, the head of the Hunu lion can be considered very close to those of Arslantaş and Sevdiliköy. Nonetheless, it cannot be ignored that the design of the limbs shows significant differences. First of all, their shapes are mostly carved rather than rendered in relief. The lower torso is only visible by means of a faint incised line that shapes its outline. The shoulder is defined by an unnatural spiral form, an attempt to balance anatomic details and decorative patterns. The external side of the right foreleg has a triangular shape with the muscles stressed by further incisions. The same pattern occurs on the front side of the legs. Despite the fact that their outline is softly and naturalistically defined, with incised lines that emphasize the upper shoulder curvatures, the overelaboration of the muscle details produces once again an extravagant result. The view of the lower legs is even more awkward. Indeed, this part protrudes, creating a sort of unnatural break in the front limbs and giving the idea that the animal is crouched. Moreover, this protuberance was completely designed to reproduce the animal paws, with the result that these are as long as the upper legs. Once again, the paws are defined by four elongated and slim claws. 6 If not otherwise specified the nomenclature of the following comparisons is always based on the numeration provided by Orthmann (1971). 2.2. Comparisons In the wider framework of the Syro-Anatolian figurative art of the late 2nd-early 1st millennium BC, the lions from Elbistan show on the one hand their adherence to specific artistic canons and, on the other hand, the intro- duction of original aspects.6 First of all, it should be noted that the lions from Arslantaş, Sevdiliköy, and Hunu are unique in their arrange- ment (Fig. 1). As a matter of fact, lion sculptures were usually integrated into specific structures with the intent of 6 If not otherwise specified the nomenclature of the following comparisons is always based on the numeration provided by Orthmann (1971). d the nomenclature of the following comparisons is always based on the numeration provided by Orthmann 54 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli guarding the gates of cities, palaces, temples or citadels (Mazzoni 2013: 470-471; Pucci 2015: 59-70). These ῾prop- er᾽ portal lions were indeed always sculpted on one side only, while their structural parts were left uncarved. The forelegs and the heads were protruding and three-dimensionally rendered, but the rest of the figures were intended for a lateral view with a bidimensional approach that implied the visibility of the hindlegs on the carved sides only. The sculptures from Arslantaş, Sevdiliköy, and Hunu were instead completely free-standing and three-dimension- ally shaped for a four-side view. They were not meant to be integrated into a structure, rather they were ῾virtual᾽ portal lions, themselves representing a structure. A couple of similar cases can be taken into consideration. The lion from Çolaklıköy, found out of its context in the namesake village, is probably the best comparable example (Tunca 1976). It was sculpted on its four sides exactly in the same manner as the Elbistan lions and its dimensions are comparable with Hunu. Nonetheless, the right-side section of the back was flattened with a rectangular hole and its rear end was squared off vertically. Hence, it should have been used with some structural purpose. Actually, this lion seems to be structurally akin to the figures adorning the entrance of the Kapara Temple-Palace at Tell Halaf (Moortgat 1955: 110-114). Indeed, the two lions of the four-bay portico were also sculpted on their four sides, serv- ing architecturally as statue-stands of the hilani façade (Gilibert 2014: 40-44). Another all-round lion is Hama C/1. It is fully three-dimensionally shaped even in the part underneath the body and between the legs. 2.2. Comparisons Unfortu- nately, the lion is the result of a very invasive restoration and not a few doubts about its original design have been raised (Orthmann 1971: 102-103; Riis, Buhl 1990: 50-52). The wide set of unfinished lions coming from quarries and workshops also show, at a first glance, similarities with the sculptures from Arslantaş and Sevdiliköy. These are especially evident for some of the standing lions from Yesemek (Duru 2012: 68-71, lev. 8-11), as well as examples from Sıkızlar, Zilfe and Demirciler (Mazzoni 1986; Mazzoni 2011: 141-143; Carter 1996: 292-293, 304-305; Konyar 2009: 178, 185-186). Actually, affinities are mostly related to their stiff posture and cubic proportions, but exact comparisons are hard to make and mislead- ing considering their incomplete nature. In fact, a more careful analysis shows that these sculptures were always thought to be ῾proper᾽ portal lions, since they were outlined on one side only and, when visible, their hindlegs were both carved on the same side. As mentioned, the posture of the forelegs of the lion from Hunu is very unnatural and atypical. It recalls the small, crouched lion displayed in the Gaziantep Museum, which interestingly seems to originally come from the Elbistan region as well (Balcıoğlu 2009). Actually, squat lions are very common in the Syro-Anatolian art, such as that visible on relief E/1 of the Herald’s Wall and H/2 of the King’s Gate at Karkemiš (Hogarth 1914: pl. B10a; Woolley 1952: pl. B55a). Moreover, crouched lions are frequently depicted as deity stands, as at Darende and with the stele B/4 from Malatya (Hawkins 2000: 304-305, 328, pl. 145-146, 164). h When we turn to their iconographic and stylistic details, further peculiarities emerge. The head of the lions, completely retreated into the shoulder and without any trace of neck, provides these figures with a unique pos- ture. The upper curved outline of the body of the lions from Arslantaş and Sevdiliköy is also very uncommon. It recalls the shape of the portal lion Malatya A/2 (Dalaporte 1940: pl. XVI-XVII; Orthmann 1971: 97-98). In contrast, the back rump-shape of the lion from Hunu is more canonical and comparable with the specimens C/1-2 and C/4-5 from the Lion’s Pit at Zincirli (Luschan 1902: Taf. 46-47), as well as with the portal lion Malat- ya A/1 (Dalaporte 1940: pl. XVIII-XXI) and the lion base Karkemiš H/11 (Woolley 1921: Pl. B21; Orthmann 1971: 41-42). 2.2. Comparisons The horizontal movement of the tail of the lions from Arslantaş and Sevdiliköy is also without comparisons. Indeed, the typical Iron Age tails curl up between the legs of the animals but always moving from up to down (Akurgal 1949: 68-71). The closest examples are attested at Ain Dara, on the portal lions A/1-2 (Orthmann 1971: 58, Taf. 1a) and on the reliefs of the outer façade of the temple terrace (Abū Assāf 1990: pl. 19a, 22a). Here, the lions’ tails run horizontally under the bodies of the animals but always curl down. h The mane covering the entire body of the lions from Arslantaş is also unique. Indeed, when attested, the fur covers the frontal parts around the heads of the animals only, as a proper mane, while in a few other cases it extends over the bellies of the lions (Akurgal 1949: 70-73). Moreover, at Arslantaş the pattern is also atypical. Some affinities can be seen with the leaf-shape mane of the sculptures of the Temple-Palace at Tell Halaf (Moort- 55 Hic sunt leones gat 1955: 110-114, Taf. 120-121, 128) and with the overlapping irregular lozenge-shape mane of the antithetical lions on the pedestal wall relief of the cella of the Storm-God Temple at Aleppo (Gonnella et al. 2005: 108-109, Abb. 152; Kohlmeyer 2013: 522). y ) More affinities with the repertoire of Syro-Anatolian art can be found when we turn to the anatomic details of the lions. The head shape of the lions from Arslantaş, with their open wide mouths and protruding upper orbit- al eye parts, finds comparison with the figures of the Temple-Palace at Tell Halaf (Cholidis, Martin 2010: 346- 354; Moortgat 1955: 113-114, Taf. 127-128). The solid and cubic form of the head of the lion from Sevdiliköy is instead very similar to those of the animals carved on reliefs B/11 and B/12 of the Outer Citadel Gate at Zincirli (Luschan 1902: Taf. 44). Moreover, its squared and large nose is close to those carved on the lion base Karkemiš H/11 (Woolley 1921: Pl. B21). The emphatic cheek muscles of Arslantaş and Sevdiliköy characterize many of the sculptures of the Herald’s Wall at Karkemiš (E/3, E/4 and E/9) (Hogarth 1914: pl. B11a-b, B14b; Orthmann 1971: 31-32), as well as the relief with the antithetical lions from Aleppo (Gonnella et al. 2005: 96, 101, 108-109, Abb. 133, 141, 151-152). 2.2. Comparisons The small triangular and retreated ear of Hunu is also comparable with examples from the Her- ald’s Wall at Karkemiš (E/1 and E/3) (Hogarth 1914: pl. B10a, B11a), as well as with reliefs from Aleppo (Gonnel- la et al. 2005: 108-109, Abb. 151-152). The big, raised ears integrated into the band around the head of Sevdiliköy recall instead the portal lions A/1-2 from Ain Dara (Orthmann 1971: Taf. 1a). Again, it should be considered that the way they hang laterally in a large patch-shape is without comparisons. h y y g y g p p p The band that surrounds the heads of the lions from Arslantaş and Sevdiliköy is quite characteristic of Syro- Anatolian art. It can be seen again at the Herald’s Wall at Karkemiš (E/1, E/3, E/4, E/6 and E/9) (Hogarth 1914: pl. B10a, B11a-b, B13a, B14b), as well as in the reliefs from the pedestal wall at Aleppo (Gonnella et al. 2005: 96, Abb. 133). The more protruding curve around the head of Hunu is instead more similar to those on the portal lions Malatya A/2 (Delaporte 1940: pl. XVI-XVII) and Ain Dara A/1-2 (Orthmann 1971: Taf. 1a), as well as on the lion base Karkemiš H/11(Woolley 1921: Pl. B21). Many sculptures and reliefs present rounded and well-defined fore and back shoulders. Not in a few cases the rendering of the anatomic details is out of proportion, recalling those of Arslantaş and Sevdiliköy. This is especially visible on some of the reliefs of the Herald’s Wall at Karkemiš (E/1 and E/4) (Hogarth 1914: pl. B10a, B11b) and again on the antithetical lions from Aleppo (Gonnella et al. 2005: 108-109, Abb. 152). The peculiar sharp angle that characterizes the joints of the forelegs at both Arslantaş and Sevdiliköy finds a good comparison with the unnatural posture of the lion carved on relief A/9a and the bulls reproduced on reliefs A/3 and A/4 at Malatya (Delaporte 1940: pl. XIX, XXII; Orthmann 1971: 91-92). The spiral-shape of the shoulder of the lion from Hunu is however more uncommon. As a decorative element, the spiral occurs for instance on the sphinx protome Zin- cirli K/8 (Orthmann 1971: 73, Taf. 67b), while a pattern similar to Hunu can be seen on the bulls decorating the sculpted base from Domuztepe (Çambel 1999: 94, pl. 2.3. Style and Dating Remarkably, the most reliable and comprehensive study on the development of lion iconography and style between the end of the 2nd and the beginning of the 1st millennium BC is still nowadays provided by E. Akurgal (1949: 57-75). However, more recently S. Mazzoni (2000: 1046-1048; 2013: 477) has added new cases and thoughts to the topic, still con- firming the arguments and the general development proposed by the Turkish scholar.hi With the exception of some stylistic details, the lions from Arslantaş and Sevdiliköy are almost identical. The first shows rounded edges, soft curves and more attention to detail, i.e. the mane and the paws, while the second is more squared and solid, but in any case they are undeniably very close to each other. They can certainly be the product of the same workshop or even of the same sculptor or artist and their dating should necessarily be the same. As far as the lion from Hunu is concerned, its unfinished status creates not a few problems for its evaluation. The sculpture is clearly smaller than those from Arslantaş and Sevdiliköy, but its general arrangement and above all the carving and rendering of its head is nearly the same. Despite the fact that the design of the forelegs is clearly different, the contem- poraneity of the whole group is probably the preferred assumption. Indeed, it cannot be ruled out that they all belong to the same workshop, but that the carving of the lion from Hunu followed a different path and the rendering of its finished front side was made by a different sculptor than the one responsible for Arslantaş and Sevdiliköy. if Considering that the heads of the lions from Arslantaş and Hunu are deeply damaged and that the one from Sevdiliköy is also not in perfect condition, only a few elements can be used for an appropriate dating of the sculp- tures. Despite the fact that the movement of the tails of the lions from Arslantaş and Sevdiliköy is unusual, it needs to be noted that tails curling up between the legs of the animals are typical of the pre-Assyrian Iron Age, while during the Assyrian period they instead start curling onto their rears (Akurgal 1949: 68-69). 2.3. Style and Dating The lions from the Elbistan plain have never been concretely integrated into the development of the Syro- Anatolian art. W. Orthmann (1971: 118) included them in his collection, saying that due to their posture they do not specifically belong to any group and because of their coherent characteristics they might all be attributed to the same workshop. Despite this, he tentatively assigned Sevdiliköy to his style II and Hunu to style III, with- out any specific mention of Arslantaş (Orthmann 1971: 486, 533). Before him, T. and N. Özgüç (1949: 63-64) instead proposed a dating for the lions from Arslantaş to the 11th or 12th century BC, since they merge rounded and smoothed surfaces typical of the Hittite period with more squared details of some later sculptures. Following the same idea, G. Eralp (1995: 118-119) also agreed that both Arslantaş and Sevdiliköy belong to an early phase of Late Hittite art. In contrast, S. Mazzoni (1997: 366) considered all sets of lions from Elbistan as belonging to a consistent group of free-standing sculptures dated to the 9th century BC, together with Havuzköy, Çolaklıköy, Tell Halaf, and Hama. More recently, A. Gilibert (2015: 143) found the comparison with the sculptures from Ain Dara more suitable, setting the dating of the lions to the 11th century BC. Moreover, V. Blanchard (2019: 191-193) considered the lions from Arslantaş and Sevdiliköy as evidence of the activities carried out across the Elbistan terri- tory during the 12th century BC. y g y It is undeniably challenging to situate the case of the Elbistan lions within the already complex lines of devel- opment of Syro-Anatolian art. Indeed, the identification of a general development of this form of art is difficult to trace, especially in the absence of reliable contexts and considering the differences in style occurring at contempo- rary sites (Orthmann 2002: 153-155; Manuelli 2016: 28-29). The lions from the Elbistan plain represent a prop- er stylistic group without any trace of a clear internal development. This increases the difficulties of delivering an appropriate cultural and chronological assessment. It should also be considered that the Syro-Anatolian lions have never been the focus of any detailed study, and curiously Orthmann has also not provided any specific analysis of them as individual figures. 2.2. Comparisons 122-123).h p p Ç p The pattern made with incised lines that reproduce the muscled shape of the forelegs at Hunu is not rare in the repertoire of the Syro-Anatolian art. Comparisons can be made with the lions decorating the reliefs of the terrace of the temple at Ain Dara (Abū Assāf 1990: pl. 19a, 22a), as well as with the relief H/2 from the King’s Gate at Karkemiš (Woolley 1952: pl. B55a; Orthmann 1971: 31-33). As far as the paws of the three sets of lions is concerned, they certainly share the general characteristic of end- ing with elongated claws. However, only the case from Arslantaş offers the possibility of drawing specific compari- sons. Interestingly, T. and N. Özgüç (1949: 63-64, Abb. 18-19), stated that when they visited the site the forepaws of lion A were visible. They describe the four claws on each paw as flat on their top, curved on the front and with some linear decorations, suggesting a very suitable comparison with the early sculptures of the Lions’ Pit at Zincirli (C/1-3) (Luschan 1902: Taf. 46-47). A further association can be made with the earlier group of reliefs from Hama (Riis, Buhl 1990: 40-42, fig. 17). The back paw still visible on the internal side of lion A shows an even more stretched and slim shape of the paws that seems to be comparable with those of the lions of the Temple-Palace at Tell Halaf (Cholidis, Martin 2010: 346-354; Moortgat 1955: 110-114, Taf. 121-122, 129). 56 Francesco Di Filippo, Federico Manuelli 2.3. Style and Dating Concerning the distinctive mane of Arslantaş, it must be stressed that the stylized spade or leaf pattern is more typical of the Hittite lions, while it developed into a flame-shape mostly during the Iron Age (Akurgal 1949: 59; Mazzoni 2000: 1046). The soft rendering of the cheeks as well as the emphatic muscles of the necks and the shoulders also echo the Hittite imperial period, while the stiff and static posture of the figures is more characteristic of the earliest Iron Age sculptures (Özgüç, Özgüç 1949: 64). Following the same dichotomy, the small retracted ear of the lion 57 Hic sunt leones from Hunu recalls Hittite prototypes, but the wide and rigid open mouth is instead closer to later examples (Orth- mann 1971: 118; Akurgal 1949: 70-72). The unusual rendering of the forelegs of Hunu is certainly close to Assyr- ian examples, but similar patterns are already visible for instance at Ain Dara (Mazzoni 2000: 1046). p p y ( A special mention is due to the paws of the lions from Arslantaş. Unfortunately, the rendering of their lateral part, which represents one of the most distinguishing features of the lion development, is not detectable. It might be assumed that the forepaws displayed a fifth lateral crouched claw. The back paw could have also been character- ized by the same pattern. Indeed, neither the schematic double-thickened claws with single spiral of the earliest Iron Age sculptures, nor the stylized multiple-spiral shape of the later ones are here reproduced (Mazzoni 2013: 477; Akurgal 1949: 68). g ) In general, comparisons have shown that the main set of affinities occur with the lions carved on the reliefs of the Herald’s Wall at Karkemiš, as well as with those of the cella of the Storm-God Temple at Aleppo, which can now be quite confidently set at the beginning of the 10th century BC (Mazzoni 1997: 266-267; Gilibert 2011: 116- 117). Nonetheless, further features usually associated with the continuity of the Hittite artistic traits into the earli- est Iron Age sculptures are also observable. 2.3. Style and Dating Specific comparisons have been made with the sculpted lions and the reliefs of the terrace at the Ain Dara temple, dated to the 11th century BC (Nóvak 2012: 48; Mazzoni 2013: 473; Gilibert 2015: 143), as well as with some of the sculptures reused in the Lions Gate at Malatya, which originally belong to the 12th century BC (Mazzoni 1997: 292; Manuelli 2019). Later parallels can also be seen in certain traits of the lions at the entrance of the Temple-Palace at Tell Halaf, for which a dating to the late 10th century BC can be assumed (Mazzoni 2013: 480; Gilibert 2014: 36). In conclusion, comparisons allow us to confidently situate the lions from the Elbistan plain between the 11th and the beginning of the 10th century BC, although an earlier dating to the 12th or a later one to the advanced 10th centuries BC cannot be fully excluded. 3.1. Method and Aims 14: Distribution of settlements in the Elbistan basin with the indication of the region’s planar surfaces within the interval from 0° to 6° slope. 3.1. Method and Aims As mentioned, the portal lions of Arslantaş, Sevdiliköy, and Hunu represent unique and original evidence of the cultural milieu at the beginning of the Iron Age. However, their uniqueness, as well as the absence of inscrip- tions and associated archaeological context, have also inevitably misled their interpretation and the reconstruc- tion of their historical significance. Yet when these monuments are plotted on a map alongside other archaeologi- cal evidence of the Elbistan region, their positioning is tremendously evocative (Fig. 14). They are located either at the foot slopes of the surrounding towering massifs (Sevdiliköy and Hunu) or on top of the natural depression between the foothills of the Anti-Taurus and the rough mountains dividing the Ceyhan river’s headwaters and the Tohma river basin (Arslantaş). In the light of these premises, a legitimate question arises: how can their outstanding topographic prominences be used to better understand the monuments’ function and historical significance? And consequently: how can we formally deal with the problem of conferring a specific significance to a place beyond our simple subjective percep- tion? Here we discuss how to define an explanatory formal model capable of outlining the main landscape features of the Elbistan basin and its archaeological evidence. This relies on an array of quantitative methods and tech- niques that are nowadays quite common in GIS Science (Tilley 1994; Llobera 2001; De Reu et al. 2013). In order to answer basic geo-historical questions, such as ‘where?’ and ‘why there?’ (Baker 2003: 37-44), two different yet interconnected approaches contribute to defining this model. First, the archaeological spatial pattern of the Elbistan basin will be outlined in order to evaluate if the lions’ positioning might share affinities and dis- crepancies with the distribution of settlements within the same region. Second, the function and the topographic prominence of the lions will be evaluated within Elbistan’s historical mobility network, as well as in relation to the geomorphological characters of their surrounding territories. 58 Francesco Di Filippo, Federico Manuelli Fig. 14: Distribution of settlements in the Elbistan basin with the indication of the region’s planar surfaces within the interval from 0° to 6° slope. Fig. 14: Distribution of settlements in the Elbistan basin with the indication of the region’s planar surfaces within the interval from 0° to 6° slope. Fig. 3.2. On the Lions’ Positioning (‘Where?’) The Elbistan region is an intermontane plain located in the western part of the eastern Taurides. It repre- sents an embedded basin with a generally level floor almost entirely ringed by steep mountains reaching altitudes between 2000 and 3000 meters. The extensive ranges of the Binboğa massif enclose the western side of the plain, while the steep slopes of the Hizanlı and Nurhak mountains mark natural borders respectively to the north and to the east. The southern edge of the Elbistan plain is bordered by the Berit Dağ massif. In contrast to the rugged geo- morphological nature of the surrounding mountains, the basin floor shows a high degree of uniformity. Its lower parts lie at about 1100 meters a.s.l. and, with the exclusion of the Soluk hills in the middle of the plain, the land is generally flat or gently sloping.h This sharp distinction between plain and mountains also marks a fundamental characteristic in Elbistan’s long-lasting settlement history. Indeed, all the pre-classical mounds so far known seem to be located exclusively within the floor-plain (Çifçi, Greaves 2010). In order to formally support this assumption and to minimize subjec- tive decisions, we first focus on the problem of the quantitative definition of plain areas within the study region. Despite the fact that several geomorphometric semi-automated approaches might help in recognizing such a land- 59 Hic sunt leones form (e.g., r.geomorphon or Topographic Position Index), we still preferred to follow a simpler method based on empirical observations. Indeed, since the slope gradient of the terrain can be assumed as the basic criterion to out- line relatively flat areas, it can be estimated that all terrains within the range between 0 and 6 degrees of slope can be safely considered plains. At first glance, this threshold value could appear subjective. It actually relies on the synthesis of different criteria that have been inferred from the survey of physical terrain conditions and comparable factors resulting from secondary literature (Veselský et al. 2015: 802-803; Di Filippo, Mori 2018: 51-53). This allowed us to assemble a set of continuous planar surfaces derived from an SRTM DEM at 30 meters res- olution through the multi-scale analysis of slope gradient geomorphometric variables.7 The results, displayed in Fig. 14, confirm that all the identified settlements do indeed lay inside the Elbistan basin flat area. 7 GRASS GIS r.param.scale module, morphometric parameter = slope (Veselský et al. 2015: 802-803). 3.2. On the Lions’ Positioning (‘Where?’) Moreover, 24 of the 28 sites are located at no more than 2000 meters distance from the plain border. This limit is represented by the proper landform perimeter or by further reliefs inside the plain itself, i.e. the Soluk hills at the center of the basin.h This picture is truly revealing when we focus on the geology of the Elbistan basin, which allows us to out- line an even more solid relationship between landscape and the distribution of sites. Almost all the settlements are located in geological areas characterized by alluvial plain sediments originating from recent stream beds (Fig. 15). More precisely, these settlements spread out across the geological interface between recent stream beds (Holocene fluvial sediments) and more ancient sequences of Pleistocene sediments composed of conglomerate, sandstone, silt- stone, and mudstones with abundant calcretes, which generally lay at a higher elevation (Yusufoğlu 2013). Fig. 15: Relationship between the sites’ distribution and geology in the study area. Adapted from: Yusufoğlu 2013: 311, courtesy of Halil Yusufoğlu. Fig. 15: Relationship between the sites’ distribution and geology in the study area. Adapted from: Yusufoğlu 2013: 311, courtesy of Halil Yusufoğlu. Fig. 15: Relationship between the sites’ distribution and geology in the study area. Adapted from: Yusufoğlu 2013: 311, courtesy of Halil Yusufoğlu. Fig. 15: Relationship between the sites’ distribution and geology in the study area. Adapted from: Yusufoğlu 2013: 311, courtesy of Halil Yusufoğlu. 60 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli This preference in the choice of settlement location can be safely considered the main feature of the sites’ dis- tribution in the Elbistan plain. The majority of the settlements (17) are located less than 300 meters apart from the geological interface formed by recent fluvial sediments cut into the more ancient Pleistocene sequences. In other words, most of the archaeological sites, and more significantly those with the longest occupation sequences, are located on natural elevations such as rocky spurs, platforms, or terraces, overlooking the underlying watercours- es (e.g., Merikli, Tedevin, Til Afşin, and Yassıhöyük; cf. Brown 1967). As a consequence, there is also a notice- able relationship between the mounds’ distribution and the network of watercourses (Konyar 2011; Çifçi, Greaves 2010). If one considers exclusively the permanent streams, i.e. those with significant and steady flows, 18 settle- ments are located at no more than 750 meters from them. 3.2. On the Lions’ Positioning (‘Where?’) Yet if the overall hydrologic system is taken into consid- eration, also including the minor streams, seasonal tributaries, and possibly springs, all the locations are character- ized by a considerable availability of water resources. Summing these data up, the archaeological settlements in the Elbistan basin follow a clear and coherent distri- bution pattern. The sites, dated from the Chalcolithic to the Iron Age, seem to be deeply influenced by the natural environment. They are located exclusively in the basin’s relatively flat area, within the ecological interface character- ized by gradients from 0 to 6 slope degrees. Moreover, they lie at the fringes of recent geological areas originating from fluvial accumulation of Holocene stream beds, showing a tight relationship with the network of the basin’s watercourses. In this context, the fact that the sculpted lions are located along the outer limit of Elbistan’s inhabited region marks a fundamental difference from the proper settlements, and the two sets of archaeological evidence clearly follow distinctive, seemingly opposite distribution patterns. f Furthermore, the picture also helps to identify a sharp distinction between the different sets of lions. Indeed, Sevdiliköy is located at more than 16 kilometers from the closest Holocenic formation and more than 2 kilometers apart from the nearest major watercourse. A similar trend seems to be shown at Arslantaş. Despite the fact that compared to Sevdiliköy the location is closer to the basin’s recent fluvial sediments (c. 3 kilometers), it still lies about 4800 meters from the steady stream. Hunu, conversely, shows some discrepancies from either Arslantaş or Sevdiliköy. The findspot of Hunu’s lion (i.e., the village of Arıtaş Höyük) exactly mirrors the main features of the other settlements of the region. It is located at the intersection of the geological interface between recent stream beds and more ancient sediments (Dumankaya, Topaloğlu 2017: Fig. 7), at the western limit of the relatively flat surface of the basin (c. 1293 meters a.s.l.), and in close connection with one of the major western branches of the Hurman river. In short, it possesses all the features emphasized for the ‘living’ settlements rather than those shared by Arslantaş and Sevdiliköy.h The comparison of the distribution pattern of settlements and portal lions seems to show that the lack of any contemporary remains in both Arslantaş and Sevdiliköy is not due to either the dearth of archaeological research or unknown historical circumstances. 3.2. On the Lions’ Positioning (‘Where?’) Instead, it confirms that these lions are open-air monuments, intentionally conceived to be erected in the open country, without any direct relationship with any coevally inhabited center nearby. 8 The place-name Alhazli/y has disappeared nowadays, but was mentioned until the beginning of the 20th century AD (Ramsey 1890: 273; Naval Staff 1919: route 86; Maunsell 1902; Kiepert 1904-1907; 1913). 9 Computed pathways have been balanced through travel times given by Sterret (1888: 299). For this region, a perfect match between real and generated pathways occurs when the spreading algorithm (r.walk) considers exclusively the rate of change of the original DEM (i.e., slope) and no friction parameters are provided. 3.3. On the Lions’ Function (‘Why There?’) The reconstruction of ancient mobility through computational techniques (i.e., the Cost Surface Analysis and the Least Cost Pathways) is nowadays routinely applied in the field of historical research (Llobera 2000; Surface- Evans, White 2012; Polla, Verhagen 2014). Its primary purpose concerns the development of ‘predictive models’ that may help us to understand the nature and development of ancient road systems. Even more significantly, as a recon- naissance tool, this array of techniques provides a means by which geographically-based research problems can be examined, serving as a method of hypothesis validation (Newhard, Levine, Rutherford 2008: 99-101; Bevan 2011).h Therefore, the analysis here proposed starts by examining an aspect that has already been highlighted by the reports of travelers and explorers who crossed the Elbistan plain at the turn of the 20th century AD: the impor- 61 Hic sunt leones tance of the places where the lions were erected in the framework of an interregional communication system. For instance, Arslantaş is reported to mark the southern border of the Sivas province (Hogarth, Munro 1893: 644) or, more often, is described as a landmark along the path that leads to Derende (Ramsay, Hogarth 1893: 96; Maun- sell 1902). As far as Sevdiliköy is concerned, the existence of the monument was still unknown at the time of the above-mentioned early surveys. Nonetheless, the location was already known at the end of the 19th century AD as Alhazli/y and was referred to as the fundamental hub in the network of communication to the Malatya region (Ramsey 1890: 273; Maunsell 1902; Kiepert 1913).8 Interestingly, these early descriptions agree significant- ly, depicting the locations of these open-air monuments as liminal spaces at the fringes of the inhabited plain on crossroads along the major pathways connecting Elbistan with the Tohma Su and Malatya basins. g j p y g y Taking this as a starting point, we have processed a set of computed routes through the combination of the r.walk and r.drain modules implemented in GRASS GIS. Considering the computed mobility network that joins the main Iron Age center of the plain (Karahöyük), to the coeval sites located just beyond the massifs ringing the Elbistan’s basin (Meriggi 1966; Kontani et al. 2012), the peculiar location of the lion sculptures stands out (Fig. 3.3. On the Lions’ Function (‘Why There?’) It is worth noting that these generated tracks correspond quite precisely to those described by ancient geographers and travelers. The first track follows a system of narrow valleys that connects Sevdiliköy to the course of the Tohma Su, near Kötükale, from where a road is reported to reach Malatya (Ramsey 1890: 273). A second route, apparently the easier one, crosses the mountains from Alhazli (roughly corresponding to modern Sevdiliköy) and reaches Arga (modern Akçadağ), at the western foot slopes of the Malatya plain (Ramsey 1890: 273-274; Naval Staff 1919: route 86 alternative). The last track allows a crossing of the Nurhak range via the mountain pass of Ola Kaya (Sterret 1888: 299; Ram- sey 1890: 273), or across the southern pass of Devrent Gedick (Naval Staff 1919: route 86).10 These variances join again descending towards the region of Polat (i.e., Ören Höyük, see Meriggi 1966).11 g g g y gg Moreover, the topographic relevance of Sevdiliköy in the framework of the interaction of the Elbistan plain with the eastern regions is especially visible when considering its geomorphometric variables (Fig. 17). The Sevdi- liköy crossroad is precisely located at the eastern margin of the Elbistan plain, where routes running eastward con- verge on the banks of the Söğütlü stream, some 2 kilometers south of the rock outcrop where the lion sculptures stand. This crossroad lies within an area of small plains located at the foot slopes of the Nurhak range. Just beyond it, a natural break of the Sügültü river gives access to a system of narrow U-shaped valleys that traverses the range in its entirety.12 This ‘entrance’ represents the most suitable natural passageways across the mountains. l h g As far as Arslantaş is concerned, the site is located on the wide and almost flat saddle that links the Elbi- stan plain to the Tohma Su basin. From the south, this geological formation is accessed following the course of the Sarsap stream, along a route flanked by the impressive, albeit quite enigmatic, Dikilitaş monument, ‘a rectangular monolith 5 meters high, set erect in a large block on a spur of the Karajik Dagh, visible from every part of the Elbistan Ovasi’ (von der Osten 1930: 108 and Fig. 116). 10 The Ola Kaya peak appears in the form Alaja in the map-series ‘Eastern Turkey in Asia’ (Maunsell 1902). Today it should be the Öğlekayasi Tepesi (N 38°12′57″ E 37°49′25″, https://www.geonames.org/10483853/oeglekayasi-tepesi.html). The Devret Gedick modern toponym features as Derbent Dağı (N 38°11′00″ E 37°45′00″, https://www.geonames.org/317081/derbent-dagi.html). 11 The computed routes do not stop at Polat but continue up to Oren Höyük. It is the first site with a documented Iron Age sequence after the steep descent (Meriggi 1966). 12 Two different quantitative approaches for semi-automated recognition of landform classes, r.geomorphon (Jasiewicz, Stepinski 2013) and the Topographic Position Index (Jenness 2006; De Reu et al. 2013), roughly outline the same results for the study area (Fig. 17). 3.3. On the Lions’ Function (‘Why There?’) 16).9 Taking this as a starting point, we have processed a set of computed routes through the combination of the r.walk and r.drain modules implemented in GRASS GIS. Considering the computed mobility network that joins the main Iron Age center of the plain (Karahöyük), to the coeval sites located just beyond the massifs ringing the Elbistan’s basin (Meriggi 1966; Kontani et al. 2012), the peculiar location of the lion sculptures stands out (Fig. 16).9 In a general framework, Arslantaş and Sevdiliköy are both considerably far away from any plausible inner road system of the plain, which instead seems to accord quite precisely with the web of inner watercourses. In contrast, h h d b h h l d f h l h hb In a general framework, Arslantaş and Sevdiliköy are both considerably far away from any plausible inner road system of the plain, which instead seems to accord quite precisely with the web of inner watercourses. In contrast, they are touched by the routes that lead from the plain to the neighboring regions. Fig. 16: Mobility pattern from Elbistan toward nearby basins with the indication of computed pathways and related isochrones. 62 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli Only a few general considerations can be raised about Hunu. Its location, lying at the foot slopes of the steep Binboğa range, clearly has no close relationship with the main route leading out of the Elbistan basin. Even by forcing the algorithm parameters in order to find a suitable track across the Binboğa (e.g., from Hunu to Kemer), bypassing the massif turns out to be more convenient in terms of travel time than crossing its arduous mountain tracks. g Entering now into more detail of the locations of Arslantaş and Sevdiliköy, the perceptions of the early 20th century travelers mentioned above seems to be fully confirmed by the computational analysis. h The lions from Sevdiliköy are located at the intersection of an array of computed routes leading eastward and crossing the Nurhak range in the direction of the plains of Sultan Suyu and Malatya. This area shows the char- acteristics of a proper crossroad, since it is located just before the paths which allow passage through the moun- tains. From Sevdiliköy three different routes allow a crossing of the Nurhak range. 3.3. On the Lions’ Function (‘Why There?’) Since the discovery of the Arslantaş lions, early topographic maps (Maunsell 1902; Kiepert 1904-1907) and itinerary accounts have reported the presence of this monument and its significance as a primary node of the communication routes crossing this stretch of the Anti-Taurus. The computational analysis interestingly confirms the perception of 19th century AD reports. It also supports the supposition that, at least during the past century, Arslantaş constituted a hub from which a set of individual roads departed to different locations into the Tohma Su basin, such as Gürün and Ispekçür (Fig. 18). The topographic prominence of Arslantaş is defined by completely different landscape features when compared to Sevdiliköy, showing different geo-morphometric variables. These do not allow us to identify meaningful land- form features that may have prompted the efforts to build the monument in that spot. A potential key to the inter- pretation comes instead from a further geographical description provided by an unknown explorer at the begin- 63 Hic sunt leones Fig. 17: Location and morphology of the ‘natural passageways’ through the Nurhak range. Fig. 17: Location and morphology of the ‘natural passageways’ through the Nurhak range. Fig. 18: The role of Arslantaş within the intermontane system of communication between the 19th and 20th centuries AD (approxi- mate scale 1:550,000). Adapted from Maunsell 1902 (left) and Kiepert 1913 (right). Fig. 18: The role of Arslantaş within the intermontane system of communication between the 19th and 20th centuries AD (approxi- mate scale 1:550,000). Adapted from Maunsell 1902 (left) and Kiepert 1913 (right). 64 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli ning of the 20th century AD, which says: ‘Arslan Tash, alt. 4,900 ft.; carved stone lions of Hittite origin. Here is the watershed between the Jihan and the Euphrates’ (Naval Staff 1919: route 85). On this basis, once we identify the Elbistan watershed areas (hydrologic basins) through the r.watershed analy- sis (GRASS GIS) and plot them on the map alongside the location of the Iron Age sites, the topographic relevance of the Arslantaş lions emerges (Fig. 19). They are located precisely on the major drainage divide between the Toh- ma Su and Ceyhan hydrological basins. In general geographical terms, water basins are areas that act like funnels by collecting and draining off pre- cipitation into common outlets (e.g., rivers or lakes). 4. DISCUSSION: THE ROLE AND MESSAGE OF THE LIONS 4. DISCUSSION: THE ROLE AND MESSAGE OF THE LIONS For a full understanding of the role covered by the lions of the Elbistan plain, two further factors and charac- teristics should be briefly considered. First, they represent a gate, if only symbolically, entailing the existence of a threshold or passage. Second, the iconographic choice of lion images is meaningful. g g g g The symbolic role of gates in the public architecture of the Ancient Near Eastern world has been stressed by many scholars, mostly by analysing textual and iconographic sources (May 2014). In the Syro-Anatolian region of the late 2nd and early 1st millennium BC, city-gates, temple-gates or palace-gates were the main public spaces inside the settlements where ceremonies and rites were performed and the royal power displayed (Wilhelm 2011: 103- 105; Miller 2012; Manuelli, Mori 2016: 211, 227). This is further underlined when we consider the message con- veyed by the image of lions. The link between lion iconography and ideological and religious aspects is well known since the end of the 4th millennium BC in Mesopotamia (Peyronel 2019). This is even more evident during the Hittite period, when the lion became the embodiment of the physical, military and political power of the kingship, reinforcing the relationship between the deities and the king (Collins 1998). In this context, the monumental guardians placed at the gates of the Hittite capital, and later wholesale adopted by the Neo-Hittite kingdoms, stood as proper symbols of the royal power (Collins 2004: 84; Pucci 2015: 62-63). They represented the subjugation of the wild world by the royalty and the town. As boundaries between the savage nature and the cities, their passage entailed the act of coming under the care of the gods and the control of the rulers (Mazzoni 1997: 294-295). The fact that these spaces were guarded by lions implied that in passing these figures one switched from being threatened by them to being protected by the self-same guardians, having moved from outside, i.e. the place of confrontation and threat, to the inside, i.e. the space the lions dominated and protected (Strawn 2001: 315-316). 13 Actually, while kapı is the common Turkish word for ‘gate’, the origin of kürki should probably be traced back to the Armenian term kürk/gürk which means ‘statue’ or ‘idol’ (Scheinhardt 1979). It leads to the equally fascinating meaning ‘the gate of the statue/idol’. 3.3. On the Lions’ Function (‘Why There?’) This implies that Arslantaş is characterized as an area that, in spite of the weather conditions, is always virtually free from significant flooding hazards. Looking at the portal lions in the context of the mobility network to and from the Elbistan plain, it is reasonable to expect that they pos- sess such a characteristic. Moreover, since adjacent watersheds are separated from each other by some physical features at higher eleva- tions (i.e., ridges), this computational process allows us to highlight a piece of further significant landscape evi- dence. In terms of landforms, the wide saddle where Arslantaş is located is defined by such a narrow slope gradient that even the r.param.scale algorithm could not fully distinguish it from the proper plain of the basin (Fig. 14). Yet as the watershed analysis clearly indicates, Arslantaş is precisely on top of an important drainage divide, a physical feature allegedly represented by a nearly imperceptible ridge, just slightly higher than the surrounding landscape. This distinctive environmental trait must have profoundly impacted this place’s significance, conferring it a power that persisted across the millennia relating to the use of surface water by local communities. As in the case of other morphologically and geologically distinct localities, such as rivers, mountain peaks, sinkholes, or springs, this area holds the geographical characteristics of a borderland. It is not surprising that at the end of the 19th century AD, Arslantaş was still referred to as the southern edge of the Sivas province (Hogarth, Munro 1893: 644). It is highly reasonable that at the beginning of the 1st millennium BC, this evocative place was already similarly perceived as the northern border of the Elbistan cultural landscape. Fig. 19: The Elbistan main hydrologic basins system. Fig. 19: The Elbistan main hydrologic basins system. 65 Hic sunt leones 4. DISCUSSION: THE ROLE AND MESSAGE OF THE LIONS 66 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli In a wider context, the process of displaying political and ideological power through the erection of landscape stone monuments was already fully established in Anatolia during the Hittite imperial period (Glatz 2021: 158- 174). In this context, rock sculptures and inscriptions were located in geologically evocative places and served the specific purpose of establishing colonial claims to borderland territories (Harmanşah 2017: 40-43). i During the Iron Age, a proliferation of stone monumental art is instead attested especially through the archi- tectural decoration of outer facades and gates as well as the innermost spaces of the temples. These monuments were marked by a strong continuity with the Hittite prototypes, revealing the intent of the new ruling classes to adhere to an already codified idea of political power (Mazzoni 2013: 472-473; Feldman 2014: 67-72). As a matter of fact, with the breakdown of the Hittite supremacy, the disputes over the inner frontiers ceased and the appro- priation of places of power was only partially resumed by the elites of the newly created Iron Age regional states (Harmanşah 2011: 57-61). Indeed, the free-standing lions of the Elbistan plain follow new and original trajectories compared to the ancestral heritage of the Hittite Empire. They follow a uniform line of development evidenced by the adoption of the same uninscribed figurative repertoire, which indicates a program conceived as a coherent whole. Moreover, if during the Late Bronze Age places with strong, distinct geological traits such as sinkholes, springs, mountain peaks, caves, or rock outcrops were preferred, the new program overturns this convention in favor of new locations whose significance was embedded in the physical features of the landscape itself as well as emphasized by the presence of specific networks of communication. p y p pi Considering the ‘virtual’ nature of the gates guarded by protecting figures and the fact that they were always erected at important road junctions, the interpretation of the lions of the Elbistan plain as proper boundary monu- ments, instead of simple landmarks, seems to be highly reasonable. The concept does not of course imply the adop- tion of a modern notion of border, in the sense of an imagined cartographic feature embracing a finite, abstract, and quantifiable geo-political entity (Harmanşah 2017: 38-40). 4. DISCUSSION: THE ROLE AND MESSAGE OF THE LIONS Despite the ‘virtual’ nature of the portal lions from the Elbistan plain, and the fact that they are not concretely asso- ciated with any city, citadel, palace or temple gates, the message behind their erection could not have been any different.h f The landscape analysis proposed here has also allowed us a better evaluation of this message, demonstrating that the position of the lions is not random. Their locations share all the earmarks of liminal spaces, borderlands, places of significance within the contemporary geographical perception, where the hubs of the interregional com- munication network intersected prominent features of the physical landscape. It is therefore not surprising that these passages were shaped into the form of open-air monumental gateways, thus blending the concepts of thresh- old and border alike.hh The Arlantaş lions were located in the open country, in a highly strategic position. This location corresponds to the highest spot of the plateau where a set of interregional routes converges, on the natural border formed by the drainage divide between the major hydrological basins of the Tohma Su and Ceyhan. Likewise, the lions from Sevdiliköy have been erected in an area of great topographic prominence, corresponding to the eastern margin of the Elbistan plain, on a natural borderland that allows monitoring of the entrance of the pathways to and from Malatya. Interestingly, the inhabitants of the Turko-Kurdish community of the Sevdili village refer to the out- crop where the lions have been found with the toponym Kürki Kapı, which they assume means ‘the gate of the mountain’.13 As far as the Hunu lion is concerned, the fact that the sculpture comes from a site where only Roman remains have been discovered cannot be neglected (Dumankaya, Topaloğlu 2017: 291). Considering that the lion has cer- tainly been reused, it seems highly reasonable to state that Arıtaş Höyük does not represent its primary archaeo- logical context, rather the sculpture was displaced there in antiquity. It is in any case clear that its original location could not be far away from the site of discovery, leading to the assumption that the lion was positioned, probably together with a second sculpture, on a spot which gave access to the Elbistan territory from the west in the vicinity of Tanır or Afşin. 4. DISCUSSION: THE ROLE AND MESSAGE OF THE LIONS Indeed, they were not necessarily thought to be placed along a line on a map separating two different controlled and organized areas; rather they marked impor- tant places within the cultural landscape, entailing a deep knowledge and awareness of the territory. In the light of this, an interesting parallel can be found, even on a smaller scale, with the case of the early Neo- Assyrian expansion that preceded the provincialization period of the 8th century BC. This phenomenon was boldly outlined by M. Liverani (1988), following the assumption that the Assyrian territorial control of fringe zones ini- tially did not spread systematically and uniformly (the ‘oil-stain’ paradigm), but rather was based on the control of a host of communication nodes ordered into a network system. The Assyrians’ repeated campaigns then had the pur- pose of either thickening the meshes of this pre-existing network or setting up further branches of the system at a greater distance. Despite the obvious differences, it might be speculated that the monumentalization of some signifi- cant places in the Elbistan plain may represent an analogous attempt at thickening the webs of a pre-existing road system, the borders of which were not outlined by imagined cartographic features and finite geo-political spaces.ih i With this in mind, a final obvious question arises: how and by whom were these lions erected? There can be no doubt that these sculptures were made by skilled stonemasons and sculptors. At the same time, it goes without saying that quarrying, sculpting and transporting the rough-hewed blocks, as well as performing their final carving and posi- tioning, required a considerable manpower (Seeher 2009: 120-125; Summers, Özel 2012: 515-516). Hence, it is hard not to think that this work was organized and executed under the auspices of a certain prominent political power. In conclusion, two main hypotheses can be discussed in this regard. Considering their dating, it can be sup- posed that the lions represent either the evidence of the existence of an independent local power in the 12th centu- ry BC or the consequence of the expansion of the kingdom of Malizi after its conquest of the Elbistan region from the c. 11th century BC. 5. HISTORICAL PERSPECTIVES g The stele was found during the one-year intensive investigation conducted at the Elbistan-Karahöyük set- tlement in 1947 (Özgüç, Özgüç 1949: 66-72). It was erected in a large open area and found in association with ash deposits and animal bones, leading to the interpretation that this was a public space characterized by cult activities and feastings (Harmanşah 2011: 65-68).14 Remarkably, the hieroglyphic Luwian signs POCULUM. PES.L.67 (REGIO) carved on the stele designate, without any doubt, the land of Elbistan and probably the city of Karahöyük itself, but the reading of this toponym is completely unknown (Hawkins, Weeden 2017: 289). As mentioned, the supposition that Karahöyük was the epicenter of a political entity based in the Elbistan territory is further supported by the evidence that the site stands out as the single largest documented mound in the whole region (Çifçi, Greaves 2010: 93). g Ç The presence of the stele at the site as well as its dating, context of discovery and subject support the pres- ence of a local authority in this region during the 12th century BC with its capital at Karahöyük. With this is mind, the lion sculptures positioned at the borders of the Elbistan plain can certainly represent material evidence of this political power, marking the access to and from its territory. This is also supported by the fact that, as men- tioned, the lions show a proper style of their own and are marked by some specific characteristics that have been not observed so far in any other of the renowned Syro-Anatolian sculptural cycles. y y p y Alternatively, a further although remote possibility is to associate the erection of the lions with the events that affected the nearby kingdom of Malizi/Melid (Hawkins 2000: 282-329). The latter had its capital at the site of Arslantepe and its domain extending to the Malatya plain and the surrounding western valleys, north-eastward of Elbistan (Di Filippo, Mori 2019). Two quite similar bas-reliefs brought to light at Arslantepe and both dated to the 12th century BC, i.e. MALATYA 9 and MALATYA 10, respectively show the Storm God of the city POCU- LUM and of the city Malizi receiving libations (Hawkins 2000: 320-322; Hawkins, Weeden 2017: 289; Manuel- li 2019). 5. HISTORICAL PERSPECTIVES The role of the Elbistan plain in the political scenario of the last centuries of the 2nd millennium BC is gener- ally considered, by the scholarship, strictly connected to the issue concerning the genealogical line of the ‘Great 67 Hic sunt leones King’ Ir-Tešub, named on the stele found at the site Karahöyük (Özgüç, Özgüç 1949: 69-72). Paleographically, this dates to sometimes in the 12th century BC, since it shows similarities with the late-13th century BC inscrip- tions known from the southern Anatolian plateau. This lead scholars to assume a derivation of Ir-Tešub from the ruling dynasty at Tarḫuntašša (Hawkins 2000: 287-289; Hawkins, Weeden 2016: 10-11; Hawkins, Weeden 2017: 288-289). On the other hand, many arguments have also recently been raised to support a possible association of this ruler with the genealogy of kings of Karkemiš (Giusfredi 2010: 41-43; Harmanşah 2011: 65-69; Bryce 2012: 85-87; Simon 2013: 824-826). King’ Ir-Tešub, named on the stele found at the site Karahöyük (Özgüç, Özgüç 1949: 69-72). Paleographically, this dates to sometimes in the 12th century BC, since it shows similarities with the late-13th century BC inscrip- tions known from the southern Anatolian plateau. This lead scholars to assume a derivation of Ir-Tešub from the ruling dynasty at Tarḫuntašša (Hawkins 2000: 287-289; Hawkins, Weeden 2016: 10-11; Hawkins, Weeden 2017: 288-289). On the other hand, many arguments have also recently been raised to support a possible association of this ruler with the genealogy of kings of Karkemiš (Giusfredi 2010: 41-43; Harmanşah 2011: 65-69; Bryce 2012: 85-87; Simon 2013: 824-826). ) It is noteworthy that scholars working on this topic have almost exclusively focused their attention on under- standing which cultural or political entity, i.e. Tarḫuntašša or Karkemiš, might have indirectly shown its power through this monument, entailing an influence or control over the Elbistan plain during the 12th century BC. But the Karahöyük stele also testifies to the irrefutable existence of a local authority during this period. The inscrip- tion is dedicated to the Storm God of the land POCULUM by the local ruler Armananis, called the ‘Lord of the Pithos-Men’, and it commemorates the visit into this territory by the above-mentioned ‘Great King’, further describing the condition of the land and the donation of cities at the time of this event (Hawkins 2000: 288-295). 14 Despite providing new important data about the Iron Age occupation at the site, the new round of investigations conducted at Elbistan-Karahöyük since 2015 have not yet supplied specific information related to the context of discovery of the stele (Uysal, Çifçi 2019: 411-412). 5. HISTORICAL PERSPECTIVES The fact that the Storm God of the city POCULUM, which most probably corresponds to Karahöyük itself, was worshiped by a local ruler of the kingdom of Malizi on an official monument found in its capital is of course remarkable. It entails on the one hand the relevance that the city of Karahöyük and its territory had in the scenario of the Syro-Anatolian states at the beginning of the Iron Age, and on the other hand the strong cultural, religious, and political relationships linking the regions of Elbistan and Malatya.h g p p g g y The so-called stele from Izgın shows instead how things had already changed during the 11th century BC. It has been found reused as a headstone in the cemetery of the namesake village, ca. 2 km southwest of Karahöyük itself. It describes the extension of the borders of the kingdom of Malizi, celebrating the building of new cities and the settlement of people by a local ruler called Taras (Hawkins 2000: 314-318). Considering its location it seems 68 Francesco Di Filippo, Federico Manuelli Francesco Di Filippo, Federico Manuelli quite obvious that the stele commemorates the colonization and annexation of the Elbistan region into the Malizi realm (Alparslan 2017: 214; Hawkins, Weeden 2017: 289). The perpetuation of this relationship, or at least the fact that from time to time the kingdom of Malizi extended its domain into this region, is testified by the later inscription from Tanır. Located on a natural rock in proximity to a spring source and along an important pass that connects the Elbistan plain to the west, the inscription, tentatively dated to the 9th-8th century BC, mentions the name of the city Malizi and contains a badly preserved group of signs probably indicating the name of one of its rulers (Doǧan-Alparslan, Alparslan 2013). In summary, despite the fact that it fails to provide an adequate historical context and explanation for the uniqueness of this phenomenon, the hypothesis that the lions erected around the Elbistan plain were a manifesta- tion of the control exerted by the kingdom of Malizi over this territory from the 11th century BC onwards cannot be completely excluded. ACKNOWLEDGMENTS We are deeply thankful to Tayfun Bilgin for having given us the permit to include in this article photographs from his archive of Arlantaş, Sevdiliköy, and Hunu (www.hittitemonuments.com, v. 1.72). We are equally indebted to Gül Eralp Kania for having allowed us to use the original pictures of her study on Sevdiliköy, and to Halil Yusufoğlu for the use of the map reproduced at Fig. 15. Special appreciation is also due to the John Henry Haynes archive of the Special Collections of the Fine Arts Library at Harvard University for the permit to publish the photograph reproduced at Fig. 1. We are grateful to Rana Zaher for having accepted to and succeeded in draw- ing the lions from available published material. For their help and precious advice offered during several phases of this study, we would like to thank Metin Batıhan, Elizabeth Carter, Ali Çifçi, Valentina D’Amico, Christian W. Hess and Mehmet Usta. Part of this research has been realized in the framework of the project ‘Beyond the Crisis’ funded by the German Research Foundation (DFG project #324049112). Francesco Di Filippo wrote the land- scape analysis and Federico Manuelli the part on iconography and style, both authors contributed to writing the introduction, the discussion, and the conclusions. Maps have been generated by Francesco Di Filippo using QGIS. 5. HISTORICAL PERSPECTIVES h p The fact that the sculptures show only few iconographic and stylistic similarities with the set of lions’ represen- tations known from Arslantepe can certainly suggest the existence of some more provincial production, as well as the employment of different craftsmen and workshops. Indeed, differences in iconographic models and details are in general recognizable when the images carved on the reliefs from Arslantepe are compared with contemporary artworks visible on the monuments coming from the territory around the site, such as Ispekçür and Darende (Poli 2008: 258-264). To conclude, there are more than a few points concerning the fascinating phenomenon of the free-standing lions of the Elbistan plain that still need further explanation. However, this study has established their promi- nence and uniqueness within the artistic scenario of the Syro-Anatolian world of the late-2nd and early-1st millen- nium BC, showing their important role as boundary monuments marking essential places of the cultural land- scape. Moreover, the analysis has clearly demonstrated how computational spatial models can efficiently be applied to iconographic and stylistic aspects in order to provide fresh new data that help to answer complex and unsolved archaeological and historical questions. Abū Assāf A. 1990, Der Tempel von ‘Ain Dara (Damaszener Forschungen 3), Mainz am Rhein, Verlag Philipp von Zabern. Akurgal E. 1949, Späthethitische Bildkunst, Ankara, Archäologisches Institut der Universität. Alparslan M. 2017, The East: Upper Land, Išuwa-Malitiya, Azzi-Hayaša Philology, in M. Weeden, L. Z. Ullmann (eds), Hittite Landscape and Geography, Leiden-Boston, Brill: 209-218 BIBLIOGRAPHY 69 Hic sunt leones Baker A. R. H. 2003, Geography and History, Cambridge, Cambridge University Press. Balcıoğlu B. 2009, Gaziantep Arkeoloji Müzesi Geç Hitit Dönemi Taş Eserleri, Gaziantep, Gaziantep Arkeoloji Müzesi. Bilgin T. 2021, Hittite Monuments, v. 1.72, http://www.hittitemonuments.com/ (retrieved on 2021-03-03). Blanchard V. 2019, Arslantepe, capitale du royaume de Malizi/Melid, in V. Blanchard (ed), Royaumes oubliés de l’empire hittite aux araméens, Paris, Lienart: 186-203. Bevan A. 2011, Computational Models for Understanding Movement and Territory, in V. Mayoral Herrera, S. Celestino Pérez (eds), Tecnologías de Información Geográfica y Análisis Arqueológico Del Territorio: Actas Del V Simposio Internacional de Arqueología de Mérida, Mérida, Consejo Superior de Investigaciones Científicas: 383- 394 Brown G. H. 1967, Prehistoric Pottery from the Antitaurus, Anatolian Studies 17: 123-164 Bryce T. R. 2012, The World of the Neo-Hittite Kingdoms: a political and military history, Oxford-New York, Oxford University Press.h Carter E. 1996, The Kahramanmaraş Archaeological Survey Project: A Preliminary Report On the 1994 Season, Araştırma Sonuçları Toplantısı 13: 289-306. Çambel H. 1999, Corpus of Hieroglyphic Luwian Inscriptions, Vol. 2: Karatepe-Aslantaş, Be Corpus of Hieroglyphic Luwian Inscriptions, Vol. 2: Karatepe-Aslantaş, Berlin, Walter de Gruyte Charles B. B. 1911, Hittite inscriptions; certain newly discovered inscriptions, together with revised copies of a number of hitherto known and still in situ, representing a portion of the results of the Cornell expedition to Asia Minor and the Assyro-Babylonian Orient, by Benson Brush Charles, Ithaca, La Vergne. y y y g Cholidis M., Martin L. 2010, Tell Halaf V. Im Krieg zerstörte Denkmäler und Ihre Restaurierung, Berlin, De Gruyter. Çifçi A., Greaves A. M. 2010, Settlement Patterns in the Second and First Millennia BC Elbistan Plain, Anatolica 36: 89-110. Collins B. J. 1998, Hattušili I, The Lion King, Journal of Cuneiform Studies 50: 15-20.h Collins B. J. 2004, The Politics of Hittite Religious Iconography, in M. Hutter, S. Hutter-Braunsar (eds), Offizielle Religion, lokale Kulte und individuelle Religiosität: Akten des religionsgeschichtlichen Symposiums “Kleinasien und angrenzende Gebiete vom Beginn des 2. Bis zur Mitte des 1. Jahrtausends v. Chr.” (Bonn, 20.–22. Februar 2003) (AOAT 318), Münster, Ugarit-Verlag: 83-115. g g Delaporte L. 1940, Malatya: fouilles de la mission archéologique française. Arslantepe, La Porte des Lions, Paris, De Boccard. De Reu J., Bourgeois J., Bats M., Zwertvaegher A., Gelorini V., De Smedt Ph., Chu W., Antrop M., De Maeyer Ph., Finke P., Van Meirvenne M., Verniers J., Crombé Ph. Eralp G. 1995, Sevdiliköy Geç Hitit Aslanı, in A. Erkanal, H. Erkanal, H. Huryilmaz, A. T. Okse, N. Cinardali, S. Gunel, H. Tekin, B. Uysal, D. Yalcikli (eds), In memoriam I. Metin Akyurt, Bahattin Devam ani kitabi: eski yakin dogu kültürleri üzerine incelemeler, İstanbul, Arkeoloji ve Sanat Yayınları: 115-120. BIBLIOGRAPHY 2013, Application of the topographic position index to heterogeneous landscapes, Geomorphology 186: 39-49. g p p gy Di Filippo F., Mori L. 2018, How Difficult? Mountain Roads and Pathways Reaching Ancient Melid (Malatya) in South-Eastern Anatolia: A Reconsideration, Studi Miceneo ed Egeo-Anatolici NS 4: 41-62.h Di Filippo F., Mori L. 2019, The Malatya Plain in the Network of Interregional Relations in the Late Bronze and Iron Ages, in N. Durak, M. Frangipane (eds), Arslantepe. Proceedings of the I. International Archaeology Sympo- sium, Malatya, İnönü University Press: 173-184. Doǧan-Alparslan M., Alparslan M. 2013, A New Luwian Rock Inscription from Kahramanmaraş, in A. Mouton, I. Rutherford, I. Yakubovich (eds), Luwian Identities. Culture, Language and Religion Between Anatolia and the Aegean, Leiden-Boston, Brill: 215-232. Dumankaya O., Topaloğlu Y. 2017, Kahramanmaraş Afşin ve Elbistan Ilçeleri 2016 Yılı Yüzey Araştırması, Araştirma Sonuçları Toplantısı 35: 285-310. Duru R. 2012, Yesemek. Eski Önasya Dünyası´nın en büyük heykel alteyesi, Gaziantep: Gaziantep Büyükşehir Bel- ediyesi Yayınları. Eralp G. 1995, Sevdiliköy Geç Hitit Aslanı, in A. Erkanal, H. Erkanal, H. Huryilmaz, A. T. Okse, N. Cinardali, S. Gunel, H. Tekin, B. Uysal, D. Yalcikli (eds), In memoriam I. Metin Akyurt, Bahattin Devam ani kitabi: eski yakin dogu kültürleri üzerine incelemeler, İstanbul, Arkeoloji ve Sanat Yayınları: 115-120. 70 Francesco Di Filippo, Federico Manuelli Feldman M. H. 2014, Communities of Style: Portable Luxury Arts, Identity and Collective Memory in the Iron Age Levant, Chicago, University of Chicago Press. g y g Gilibert A. 2011, Syro-Hittite Monumental Art and the Archaeology of Performance. The Stone Reliefs at Carchemish and Zincirli in the Earlier First Millennium BCE, Berlin-New York, De Gruyter. Gilibert A. 2014, Death, Amusement and the City: Civic Spectacles and the Theatre Palace of Kapara, King of Gūzanā, KASKAL 10: 35-68. Gilibert A. 2015, Religion and Propaganda under the Great Kings of Karkemiš, in A. D’Agostino, V. Orsi, G. Torri (eds), Sacred Landscape of Hittites and Luwians: proceedings of the International Conference in Honour of Franca Peccchioli Daddi: Florence, February 6th-8th 2014, Firenze, Firenze University Press: 137-155. y y Giusfredi F. 2010, Sources for a Socio-Economic History of the Neo-Hittite States, Heidelberg, Universitätsverlag Winter.h Glatz C. 2021, The Making of Empire in Bronze Age Anatolia. Hittite sovereign Practice, Resistance, and Negotiation, Cambridge, Cambridge University Press. Gonnella J., Khayyata W., Kohlmeyer K., al-Ghafour A. A. 2005, Die Zitadelle von Aleppo und der Tempel des Wet- tergottes: neue Forschungen und Entdeckungen, Münster, Rhema Verlag. BIBLIOGRAPHY Harmanşah Ö. 2011, Moving Landscapes, Making Place: Cities, Monuments and Commemoration at Malizi/ Melid, Journal of Mediterranean Archaeology 24: 55-83. h Ö d h h l l l d d h l f l Harmanşah Ö. 2017, Borders are rough-hewn: monuments, local landscapes and the politics of place in a Hittite borderland, in E. L. Baysal, L. Karakatsanis (eds), Bordered Places \| Bounded Times. Cross-Disciplinary Perspec- tives on Turkey, London, British Institute at Ankara Monograph: 37-51. Hawkins J. D. 2000, Corpus of Hieroglyphic Luwian Inscriptions: Vol. 1. Inscriptions of the Iron Age: Part 1. Text: Introduction, Karatepe, Karkamis, Tell Ahmar, Maras, Malatya, Commagene, Berlin-Boston, De Gruyter. J p f g yp p p f g Introduction, Karatepe, Karkamis, Tell Ahmar, Maras, Malatya, Commagene, Berlin-Boston, De Gruyter. Hawkins J. D., Weeden M. 2016, Sketch history of Karkemish in the earlier Iron Age (Iron I-IIB), in T. J. Wilkin- son, E. Peltenburg (eds), Carchemish in Context: The Land of Carchemish Project, 2006-2010, Oxford-Philadel- phia, Oxbow Books: 9-21. p Hawkins J. D., Weeden M., 2017, Kizzuwatna and the Euphrates States: Kummaha, Elbistan, Malatya Philology, in M. Weeden, L. Z. Ullmann (eds), Hittite Landscape and Geography, Leiden-Boston, Brill: 281-294. Hogarth D. G. 1914, Carchemish: Report on the Excavations at Djerabis. Part I. Introductory, London, Trustees of the British Museum. Hogarth D. G., Munro J. A. R. 1893, Modern and Ancient Roads in Eastern Asia Minor, Royal Geographical Soci- ety 3: 643-744.i Jasiewicz J., Stepinski T. 2013, Geomorphons – a pattern recognition approach to classification and mapping of landforms, Geomorphology 182: 147-156. Jenness J. 2006. Topographic Position Index (TPI) v. 1.3a, http://www.jennessent.com/arcview/tpi.htm (retrieved on 2021-1-29) Kiepert R. 1904-19077. Malatja. Karte von Kleinasien, Berlin, Dietrich Reimer (Ernst Vohsen). Kiepert R. 1913. Malatja. Karte von Kleinasien, Berlin, Dietrich Reimer (Ernst Vohsen).h Kohlmeyer K. 2013, The Temple of the Storm-God of Aleppo, in W. Orthmann, P. Matthiae, M. al-Maqdissi (eds), Archéologie et Histoire de la Syrie I. La Syrie de l’époque néolithique à l’âge du fer, Harrasowitz, Wies- baden: 511-524. Kökten I. K. 1960, Anadolu Maraş vilayetinde tarihten dip tarihe gidiş, Türk Arkeoloji Dergisi X/1: 42-52. Kontani R., Sudo H., Yamaguchi Y., Hayakawa Y., Kulakoğlu F., Emre K. 2012, Preliminary Report for the Archae- ological Survey in Kayseri Province, Turkey (KAYAP), Fourth Season (2011), Bulletin of the Okayama Orient Museum 26: 15-29. Konyar E. Liverani M. 1988, The Growth of the Assyrian Empire in the Habur/Middle Euphrates Area: A New Paradigm, State Archives of Assyria Bulletin 2/2: 81-98. BIBLIOGRAPHY 2009, Kahramanmaraş Yüzey Araştırması 2007, Araştirma Sonuçları Toplantısı 26: 175-186.r 2009, Kahramanmaraş Yüzey Araştırması 2007, Araştirma Sonuçları Toplantısı 26: 175-186. Konyar E. 2011, Surveys in Kahramanmaraş in 2010, ANMED: News of Archaeology from Anatolia’s Mediterranean Areas 9: 174-179.h Liverani M. 1988, The Growth of the Assyrian Empire in the Habur/Middle Euphrates Area: A New Paradigm, State Archives of Assyria Bulletin 2/2: 81-98. 71 Hic sunt leones Llobera M. 2000, Understanding movement: a pilot model towards the sociology of movement, in G. Lock (ed), Beyond the Map. Archaeology and Spatial Technologies, Amsterdam, IOS Press: 65-84. y p gy p g Llobera M. 2001, Building Past Landscape Perception with GIS: Understanding Topographic Prominence, Journal of Archaeological Science 28: 1005-1014. Manuelli F. 2016, What remains when contact breaks off? Survival of knowledge and techniques in material cul- ture of the peripheral regions of the Hittite Empire after its dissolution, in E. Foietta, C. Ferrandi, E. Quirico, F. Giusto, M. Mortarini, J. Bruno, L. Somma (eds), Cultural & Material Contacts in the Ancient Near East. Pro- ceedings of the International Workshop, 1-2 December 2014, Torino, Torino, Apice Libri: 26-35. Manuelli F. 2019, Carving the memory, altering the past. PUGNUS-mili and the earlier Iron Age rulers at Arslantepe/Malizi (South-Eastern Turkey), in R. Lafer, H. Dolenz, M. Luik (eds), Antiquitates variae. Fest- schrift für Karl Strobel zum 65. Geburtstag, Rahden/Westf, VML Verlag: 227-241.hi ft f g g Manuelli F., Mori L. 2016, “The King at the Gate”. Monumental Fortifications and the Rise of Local Elites at Arslantepe at the End of the 2nd Millennium BCE, Origini 39/1: 209-241.fi t Manuelli F., Mori L. 2016, “The King at the Gate”. Monumental Fortifications and the Rise of Local Elites at Arslantepe at the End of the 2nd Millennium BCE, Origini 39/1: 209-241. Maunsell, F. R., 1902. Malatia. Eastern Turkey in Asia (IDWO 1522), London, Great Britain War Office, http:// Maunsell, F. R., 1902. Malatia. Eastern Turkey in Asia (IDWO 1522), London, Great Britain War Office, http:// magma.cnr.it/resource/map/eta/ (retrieved on: 2021-1-10). 986, A sculptures quarry at Sikizlar, Annales Archéologiques Arabes Syriennes 36-37: 268-275. ’ l h l b l d A h l Mazzoni S. 1986, A sculptures quarry at Sikizlar, Annales Archéologiques Arabes Syriennes 3 Mazzoni S. 1997, L’arte siro-ittita nel suo contesto archeologico, Contributi e Materiali di Archeologia Orientale 7: 287-327. Mazzoni S. 2000, Crisis and Change: The Beginning of the Iron Age in Syria, in P. BIBLIOGRAPHY Matthiae, A. Enea, L. Peyro- nel, F. Pinnock (eds), Proceedings of the First International Congress on the Archaeology of the Ancient Near East, Roma, Università di Roma La Sapienza: 1043-1055. p , Urban landscape of the Syro-Hittite capitals: models in contact, Mesopotamia 46: 137-155h Mazzoni S. 2011, Urban landscape of the Syro-Hittite capitals: models in contact, Mesopot Mazzoni S. 2013, Arts and Cross-Cultural Communication in the Early 1st Millennium: The Syro-Anatolian Con- tact, in Across the Border: Late Bronze-Iron Age Relations between Syria and Anatolia, in K. A. Yener (ed), Across the border: Late bronze-iron age relations between Syria and Anatolia: proceedings of a symposium held at the Research Center of Anatolian studies, Koç University, Istanbul, May 31 - June 1, Leuven-Paris-Walpole, Peeters: 465-492. May N. N. 2014, Gates and their Function in Mesopotamia and Israel, in N. N. May, U. Steiner (eds), The Fabric of Cities. Aspects of Urbanism, Urban Topography and Society in Mesopotamia, Greece and Rome, Leiden-Boston, Brill: 77-121. May N. N. 2014, Gates and their Function in Mesopotamia and Israel, in N. N. May, U. Steiner (eds), The Fabric of Cities. Aspects of Urbanism, Urban Topography and Society in Mesopotamia, Greece and Rome, Leiden-Boston, Brill: 77-121. Meriggi P. 1966, Quinto viaggio anatolico, Oriens Antiquus 5: 67-107. p f p g p y y p Meriggi P. 1966, Quinto viaggio anatolico, Oriens Antiquus 5: 67-107.i Q Meriggi P. 1975, Manuale di eteo geroglifico II/3, Roma, Edizioni dell’Ateneo.h ( ( Q Meriggi P. 1975, Manuale di eteo geroglifico II/3, Roma, Edizioni dell’Ateneo.h i Miller J. L. 2012, The (City-)Gate and the Projection of the Royal Power in Hatti, in G. Wilhelm (ed), Organiza- tion, Representation, and Symbols of Power in the Ancient Near East. Proceedings of the 54th Rencontre Assyri- ologique Internationale at Würzburg 20–25 July 2008, Winona Lake, Eisenbrauns: 675-685. Moortgat A 1955 Tell Halaf III Die Bildwerke Berlin De Gruyter tion, Representation, and Symbols of Power in the Ancient Near East. Proceedings of the 54th Rencontre Assyri- ologique Internationale at Würzburg 20–25 July 2008, Winona Lake, Eisenbrauns: 675-685. Moortgat A. 1955, Tell Halaf III. Die Bildwerke, Berlin, De Gruyter. Moortgat A. 1955, Tell Halaf III. Die Bildwerke, Berlin, De Gruyter.f Naval Staff 1919, A Handbook of Asia Minor Volume IV/2. Cilicia, Antitaurus, and North Syria, London, Intelli- gence Department. Newhard J. M. L., Levine N., Rutherford A. BIBLIOGRAPHY 2008, Least-cost pathway analysis and inter-regional interaction in the Göksu valley,Turkey, Anatolian Studies 58, 87-102. Nóvak M. 2012, The Temple of ‘Ain Dāra in the Context of the Imperial and Neo-Hittite Architecture and Art, in J. Kamlah (ed), Temple Building and Temple Cult. Architecture and Cultic Paraphernalia of temples in the Levant (2.-1. Mill. B.C.E.), Proceedings of the Conference on the Occasion of the 50th Anniversary of the Institute of Bibli- cal Archaeology at the University of Tübingen (28-30 May 2010), Wiesbaden, Harrassowitz: 41-54. gy y f g y Orthmann W. 1971, Untersuchungen zur Späthethitschen Kunst, Bonn, Saarbrücken. gy y f g y Orthmann W. 1971, Untersuchungen zur Späthethitschen Kunst, Bonn, Saarbrücken. Orthmann W. 2002, Die Bildkunst Übergang von der Groβreichszeit zur späthethitischen Periode, in E. A. Braun- Holzinger, H. Matthäus (eds), Die nahöstlichen Kulturen und Griechenland an der Wende von 2. zum 1. Jah- rtausend v. Chr. Kontinuität und Wandel von Strukturen und Mechanismen kultureller Interaktion, Paderborn, Bibliopolis: 153-159. 72 72 Francesco Di Filippo, Federico Manuelli Özgüç T., Özgüç N. 1949, Karahöyük hafriyatı raporu 1947: Türk Tarih Kurumu tarafından yapılan, Ankara, Türk Tarih Kurumu tarafından Basımevi. Peyronel L. 2019, Il ruggito del leone. Qualche osservazione sulle immagini ferine nel mondo siriano del III mil- lennio a.C., in S. Valentini, G. Guarducci (eds), Between Syrian and the Highlands. Studies in Honor of Giorgio Buccellati and Marilyn Kelly-Buccellati (Studies on the Ancient Near East and the Mediterranean 3), Roma, Arbor Sapientiae: 323-334. p Poli P. 2008, Traduzione, assimilazione, interpretazione di elementi iconografici ed ideologici da tradizio- ni diverse. Il caso di Malatya, in B. Bellucci, E. Jucci, A. Rizza, B. M. Tomassini Pieri (eds), Traduzione di tradizioni e tradizioni di traduzione. Atti del quarto incontro “Orientalisti” (Pavia, 19-21 Aprile 2007), Milano, Qu.A.S.A.R.: 249-276. Polla S., Verhagen P. 2014, Computational approaches to the study of movement in archaeology. Theory, practice and interpretation of factors and effects of long term landscape formation and transformation, Berlin, de Gruyter. f Pucci M. 2015, Founding and Planning a new Town: the southern Town Gate at Zincirli, in P. Ciafardoni, D. Giannessi (eds), From the Treasures of Syria. Essays on Art and Archaeology in Honour of Stefania Mazzoni, Lei- den, Nederland Institute for the Near East: 35-74.h Ramsay W. M. 1890, The Historical Geography of Asia Minor, London, RGS. h Ramsay W. M., Hogarth D. G. BIBLIOGRAPHY 1893, Pre-Hellenic Monuments of Cappadocia (2), Recueil de travaux relatifs à la philologie et à l’archéologie égyptiennes et assyriennes 15: 89-97. Riis P. J., Buhl M.-L. 1990, Les objets de la période dite syro-hittite (Age du Fer), Hama II/2, Copenhagen, Fonda- tion Carlsberg. g Scheinhardt H. 1979, Typen türkischer Ortsnamen, Heidelberg, Carl Winter. g Scheinhardt H. 1979, Typen türkischer Ortsnamen, Heidelberg, Carl Winter. Seeher J. 2009, Die Techniken der Steinbearbeitung in der hethitischen Architektur des 2. Jahrtausends v. Chr, in M. Bachmann (ed), Bautechnik im antiken und vorantiken Kleinasien: Internationale Konferenz 13.–16. Juni 2007 in Istanbul, Istanbul, Ege Yayınları: 119-156. Simon Z. 2013, Wer war Großkönig I(a)+ra/i-TONITRUS der Karahöyük-Inscchrift?, in L. Feliu, J. Llop, A. Mill- er Albà, J. Sanmartín (eds), Time and History in the Ancient Near East. Proceedings of the 56th Rencontre Assyri- ologique Internationale at Barcelona, 26-30 July 2010, Winona Lake, Eisenbrauns: 823-832. J. R. S. 1888, An Epigraphical Journey in Asia Minor, Boston, Damrell & Upham. Surface-Evans S. L., White D. A. 2012, Least Cost Analysis of Social Landscapes, Salt Lake City, The University of Utah Press. Strawn B. A. 2001, “What is Stronger Than Lions”? Leonine Image and Metaphor in the Hebrew Nibble and the Ancient Near East (PhD Thesis), Faculty of Theological Seminary, Princeton University. Öh hh Summers G. D., Özel E. 2012, The Hittite Stone and Sculpture Quarry at Karakız Kasabası and Hapis Boğazı in the District of Sorgun, Yozgat, Central Anatolia, American Journal of Archaeology 116/3: 507-519. hh Summers G. D., Özel E. 2012, The Hittite Stone and Sculpture Quarry at Karakız Kasabası and Hapis Boğazı in the District of Sorgun, Yozgat, Central Anatolia, American Journal of Archaeology 116/3: 507-519. Tilley C. 1994, A phenomenology of landscape: places, paths and monuments, Oxford, Berg Publishers. Tunca Ö. 1976, The Lion of Çolaklıköy, Anatolian Studies 26: 113-16. h Uysal B., Çifçi A. 2019, Elbistan Karahöyük Kazısı, Kazı Sonuçları Toplantısı 41/1: 411-422. Veselský M., Bandura P., Burian L., Harciníková T., Bella P. 2015, Semi-automated recognition of planation sur- faces and other flat landforms: a case study from the Aggtelek Karst, Hungary, Open Geosciences 1: 799-811. l y gg g y p Von der Osten H. H. 1930, Explorations in Hittite Asia Minor, 1929 (OIC 8), Chicago, University of Chicago Press. Von Luschan F. 1902, Ausgrabungen in Sendschirli. Vol. 3, Thorsculpturen. Berlin, Georg Reimer. l gg g p Von der Osten H. H. BIBLIOGRAPHY 1930, Explorations in Hittite Asia Minor, 1929 (OIC 8), Chicago, Unive p ( ) g y g Von Luschan F. 1902, Ausgrabungen in Sendschirli. Vol. 3, Thorsculpturen. Berlin, Georg Reimer. p g Von Luschan F. 1902, Ausgrabungen in Sendschirli. Vol. 3, Thorsculpturen. Berlin, Georg Reim h Wilhelm G. 2011, Combining textual and archaeological evidence of urban structures: Hattusa and Sarissa, Meso- potamia 46: 103-108.h Woolley C. L. 1921, Carchemish: Report on the Excavations at Djerabis. Part II. The Town Defenses, London, Trus- tees of the British Museum.h Woolley C. L. 1952, Carchemish: Report on the Excavations at Djarabis. Part III. The Excavations in the Inner Town; The Hittite Inscriptions, London, Trustees of the British Museum. h Yusufoğlu H. 2013, An intramontane pull-apart basin in tectonic escape deformation: Elbistan Basin, Eastern Tau- rides, Turkey, Journal of Geodynamics 65: 308-329.
https://openalex.org/W2563336411	https://www.matec-conferences.org/10.1051/matecconf/20178701014/pdf	English	null	Effectiveness of Mangrove Forest as Coastal Protection along the West Coast of Northern Peninsular Malaysia	MATEC web of conferences	2,016	cc-by	3,860	1 Introduction The effectiveness of mangrove forest as coastal protection in front of the coastal area was evidenced during the IOT 2004 [8]. The coastal properties without any protection were completely destroyed by the tsunami wave. Figure 1 illustrates how mangrove forest helps to dissipate the wave energy, hence, reducing the impact along the coastal area [9]. The importance and benefits offered by mangroves have encouraged increasing number of related research. There had been numerous mangrove rehabilitation programs and replanting initiatives for protection at affected area organized by various agencies. Effectiveness of Mangrove Forest as Coastal Protection along the West Coast of Northern Peninsular Malaysia Department of Civil and Environmental Engineering, Universiti Teknologi PETRONAS, 32610 Seri Iskandar, Perak. Abstract. The effectiveness of mangrove forest as coastal protection was evidenced during Indian Ocean Tsunami (IOT) on 24th December 2004. This paper presents the assessment on the characteristics and distribution of the mangrove forest along three northern states (Perlis, Kedah and Perak) affected by IOT 2004. At each site, the characteristics of individual tree were recorded including their species, height and diameter at breast height (DBH). A total of 52 matured trees were sampled that encompassing four species; A. marina (n=21), A. alba (n=7), R. mucronata (n=20) and R. apiculata (n=4) were found with A. marina as the dominant species. The zonation of mangrove forest along the northern states revealed that Avicenniacea inhabits the frontal area. The field survey also discovered that mangrove forest along the states was generally insufficient to protect the coastal area from future tsunami. There were several rehabilitation programs that had been implemented, but not all replanting efforts were successful. Outcomes from this research provided invaluable findings for the optimum replanting scheme to ensure acceptable level of protection along Malaysia’s coastline. a Corresponding author: mustafa_hashim@utp.edu.my © The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons Attribution License 4.0 (http://creativecommons.org/licenses/by/4.0/). , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences DOI: 10.1051/matecconf/20178701014 1.1 Indian Ocean Tsunami 2004 Tsunami is one of the disastrous natural events occurred and mainly generated by the sliding of the earth's surface, namely earthquakes. Generally, tsunami is triggered by the earthquake of magnitude 6.5 and above [1]. Up to 2012, Japan was recorded as the highest frequency of tsunami occurrences with 48 events [2]. The most catastrophic event recorded in tsunami history was the Indian Ocean Tsunami which occurred on 24th December 2004 (IOT 2004). The tsunami waves reached 13 countries including Indonesia, Sri Lanka, India, Thailand, Somalia, Burma, Maldives, Malaysia, Tanzania, Bangladesh and Kenya [3]. The epicentre of the event occurred at about 30 km off the West Coast of Northern Sumatra, Indonesia. The 20 m wave height led to about 226,000 deaths and approximately USD 9.93 billion of economic losses [4,5,6]. The tsunami wave had propagated to the West Coast of Peninsular Malaysia and affected five states including Perlis, Pulau Pinang, Kedah, Perak and Selangor with 63 numbers of total deaths and missing people reported [7]. However, the optimum approach or technical strategy for the mangrove replanting scheme is still under active research. Several factors govern the performance of mangrove as coastal protection which include species of the tree, characteristics of tree (height, age, diameter of stem and root, diameter at breast height), basal area and also wave characteristics [8,10]. Thus, this paper aim to assess the characteristics and distribution of the mangrove forest along the West Coast of Peninsular Malaysia. , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences , 01014 (2017) 87 MATEC Web of Conferences DOI: 10.1051/matecconf/20178701014 Figure 1. Wave energy and its impact at coastal area for A. With mangrove, B. Without mangrove [9]. 2.3.2 Species diversity The ecological approach was used to determine the diversity of the mangrove forest. The relationship proposed by Shannon-Weiner (Eq 2) was used to derive the diversity index. i P n i i P H ln 1 ' (2) (2) Where; P = n/N n = No of individuals Figure 3. Distribution of mangrove forest in the surveyed coastlines. 8% 14% 38% 40% R. apiculata A. alba R. mucronata A. marina N = Total number of mangrove tree 2.2 Data Collection than 10 years. The age was categorized based on the average height of the trees. The height of the tree greatly influenced their capabilities to attenuate wave. However, young mangrove trees below two years old were not included in this study. They are easily being washed away and not able to withstand the wave action [8]. The mangrove forest at each site was assessed and the time, date, coordinate and tide level were recorded. The random sampling plots (30 x 30 m) were established at each site. The sampling plots were chosen as a representative of general mangrove condition for the accessed area. The selected matured trees found within the plot were identified. Measurement were carried out covering chosen trees representing different age groups. The data from selected trees were gathered include species of mangrove, basal area in m2, height of tree in m, and diameter at breast height (DBH) in cm. A. marina (40%) was the dominant species found followed by R. mucronata (38%), A. alba (14%) and R. apiculata (8%) (Figure 3). The types of mangrove found relate well with general zonation of mangrove forest. Typical zonation of the mangrove forest in Malaysia is made up of Avicennia sp. or Sonneratia sp. on the seawards most and followed by Rhizophora sp., Brugueira sp., Xylocarpus sp., and transitional dryland (Fig 4) [14, 15]. The field investigation found high number of Avicennia sp. in the visited coastal area. Avicennia sp. are able to grow seaward as they may adapt to the high salinity range from 20.9 to 31.25 ppt [16]. 2.1 Study Area West Coast of Peninsular Malaysia had been hit with IOT 2004. The area is bordered by the calm sea of Malacca Straits, which enabled the small particles of sediment (mud) to settle down at the coastal area [11]. Hence, the conditions contributed to the growth of mangrove forest in this part of coastline rather than the East Coast of Peninsular Malaysia [12]. However, these areas affected by the IOT 2004 are not fully recovered and potentially at high risk for future tsunami. In order to explore the current level of protection available at the previously affected IOT 2004 area, 10 sampling locations were assessed (Fig 2). These involved sites covering the three states, namely Perak (Parit Haji Dollah, Tg Kepah, Pantai Remis, Kuala Sepetang and Kuala Gula), Kedah (Pantai Merdeka, Kuala Sg Limau, Sg Ibus, Kuala Teriang) and Perlis (Kuala Sanglang). Figure 1. Wave energy and its impact at coastal area for A. With mangrove, B. Without mangrove [9]. Figure 1. Wave energy and its impact at coastal area for A. With mangrove, B. Without mangrove [9]. Figure 2. Locations of study area located at the coastline of Perlis, Kedah and Perak (Image adapted from Google Earth, 2016) Figure 2. Locations of study area located at the coastline of Perlis, Kedah and Perak (Image adapted from Google Earth, 2016) 2 2 DOI: 10.1051/matecconf/20178701014 , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences , 01014 (2017) 87 MATEC Web of Conferences Where; Circumference1.3 = Circumference at 1.3 m height 2.3.1. Diameter at Breast Height (DBH) Table 1. Mangrove species found at sampling plots. Location Rhizophoraceae Avicenniaceae R. mucronata R. apiculata A. alba A. marina Parit Haji Dollah 3 0 1 0 Tg Kepah 0 0 0 4 Kg Pantai Remis 0 0 2 2 Kuala Sepetang 2 3 0 2 Kg Kuala Gula 4 0 0 0 Pantai Merdeka 3 0 0 1 Kuala Sg Limau 0 0 3 1 Sg Ibus 0 0 1 3 Kuala Teriang 4 1 0 8 Kuala Sanglang 4 0 0 0 Table 1. Mangrove species found at sampling plots. To determine the DBH, the circumference of the mangrove tree at 1.3 m height was measured at site. The DBH was calculated by using the relationship (Eq 1) stated in [13]. 3. 1 nce Circumfere DBH (1) Where; Circumference1.3 = Circumference at 1.3 m height 3. 1 nce Circumfere DBH (1) (1) Where; 3.1. Mangrove Distribution Mangrove forest along the coastline of Perlis, Kedah and Perak was assessed. A total of 52 matured trees from four species and two families were identified. The species included Rhizophora mucronata, Rhizophora apiculata, Avicennia alba and Avicennia marina. Table 1 shows number of the inspected mangrove trees for the respective species. They were basically divided into three age groups; (1) three to six years, (2) seven to ten years, and (3) more Figure 3. Distribution of mangrove forest in the surveyed coastlines. 3 DOI: 10.1051/matecconf/20178701014 , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences , 01014 (2017) 87 MATEC Web of Conferences Figure 4. Typical mangrove forest zonation in Peninsular Malaysia [15]. representative area. The diversity of the species indicates how the area is able to attenuate wave. Different species have different wave attenuation capability [10]. Rhizophora sp. have larger and complex root and pneumatophores which enable them to create greater friction to wave as compared to other species [18]. DBH indicates as the best height to determine the tree growth. The DBH is also an approach to predict the dimension of the large tree [19,20]. The derivation from the relationship between DBH and height of the mangrove tree can be used to estimate the size of the tree [13,19]; and it indicated positive correlation as height increases with DBH (Fig 6). It is important to note that the DBH of the tree should increase with height as it grows [19]. Figure 4. Typical mangrove forest zonation in Peninsular Malaysia [15]. The Pearson correlation analysis is used to reveal how strong the relationship between the variables. The value for the correlation ranges from -1 to +1 (-1 ≤ r ≤ 1). Values close to +1 indicate the variables having strong positive relationship, while values close to -1 show strong negative relationship. However, value of zero implies no relationship. 3.2 Condition of Mangrove Forest y = 3.1317x - 6.7114 r = 0.71 n = 20 -5.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 0.00 2.00 4.00 6.00 8.00 10.00 H (m) DBH (cm) Rhizophora mucronata y = 1.3733x - 6.5452 r = 0.96 n = 4 0.00 5.00 10.00 15.00 20.00 25.00 0 5 10 15 20 25 H (m) DBH (cm) Rhizophora apiculata y = 2.2888x - 4.4639 r = 0.67 n = 21 0.00 10.00 20.00 30.00 40.00 0.00 5.00 10.00 15.00 H (m) DBH (cm) Avicennia marina y = 1.7471x - 1.5106 r = 0.78 n = 7 0.00 5.00 10.00 15.00 20.00 0.00 2.00 4.00 6.00 8.00 10.00 H (m) DBH (cm) Avicennia alba y = 3.1317x - 6.7114 r = 0.71 n = 20 -5.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 0.00 2.00 4.00 6.00 8.00 10.00 H (m) DBH (cm) Rhizophora mucronata y = 1.3733x - 6.5452 r = 0.96 n = 4 0.00 5.00 10.00 15.00 20.00 25.00 0 5 10 15 20 25 H (m) DBH (cm) Rhizophora apiculata y = 2.2888x - 4.4639 r = 0.67 n = 21 0.00 10.00 20.00 30.00 40.00 0.00 5.00 10.00 15.00 H (m) DBH (cm) Avicennia marina Figure 6. Correlation analysis between the height of the tree, H (m) and diameter at breast height, DBH (cm). Peninsular Malaysia by the relevant authorities from 2005 to 2010 (Table 2) [15]. The replantation is hoped to protect the coastal properties from wave actions. Peninsular Malaysia by the relevant authorities from 2005 to 2010 (Table 2) [15]. The replantation is hoped to protect the coastal properties from wave actions. 3.2 Condition of Mangrove Forest The diversity index is one of the approaches used to measure the diversity of the mangrove forest [17]. The derivation attributed from two facets including species richness and species evenness. The species diversity, evenness and richness are shown in Figure 5. It is revealed that R. apiculata recorded a strong correlation (n=4; r=0.96) as compared to other species. However, analysis concluded that the growth of mangrove forest synchronises well with DBH. All the correlation coefficient values (r) are above 0.5 which indicate moderate correlation. Hence, the growth of mangrove trees are considered satisfactory. The species diversity along the coastlines was estimated to range from 0 to 1.08. The highest species diversity found was in Kuala Sepetang with a value of 1.08. The value attributed from 3 species including R. mucronata, R. apiculata and A. marina. Meanwhile, the area with zero diversity indicated only one species at the Figure 5. Species diversity of mangrove forest in the surveyed coastlines Parit Haji Dollah Tg Kepah Kg Pantai Remis Kuala Sepetang Kg Kuala Gula Pantai Merdeka Kuala Sg Limau Sg Ibus Kuala Perlis Kuala Teriang Sp Diversity 0.56 0 0.69 1.08 0 0.56 0.56 0.56 0 0.86 0 0.2 0.4 0.6 0.8 1 1.2 Species diversity Figure 5. Species diversity of mangrove forest in the surveyed coastlines 4 4 , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences DOI: 10.1051/matecconf/20178701014 , 01014 (2017) 87 MATEC Web of Conferences Figure 6. Correlation analysis between the height of the tree, H (m) and diameter at breast height, DBH (cm). 3.3 Mangrove as Coastal Protection Mangrove forest provides countless benefits, not only to human but for ecosystem as well [21]. Unfortunately, despite the benefits offered, mangrove forest had been exploited extensively. The major factor of decreasing mangrove was due to land conversion [22]. The mangrove forest had been explored for shrimp aquaculture due to the suitability of its surrounding soil and water quality. Table 2. Locations for the rehabilitation programme in Peninsular Malaysia after IOT 2004 [15]. State Location Perlis Kuala Sanglang Kedah Kg. Huma, Tg. Dawai Kota Kuala Muda P. Pinang Pantai Malindo Perak Lekir, Manjung Selangor Bagan Nakhoda Omar Analysis shows that Malaysia is exposed to some level of risk of tsunami especially at Sabah and West Coast of Peninsular Malaysia including Perlis, Penang, Kedah, Perak and Selangor (Fig 7) [15]. Figure 7. Map showing the tsunami risk area in Malaysia [15]. However, recent visit in April 2016 revealed that some of the replantation programmes were not successful. As an example, the mangroves that were replanted in Lekir, Perak in 2008 showed a retarded growth. The villagers commented that the mangrove saplings had been planted for two times, but failed. The saplings only survived for a year and later washed away after the subsequent monsoon season at the end of the year. Currently, the third attempt with two years old R. mucronata were planted in early 2015. However, only few trees were found to still survive, while most of the replanted trees were destroyed by the monsoon. Remnants of the destroyed matured trees were also observed at Sg. Haji Dollah which is located south of Lekir (Fig 8). These mangroves were struggling in front of the rock revetment. It is suggested that the presence of Figure 7. Map showing the tsunami risk area in Malaysia [15]. After the IOT 2004, about 1295.80 ha (4,395,472 number of mangrove trees) had been replanted in 5 , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences DOI: 10.1051/matecconf/20178701014 revetment has led to the degradation of these trees. Mangrove trees planted behind the revetment was found to survive well. of coastal protection measure for this area. Interview with local villagers described further details on the mangrove forest at site. According to one of the senior citizens residing close to the beach, the mangrove replanting was done since 2005, after the IOT 2004. The three years old R. mucronata were planted with about 1.3 m spacing. 4 Conclusion the aftermath of the 2004 Indian Ocean earthquake and tsunami, Risk Management Solutions, Newark (2006). The IOT 2004 was the most outrageous event occurred in Malaysia with 63 numbers of total deaths. It was evidenced from the incident that the mangrove forest had contributed significantly in reducing the impacts from the wave. The unique configuration of mangrove trees enabled them to dissipate wave energy. [5] [5] A. D. Hawkes, M. Bird, S. Cowie, C. Grundy-Warr, B. P. Horton, A. T. S. Hwai, C. Macgregor, J. Nott, J. Rigg, J. E. Ong, L. Law, R. Robinson, M. Tan- Mullins, T. T. Sa, Z. Yasin and L. W. Aik, Sediments deposited by the 2004 Indian Ocean tsunami along Malaysia-Thailand Peninsula, Marine Geology, 169-190 (2007) The field investigation shows that A. marina was found as the dominant species along the West Coast of Peninsular Malaysia. The zonation of mangrove forest with Avicenniaceae as frontal species contributed to this result. [6] P. C. Athukorala, Disaster, generosity and recovery: Indian Ocean Tsunami, Australian national University, Canberra (2012) However, recent field investigation also discovered that there was no sufficient mangrove belt to provide effective protection especially at the South of Perak. The replantation of mangrove forest by the authorities had experienced failures twice due to the monsoons. However, further north of Perak, the distribution of mangrove forest was found to be increasing. It is discovered that the area is one of the tourist areas under the jurisdiction of the local government. Therefore, the area was given extra attention as compared to the South of Perak. Analysis recommended that future focus should be given to the area with less mangrove forest distribution. The optimum replanting scheme should be implemented immediately to ensure safety of Malaysia’s coastline from future natural disaster. This field study provides useful database which can be used for future comprehensive assessment with the aid of GIS technology and remote sensing analysis. [7] K. Rostam, I. Herman and A. B. M. Nor, Kawasan penempatan semula mangsa tsunami di Malaysia: Analisis kualiti kejiranan dan kesejahteraan isi rumah, Malaysian Journal of Society and Space, 5, 33-43 (2009) [8] D. M. Patel, V. M. Patel, K. Bhupesh and K. A. Patel, Peformance of mangrove in tsunami resistance, International Journal of Emerging Technology and Research, 1, 29-32 (2014) [9] Y. Mazda, M. Magi, M. Kogo and P. N. 3.3 Mangrove as Coastal Protection The villagers were also briefed on how to grow the mangrove tree. They were advised to plant the mangrove seedlings from the mature tree at the nursery area. This awareness campaign by the authorities allowed the mangrove to grow well at the Kuala Teriang. Among the successful replanting efforts include those in Kuala Teriang, Langkawi, Kedah. The area was one of the most affected area hit by the tsunami in Peninsular Malaysia. Healthy mangrove forest grows along the coastal area from Sg Melaka to Sg Kuala Teriang (Fig 9). In front of the mangrove forest, there was an approximately 100 m long geotextile fabric tube placed in front of beach. The geotubes had been laid as another line Figure 8. Replantation program implemented at Lekir, Perak, A. The replantation area in Tg Kepah. B. Fall-out of matured mangrove tree. (Image adapted from Google Earth, 2016; [15]) Figure 8. Replantation program implemented at Lekir, Perak, A. The replantation area in Tg Kepah. B. Fall-out of matured mangrove tree. (Image adapted from Google Earth, 2016; [15]) Figure 9. Geotube in front of mangrove forest at Kuala Teriang. (Image adapted from Google Earth, 2016) Figure 9. Geotube in front of mangrove forest at Kuala Teriang. (Image adapted from Google Earth, 2016) 6 6 DOI: 10.1051/matecconf/20178701014 , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences , 01014 (2017) 87 MATEC Web of Conferences 4 Conclusion Hong, Mangroves as a coastal protection from waves in the Tong King delta, Vietnam, Mangroves and Salt Marshes, 1, 127-135 (1997) [10] Y. Mazda and M. Magi, Wave reduction in mangrove forest dominated by Sonneratia sp., Wetlands Ecology and Management, 14, 365-378 (2006) Acknowledgement [11] F. Alemayehu, O. Richard, K. M. James and O. Wasonga, Assessment of Mangrove Covers Change and Biomass in Mida Creek, Kenya, Journal of Forestry, 4, 398-413 (2014) This research is fully supported by FRGS grant (FRGS/2/2013/TK03/UTP/02/1). The authors also wish to express special thanks to the Centre of Post Graduate Studies, Sustainable Resources MOR and Department of Civil and Environmental Engineering, Universiti Teknologi PETRONAS for their support and encouragement. [12] N. A. Kamaruddin, S. Fujii and B. R. Shivakoti, Applicability of satellite monitoring on mangrove forests in Malaysia: A review of potential benefits and challenges, International Journal of Research, 2, 185-204 (2015) [13] T. Buba, Relationships between Stem Diameter at Breast Height (DBH), Tree Height, Crown Length, and Crown Ratio of V. paradoxa C.F. Gaertn in the Nigerian Guinea Savanna, African Journal of Biotechnology, 12, 3441-3446 (2013) References [1] P. Whitmore, U. t. Brink, M. Carpolo, V. Huerfano- Moreno, W. Knight, W. Sammler and A. Sandrik, NOAA/West Coast Alaska tsunami warning center atlantic ocean response criteria, Science of Tsunami Hazards, 86-107 (2009). [14] K. S. Sinfuego and I. E. Buot Jr, Mangrove zonation and utilization by the local people in Ajuy and Pedada Bays, Panay Island, Philippines, Journal of Marine and Island Cultures, 3 (2014) [2] A. Suppasri, T. Futami, S. Tabuchi and F. Imamura, Mapping of historical tsunamis in the Indian and Southwest Pacific Oceans, International Journal of Disaster Risk Reduction, 62-71 (2012) [15] Y. Y. Hwai and S. Rosli, Management of mangrove forests: Policies and prescriptions, Jabatan Perhutanan Semenanjung Malaysia, Putrajaya (2012) [3] M. Z. Asmawi and A. N. Ibrahim, The impacts of tsunami on the well-being of the affected community in Kuala Muda, Kedah, Malaysia, Journal of Clean Energy Technologies, 1, 246-250 (2013) [16] H. Marisa and Sarno, Three Species Zonation of Sonneratia; based on Salinity in River Calik, South Sumatera, in International Conference of PLant, Marine and Environmental Sceinces (PMES-2015), Kuala Lumpur (2015) [4] I. Fehr, P. Grossi, S. Hernandez, T. Krebs, S. McKay, R. Muir-Wood, A. Pomonis, D. C. Souch, D. Windeler and Y. Xie, Managing tsunami risk in 7 , 01014 (2017) 87 ENCON 2016 MATEC Web of Conferences , 01014 (2017) 87 MATEC Web of Conferences DOI: 10.1051/matecconf/20178701014 [17] S. Das, P. S. Lyla and S. A. Khan, Marine microbial biodiversity: Present status and advanced statistical paradigms, Conservation Biology in Asia, 368-385 (2006) [18] K. Jusoff, Malaysian mangrove forests and their significance to the coastal marine environment, Pollution Journal Environmental Study, 22, 979- 1005 (2013). [19] A. Sumida, T. Miyaura and H. Torii, Relationships between stem diameter at breast height revisited: Analyses of stem growth using 20-year data of an even-aged Chamaecyparis obtusa stand, Tree Physiology Advance, 1-13 (2013) [20] D. W. Aheto, A. A. O. Aduomih and E. A. Obodai, Structural parameters and above-ground biomass of mangrove tree species around the Kakum river estuary of Ghana, Annals of Biological Research, 2, 504-514, (2011) [21] A. M. Hashim and S. M. P. Catherine, A laboratory study on wave reduction by mangrove forests, APCBEE Procedia, 27-32, (2013) [22] M. Simard, V. H. Rivera-Monroy, J. E. Mancera- Pineda, E. Castaneda-Moya and R. R. References Twilley, A systematic method for 3D mapping of mangrove forests based on Shutle Radar Topography Mission elevation data, ICEsat/GLAS wavefoms and field data: Application to Cienaga Gande de Santa Marta, Colombia, Remote Sensing of Environment, 112, 2131-2144, (2008) 8
https://openalex.org/W4225320261	https://www.biorxiv.org/content/biorxiv/early/2021/12/13/2021.12.12.472242.full.pdf	English	null	Dynamics of CTCF- and cohesin-mediated chromatin looping revealed by live-cell imaging	Science	2,022	cc-by	8,219	ARTICLE PREPRINT 1 Department of Biological Engineering, Massachusetts Institute of Technology; Cambridge, 02139, USA; 2 Department of Physics, Massachusetts Institute of Technology; Cambridge, MA 02139, USA; 3 Department of Molecular and Cell Biology, University of California, Berkeley; Berkeley, CA 94720, USA; 4 Institute for Medical Engineering and Sciences, Massachusetts Institute of Technology; Cambridge, MA 02139, USA; 5 Institut Curie; Paris 75005, France; 6 Max Planck Institute of Molecular Cell Biology & Genetics; Dresden, Germany; 7 Center for Systems Biology Dresden; Dresden, Germany; 8 Cluster of Excellence Physics of Life, TU Dresden; Dresden, Germany; 1 Department of Biological Engineering, Massachusetts Institute of Technology; Cambridge, 02139, USA; 2 Department of Physics, Massachusetts Institute of Technology; Cambridge, MA 02139, USA; 3 Department of Molecular and Cell Biology, University of California, Berkeley; Berkeley, CA 94720, USA; 4 Institute for Medical Engineering and Sciences, Massachusetts Institute of Technology; Cambridge, MA 02139, USA; 5 Institut Curie; Paris 75005, France; 6 Max Planck Institute of Molecular Cell Biology & Genetics; Dresden, Germany; 7 Center for Systems Biology Dresden; Dresden, Germany; 8 Cluster of Excellence Physics of Life, TU Dresden; Dresden, Germany; 1 Department of Biological Engineering, Massachusetts Institute of Technology; Cambridge, 02139, USA; 2 Department of Physics, Massachusetts Institute of Technology; Cambridge, MA 02139, USA; 3 Department of Molecular and Cell Biology, University of California, Berkeley; Berkeley, CA 94720, USA; 4 Institute for Medical Engineering and Sciences, Massachusetts Institute of Technology; Cambridge, MA 02139, USA; 5 Institut Curie; Paris 75005, France; 6 Max Planck Institute of Molecular Cell Biology & Genetics; Dresden, Germany; 7 Center for Systems Biology Dresden; Dresden, Germany; 8 Cluster of Excellence Physics of Life, TU Dresden; Dresden, Germany; † MG HBB d SGH t ib t d ll d li t th i fi t C d l id@ it d h @ i b d ; , y; †: MG, HBB and SGH contributed equally and may list their own name first. Correspondence: leonid@mit.edu, zechner@mpi-cbg.de, ashansen@mit.edu Animal genomes are folded into loops and topologically associating domains (TADs) by CTCF and cohesin, but whether these loops are stable or dynamic is unknown. Here, we directly visualize chromatin looping at the Fbn2 TAD in mouse embryonic stem cells using super-resolution live-cell imaging and quantify looping dynamics by Bayesian inference. Our results are consistent with cohesin-mediated loop extrusion in cells, and with CTCF both stopping and stabilizing cohesin. Surprisingly, the Fbn2 loop is both rare and dynamic, with a looped fraction of ~3-6.5% and a median loop lifetime of ~10-30 minutes. Instead of a stable loop, our results establish a highly dynamic view of TADs and loops where the Fbn2 TAD exists predominantly in a partially extruded conformation. This dynamic and quantitative view of TADs may facilitate a mechanistic understanding of their functions. dynamics of chromatin looping (16). Furthermore, previous studies have yielded conflicting results as to whether loops are well-defined in single cells (10–15), perhaps due to the difficulty associated with distinguishing bona fide CTCF- and cohesin-mediated loops from mere proximity that emerges stochastically (16). Recent pioneering work has visualized enhancer-promoter interactions (17, 18) and long-range V(D)J-chromatin interactions (19) in live cells. However, the dynamics of loop extrusion and the lifetime of CTCF/cohesin loops have not yet been quantified in living cells, which we therefore set out to do. dynamics of chromatin looping (16). Furthermore, previous studies have yielded conflicting results as to whether loops are well-defined in single cells (10–15), perhaps due to the difficulty associated with distinguishing bona fide CTCF- and cohesin-mediated loops from mere proximity that emerges stochastically (16). Recent pioneering work has visualized enhancer-promoter interactions (17, 18) and long-range V(D)J-chromatin interactions (19) in live cells. However, the dynamics of loop extrusion and the lifetime of CTCF/cohesin loops have not yet been quantified in living cells, which we therefore set out to do. INTRODUCTION Mammalian genomes are folded into loops and domains known as Topologically Associating Domains (TADs) by the proteins CTCF and cohesin (1). Mechanistically, cohesin is thought to load on DNA and bidirectionally extrude loops until it is blocked by CTCF such that CTCF establishes TAD boundaries (2–7). Functionally, CTCF- and cohesin-mediated looping and TADs play critical roles in multiple nuclear processes including regulation of gene expression, somatic recombination, and DNA repair (8). For example, TADs are thought to regulate gene expression by increasing the frequency of enhancer-promoter interactions within a TAD, and decreasing enhancer-promoter interactions between TADs (9). However, to understand how TADs and loops are formed and maintained, and how they function, it is necessary to understand whether TADs are stable or dynamic structures and to quantify the dynamics and lifetime of CTCF/cohesin-mediated loops. . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: ioRxiv preprint RESULTS As expected, the 3D distance was significantly increased in C65 (Fig. 1E). Interestingly, from Mean-Squared Displacement (MSD) analysis, chromatin showed Rouse polymer dynamics with a scaling of MSD~t0.5 for all three clones (24) (Fig. 1F). We conclude that our approach faithfully reports on CTCF looping dynamics in live cells without noticeable artifacts. To validate our system for tracking Fbn2 loop dynamics, we carried out a series of control experiments. First, we confirmed using Micro-C (22, 23) that our locus labeling approach did not measurably perturb the Fbn2 loop (Fig. 1A). Second, as a positive looping control we deleted To elucidate the specific contribution of CTCF and cohesin, we generated cell lines to acutely deplete the cohesin subunit RAD21, CTCF, and the cohesin unloader WAPL. We used genome-editing to endogenously tag these Fig. 1. Endogenous labeling and tracking of the Fbn2 loop with super-resolution live cell imaging. (A) Fluorescent labeling of Fbn2 loop anchors does not perturb the Fbn2 TAD. mESC Micro-C contact map comparing the parental untagged (C59, top left) and tagged (C36, bottom right) cell lines. Red triangles: CTCF binding with orientation. C36 ChIP-seq shows CTCF (GSM3508478) and cohesin (Smc1a; GSM3508477) binding as compared to Input (GSM3508475). RNA-seq (GSE123636) and transcript annotation tracks (GRCm38) show that Fbn2 is transcriptionally inactive. Genome coordinates: mm10. (B) Overview of tagging and readout using 3D distance. (C) Overview of the genome-edited cell lines (left) and a representative maximum intensity projection (MIP) of a cell showing two pairs of “dots” (right). (D) Representative 3D trajectory over time of a dot pair. MIPs of the 3D voxels centered on the mScarlet dot (top) and 3D distances between dots (bottom) are shown. (E) 3D distance probability density functions of dot pairs (n=32,171; n=46,163; n=13,566 distance measurements for C27, C36, C65 respectively) (F) Localization error corrected Mean Squared Displacement (MSD) plots (n=358; n=491; n=147 trajectories in C27; C36; C65 respectively). Fig. 1. Endogenous labeling and tracking of the Fbn2 loop with super-resolution live cell imaging. (A) Fluorescent labeling of Fbn2 loop anchors does not perturb the Fbn2 TAD. mESC Micro-C contact map comparing the parental untagged (C59, top left) and tagged (C36, bottom right) cell lines. Red triangles: CTCF binding with orientation. C36 ChIP-seq shows CTCF (GSM3508478) and cohesin (Smc1a; GSM3508477) binding as compared to Input (GSM3508475). RNA-seq (GSE123636) and transcript annotation tracks (GRCm38) show that Fbn2 is transcriptionally inactive. Genome coordinates: mm10. RESULTS To visualize the dynamics of CTCF/cohesin looping, we chose as our model system the loop holding together the two CTCF-bound boundaries of the 505 kb Fbn2 TAD in mouse Embryonic Stem Cells (mESCs). This TAD is verified to be CTCF dependent (20) and relatively simple as it contains a single gene, Fbn2, which is not expressed in mESCs (Fig. 1A). We used genome-editing to homozygously label the left and right CTCF sites of the Fbn2 TAD with TetO and Anchor3 arrays, which we then visualized by co-expressing the fluorescently tagged Though recent advances in single-cell genomics and fixed-cell imaging have made it possible to generate static snapshots of 3D genome structures in single cells (10–15), live-cell imaging is required to understand the 1 1 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT binding proteins TetR-3x-mScarlet and EGFP-OR3 (21) (clone C36) (Fig. 1B-D). We developed a comprehensive image analysis framework, ConnectTheDots, to extract trajectories of 3D loop anchor positions from the acquired movies (Fig. S1). By optimizing 3D super-resolution live- cell imaging conditions (16), we could track Fbn2 looping dynamics at 20 second resolution for over 2 hours (Fig. 1D). After DNA replication in S/G2 phase, it is no longer possible to reliably distinguish intrachromosomal from sister-chromosomal interactions (16). We therefore developed and validated a convolutional neural network to filter out replicated and low-quality dots (Fig. S2). Thus, we only consider G1 and early S-phase cells. the 505 kb between the CTCF sites, generating clone C27 (Fig. 1C). As expected, this significantly reduced the 3D distance (Fig. 1E). Third, as a negative control for CTCF- mediated looping, we generated clone C65 (Fig. 1C) by homozygously deleting the 3 CTCF motifs in the Fbn2 TAD (L1, L2, R1; Fig. 1A) and validated that this resulted in loss of CTCF binding and cohesin co-localization by ChIP-Seq (Fig. S3). RESULTS First, we achieved near-complete depletion of RAD21 and CTCF within 2 hours, while WAPL depletion took 4 hours and was less complete (Fig. S4). Second, long-term depletion of RAD21 and CTCF led to cell death as expected for loss of essential proteins (Fig. S5), whereas WAPL depletion caused only a minor growth defect and occasionally yielded visible compacted (‘vermicelli’) chromosomes after sustained depletion (26). Third, we quantified the protein abundances in the AID cell lines without IAA and note that they are lower likely due to leaky protein depletion (Fig. S6). Fourth, we used Micro-C to verify that RAD21 and CTCF depletion led to loss of the Fbn2 loop or corner peak as expected (27–30) (Fig. 2A) and used ChIP-Seq to verify disrupted CTCF and cohesin chromatin binding (Fig. S7). In contrast, WAPL depletion increased corner peak strength (29–31) (Fig. 2A). Thus, our validated AID lines enable efficient and acute protein depletion. increased corner peak strength (29–31) (Fig. 2A). Thus, our validated AID lines enable efficient and acute protein depletion. We next studied the specific roles of RAD21, CTCF, and WAPL in loop extrusion in vivo by imaging our validated AID lines (Fig. 2B-C). Consistent with RAD21 being required for loop extrusion, RAD21 depletion strongly increased the 3D distances (Fig. 2B-C). Consistent with CTCF being the boundary factor required for Fbn2 loop formation (Fig. 1B), but not for loop extrusion, CTCF depletion increased 3D distances albeit significantly less than RAD21 depletion (7) (Fig. 2B-C). Finally, consistent with prior observations that WAPL depletion increases cohesin’s residence time and abundance on chromatin (26), potentially allowing it to extrude longer and more stable loops (29, 31), WAPL depletion decreased the 3D distances (Fig. 2B-C; ΔWAPL is attenuated due to lower CTCF and Fig. 2. Degradation of CTCF, cohesin, and WAPL reveal their role in loop extrusion and looping-mediated chromosome compaction. (A) Micro-C data for the AID-tagged clones for RAD21 (left), CTCF (middle), and WAPL (right), showing control data (no IAA treatment; top half) and protein degradation data (3 hours post IAA; bottom half). Schematics illustrating the expected effect on DNA are shown above each subpanel. (B) Representative trajectories with (colored lines) or without IAA treatment (gray lines) for each AID-tagged clone. RESULTS (C) 3D distance probability density functions of dot pairs (n=45,379; n=10,469; n=18,153 distance measurements for ΔRAD21 (2 hr), ΔCTCF (2 hr), ΔWAPL (4 hr) depletion conditions respectively, and n=17,605; n=11,631; n=21,001 for the same clones without treatment). (D) Localization error corrected MSD plots for the AID-tagged clones (left) (n=537; n=137; n=215 trajectories in ΔRAD21 (2 hr), ΔCTCF (2 hr), ΔWAPL (4 hr) depletion conditions respectively, and n=183; n=151; n=257 without treatment (gray lines)). The effective tether length is obtained by computing the ratio of the steady- state variance of each clone to the value in the RAD21-depletion condition (note that 2<R²> is also the asymptotic value of the MSD; see also Supplementary Material). (E) Representative 3D polymer conformation from simulations mimicking the RAD21 (95% cohesin depletion) (left) and CTCF (100% CTCF depletion) (right) depletion conditions. Simulated chromatin segments in loops are colored red and unextruded segments are blue. (F) Mean cohesin separations and processivities in simulations are obtained by matching Micro-C contact frequency decay versus genomic distance. Conversion factors for simulated monomer sizes and time steps to nanometers and seconds are obtained by calibrating the ΔCTCF simulations to the 2-hour CTCF depletion MSD plots. The extrusion rate is for two-sided extrusion. 3 Fig. 2. Degradation of CTCF, cohesin, and WAPL reveal their role in loop extrusion and looping-mediated chromosome compaction. (A) Micro-C data for the AID-tagged clones for RAD21 (left), CTCF (middle), and WAPL (right), showing control data (no IAA Fig. 2. Degradation of CTCF, cohesin, and WAPL reveal their role in loop extrusion and looping-mediated chromosome , cohesin, and WAPL reveal their role in loop extrusion and looping-mediated chromosom Fig. 2. Degradation of CTCF, cohesin, and WAPL reveal their role in loop extrusion and Fig. 2. Degradation of CTCF, cohesin, and WAPL reveal their role in loop extrusion and looping mediated chromosome compaction. (A) Micro-C data for the AID-tagged clones for RAD21 (left), CTCF (middle), and WAPL (right), showing control data (no IAA treatment; top half) and protein degradation data (3 hours post IAA; bottom half). Schematics illustrating the expected effect on DNA are shown above each subpanel. (B) Representative trajectories with (colored lines) or without IAA treatment (gray lines) for each AID-tagged clone. RESULTS (B) Overview of tagging and readout using 3D distance. (C) Overview of the genome-edited cell lines (left) and a representative maximum intensity projection (MIP) of a cell showing two pairs of “dots” (right). (D) Representative 3D trajectory over time of a dot pair. MIPs of the 3D voxels centered on the mScarlet dot (top) and 3D distances between dots (bottom) are shown. (E) 3D distance probability density functions of dot pairs (n=32,171; n=46,163; n=13,566 distance measurements for C27, C36, C65 respectively) (F) Localization error corrected Mean Squared Displacement (MSD) plots (n=358; n=491; n=147 trajectories in C27; C36; C65 respectively). 2 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT factors with mAID in the C36 line, allowing for degradation with Indole-3-acetic acid (IAA) (25) (Fig. S4) and validated these cell lines. First, we achieved near-complete depletion of RAD21 and CTCF within 2 hours, while WAPL depletion took 4 hours and was less complete (Fig. S4). Second, long-term depletion of RAD21 and CTCF led to cell death as expected for loss of essential proteins (Fig. S5), whereas WAPL depletion caused only a minor growth defect and occasionally yielded visible compacted (‘vermicelli’) chromosomes after sustained depletion (26). Third, we quantified the protein abundances in the AID cell lines without IAA and note that they are lower likely due to leaky protein depletion (Fig. S6). Fourth, we used Micro-C to verify that RAD21 and CTCF depletion led to loss of the Fbn2 loop or corner peak as expected (27–30) (Fig. 2A) and used ChIP-Seq to verify disrupted CTCF and cohesin chromatin binding (Fig. S7). In contrast, WAPL depletion factors with mAID in the C36 line, allowing for degradation with Indole-3-acetic acid (IAA) (25) (Fig. S4) and validated these cell lines. ARTICLE PREPRINT cohesin abundances in this clone and less efficient depletion; Fig. S4, S6). contributing to the longer effective tether in ΔCTCF. The increased tether length after CTCF depletion is thus consistent with CTCF serving as an extrusion boundary. In summary, we find that on average just over half of the Fbn2 TAD is extruded into loops. contributing to the longer effective tether in ΔCTCF. The increased tether length after CTCF depletion is thus consistent with CTCF serving as an extrusion boundary. In summary, we find that on average just over half of the Fbn2 TAD is extruded into loops. To quantify the effect of loop extrusion in our AID lines we turned to polymer physics theory. The Rouse model predicts a linear relationship between chain length and mean squared distance (<R²>) between the fluorescent labels (dashed lines in Fig. 2D, Fig. S8). This relationship allows us to assign an “effective tether length” to each experimental condition by taking ΔRAD21 as a reference value, and assuming that ΔRAD21 represents the fully unextruded state with a genomic separation of 515 kb. We find an effective tether of ~200 kb in wildtype (C36) and ~280 kb in ΔCTCF, corresponding to ~39% and ~54% of the full genomic separation respectively. Thus, conversely, the genomic separation between the two labels shortens by ~46% due to extrusion alone (ΔRAD21 vs. ΔCTCF) and ~61% due to extrusion with boundaries (ΔRAD21 vs. C36). We refer to the latter percentages as the fraction (of the Fbn2 region) extruded but caution that extrusion of the loop anchors themselves complicates this picture, specifically Having demonstrated that CTCF and cohesin regulate Fbn2 compaction (Fig. 2A-D), we next sought to develop an inference method to quantify loop dynamics. To generate ground truth data to benchmark such a method, we used the Micro-C contact maps (Fig. 2A), the absolute 3D distances (Fig. 2C), and the MSDs (Fig. 2D) in the C36, ΔCTCF, and ΔRAD21 conditions to constrain and parameterize 3D polymer simulations incorporating loop extrusion. Consistent with our ΔRAD21 data, our polymer simulations resulted in chromosome decompaction after near-complete RAD21 depletion (Fig. 2E) and accurately matched our experimental data (Fig. 2F). We then used these simulations to benchmark a method to infer looped states. To infer looped state dynamics, we developed To infer looped state dynamics, we developed Fig. 3. Bayesian Inference of Looping Dynamics (BILD) reveals rare and dynamic CTCF loops. ARTICLE PREPRINT (A) Example trajectory from polymer simulations with loop extrusion. Extrusion shortens the effective tether (unextruded length; red) between the CTCF sites; The tether is minimal when cohesin is stalled at CTCF on both sides, which we take as ground truth looping events (black bar). The inference captures these accurately, but raises false positive detections occasionally (pink bars). We limit false positive detections by introducing an evidence bias ΔE (purple bar). (B) Schematic overview over BILD. Building on the analytical solution to the Rouse model, we employ an evidence-based optimization scheme to determine the optimal looping profile in two steps. (C) BILD procedure. In a first step, we maximize the evidence over the number of switches in the profile. In the second step we then find the best positions for these switches by maximizing the posterior. (D) Illustrative examples of inferred profiles on real trajectory data. (E) Kaplan-Meier survival curves rescaled by the inferred looped fraction. Gray lines are maximum likelihood fits of a single exponential to the data, accounting for censoring (Supplementary Material). (F) Fraction of time the Fbn2-locus spends in the fully looped conformation for each of our experimental conditions. Error bars are bootstrapped 95% confidence intervals. (G) Median loop lifetimes from the Kaplan-Meier survival curves (squares) or exponential fits (crosses). Confidence intervals are determined from the confidence intervals on the Kaplan-Meier curve and the likelihood function of the exponential fit, respectively. Where the upper confidence limit on the survival curve did not cross below 50% an arrowhead indicates a semi-infinite confidence interval. 4 Fig. 3. Bayesian Inference of Looping Dynamics (BILD) reveals rare and dynamic CTCF loops. (A) Example trajectory from polymer simulations with loop extrusion. Extrusion shortens the effective tether (unextruded length; red) between the CTCF sites; The tether is minimal when cohesin is stalled at CTCF on both sides, which we take as ground truth looping events (black bar). The inference captures these accurately, but raises false positive detections occasionally (pink bars). We limit false positive detections by introducing an evidence bias ΔE (purple bar). (B) Schematic overview over BILD. Building on the analytical solution to the Rouse model, we employ an evidence-based optimization scheme to determine the optimal looping profile in two steps. (C) BILD procedure. In a first step, we maximize the evidence over the number of switches in the profile. RESULTS (C) 3D distance probability density functions of dot pairs (n=45,379; n=10,469; n=18,153 distance measurements for ΔRAD21 (2 hr), ΔCTCF (2 hr), ΔWAPL (4 hr) depletion conditions respectively, and n=17,605; n=11,631; n=21,001 for the same clones without treatment). (D) Localization error corrected MSD plots for the AID-tagged clones (left) (n=537; n=137; n=215 trajectories in ΔRAD21 (2 hr), ΔCTCF (2 hr), ΔWAPL (4 hr) depletion conditions respectively, and n=183; n=151; n=257 without treatment (gray lines)). The effective tether length is obtained by computing the ratio of the steady- state variance of each clone to the value in the RAD21-depletion condition (note that 2<R²> is also the asymptotic value of the MSD; see also Supplementary Material). (E) Representative 3D polymer conformation from simulations mimicking the RAD21 (95% cohesin depletion) (left) and CTCF (100% CTCF depletion) (right) depletion conditions. Simulated chromatin segments in loops are colored red and unextruded segments are blue. (F) Mean cohesin separations and processivities in simulations are obtained by matching Micro-C contact frequency decay versus genomic distance. Conversion factors for simulated monomer sizes and time steps to nanometers and seconds are obtained by calibrating the ΔCTCF simulations to the 2-hour CTCF depletion MSD plots. The extrusion rate is for two-sided extrusion. 3 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT ARTICLE PREPRINT We then infer when the switches occur by maximum a posterior estimation (Fig. 3C). To reduce false positives, we introduced an evidence margin ΔE, which reduces the sensitivity of loop detection but renders the detected looping segments more reliable (Fig. 3A, Fig. S10). Our final inference scheme, BILD, accurately inferred both the looped fraction and loop lifetime when applied to our 3D polymer simulation data with experimentally realistic localization uncertainty (Fig. S10-11). We next used BILD to infer looping in our experimental trajectory data (Fig. 3D-G). BILD revealed that the Fbn2 TAD is fully looped ~6.5% (~3%) of the time, but spends ~93.5% (97%) of the time in a fully Wondering how to reconcile a low looped fraction of ~3% with a clear and strong corner peak in the Micro-C map, we set up polymer simulations with loop extrusion. We found that CTCF-mediated stabilization of cohesin was Fig. 4. Comprehensive picture of the Fbn2 TAD. (A) Comparison of Micro-C data for the C36 line to in silico Micro-C of our best-fit simulation, map (left) and contact probability scaling (right). (B) BILD applied to the same simulation (green), comparing to C36 experimental data (blue). (C) Number of cohesins forming the looped state in the simulation (n = 18,789). (D) “Anatomy” of the Fbn2 TAD. Quantitative description of the Fbn2 TAD using both real data (blue) and our best-fit simulation (green). Cohesin processivity and density and CTCF stalling probability and lifetime boost are simulation parameters. Fraction of time in different conformations was extracted from simulation ground truth, using effective tether lengths of 1.1 kb and 505 kb as cutoffs to define “fully looped” and “fully unlooped” respectively. Fraction of TAD unextruded was obtained as mean tether length over the full simulation. Experimental values are from Figs. 2, 3. Fig. 4. Comprehensive picture of the Fbn2 TAD. (A) Comparison of Micro-C data for the C36 line to in silico Micro-C of our best-fit simulation, map (left) and contact probability scaling (right). (B) BILD applied to the same simulation (green), comparing to C36 experimental data (blue). (C) Number of cohesins forming the looped state in the simulation (n = 18,789). (D) “Anatomy” of the Fbn2 TAD. Quantitative description of the Fbn2 TAD using both real data (blue) and our best-fit simulation (green). Cohesin processivity and density and CTCF stalling probability and lifetime boost are simulation parameters. ARTICLE PREPRINT unlooped or partially extruded conformation (Fig. 3F). We use brackets to indicate the looped fraction after false positive correction (Fig. S11). In contrast, we observed a minimal looped fraction of ~2% (~0%) in ΔRAD21 and ΔCTCF, and ~4% (~1%) in C65, whereas the looped fraction was significantly increased to ~10% (~6%) in ΔWAPL, consistent with WAPL unloading cohesin from chromatin (26). Finally, we estimated the lifetime of the looped state (Fig. 3E, G). Lifetime estimation is challenging due to ‘censoring’, which occurs if the trajectory either begins or ends in the looped state. To correct for censoring, we used the Kaplan-Meier estimator of the survival probability (Fig. 3G). We also provide an orthogonal estimate of the median lifetime from an exponential fit, which generally agrees with the Kaplan- Meier estimate. Together, these give an estimate of the median loop lifetime of ~10-30 min in the wildtype C36 line (Fig. 3G, Fig. S11D). Though our estimates are associated with uncertainty and though we cannot exclude the existence of a very rare but long-lived loop state, these results nevertheless reveal the fully looped CTCF-cohesin complex state to be both rare (~3%) and quite dynamic (median ~10-30 min; mean ~15-45 min). Thus, during an average ~12 hour mESC cell cycle, the looped state will occur ~1-2 times lasting cumulatively ~20-45 min, but the remaining ~11.5 hours will be in the partially extruded or unlooped state. Bayesian Inference of Looping Dynamics (BILD). We coarse-grain the possible conformations of the two CTCF sites into two states: 1) a state of sustained contact (the ‘looped state’), where presumably a CTCF-cohesin complex holds together the Fbn2 loop and 2) all other possible conformations including partially extruded conformations, random contacts, and the fully unlooped conformation (the ‘unlooped state’; Fig. 3A-B, Fig. S9). We model the unlooped state (labelled θ=0) as a free Rouse polymer and calibrate it to our ΔCTCF data. For the looped state we introduce a switchable bond between the two CTCF sites (θ=1; Fig. 3B), whose strength we set to reproduce the 10 kb distance between the fluorophores, using ΔRAD21 as reference for a free 515 kb chain. Thus, our biological controls allow us to define the looped state as the state that disappears in the ΔCTCF condition. We infer the number of switches k between the two states by maximizing the model evidence E. ARTICLE PREPRINT In the second step we then find the best positions for these switches by maximizing the posterior. (D) Illustrative examples of inferred profiles on real trajectory data. (E) Kaplan-Meier survival curves rescaled by the inferred looped fraction. Gray lines are maximum likelihood fits of a single exponential to the data, accounting for censoring (Supplementary Material). (F) Fraction of time the Fbn2-locus spends in the fully looped conformation for each of our experimental conditions. Error bars are bootstrapped 95% confidence intervals. (G) Median loop lifetimes from the Kaplan-Meier survival curves (squares) or exponential fits (crosses). Confidence intervals are determined from the confidence intervals on the Kaplan-Meier curve and the likelihood function of the exponential fit, respectively. Where the upper confidence limit on the survival curve did not cross below 50% an arrowhead indicates a semi-infinite confidence interval. 4 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT ARTICLE PREPRINT Fraction of time in different conformations was extracted from simulation ground truth, using effective tether lengths of 1.1 kb and 505 kb as cutoffs to define “fully looped” and “fully unlooped” respectively. Fraction of TAD unextruded was obtained as mean tether length over the full simulation. Experimental values are from Figs. 2, 3. 5 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT other similarly sized loops in mESCs are likely weaker than Fbn2 (Fig. S14). Our results thus rule out static models of TADs, where TADs exist in either a fully unlooped state or a fully looped state stably bridged by one cohesin (Fig. 1B). Instead, we show that the Fbn2 TAD most often exists in a partially extruded state formed by a few cohesins in live cells (~92%; Fig. 4D), and that when the rare looped state is formed, it is both transient (~10-30 min median lifetime; Fig. 4B) and sometimes bridged by multiple cohesins (Fig. 4C). Overall, frequent cohesin-mediated contacts within a TAD rather than rare CTCF-CTCF loops may therefore be more important for regulatory interactions, such as those between enhancers and promoters. Thus, instead of stable loops, we observe a much more dynamic and transitory picture of TADs in live cells (Fig. 4D), which may also help explain cell-to-cell variation in 3D genome structure, and consequently stochasticity in downstream processes such as gene expression and cell differentiation. necessary to reproduce both these features in our simulations (Fig. 4, S12; Supplementary Material), consistent with recent reports (32, 33) (Fig. 4, S12; Supplementary Material). We confirmed this effect using iFRAP of cohesin, finding that CTCF depletion decreases cohesin’s residence time (Fig. S13). Incorporating this effect, we then simulated loop extrusion with a cohesin density of 1/240 kb and processivity of 150 kb (processivity = lifetime * extrusion speed). Acknowledgements We thank Drs. Robert Tjian and Xavier Darzacq for hosting early parts of this work, Dr. Jeffrey Alexander for providing the RNA destabilization elements, Lydia Joh for assistance with cloning, and the Hansen, Zechner, and Mirny labs for insightful discussions. Funding: This work was supported by NIH grants R00GM130896 (ASH), DP2GM140938 (ASH), R33CA257878 (ASH), UM1HG011536 (ASH and LM), R01GM114190 (LM), NSF grant 2036037 (ASH), the Mathers’ Foundation (ASH), a Pew- Stewart Cancer Research Scholar grant (ASH), Chaires d'excellence Internationale Blaise Pascal (LM), American-Italian Cancer Foundation research fellowship (MG), and core funding from the Max Planck Institute of Molecular Cell Biology and Genetics (CZ). Author contributions: ASH conceived and initiated the project. HBB, MG, SGH, LM, CZ, ASH designed the project. ASH performed genome-editing and generated the cell lines. GMD cloned plasmids. MG, AJ, CC, and ASH characterized and validated the cell lines. THSH performed Micro-C. CC performed ChIP-Seq. MG, AJ, and HBB optimized imaging experiments with input from ASH. MG, AJ, and HBB collected the imaging data with acquisition led by MG and AJ. MG and AJ performed control experiments and characterized the AID cell lines. HBB developed image processing pipeline, the CNN, and analyzed the imaging data with input from ASH, SGH, MG, and AJ. HBB performed polymer simulations with input from SGH and LM. MG, AJ, HBB, and ASH annotated trajectory data. SGH and CZ designed BILD with input from HBB, LM, and ASH. SGH developed and benchmarked BILD, applied BILD to trajectory data, and developed MSD analysis with input from HBB and CZ. HBB and SGH analyzed polymer simulations. ASH, LM, and CZ supervised the project. HBB, MG, SGH, AJ, and ASH drafted the manuscript and figures. All authors edited the manuscript and figures. Competing interests: Authors declare that they have no competing interests. Data and materials availability: All the data and code are available as described in the supplementary materials. The GEO accession number is GSE187487. ARTICLE PREPRINT When cohesin reaches a CTCF site, it has a probability of 12.5% to stall, which, using the estimate of 50% CTCF occupancy (34), translates to a ~25% capture efficiency of CTCF. Once stalled by CTCF, cohesin is stabilized 4-fold beyond its base lifetime of ~20 min (35) (Fig. S13), facilitating the formation of longer loops. These simulations reproduced both our experimental Micro-C maps (Fig. 4A) and the median loop lifetime and low looped fraction (Fig. 4B). Together, these results allow us to paint a comprehensive mechanistic picture of the Fbn2 TAD (Fig. 4C-D): most of the time (~92%), the TAD is partially extruded, with ~57-61% of the Fbn2 region captured in ~1- 3 extruding cohesin loops, while ~39-43% remain unextruded. The fully unlooped conformation, as it would be found in the absence of cohesin, occurs only ~6% of the time, while the fully looped state is even more rare at ~3% (~2% in simulations) and has a median lifetime of ~10-30 min. Interestingly, our simulations reveal that the looped state is sometimes held together by multiple cohesins (Fig. 4C), which also explains why the loop lifetime can be substantially shorter than the CTCF-stabilized cohesin lifetime (Fig. S13). Nevertheless, we stress that both the mechanistic assumptions of our polymer simulations and the experimental data constraining them are associated with uncertainty, resulting in uncertainty of the inferred parameters (Fig. 4D). We also note that TADs smaller than the 505 kb Fbn2 TAD as well as TADs with stronger CTCF boundaries may have a higher looped fraction (Giorgetti and co-workers, personal communication). Furthermore, we propose that our absolute quantification of the Fbn2 looped fraction may now allow calibrated inference of absolute looped fractions genome-wide, based on Micro-C (15). REFERENCES Renom, F. Bantignies, G. Cavalli, M. Nollmann, Single-cell absolute contact probability detection reveals chromosomes are organized by multiple low-frequency yet specific interactions. 1. J. A. Beagan, J. E. Phillips-Cremins, On the existence and functionality of topologically associating domains. Nature Genetics (2020), doi:10.1038/s41588-019-0561-1. Nature Communications. 8, 1753 (2017). Nature Communications. 8, 1753 (2017). 16. H. B. Brandão, M. Gabriele, A. S. Hansen, Tracking and interpreting long-range chromatin interactions with super- resolution live-cell imaging. Current Opinion in Cell Biology. 70, 18–26 (2021). 2. M. Ganji, I. A. Shaltiel, S. Bisht, E. Kim, A. Kalichava, C. H. Haering, C. Dekker, Real-time imaging of DNA loop extrusion by condensin. Science. 360, 102–105 (2018). C. H. Haering, C. Dekker, Real-time imaging of DNA loop extrusion by condensin. Science. 360, 102–105 (2018). p g g g p resolution live-cell imaging. Current Opinion in Cell Biology. 70, 18–26 (2021). 3. S. Golfier, T. Quail, H. Kimura, J. Brugués, Cohesin and condensin extrude DNA loops in a cell cycle-dependent manner. eLife. 9, e53885 (2020). 17. H. Chen, M. Levo, L. Barinov, M. Fujioka, J. B. Jaynes, T. Gregor, Dynamic interplay between enhancer–promoter topology and gene activity. Nature Genetics. 50, 1296–1303 (2018). 4. A. S. Hansen, CTCF as a boundary factor for cohesin- mediated loop extrusion: evidence for a multi-step mechanism. Nucleus. 11, 132–148 (2020). 18. J. M. Alexander, J. Guan, B. Li, L. Maliskova, M. Song, g Y. Shen, B. Huang, S. Lomvardas, O. D. Weiner, Live-cell imaging reveals enhancer-dependent Sox2 transcription in the absence of enhancer proximity. eLife. 8, e41769 (2019). 5. G. Fudenberg, M. Imakaev, C. Lu, A. Goloborodko, N. Abdennur, L. A. Mirny, Formation of Chromosomal Domains by Loop Extrusion. Cell Reports (2016), doi:10.1016/j.celrep.2016.04.085. 19. N. Khanna, Y. Zhang, J. S. Lucas, O. K. Dudko, C. Murre, Chromosome dynamics near the sol-gel phase transition dictate the timing of remote genomic interactions. Nature 6. Y. Kim, Z. Shi, H. Zhang, I. J. Finkelstein, H. Yu, H h i DNA b l i S i 36 6. Y. Kim, Z. Shi, H. Zhang, I. J. Finkelstein, H. Yu, Human cohesin compacts DNA by loop extrusion. Science. 366, 1345–1349 (2019). , ( ) 20. E. de Wit, E. S. M. Vos, S. J. B. Holwerda, C. Valdes- Quezada, M. J. A. M. Verstegen, H. Teunissen, E. Splinter, P. J. Wijchers, P. H. L. Krijger, W. de Laat, CTCF Binding Polarity Determines Chromatin Looping. Molecular Cell. 60, 676–684 (2015). REFERENCES ( ) 7. I. F. Davidson, B. Bauer, D. Goetz, W. Tang, G. Wutz, J.-M. Peters, DNA loop extrusion by human cohesin. Science. 366, 1338–1345 (2019). 8. A. M. Oudelaar, D. R. Higgs, The relationship between genome structure and function. Nature Reviews Genetics. 22, 154–168 (2021). 21. T. Germier, S. Kocanova, N. Walther, A. Bancaud, H. A. Shaban, H. Sellou, A. Z. Politi, J. Ellenberg, F. Gallardo, K. 9. O. Symmons, V. V. Uslu, T. Tsujimura, S. Ruf, S. Nassari, W. Schwarzer, L. Ettwiller, F. Spitz, Functional and topological characteristics of mammalian regulatory domains. G R h (2014) d i 10 1101/ 163519 113 9. O. Symmons, V. V. Uslu, T. Tsujimura, S. Ruf, S. Nassari, W. Schwarzer, L. Ettwiller, F. Spitz, Functional and topological characteristics of mammalian regulatory domains. Genome Research (2014), doi:10.1101/gr.163519.113. 10. M. V Arrastia, J. W. Jachowicz, N. Ollikainen, M. S. , , , g, , Bystricky, Real-Time Imaging of a Single Gene Reveals Transcription-Initiated Local Confinement. Biophysical Journal. 113, 1383–1394 (2017). 22. T.-H. S. Hsieh, C. Cattoglio, E. Slobodyanyuk, A. S. Hansen, O. J. Rando, R. Tjian, X. Darzacq, Resolving the 3D Landscape of Transcription-Linked Mammalian Chromatin Folding Molecular Cell (2020) Bystricky, Real-Time Imaging of a Single Gene Reveals Transcription-Initiated Local Confinement. Biophysical Journal. 113, 1383–1394 (2017). Genome Research (2014), doi:10.1101/gr.163519.113. 10. M. V Arrastia, J. W. Jachowicz, N. Ollikainen, M. S. ( ) 22. T.-H. S. Hsieh, C. Cattoglio, E. Slobodyanyuk, A. S. Hansen, O. J. Rando, R. Tjian, X. Darzacq, Resolving the 3D Landscape of Transcription-Linked Mammalian Chromatin Curtis, C. Lai, S. A. Quinodoz, D. A. Selck, R. F. Ismagilov, M. Guttman, Single-cell measurement of higher-order 3D genome organization with scSPRITE. Nature Biotechnology (2021), doi:10.1038/s41587-021-00998-1. Folding. Molecular Cell (2020), doi:https://doi.org/10.1016/j.molcel.2020.03.002. 23. A. S. Hansen, T.-H. S. Hsieh, C. Cattoglio, I. Pustova, R. Saldaña-Meyer, D. Reinberg, X. Darzacq, R. Tjian, Distinct Classes of Chromatin Loops Revealed by Deletion of an RNA- Binding Region in CTCF. Molecular Cell (2019), doi:https://doi.org/10.1016/j.molcel.2019.07.039. 11. S. Quentin, J. Daniel, C. Jia-Ming, C. D. I., P. G. L., B. Boyan, S. Tom, G. Julian, J. Caroline, N. Marcelo, B. Frédéric, C 11. S. Quentin, J. Daniel, C. Jia-Ming, C. D. I., P. G. L., B. Boyan, S. Tom, G. Julian, J. Caroline, N. Marcelo, B. Frédéric, C. Giacomo, TADs are 3D structural units of higher-order chromosome organization in Drosophila. Science Advances. 4, eaar8082 (2021). DISCUSSION Our findings reveal the CTCF/cohesin-mediated looped state that holds together CTCF boundaries of TADs to be rare, dynamic, and transient. A key limitation of our study is that it represents just one loop in one cell type. Nevertheless, the Fbn2 loop is among the strongest quartile of “corner peaks” in Micro-C maps, suggesting that most 6 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT REFERENCES 12 T J Stevens D Lando S Basu L P Atkinson Y Cao 11. S. Quentin, J. Daniel, C. Jia-Ming, C. D. I., P. G. L., B. Boyan, S. Tom, G. Julian, J. Caroline, N. Marcelo, B. Frédéric, C. Giacomo, TADs are 3D structural units of higher-order chromosome organization in Drosophila. Science Advances. 4, eaar8082 (2021). 12. T. J. Stevens, D. Lando, S. Basu, L. P. Atkinson, Y. Cao, S. F. Lee, M. Leeb, K. J. Wohlfahrt, W. Boucher, A. O’Shaughnessy Kirwan J Cramard A J Faure M Ralser E Giacomo, TADs are 3D structural units of higher-order chromosome organization in Drosophila. Science Advances. 4, eaar8082 (2021). 24. H. Hajjoul, J. Mathon, H. Ranchon, I. Goiffon, J. Mozziconacci, B. Albert, P. Carrivain, J.-M. Victor, O. Gadal, K. Bystricky, A. Bancaud, High-throughput chromatin motion tracking in living yeast reveals the flexibility of the fiber throughout the genome. Genome Research . 23, 1829–1838 (2013). ( ) 12. T. J. Stevens, D. Lando, S. Basu, L. P. Atkinson, Y. Cao, S F Lee M Leeb K J Wohlfahrt W Boucher A 12. T. J. Stevens, D. Lando, S. Basu, L. P. Atkinson, Y. Cao, S. F. Lee, M. Leeb, K. J. Wohlfahrt, W. Boucher, A. O’Shaughnessy-Kirwan, J. Cramard, A. J. Faure, M. Ralser, E. S. F. Lee, M. Leeb, K. J. Wohlfahrt, W. Boucher, A. O’Shaughnessy-Kirwan, J. Cramard, A. J. Faure, M. Ralser, E. Blanco, L. Morey, M. Sansó, M. G. S. Palayret, B. Lehner, L. Di Croce, A. Wutz, B. Hendrich, D. Klenerman, E. D. Laue, 3D Blanco, L. Morey, M. Sansó, M. G. S. Palayret, B. Lehner, L. Di Croce, A. Wutz, B. Hendrich, D. Klenerman, E. D. Laue, 3D structures of individual mammalian genomes studied by single- cell Hi-C. Nature. 544, 59–64 (2017). structures of individual mammalian genomes studied by single- cell Hi-C. Nature. 544, 59–64 (2017). 25. T. Natsume, T. Kiyomitsu, Y. Saga, M. T. Kanemaki, Rapid Protein Depletion in Human Cells by Auxin-Inducible Degron Tagging with Short Homology Donors. Cell Reports. 15, 210–218 (2016). 25. T. Natsume, T. Kiyomitsu, Y. Saga, M. T. Kanemaki, Rapid Protein Depletion in Human Cells by Auxin-Inducible Degron Tagging with Short Homology Donors. Cell Reports. 15, 210–218 (2016). 26. A. Tedeschi, G. Wutz, S. Huet, M. Jaritz, A. Wuensche, E. Schirghuber, I. F. Davidson, W. Tang, D. A. Cisneros, V. Bhaskara, T. Nishiyama, A. Vaziri, A. Wutz, J. Ellenberg, J.-M. ARTICLE PREPRINT Aiden, Cohesin Loss Eliminates All Loop Domains. Cell (2017), doi:10.1016/j.cell.2017.09.026. Factor WAPL Restricts Chromatin Loop Extension. Cell. 169, 693-707.e14 (2017). doi:10.1016/j.cell.2017.09.026. 28. E. P. Nora, A. Goloborodko, A. L. Valton, J. H. Gibcus, A. Uebersohn, N. Abdennur, J. Dekker, L. A. Mirny, B. G. Bruneau, Targeted Degradation of CTCF Decouples Local Insulation of Chromosome Domains from Genomic Compartmentalization. Cell (2017), doi:10.1016/j.cell.2017.05.004. ( ) 32. Y. Li, J. H. I. Haarhuis, Á. Sedeño Cacciatore, R. Oldenkamp, M. S. van Ruiten, L. Willems, H. Teunissen, K. W. Muir, E. de Wit, B. D. Rowland, D. Panne, The structural basis for cohesin–CTCF-anchored loops. Nature. 578, 472–476 (2020). 33. G. Wutz, R. Ladurner, B. G. St Hilaire, R. R. Stocsits, K. Nagasaka, B. Pignard, A. Sanborn, W. Tang, C. Várnai, M. P. Ivanov, S. Schoenfelder, P. van der Lelij, X. Huang, G. , , , y, Bruneau, Targeted Degradation of CTCF Decouples Local Insulation of Chromosome Domains from Genomic Compartmentalization. Cell (2017), doi:10.1016/j.cell.2017.05.004. 29. G. Wutz, C. Várnai, K. Nagasaka, D. A. Cisneros, R. R. 29. G. Wutz, C. Várnai, K. Nagasaka, D. A. Cisneros, R. R. Stocsits, W. Tang, S. Schoenfelder, G. Jessberger, M. Muhar, M. J. Hossain, N. Walther, B. Koch, M. Kueblbeck, J. Ellenberg, J. Zuber, P. Fraser, J. Peters, Topologically associating domains and h ti l d d h i d l t d b CTCF 29. G. Wutz, C. Várnai, K. Nagasaka, D. A. Cisneros, R. R. Stocsits, W. Tang, S. Schoenfelder, G. Jessberger, M. Muhar, M. J. Hossain, N. Walther, B. Koch, M. Kueblbeck, J. Ellenberg, J. Dürnberger, E. Roitinger, K. Mechtler, I. F. Davidson, P. Fraser, E. Lieberman-Aiden, J.-M. Peters, ESCO1 and CTCF enable formation of long chromatin loops by protecting cohesinSTAG1 from WAPL. eLife. 9, e52091 (2020). Stocsits, W. Tang, S. Schoenfelder, G. Jessberger, M. Muhar, M. J. Hossain, N. Walther, B. Koch, M. Kueblbeck, J. Ellenberg, J. Zuber, P. Fraser, J. Peters, Topologically associating domains and chromatin loops depend on cohesin and are regulated by CTCF, Zuber, P. Fraser, J. Peters, Topologically associating domains and chromatin loops depend on cohesin and are regulated by CTCF, WAPL, and PDS5 proteins. The EMBO Journal (2017), doi:10.15252/embj.201798004. 34. C. Cattoglio, I. Pustova, N. Walther, J. J. Ho, M. Hantsche-Grininger, C. J. Inouye, M. J. Hossain, G. M. Dailey, J. Ellenberg, X. Darzacq, R. Tjian, A. S. Hansen, Determining cellular CTCF and cohesin abundances to constrain 3D genome models. REFERENCES Peters, Wapl is an essential regulator of chromatin structure and chromosome segregation N t 501 564 568 (2013) 13. E. H. Finn, G. Pegoraro, H. B. Brandão, A.-L. Valton, M. E. Oomen, J. Dekker, L. Mirny, T. Misteli, Extensive Heterogeneity and Intrinsic Variation in Spatial Genome Organization. Cell. 176, 1502-1515.e10 (2019). 26. A. Tedeschi, G. Wutz, S. Huet, M. Jaritz, A. Wuensche, E. Schirghuber, I. F. Davidson, W. Tang, D. A. Cisneros, V. Bhaskara, T. Nishiyama, A. Vaziri, A. Wutz, J. Ellenberg, J.-M. Peters, Wapl is an essential regulator of chromatin structure and chromosome segregation. Nature. 501, 564–568 (2013). 14. B. Bintu, L. J. Mateo, J.-H. Su, N. A. Sinnott-Armstrong, M. Parker, S. Kinrot, K. Yamaya, A. N. Boettiger, X. Zhuang, Super-resolution chromatin tracing reveals domains and cooperative interactions in single cells. Science. 362, eaau1783 (2018). 27. S. S. P. Rao, S. C. Huang, B. Glenn St Hilaire, J. M. Engreitz, E. M. Perez, K. R. Kieffer-Kwon, A. L. Sanborn, S. E. Johnstone, G. D. Bascom, I. D. Bochkov, X. Huang, M. S. Shamim, J. Shin, D. Turner, Z. Ye, A. D. Omer, J. T. Robinson, T. Schlick, B. E. Bernstein, R. Casellas, E. S. Lander, E. L. 15. D. I. Cattoni, A. M. Cardozo Gizzi, M. Georgieva, M. Di Stefano, A. Valeri, D. Chamousset, C. Houbron, S. Déjardin, J.- B. Fiche, I. González, J.-M. Chang, T. Sexton, M. A. Marti- 7 . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint . CC-BY 4.0 International license available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprint this version posted December 13, 2021. ; https://doi.org/10.1101/2021.12.12.472242 doi: bioRxiv preprint ARTICLE PREPRINT eLife. 8, e40164 (2019). g , y , , y, Ellenberg, X. Darzacq, R. Tjian, A. S. Hansen, Determining cellular CTCF and cohesin abundances to constrain 3D genome models. eLife. 8, e40164 (2019). 30. T.-H. S. Hsieh, C. Cattoglio, E. Slobodyanyuk, A. S. Hansen, X. Darzacq, R. Tjian, bioRxiv, in press, doi:10.1101/2021.07.14.452365. 30. T.-H. S. Hsieh, C. Cattoglio, E. Slobodyanyuk, A. S. Hansen, X. Darzacq, R. Tjian, bioRxiv, in press, doi:10.1101/2021.07.14.452365. 31. J. H. I. Haarhuis, R. H. van der Weide, V. A. Blomen, J. O. Yáñez-Cuna, M. Amendola, M. S. van Ruiten, P. H. L. Krijger, H. Teunissen, R. H. Medema, B. van Steensel, T. R. Brummelkamp E de Wit B D Rowland The Cohesin Release 35. A. S. Hansen, I. Pustova, C. Cattoglio, R. Tjian, X. Darzacq, CTCF and cohesin regulate chromatin loop stability with distinct dynamics. eLife. 6, e25776 (2017). O. Yáñez-Cuna, M. Amendola, M. S. van Ruiten, P. H. L. Krijger, H. Teunissen, R. H. Medema, B. van Steensel, T. R. Brummelkamp, E. de Wit, B. D. Rowland, The Cohesin Release 8
https://openalex.org/W3013383869	https://www.nature.com/articles/s41467-020-19510-5.pdf	English	null	Organized cannabinoid receptor distribution in neurons revealed by super-resolution fluorescence imaging	bioRxiv (Cold Spring Harbor Laboratory)	2,020	cc-by	13,838	1 iHuman Institute, ShanghaiTech University, 201210 Shanghai, China. 2 University of the Chinese Academy of Sciences, 100049 Beijing, China. 3 School of Life Science and Technology, ShanghaiTech University, 201210 Shanghai, China. 4 Department of Medical Neuroscience, Dalhousie University, Halifax, NS B3H 4R2, Canada. 5These authors contributed equally: Hui Li, Jie Yang, Cuiping Tian. ✉email: zhongsh@shanghaitech.edu.cn NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications ARTICLE ARTICLE Organized cannabinoid receptor distribution in neurons revealed by super-resolution ﬂuorescence imaging The discovery of the MPS, and the many studies that have characterized cellular structures across different cell types9,10, demonstrate the power of super-resolution imaging for uncovering intracellular structures at the nano-scale. Recently, using a type of super-resolution imaging called STORM, Zhou et al.11 proposed a model where CB1 forms a periodic pattern when activated. They found a ~190 nm periodic pattern of CB1 in cultured hippocampal neurons exclusively under the adminis- tration of agonists. A previous study, using a similar imaging technology, also revealed distinct CB1 structures across different cell types in the brain, but showed no sign of a periodic pattern12. Thus, the intracellular organization of CB1 in neurons remains unclear. To avoid artifacts caused by ﬁxation procedures, we performed live super-resolution imaging on cultured neurons. To this end, we used structured illumination microscopy (SIM), a type of super-resolution imaging that is suitable for investigating the live structure of cellular molecules at high spatial resolution15. Using SIM microscopy, with a spatial resolution around 122 nm (Supplementary Fig. 2e), we revealed clusters of CB1 in cultured neurons, and those clusters appeared highly organized as hotspots in some regions of axons (Supplementary Fig. 2f–h). Again, CB1 exhibited both high and low periodic hotspots (Supplemen- tary Fig. 2f–h). The spatial distance between periodic hotspots was ~190 nm, which was comparable to our results with antibody labeling in ﬁxed preparations (Fig. 1g). Notably, the regular- ity of the CB1 structure in cultured neurons was higher than that in the brain (Fig. 1h). Therefore, we conclude that there is a commonality of the semi-periodic feature of CB1 in neurons, an unexpected characteristic of GPCRs in the cellular membrane. Herein, by employing another type of super-resolution ima- ging, called stimulated emission depletion (STED), we system- atically investigated the nano-structure of CB1 and other GPCRs in brain tissues and primary cultured neurons. Our results revealed a periodic structure of CB1 clusters along the axons of inhibitory interneurons. These CB1 clusters were organized into “hotspots” ~190 nm apart. Using dual-color STED imaging and cellular biology techniques, we demonstrated that the CB1 hot- spots were associated with the MPS. The CB1 hotspots exhibited conﬁned dynamics, which were reduced by receptor activation. Thus, our current studies demonstrate that the ~190 nm periodic structure of the cytoskeleton appears to be the backbone for intracellular signaling to occur. CB1 is associated with MPS in neurons. Organized cannabinoid receptor distribution in neurons revealed by super-resolution ﬂuorescence imaging Hui Li1,5, Jie Yang1,2,3,5, Cuiping Tian1,5, Min Diao1, Quan Wang1,2,3, Simeng Zhao 1, Shanshan Li1, Fangzhi Tan 1, Tian Hua 1, Ya Qin3, Chao-Po Lin3, Dylan Deska-Gauthier4, Garth J. Thompson 1, Ying Zhang4, Wenqing Shui1,3, Zhi-Jie Liu 1,3, Tong Wang3 & Guisheng Zhong 1,3✉ G-protein-coupled receptors (GPCRs) play important roles in cellular functions. However, their intracellular organization is largely unknown. Through investigation of the cannabinoid receptor 1 (CB1), we discovered periodically repeating clusters of CB1 hotspots within the axons of neurons. We observed these CB1 hotspots interact with the membrane-associated periodic skeleton (MPS) forming a complex crucial in the regulation of CB1 signaling. Fur- thermore, we found that CB1 hotspot periodicity increased upon CB1 agonist application, and these activated CB1 displayed less dynamic movement compared to non-activated CB1. Our results suggest that CB1 forms periodic hotspots organized by the MPS as a mechanism to increase signaling efﬁcacy upon activation. 1 iHuman Institute, ShanghaiTech University, 201210 Shanghai, China. 2 University of the Chinese Academy of Sciences, 100049 Beijing, China. 3 School of Life Science and Technology, ShanghaiTech University, 201210 Shanghai, China. 4 Department of Medical Neuroscience, Dalhousie University, Halifax, NS B3H 4R2, Canada. 5These authors contributed equally: Hui Li, Jie Yang, Cuiping Tian. ✉email: zhongsh@shanghaitech.edu.cn 1 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 G -protein-coupled receptors (GPCRs) are a large family of membrane proteins that play important roles in cellular functions by initializing a variety of intracellular processes via neurotransmitters and hormone signaling. To fulﬁll their function, GPCRs are likely anchored by cellular skeletal structures facilitating their interactions with intracellular protein complexes. Therefore, their membrane organization and relationship with skeleton proteins, which are still largely unknown, is likely critical for their cellular function. In addition, it would be important to know if diverse GPCRs share similar structural organizations. distributed in the axons of inhibitory interneurons, especially in those of cholecystokinin (CCK)-positive inhibitory interneurons (Supplementary Fig. 1c, d)12. CB1 did express much lower in myelinated axons and in the axons of excitatory neurons (Sup- plementary Fig. 1e, f). p y g While we conﬁrmed that CB1 was distributed in the axon shaft, the nature of its distribution was unknown. STED imaging is known to be well suited for studying the nano-scale structure of cellular components in ﬁxed preparations9,13. Therefore, we undertook STED imaging, with a spatial resolution around 70 nm (Supplementary Fig. Organized cannabinoid receptor distribution in neurons revealed by super-resolution ﬂuorescence imaging 2a), of immunolabeled hippocampal tissues to examine CB1 organizations along the axon shaft. CB1 displayed hotspots both with and without apparent periodicity within axons of the same axonal segment (Fig. 1a–c). We quantiﬁed the degree of periodicity using one-dimensional (1D) autocorrelation analysis by projecting the signals to the longitudinal axis of the axon and calculating the average 1D autocorrelation function over many axon segments8,14. The 1D autocorrelation amplitude, deﬁned as the average amplitude of the peaks at ~190 nm, quantiﬁes the degree of periodicity of the CB1 hotspots8,14. Most GPCRs are expressed at low levels in native tissue. Typically, investigators must overexpress them in cultured cells in order to study their structures and functions. The cannabinoid receptor 1 (CB1) is one of the highest-expressed GPCRs in the central nervous system, making it feasible to study in its natural state1. CB1 is important for many biological functions such as pain, mood and memory2–5. Recent structural studies have revealed the isolated atomic arrangement of CB1 6,7. However, these studies have primarily used X-ray crystallography or cryo-electron microscopy techniques that require high concentrations of well- puriﬁed proteins in detergent solution. CB1 structure has rarely been studied within cells in combination with signaling proteins, integral membrane proteins, or other membrane-associated pro- teins. Thus, the CB1 cellular structure remains incomplete, lim- iting our ability to fully exploit its function. To identify whether CB1 possess a similar distribution in other brain regions, we performed STED imaging in the cortex, and observed a similar semi-periodicity in neuron axons (Fig. 1d–f). The distance between these rhythmic hotspots was also ~190 nm in the cortex as it was hippocampus (Fig. 1g). Next, we imaged the structure of CB1 from cultured hippocampal neurons. We found that antibody-labeled CB1 exhibited hotspots with both high and low degree of periodicity in axons of cultured neurons (Supplementary Fig. 2b–d). These results are comparable to the in vivo results above (Fig. 1a–c). g y y Super-resolution microscopy, an imaging method that over- comes the diffraction limit of conventional microscopy, has led to the discovery of the membrane-associated periodic skeleton (MPS) in the axons of neurons8. This highly ordered sub- membrane skeletal structure can play many roles in neuronal function, including acting as a ﬂexible mechanical support, organizing membrane protein distribution, and the development of axons and dendrites. Organized cannabinoid receptor distribution in neurons revealed by super-resolution ﬂuorescence imaging The semi-periodic organization of CB1 in axons raised two important questions, how is the CB1 semi-periodic organization formed and what are the complexes related to the semi-periodic organization of CB1? To answer the ﬁrst question, we need to identify the components associated with the semi-periodic complex of CB1 in native states. To this end, we performed mass spectrometry (MS) experiments with modiﬁcations to increase the enrichment of membrane proteins in six different regions of the mouse brain (Supple- mentary Fig. 3). We found that CB1 was expressed at different levels in the six brain regions and that the expression level par- alleled the AMPA speciﬁc glutamate receptor 1 and N-methyl-D- aspartate (NMDA) receptor (Supplementary Fig. 3a), which are known to crosstalk with CB1 in the central nervous system16,17. Furthermore, we observed the association of GPCR-related sig- naling molecules with CB1 (such as Gi, tyrosine-protein kinase NS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications Results CB1 exhibits semi-periodic hotspots in neurons. To study the structure of CB1 in vivo, we ﬁrst demonstrated the speciﬁcity of our CB1 antibody labeling (Supplementary Fig. 1a, b). We found, in agreement with previous ﬁndings, that CB1 is mainly URE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 2 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 1.0 0.5 0.0 1.0 0.5 0.0 - 1 - 1 f e 1.0 0.5 0.0 –0.5 –1.0 1000 0 –1000 500 –500 Lag (nm) 1.0 0.5 0.0 500 1000 0 Distance (nm) 1.0 0.5 0.0 Autocorrelation Normalized intensity Normalized intensity b c d a Hippocampus Confocal STED Cortex Confocal 0.5 µm STED Distance (nm) 0.0 0.2 0.4 0.6 0.8 HC Cortex Fixed Live Autocorrelation amplitude Brain tissue Primary culture g h 0 100 HC Cortex Fixed Live Brain tissue Primary culture 50 150 200 250 Normalized intensity Normalized intensity 1.0 0.5 0.0 –0.5 –1.0 Autocorrelation 1.0 0.5 0.0 –0.5 –1.0 Autocorrelation 1.0 0.5 0.0 –0.5 –1.0 Autocorrelation 1000 0 –1000 500 –500 Lag (nm) 1000 0 –1000 500 –500 Lag (nm) 1000 0 –1000 500 –500 Lag (nm) 500 1000 0 Distance (nm) 500 1000 0 Distance (nm) 500 1000 0 Distance (nm) 0.5 µm Fig. 1 Periodic hotspots of CB1 of varying strength in axons from brain tissue. a Representative confocal and corresponding STED images of CB1 in the hippocampus of mature C57 mice. N = 3 biological replicates. b, c Intensity plotted along the lines in the box regions (top graph). Autocorrelation analysis for the confocal and STED images (bottom graph). d–f Similar to (a–c), but in the cortex region of mature C57 mice. N = 3 biological replicates. g Histogram of CB1 spacing in different samples, live and ﬁxed hippocampal neurons, brain tissue of the hippocampus (HC) and cortex. Data are mean ± s.e. m. (N = 3 biological replicates; 70–120 axonal regions were examined per condition). p = 0.09, (no signiﬁcance), one-way ANOVA. Actual spacing (from left to right), 192 ± 0.8 nm, 192 ± 0.8 nm, 187 ± 2 nm, 190 ± 2 nm. h Amplitude of the average autocorrelation functions calculated from randomly selected axon segments in different samples. p = 0.13 (no signiﬁcance), one-way ANOVA. Actual autocorrelation amplitude (from left to right), 0.28 ± 0.02, 0.24 ± 0.03, 0.21 ± 0.01, 0.23 ± 0.02. Data in (g, h) are mean ± s.e.m. Results (N = 3 biological replicates; 70–120 axonal regions were examined per condition). Source data are provided as a Source Data ﬁle. a Hippocampus Confocal STED 0.5 µm d Cortex Confocal 0.5 µm STED d a 0.5 µm STED STED 0.5 µm 1.0 0.5 0.0 - 1 f Normalized intensity 1.0 0.5 0.0 –0.5 –1.0 Autocorrelation 1000 0 –1000 500 –500 Lag (nm) 500 1000 0 Distance (nm) 1.0 0.5 0.0 –0.5 –1.0 1000 0 –1000 500 –500 Lag (nm) 1.0 0.5 0.0 500 1000 0 Distance (nm) Autocorrelation Normalized intensity b f 1.0 0.5 0.0 Normalized intensity c 1.0 0.5 0.0 –0.5 –1.0 Autocorrelation 1000 0 –1000 500 –500 Lag (nm) 500 1000 0 Distance (nm) 1.0 0.5 0.0 - 1 e Normalized intensity 1.0 0.5 0.0 –0.5 –1.0 Autocorrelation 1000 0 –1000 500 –500 Lag (nm) 500 1000 0 Distance (nm) b c e Distance (nm) g 0 100 HC Fixed Live Brain t Primary culture 50 150 200 250 HC Cortex Brain tissue 0.0 0.2 0.4 0.6 0.8 HC Cortex Fixed Live Autocorrelation amplitude Brain tissue Primary culture h h g Fig. 1 Periodic hotspots of CB1 of varying strength in axons from brain tissue. a Representative confocal and corresponding STED images of CB1 in the hippocampus of mature C57 mice. N = 3 biological replicates. b, c Intensity plotted along the lines in the box regions (top graph). Autocorrelation analysis for the confocal and STED images (bottom graph). d–f Similar to (a–c), but in the cortex region of mature C57 mice. N = 3 biological replicates. g Histogram of CB1 spacing in different samples, live and ﬁxed hippocampal neurons, brain tissue of the hippocampus (HC) and cortex. Data are mean ± s.e. m. (N = 3 biological replicates; 70–120 axonal regions were examined per condition). p = 0.09, (no signiﬁcance), one-way ANOVA. Actual spacing (from left to right), 192 ± 0.8 nm, 192 ± 0.8 nm, 187 ± 2 nm, 190 ± 2 nm. h Amplitude of the average autocorrelation functions calculated from randomly selected axon segments in different samples. p = 0.13 (no signiﬁcance), one-way ANOVA. Actual autocorrelation amplitude (from left to right), 0.28 ± 0.02, 0.24 ± 0.03, 0.21 ± 0.01, 0.23 ± 0.02. Data in (g, h) are mean ± s.e.m. (N = 3 biological replicates; 70–120 axonal regions were examined per condition). Source data are provided as a Source Data ﬁle. NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunicati Results Then, to demonstrate the associatio complex components identiﬁed, we carri another type of imaging assay that r h i l di t ib ti f t bj t 19 It CB1-ankB WT 30 μm CB1-ankB cnr1−/− 30 μm 500 1000 0 1.0 0.5 0.0 500 1000 0 Normalized intensity 1D Cross-correlation 1.0 0.5 0.0 –0.5 600 –600 0 1.0 0.5 0.0 –0.5 Lag (nm) 600 –600 0 1D Cross-correlation Lag (nm) Distance (nm) Distance (nm) c CB1/βII-spectrin 0.5 μm βII-spectrin CB1 b a 200 kDa 52 kDa ankB CB1 IB IP: CB1-ankB CB1 ICL3-ankB CB1 C terminal-ankB * CB1 IgG Input g Ratio r 0 5 10 15 * CB1 ICL3/-ankB CB1-ankB h 10 μm ankB CB1 C-term- d f e CB1-ankB 10 μm 20 1.0 0.5 0.0 Normalized intensity pattern of CB1 with the MPS Indeed at periodic hotspots CB1 unambiguously show that CB1 forms a semi-periodic complex CB1-ankB WT 30 μm CB1-ankB cnr1−/− 30 μm 500 1000 0 1.0 0.5 0.0 500 1000 0 Normalized intensity 1D Cross-correlation 1.0 0.5 0.0 –0.5 600 –600 0 1.0 0.5 0.0 –0.5 Lag (nm) 600 –600 0 1D Cross-correlation Lag (nm) Distance (nm) Distance (nm) c CB1/βII-spectrin 0.5 μm βII-spectrin CB1 b a 200 kDa 52 kDa ankB CB1 IB IP: CB1-ankB CB1 ICL3-ankB CB1 C terminal-ankB * CB1 IgG Input g Ratio r 0 5 10 15 * CB1 ICL3/-ankB CB1-ankB h 10 μm ankB CB1 C-term- d f e CB1-ankB 10 μm 20 1.0 0.5 0.0 Normalized intensity CB1/βII-spectrin 0.5 μm βII-spectrin CB1 a b c c c 500 1000 0 1D Cross-correlation 1.0 0.5 0.0 –0.5 600 –600 0 Lag (nm) Distance (nm) 1.0 0.5 0.0 Normalized intensity 1.0 0.5 0.0 500 1000 0 Normalized intensity 1.0 0.5 0.0 –0.5 Lag (nm) 600 –600 0 1D Cross-correlation Distance (nm) CB1-ankB CB1 ICL3-ankB CB1 C terminal-ankB * Ratio r 0 5 10 15 * CB1 ICL3/-ankB CB1-ankB h 10 μm ankB CB1 C-term- 20 h CB1-ankB WT 30 μm d CB1-ankB cnr1−/− 30 μm e e d WT d WT WT f f CB1-ankB 10 μm g 200 kDa 52 kDa ankB CB1 IB IP: CB1 IgG Input g unambiguously show that CB1 forms a semi-periodic complex associated with the MPS. pattern of CB1 with the MPS. Results and Fyn), indicating the reliability of our modiﬁed MS method (Supplementary Fig. 3b). a tight association between CB1 and MPS components. To test the above view, we carried out two independent types of experiments, two-color STED imaging and proximity ligation assays (PLA)19. First, we visualized the spatial relation of CB1 and MPS molecules by dual-color STED imaging. The MPS was visualized through immunolabelling of the C-terminus of βII-spectrin, which is located at the center of each spectrin tetramer connecting adjacent actin rings14,18. We quantiﬁed the extent of colocaliza- tion between CB1 and βII-spectrin using 1D cross-correlation analysis to calculate the average 1D cross-correlation function between the two color channels over randomly chosen axon segments8. At periodic hotspots, CB1 displayed high colocaliza- tion with βII-spectrin while at nonperiodic clusters, CB1 displayed little colocalization with βII-spectrin (Fig. 2a–c). Previous studies demonstrated that ankB mediated the attach- ment of membrane proteins to the MPS, and is located in the middle region of each spectrin tetramer18,20. Thus, we reasoned that CB1’s interaction with ankB may organize the semi-periodic Next, we compared the expression extent of selected MPS components with CB1 expression. While different isoforms of ankyrin expressed across different functional domains of neurons, ankyrin-B (ankB) is predominantly distributed in the axons of neurons14,18. Our MS experiments indicated that the expression level of CB1 was highly correlated to that of ankB, but not ankyrin-R (ankR) (Supplementary Fig. 3a). Further, the expres- sion level of CB1 was correlated to that of both αII-spectrin and βII-spectrin (Supplementary Fig. 3a) while other membrane proteins seem unrelated to the expression levels of CB1 (Supplementary Fig. 3c). Earlier studies showed that MPS, like ankyrin, spectrin, displayed a highly coordinated pattern with an interval around 190 nm, which resembles the structure of CB1 in some regions of neurons revealed both by STED and SIM imaging experiments. Thus, consistent with our MS experimental results, there could be NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 3 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 ith the MPS. Indeed, at periodic hotspots, CB1 olocalization with ankB while at nonperiodic played little colocalization with ankB (Supple- –c). Furthermore, the spatial distance between tspots was comparable to both those of βII- kB (S l t Fi 4d) Th lt unambiguously show that CB1 forms a associated with the MPS. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 CB1 connects to the MPS through ankB forming semi-periodic hotspots. As such, we would expect CB1 to display conﬁned dynamics by interacting with the MPS through ankB. We investigated the dynamics of individual hotspots of CB1 with live SIM imaging. Neurons were ectopically expressed with a CB1-RFP fusion protein (RFP protein is fused to the end of CB1 C-terminal) and live SIM images were acquired 1 day after transfection. In order to identify whether CB1-RFP was recruited to the native site, we performed two-color STED imaging of CB1-RFP and βII-spectrin. CB1-RFP displayed highly periodic hotspots that colocalized with βII-spectrin, as well as nonperiodic clusters that did not colocalize with βII-spectrin (Supplementary Fig. 6a–c) These results were similar to the spatial distribution between CB1 and βII-spectrin (Fig. 2a–c), supporting that CB1-RFP is transported to CB1 native sites. We proceeded to examine the dynamics of CB1-RFP using live SIM imaging. Periodic CB1 hotspots were conﬁned to movements around their starting positions displaying conﬁned displacement changes (average 69 nm) between time frames (Fig. 3a–c). Averaged autocorrelation analysis of CB1 at different time frames showed a similar periodic distribution, indicating that CB1 clus- ters maintained their periodicity through time (Fig. 3d). The cross-correlation of neighboring time points was calculated showing that the average cross-correlation value peaked at zero point (Fig. 3e), indicating little or no systematic shift of the CB1 periodicity across different time frames. The periodic wavelength was around 190 nm at different time frames (Fig. 3f). We com- piled the moving traces of individual hotspots of CB1 that clearly indicated conﬁned movement both at short (2 min) (Fig. 3g) and longer (6 min) (Supplementary Fig. 7a–g) imaging time frames. The moving traces between the two time frames were comparable (Supplementary Fig. 7h), suggesting that CB1 displayed a con- ﬁned movement around its anchoring point, likely mediated by ankB. p y p In order to show that CB1 interacted with ankB in tetracycline- induced CB1-CHO cells, we performed immunoprecipitation experiments (Fig. 2g). To identify which parts of CB1 may mediate the interaction between CB1 and ankB, we truncated the third intracellular loop (ICL3) or C terminal of CB1. Notably, the PLA signals were signiﬁcantly decreased in both conditions, suggesting that both regions, ICL3 and C-terminal, play a role in the interaction between CB1 and ankB (Fig. 2h), and that neither region alone is sufﬁcient to mediate their interaction. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 ARTICLE Fig. 2 CB1 hotspots are connected with components of the MPS. a Two-color STED images of CB1 (magenta) and βII-spectrin (green) in the axons of cultured neurons. N = 3 biological replicates. b, c Top, enlarged images taken from white boxes from (a). 1D projection traces of βII-spectrin (green) and CB1 (magenta) signals along the axon are shown in the middle. 1D cross-correlation functions between the distributions of CB1 and βII-spectrin from CB1-positive axon segments are shown in the bottom. d–f PLA was performed in cultured neurons of WT mouse (d), cnr1−/−mouse (e), and tetracycline-induced CB1 transfected CHO cells (f) with antibody of CB1 and ankB. Cell nuclei were stained with DAPI (blue). N = 3 biological replicates. g Immunoprecipitation of CB1 and ankB in CB1-CHO cells. Samples were processed for immunoprecipitation with either anti-CB1 or IgG control antibodies. Immunoprecipitates were immunoblotted with the anti-CB1 antibody (55kD) and anti-ankB antibody (220kD). N = 2 biological replicates. h PLA was performed in HEK-293T cells coexpressing ankB and different fragments of CB1, including wild type (CB1), CB1 with truncated ICL3 loop (CB1ΔICL3), or CB1 with truncated C-terminal (CB1ΔC-term). ratio r from left to right: 14.1 ± 0.5, 6.3 ± 0.4, 5.5 ± 0.4. Data are mean ± s.e.m. (N = 3 biological replicates; 25–50 imaging regions were examined per condition). *p < 0.0001, statistical analysis was performed by unpaired two-tailed Student’s t test. Source data are provided as a Source Data ﬁle. components mediating the CB1 distribution pattern in axons, such as the nonperiodic organization in presynaptic sites. and βII-spectrin closely distribute in neurites. Indeed, in cnr1−/− mouse, PLA signal with α-adducin and βII-spectrin was not changed (Supplementary Fig. 5a, b), indicating the reliability of PLA to detect proteins within a short distance. Then, we determined whether CB1 localizes closely with MPS components in cultured neurons. PLA signals of CB1 and ankB were observed in cultured primary neurons (Fig. 2d), but no PLA signals were detected in the cnr1−/−mouse neurons (Fig. 2e). Upon the induction of CB1 in the presence of tetracycline, close physical distances were detected between CB1 and several MPS compo- nents, including ankB and βII-spectrin in CB1-CHO cell lines (Fig. 2f and Supplementary Fig. 5c, d). Together, these results from neurons and inducible CB1-CHO cell lines clearly illustrate the close physical distance between CB1 and MPS components. CB1 displays conﬁned dynamics in neurons. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 Together, these results support the view that CB1 may form a complex with cytoskeleton-related proteins, such as ankB, αII-spectrin and βII- spectrin. CB1 structure in different compartments of axons. We exam- ined whether the close relation between CB1 and cytoskeleton components locates in axons or dendrites of neurons. In order to achieve this, we ﬁrst performed PLA labeling, and then immu- nolabeled either axons or dendrites with tuj1 or MAP2 antibody, respectively. Our results show that the PLA signal between CB1 and ankB substantially diminished after the second round of immunolabeling. We further explored the PLA signal between CB1 and spectrin, which can survive the immunolabeling proce- dure. Our results clearly showed that the PLA signal located in axons but not in dendrites (Supplementary Fig. 5e–f), indicating that CB1 is associated with MPS components speciﬁcally in axons. CB1 structure in different compartments of axons. We exam- ined whether the close relation between CB1 and cytoskeleton components locates in axons or dendrites of neurons. In order to achieve this, we ﬁrst performed PLA labeling, and then immu- nolabeled either axons or dendrites with tuj1 or MAP2 antibody, respectively. Our results show that the PLA signal between CB1 and ankB substantially diminished after the second round of immunolabeling. We further explored the PLA signal between CB1 and spectrin, which can survive the immunolabeling proce- dure. Our results clearly showed that the PLA signal located in axons but not in dendrites (Supplementary Fig. 5e–f), indicating that CB1 is associated with MPS components speciﬁcally in axons. Then, we used speciﬁc antibodies to label different regions of axons, such as axon initial segments (AISs) with βIV-spectrin antibody or presynaptic sites with VGAT antibody. Our results show that select AISs express CB1 with periodic pattern with an interval around 190 nm (Supplementary Fig. 5g, h, l–n) while some AISs did not express CB1 (Supplementary Fig. 5g, i). Interestingly, presynaptic sites with VGAT labeling did express CB1 (Supplementary Fig. 5j), but most were not periodic (Supplementary Fig. 5k). Thus, our results indicated that CB1 exhibited a specialized pattern in different regions of axons. Given that ankyrin and spectrin are uniformly distributed with a regular patter around 190 nm, there are likely some additional Active CB1 associate with MPS and display less dynamics. Results Indeed, at periodic hotspots, CB1 displayed high colocalization with ankB while at nonperiodic clusters, CB1 displayed little colocalization with ankB (Supple- mentary Fig. 4a–c). Furthermore, the spatial distance between periodic CB1 hotspots was comparable to both those of βII- spectrin and ankB (Supplementary Fig. 4d). These results Then, to demonstrate the association between CB1 and the complex components identiﬁed, we carried out PLA experiments, another type of imaging assay that reliably detects the close physical distribution of two subjects19. It is known that α-adducin Then, to demonstrate the association between CB1 and the complex components identiﬁed, we carried out PLA experiments, another type of imaging assay that reliably detects the close physical distribution of two subjects19. It is known that α-adducin NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 4 4 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 CB1 were displayed with coordinated clusters with an interval of around 190 nm upon WIN application with STED imaging the immunolabeled CB1 in the primary hippocampal neurons (Fig. 4a). Next, we imaged CB1 in the presence of antagonist to test whether blocking CB1 activity had an impact on its organi- zation. Surprisingly, our results show that CB1 still exhibited a periodic pattern in some axons with CB1 antagonist, suggesting that the basal level activation of CB1 might not be necessary for the periodic organization (Fig. 4a–c). All raw STED images were included in Supplementary Fig. 8. To test the effect of CB1 acti- vation on its dynamics, we imaged the dynamics of CB1 in the primary neurons transfected with CB1-RFP in the presence of 1 p p y Then, we used speciﬁc antibodies to label different regions of axons, such as axon initial segments (AISs) with βIV-spectrin antibody or presynaptic sites with VGAT antibody. Our results show that select AISs express CB1 with periodic pattern with an interval around 190 nm (Supplementary Fig. 5g, h, l–n) while some AISs did not express CB1 (Supplementary Fig. 5g, i). Interestingly, presynaptic sites with VGAT labeling did express CB1 (Supplementary Fig. 5j), but most were not periodic (Supplementary Fig. 5k). Thus, our results indicated that CB1 exhibited a specialized pattern in different regions of axons. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 Given that ankyrin and spectrin are uniformly distributed with a regular patter around 190 nm, there are likely some additional NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 5 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 1 2 3 4 5 500 nm 1 2 3 4 5 a b x (nm) ) m n ( y –100 0 100 200 –150 0 150 g 150 0 200 y (nm) c 0 x (nm) d f e Autocorrelation amplitude 1DCross-correlation amplitude 0 s 60 s 120 s 0 200 600 –200 –600 Lag (nm) Ave-Autocorrelation 1.0 0.5 0.0 –0.5 0–60 s 0–120 s n oit ale rr o C - s s o r C D 1 mean velocity 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0 s 60 s 120 s 0.6 0.4 0.2 0 0.8 0-60 s 0 s-120 s 0 s 60 s 120 s 100 200 0 Distance (nm) 150 250 50 150 0 y (nm) 150 0 y (nm) 150 0 y (nm) 150 0 y (nm) 200 0 x (nm) 200 0 x (nm) 200 0 x (nm) 200 0 x (nm) 1.0 0.5 0.0 –0.5 0 200 600 –200 –600 Lag (nm) B1 hotspots display stable dynamics revealed by live SIM imaging. a Representative live image of transfected CB1-R 2) of the SD rat acquired by SIM. Individual CB1 hotspots are marked with purple balls, and locations of each individ ith lines. Line colors indicate trace indexes. N = 3 biological replicates. b Five CB1 hotspots shown in (a) with their re changes from (b) are around 60–70 nm between neighboring time points. d Averaged autocorrelation analysis of CB1 nts with the histogram of the autocorrelation amplitude. p = 0.99 (no signiﬁcance), one-way ANOVA. Actual autocorre 2, 0.47 ± 0.03, 0.47 ± 0.02. e Averaged cross-correlation analysis between the neighboring frames (0 vs 60 s, 0 vs 120 erties and histogram showing amplitude of average cross-correlation. p = 0.386 (no signiﬁcance), statistical analysis ed Student’s t test. Actual cross-correlation amplitude (from left to right), 0.45 ± 0.05; 0.51 ± 0.02. f The histogram of C 0.69 (no signiﬁcance, one-way ANOVA). Actual spacing (from left to right), 192 ± 0.6, 192 ± 0.7, 191 ± 0.6 nm. Data in (d gical replicates; 70–120 axonal regions were examined per condition). g Traces of the individual CB1 hotspots over tim urce Data ﬁle. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41 1 2 3 4 5 500 nm a mean velocity 1 2 3 4 5 500 nm 1 2 3 4 5 a b 150 0 200 y (nm) c 0 x (nm) mean velocity 150 0 y (nm) 150 0 y (nm) 150 0 y (nm) 150 0 y (nm) 200 0 x (nm) 200 0 x (nm) 200 0 x (nm) 200 0 x (nm) 1 2 3 4 5 b 150 0 200 y (nm) c 0 x (nm) 150 0 y (nm) 150 0 y (nm) 150 0 y (nm) 150 0 y (nm) 200 0 x (nm) 200 0 x (nm) 200 0 x (nm) 200 0 x (nm) d d Autocorrelation amplitude 0 s 60 s 120 s 0 200 600 –200 –600 Lag (nm) Ave-Autocorrelation 1.0 0.5 0.0 –0.5 0.60 0.55 0.50 0.45 0.40 0.35 0.30 0 s 60 s 120 s e 1DCross-correlation amplitude 0–60 s 0–120 s n oit ale rr o C - s s o r C D 1 0.6 0.4 0.2 0 0.8 0-60 s 0 s-120 s 1.0 0.5 0.0 –0.5 0 200 600 –200 –600 Lag (nm) e f 0 s 60 s 120 s 100 200 0 Distance (nm) 150 250 50 f x (nm) ) m n ( y –100 0 100 200 –150 0 150 g g g Fig. 3 Periodic CB1 hotspots display stable dynamics revealed by live SIM imaging. a Representative live image of transfected CB1-RFP in the primary neuron (DIV 9–12) of the SD rat acquired by SIM. Individual CB1 hotspots are marked with purple balls, and locations of each individual time points are connected with lines. Line colors indicate trace indexes. N = 3 biological replicates. b Five CB1 hotspots shown in (a) with their relative locations. c Displacement changes from (b) are around 60–70 nm between neighboring time points. d Averaged autocorrelation analysis of CB1 distributions at different time points with the histogram of the autocorrelation amplitude. p = 0.99 (no signiﬁcance), one-way ANOVA. Actual autocorrelation from left to right, 0.47 ± 0.02, 0.47 ± 0.03, 0.47 ± 0.02. e Averaged cross-correlation analysis between the neighboring frames (0 vs 60 s, 0 vs 120 s) showed similar distribution properties and histogram showing amplitude of average cross-correlation. p = 0.386 (no signiﬁcance), statistical analysis was performed by unpaired two-tailed Student’s t test. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 Actual cross-correlation amplitude (from left to right), 0.45 ± 0.05; 0.51 ± 0.02. f The histogram of CB1 spacing across time points. p = 0.69 (no signiﬁcance, one-way ANOVA). Actual spacing (from left to right), 192 ± 0.6, 192 ± 0.7, 191 ± 0.6 nm. Data in (d−f) are mean ± s. e.m. (N = 3 biological replicates; 70–120 axonal regions were examined per condition). g Traces of the individual CB1 hotspots over time. Source data are provided as a Source Data ﬁle. NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 6 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 Control CB 1 +RIM +WIN Control Autocorrelation amplitude Autocorrelation amplitude c b d 0 200 600 –200 –600 Lag (nm) Ave-Autocorrelation 1.0 0.5 0.0 –0.5 f h * 200 nm +WIN CB 1- RFP CB 1- RFP Control Displacement (nm/min) +WIN CB 1 +RIM CB 1 1 µm a 0 200 600 –200 –600 Lag (nm) Ave-Autocorrelation 1.0 0.5 0.0 –0.5 Control +WIN e –200 –100 0 100 200 200 100 0 –100 –200 x (nm) g y (nm) Mean velocity Mean velocity 0.6 0.4 0.2 0 0.8 Control +WIN +RIM * 0.6 0.4 0.2 0 0.8 Control +WIN 300 200 100 0 Control +WIN Fig. 4 The dynamics of active CB1 hotspots. a Representative STED images of CB1 in primary hippocampal neurons for control, with WIN treatment (500 nM, 10 min) and RIM (1 µM, 10 min) treatment. N = 3 biological replicates. b Average autocorrelation analysis for the STED images of CB1 in different conditions. c The histogram of autocorrelation amplitude of CB1 (from left to right), 0.28 ± 0.02, 0.38 ± 0.02, 0.25 ± 0.02. p = 0.0003, p < 0.01, one-way ANOVA. d Representative live image of transfected CB1-RFP in untreated neurons (top, “Control”), and neurons treated with WIN (bottom, “+WIN”) acquired by SIM. Individual CB1 hotspots are marked with purple balls, and locations of each individual time points are connected with lines. Line colors indicate trace indexes. N = 3 biological replicates. e Averaged autocorrelation analysis of CB1 distributions in live neurons. f The histogram of autocorrelation amplitude of CB1 without and with WIN treatment was shown. p < 0.0001, statistical analysis was performed by unpaired two-tailed Student’s t test. Actual autocorrelation amplitude (from left to right), 0.33 ± 0.02, 0.46 ± 0.01. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 g Traces the dynamics of CB1 hotspots without WIN treatment (magenta) and CB1 hotspots with WIN treatment (blue) over time. h The displacement of CB1 hotspots in different conditions. Control, 83.9 ± 2.5 nm/min. N = 291 spots. WIN, 64.6 ± 2.5 nm/min. N = 225 spots. p < 0.0001, statistical analysis was performed by unpaired two-tailed Student’s t test. Data in (c, f, h) are mean ± s.e.m. (N = 3 biological replicates; 70–120 axonal regions were examined per condition). Source data are provided as a Source Data ﬁle. Control CB 1 +WIN CB 1 +RIM CB 1 1 µm a a +RIM +WIN Control b 0 200 600 –200 –600 Lag (nm) Ave-Autocorrelation 1.0 0.5 0.0 –0.5 Autocorrelation amplitude c 0.6 0.4 0.2 0 0.8 Control +WIN +RIM b Ave-Autocorrelation 0 200 600 –200 –600 Lag (nm) Ave-Autocorrelation 1.0 0.5 0.0 –0.5 Control +WIN e Autocorrelation amplitude f * 0.6 0.4 0.2 0 0.8 Control +WIN d d CB 1- RFP Control Mean velocity e e h * Displacement (nm/min) –200 –100 0 100 200 200 100 0 –100 –200 x (nm) g y (nm) 300 200 100 0 Control +WIN 200 nm +WIN CB 1- RFP Mean velocity h g Fig. 4 The dynamics of active CB1 hotspots. a Representative STED images of CB1 in primary hippocampal neurons for control, with WIN treatment (500 nM, 10 min) and RIM (1 µM, 10 min) treatment. N = 3 biological replicates. b Average autocorrelation analysis for the STED images of CB1 in different conditions. c The histogram of autocorrelation amplitude of CB1 (from left to right), 0.28 ± 0.02, 0.38 ± 0.02, 0.25 ± 0.02. p = 0.0003, p < 0.01, one-way ANOVA. d Representative live image of transfected CB1-RFP in untreated neurons (top, “Control”), and neurons treated with WIN (bottom, “+WIN”) acquired by SIM. Individual CB1 hotspots are marked with purple balls, and locations of each individual time points are connected with lines. Line colors indicate trace indexes. N = 3 biological replicates. e Averaged autocorrelation analysis of CB1 distributions in live neurons. f The histogram of autocorrelation amplitude of CB1 without and with WIN treatment was shown. p < 0.0001, statistical analysis was performed by unpaired two-tailed Student’s t test. Actual autocorrelation amplitude (from left to right), 0.33 ± 0.02, 0.46 ± 0.01. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 g Traces the dynamics of CB1 hotspots without WIN treatment (magenta) and CB1 hotspots with WIN treatment (blue) over time. h The displacement of CB1 hotspots in different conditions. Control, 83.9 ± 2.5 nm/min. N = 291 spots. WIN, 64.6 ± 2.5 nm/min. N = 225 spots. **p < 0.0001, statistical analysis was performed by unpaired two-tailed Student’s t test. Data in (c, f, h) are mean ± s.e.m. (N = 3 biological replicates; 70–120 axonal regions were examined per condition). Source data are provided as a Source Data ﬁle. CB1 agonist, WIN. Periodic CB1 hotspots were conﬁned (Fig. 4d). CB1 displayed higher periodicity as reﬂected by the higher autocorrelation amplitude (Fig. 4d, e). Therefore, our results show that with WIN application, active CB1 moved around their ori- ginal point to a lesser extent than nonactive CB1 (Fig. 4e). treated cultures with either latrunculinB (latB) or cytochalasin D (cytoD) to disrupt cytoskeletal structure8,14. We found that fol- lowing both latB and cytoD treatments, periodic hotspots of CB1 were no longer observed (Fig. 5a, b), suggesting that the cytos- keleton is important for maintaining the CB1 structure in neu- rons. Next, we tested whether the cytoskeleton affects CB1 intracellular downstream signaling. CB1 can activate both Akt and ERK1/221,22. In primary neuron cells, pretreatment with latB resulted in decreased phosphorylation levels of both Akt and ERK1/2 in a dose-dependent manner (Supplementary Fig. 10a, b). Thus, these results suggest that the intracellular signaling of CB1 is dependent upon the MPS cytoskeleton (Fig. 5c–e). Next, we used ﬂuorescence recovery after photobleaching (FRAP) experiments to examine the recovery rate of CB1 after photobleaching to further study the dynamics of CB1 in culture neurons. We transfected the neurons with CB1-RFP. We then measured the half-time recovery rate of CB1. Our results showed that with WIN application the half-time recovery of CB1 was slower than in the control, while this effect was blocked with the application of CB1 antagonist (Supplementary Fig. 9a–c). The half-time recovery of CB1 was not affected by the application of agonists of other types of GPCRs, suggesting that the change was likely caused by CB1 activation (Supplementary Fig. 9c). Thus, our results show that CB1 displayed less dynamics with receptor activation. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 Our ﬁndings with imaging probes likely reﬂect the dynamics of CB1 in neurons as both probes in live neurons displayed a spatial pattern similar to that of CB1 antibody in ﬁxed neurons. Previous studies have shown that GPCRs form homodimers, heterodimers, or oligomers to affect their downstream signaling pathways23–26. can be activated and then some residues of ICL3 region are phosphorylated and contain more negative charges. By this way, more negative charged residues after receptor activation could attract and bind to the positive parts of the disordered domains of ankB and thus leading to a stronger connection between CB1 and ankB. This could explain the less movement of CB1 in the pre- sence of WIN. The exact molecular mechanism underlying the distinct dynamics of CB1 is an important question and deserves further investigation. Uncovering the cellular structure of GPCRs has proven to be a challenging task, as several prior studies did not observe the semi- periodic organization of CB1 as we did12,11. An early study using STORM to characterize the CB1 structure in brain tissue did not observe any sign of organized CB1 pattern in axons. This could be due to the use of different antibodies, though most is likely because they did not focus on CB1 in the axonal shaft region, but instead focused on axonal boutons12. Intriguingly, Zhou et al.11 found that CB1 displays a nonperiodic structure without the application of WIN, an agonist of CB1, and becomes periodic upon the administration of WIN. g Actually, Zhou’s data implied that there was weak CB1 peri- odicity and colocalization with spectrin at ~190 nm before ago- nist stimulation11. In the native state, we found that axons contained a semi-periodic pattern. Most synaptic sites did not exhibit periodic pattern, implying that the molecular mechanism to organize the synapses and CB1 complexes may be different. A small population of AISs had CB1 expression, while the others did not contain CB1 at all. This might be due to the fact that a small population of primary neurons in culture are inhibitory. The periodic distribution without application of WIN could be explained by the fact that constitutive released ligand, such as 2-AG, may activate CB1 to a certain extent and thus lead to the periodic pattern of CB1. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 CB1 +LatB CB1 +CytoD βII-spectrin +CytoD βII-spectrin +LatB a CB1/βII-spectrin +CytoD 0.5 μm CB1/βII-spectrin +LatB 0.5 μm b CB1 +LatB CB1 +CytoD βII-spectrin +CytoD βII-spectrin +LatB a CB1/βII-spectrin +CytoD 0.5 μm CB1/βII-spectrin +LatB 0.5 μm Nonactivated 190 nm 190 nm ERK JNK p38 AKT cAMP Gi/o AC ERK JNK p38 AKT cAMP Gi/o ERK JNK p38 AKT cAMP Gi/o AC ERK JNK p38 ERK JNK p38 AKT AKT Gi/o Gi/o ERK JNK p38 AKT Gi/o cAMP AC cAMP AC cAMP AC Activated Disrupted MPS CB1 ankB Spectrin Actin c d e b AC Fig. 5 Schematic illustration showing CB1 forming dynamic peri-periodic hotspots to increase signaling efﬁciency. a, b Two-color STED images of CB1 (magenta) and βII-spectrin (green) in neurons treated with LatB (a) and CytoD (b). N = 3 biological replicates. c Without ligand binding, MPS sets the range for CB1 distribution. d Upon ligand binding, active CB1 are recruited to the MPS and become more periodic, making downstream AKT and ERK signaling more effective. e MPS degradation leads to less strong periodic clusters of CB1 and thus less downstream signaling. Nonactivated 190 nm 190 nm c CB1 +LatB βII-spectrin +LatB a CB1/βII-spectrin +LatB 0.5 μm c c a ERK JNK p38 AKT cAMP Gi/o AC ERK JNK p38 AKT cAMP Gi/o ERK JNK p38 AKT cAMP Gi/o AC Activated d AC d CB1 +CytoD βII-spectrin +CytoD CB1/βII-spectrin +CytoD 0.5 μm b b CB1 +CytoD b ERK JNK p38 ERK JNK p38 AKT AKT Gi/o Gi/o ERK JNK p38 AKT Gi/o cAMP AC cAMP AC cAMP AC Disrupted MPS CB1 ankB Spectrin Actin e βII-spectrin +CytoD CB1/βII-spectrin +CytoD 0.5 μm e ankB Fig. 5 Schematic illustration showing CB1 forming dynamic peri-periodic hotspots to increase signaling efﬁciency. a, b Two-color STED images of CB1 (magenta) and βII-spectrin (green) in neurons treated with LatB (a) and CytoD (b). N = 3 biological replicates. c Without ligand binding, MPS sets the range for CB1 distribution. d Upon ligand binding, active CB1 are recruited to the MPS and become more periodic, making downstream AKT and ERK signaling more effective. e MPS degradation leads to less strong periodic clusters of CB1 and thus less downstream signaling. conditions. Here, we used SIM to investigate the dynamics of CB1. Discussion Vi li i Visualization of GPCRs in native tissue is necessary for under- standing the intracellular organization of these receptors in real physiology. With the recent development of super-resolution imaging methods, we now can observe the organization of GPCRs at the cellular and tissue levels12,11. It is also important to visualize the live dynamics of GPCRs in order to understand potential functional changes in physiological and pathological CB1 signaling is related to cytoskeleton. To evaluate the dependence of CB1 periodic hotspots on the cytoskeleton, we TURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 7 7 ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 In conclusion, we found that CB1 is distributed in axons as organized hotspots separated by approximately 190 nm, in a similar spatial distribution to the MPS. Especially, CB1 tends to be more organized as periodic hotspots upon agonist application suggesting that active CB1 associate more strongly with the MPS. When a GPCR is activated by an agonist, it increases the kinetics of interaction with G proteins and conducts downstream signal transduction within hotspots, which are usually conﬁned to the cytoskeletal and clathrin-forming grids27. Our results suggest that active CB1 is clustered into hotspots where G proteins or β- arrestin can easily collide with and form a relatively stable interaction with intracellular signaling components to increase signaling efﬁcacy. Our results also indicate that CB1 are anchored to the MPS, likely by ankB, forming a fundamental structural unit that may be important for the proper function of neurons. These units likely form the structural basis for hotspots where signals transfer from extracellular to intracellular compartments. Our observation of periodic hotspots of CB1 along axon shafts, and the role of the cytoskeleton in CB1’s intracellular signaling, indicate a horizon for the study of structural−functional interaction in neurons. stained with secondary antibodies as described above in the incubation solution for 1 h at room temperature. After incubation, samples were washed with PBS. Neuron samples were mounted with ProLong Gold (P36930, Life Technology) and imaged. stained with secondary antibodies as described above in the incubation solution for 1 h at room temperature. After incubation, samples were washed with PBS. Neuron samples were mounted with ProLong Gold (P36930, Life Technology) and imaged. Perfusion and immunostaining of brain sections. Brieﬂy, adult male C57BL/6 mice and CB1 knockout mice were anesthetized with sodium pentobarbital (40 mg/ kg, i.p.) with no avoidance response to foot pinch. They were then perfused with normal saline (at 37 °C) and subsequently by ice-cold 4% PFA for ﬁxation. Brains were post-ﬁxed in 4% PFA for 4 h, then dehydrated in 30% sucrose. Brains were frozen at −80 °C and then sectioned at 20-μm-thick with the freezing microtome (Leica CM1950) for immunoﬂuorescence labeling. Methods The HyD time gating was set at 0.3–6 ns for confocal and 0.5–6 ns for STED, and the power of 633 nm laser and 775 nm laser was 20–30 pW and 7–8 mW, respectively. Other acquiring parameters were similar to the above. Deconvolution of STED images was performed by Huygens software (Scientiﬁc Volume Imaging) with the Huygens classical maximum likelihood estimation (CMLE) deconvolution algorithm and theoretical Point Spread Functions. The deconvolution was performed by using Deconvolution Wizard. Deconvolution parameters were set to match the refractive index (RI) of the mounting media (RI of ProLong Gold polymerized until complete dry was 1.47) and the RI of the objective oil (1.518). Detailed deconvolution parameters were as follows, background, automatic estimation; estimate mode, lowest; area radius, 0.7; deconvolution algorithm, CMLE; maximum iteration, 40; signal-to-noise ratio, 7–10; quality threshold, 0.001; iteration mode, optimized; PSFs per brick, one PSF; brick layout, auto. For single-color imaging on ﬁxed brain slices, CB1 was labeled with Alexa Fluor 647, excited by the WLL at 633 nm and depleted by 775 nm laser. The emission spectrum was set at 657–700 nm. In order to reduce the adverse effect of aberration and scatter caused by the thickness of tissue, only those processes in close proximity (<1 μm in z axis) and parallel to cover glasses were chosen to acquire 2D images. The HyD time gating was set at 0.3–6 ns for confocal and 0.5–6 ns for STED, and the power of 633 nm laser and 775 nm laser was 20–30 pW and 7–8 mW, respectively. Other acquiring parameters were similar to the above. C57BL/6 female mice and KM mouse strains were used as embryo donors and pseudo-pregnant foster mothers, respectively. Superovulated female C57BL/6 mice (3–4 weeks old) were mated to C57BL/6 stud males, and fertilized embryos were collected from the ampullae. Different concentrations of Cas9 mRNA and sgRNAs were mixed and co-injected into the cytoplasm of one-cell stage fertilized eggs. After injection, surviving zygotes were transferred into oviducts of KM albino pseudo-pregnant females and allowed to develop to term. Deconvolution of STED images was performed by Huygens software (Scientiﬁc Volume Imaging) with the Huygens classical maximum likelihood estimation (CMLE) deconvolution algorithm and theoretical Point Spread Functions. The deconvolution was performed by using Deconvolution Wizard. Methods Deconvolution parameters were set to match the refractive index (RI) of the mounting media (RI of ProLong Gold polymerized until complete dry was 1.47) and the RI of the objective oil (1.518). Detailed deconvolution parameters were as follows, background, automatic estimation; estimate mode, lowest; area radius, 0.7; deconvolution algorithm, CMLE; maximum iteration, 40; signal-to-noise ratio, 7–10; quality threshold, 0.001; iteration mode, optimized; PSFs per brick, one PSF; brick layout, auto. p p g p Mutant mice were genotyped to ensure the deletion of target CB1 segment. To mitigate off-targets effects, mutant mice were crossed into C57Bl/6 for ﬁve generations before being used for experimental purposes. Primary culture of rat neurons. Sprague−Dawley rats of either sex at P0 were used for culturing rat hippocampal or cortex neurons. Brain tissues were isolated and digested with papain (1 mg/mL; Sigma, P3125) at 37 °C for 30 min. The digested tissues were washed with Dulbeccoʼs Modiﬁed Eagle Medium (DMEM) solution (Hyclone, SH30243.FS) three times and then transferred to culture medium containing Neurobasal medium (Thermo Fisher Scientiﬁc, 21103049) supplemented with 2% (vol/vol) B27 supplement (Thermo Fisher Scientiﬁc, 17504044) and 1% (vol/vol) Glutamax (Thermo Fisher Scientiﬁc, 35050-061). The tissues were gently triturated until no chunks of tissue were left. Dissociated cells were then counted and plated onto poly-D-lysine-coated 12-mm coverslips (12- 545-80, Fisher Brand) or 29 mm Glass (#1.5 cover glass) bottom dishes (D29-20- 1.5-N, Cellvis). Neuronal cultures were maintained in the culture medium in a humidiﬁed atmosphere with 5% CO2 at 37 °C. One-half of the medium was changed every 3 days to maintain neuron viability. y To measure the spatial resolution of the STED imaging microscope, 20 nm ﬂuorescent microspheres (F8887, Thermo Fisher Scientiﬁc) were used to measure the performance of microscope. Beads were diluted 100-fold with ddH2O and then sonicated in order to disperse bead aggregates. 1 μL beads were fully spread on a 0.17-um-thick and PLL-coated coverslip. After about ten minutes with those beads dried, coverslips were mounted on the slide with a thin layer of mounting media (ProLong Gold). The slide was kept at room temperature until the ProLong Gold was fully dried. The microspheres were excited by a 561 nm laser and depleted by a 660 nm laser. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 For the ﬂuorescence immunostainings procedure, brain sections were rinsed in PBS, permeabilized, and then blocked with blocking solution (3% w/v donkey serum and 0.3% v/v Triton X-100 in PBS) for 2 h at room temperature and then overnight at 4 °C with the primary antibody in 0.1% Triton X-100 and 1% serum in PBS. After washing with PBS, sections were incubated with the secondary antibody in 0.1% Triton X-100 and 1% serum in PBS for 2 h at room temperature. After incubation, sections were washed with PBS. Sections were mounted with ProLong Gold mounting medium for following imaging. STED imaging. Confocal and STED images were obtained at a Leica TCS SP8‐3X gated STED system (Leica Microsystems) equipped with a pulsed white light laser (WLL, tunable from 470 to 670 nm) for excitation, a 592 nm CW laser (MPB Communications), a 660 nm CW laser (MPB Communications), and a 775 nm pulsed laser (Oneﬁve) for depletion. The system includes a ×100 objective lens (Leica, HC PL APO CS2 ×100/1.40 oil), two HyD detectors, and the TCS SP8 time- gated system. Each time before imaging, the depletion laser was co-aligned with the excitation laser. By STED effect, the lateral ﬂuorescence was ﬁltered out, leaving the remaining ﬂuorescence in the center to be detected. ARTICLE ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 However, in the presence of antagonist, CB1 still remained periodic in some regions of axons in primary culture neurons and its overall periodic pattern was not different. This suggests that the periodic organization of CB1 in native state might not be caused by constitutive released ligand in culture neurons, though we could not exclude this possibility in the brain tissue. Our results using the same antibody used in Zhou’s experi- ments showed both periodic and nonperiodic hotspots in both culture neurons and native brain tissue. We observed the peri- odicity of CB1 clusters in live neurons to avoid artifacts caused by ﬁxation procedures. Further, we found that active CB1 may behave differently than native ones, as CB1 displayed less dynamic and more conﬁned movements upon WIN application. Our molecular biology and imaging results suggest that the ICL3 may participate in the association between CB1 and MPS com- ponents, such as ankB and spectrin. In the presence of WIN, CB1 NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 8 Methods Animals. C57BL/6 mice, Sprague−Dawley rat, CB1 knockout mice (Biocytogen) and Kunming (KM) mice (for knockout mice generation) were used in this study. Male C57BL/6 mice and CB1 knockout mice at 8–12 weeks were used. Animals were housed under a 12 h light/dark cycle at a room temperature of 22 ± 1 °C with 45% humidity, given ad libitum access to food and water. All experimental pro- cedures were approved by the Institutional Animal Care and Use Committee of ShanghaiTech University, China. For single or dual color 2D imaging on the primary neuron culture, samples were excited by the WLL at 488 nm (Alexa Fluor 488-labeled βII-spectrin or ankB) and 561 nm (Alexa Fluor 555-labeled CB1) and depleted by 592 and 660 nm laser, respectively. The emission spectrums were set at 493–550 nm and 566–600 nm accordingly. Channels were recorded sequentially. Acquisition parameters were optimized as follows: logical size, 1024 ×1024 pixels; pixel size, 23 nm; scan speed, 400 Hz; scan direction, unidirectional; line average 4, no frame average; optical zoom factor 5, pinhole 1 airy disc; detector gain, 100%; HyD time gating, 0.3–6 ns for confocal and 1–6 ns for STED; excitation power 10–30 pW; STED power, 11–17 mW for 592 nm and 5–8 mW for 660 nm. Generation of Cnr1 knockout mouse. For Cnr1 gene targeting, two sgRNAs were designed to target either the upstream or downstream region of its coding sequence by the CRISPR design tool (http://crispr.mit.edu) and screened for on-target activity using UCATM (Universal CRISPR Activity Assay, Biocytogen). PCR ampliﬁcation was performed to add the T7 promoter sequence to the Cas9 mRNA and sgRNAs DNA template and then the T7-Cas/sgRNA PCR products were gel puriﬁed. They were used as the template for in vitro transcription with the MEGAshortscript T7 kit (Cat. No. AM1354, Life Technologies). The Cas9 mRNA and sgRNAs products were puriﬁed with MEGAclear kit (Cat. No. AM1908, Life Technologies) and eluted with RNase-free water. For single-color imaging on ﬁxed brain slices, CB1 was labeled with Alexa Fluor 647, excited by the WLL at 633 nm and depleted by 775 nm laser. The emission spectrum was set at 657–700 nm. In order to reduce the adverse effect of aberration and scatter caused by the thickness of tissue, only those processes in close proximity (<1 μm in z axis) and parallel to cover glasses were chosen to acquire 2D images. NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications ARTICLE Immuno-complex was incubated with Protein A-Magnetic beads overnight on a rotating wheel at 4 °C from the precleared supernatant with anti-CB1 antibody covalently coupled to Protein G-Magnetic beads. The pellet was then washed 5 times in wash buffer containing (in mM), 20 Tris-HCl, 100 NaCl, 1 Ethylenediaminetetraacetic acid (EDTA), 0.5% Nonidet P- 40 (pH = 8.0). The beads were to wash away any proteins nonspeciﬁcally bound to the beads. The immunoprecipitates were mixed with the loading buffer and resolved by SDS-PAGE. Western blots were performed with relevant antibodies. Rabbit IgG was used as a negative control. In situ PLAs. The protocol for PLA can be found on Duolink in situ PLA detection Kit (Sigma) following the instructions of the supplier with slight modiﬁcations31. Neurons at DIV 14–18 were ﬁxed and incubated with primary antibodies overnight as described in the section of immunoﬂuorescence staining. On the next day neurons were washed 3× with Phosphate-Buffered Saline, 0.1% Tween® 20 Detergent (PBST). All remaining steps were carried out in a 37 °C humidiﬁed chamber. To detect CB1−cytoskeleton protein interaction, a mixture of equal amounts of anti-CB1 antibody directly linked to a plus PLA probe and cytoskeleton protein antibody (like ankyrin B or βII-spectrin) directly linked to a minus PLA probe was used. Neurons were incubated with a pair of PLA probes diluted 1:10/ each in 1× blocking solution for 1 h, before washing 3× at 10 min each with 1× Washing Buffer-A. Ligase was diluted 1:40 into 1× Ligation buffer and added to neurons for 0.5 h ligation, followed by 3 × 5 min washes with 1× Washing Buffer-A. Polymerase was then applied 1:80 to neurons in 1× ampliﬁcation stock for 2.5 h. Neurons were washed 2 × 10 min with 1× Washing Buffer-B, 1 × 10 min with 0.01× Washing Buffer-B, and 1 × 10 min with PBST. Neurons were ﬁxed again by 4% PFA for 10 min to better retain the PLA signals before incubated with axonal primary antibody like MAP2 or Tuj1 at room temperature for 2 h. Then neurons incubated with ﬂuorescent secondary antibody for 1 h at room temperature before mounting with Mounting Medium with 4′,6-Diamidino-2-Phenylindole, Dihy- drochloride (DAPI). The samples were observed in a Nikon confocal microscope equipped with an apochromatic ×60 oil-immersion objective (N.A. 1.4), and a 405, a 488 and a 561 nm laser line. ARTICLE The membranes were then incubated with HRP-conjugated secondary antibody (1:1000; Pierce) for 2 h at room temperature. Signals were visualized using enhanced chemilumines- cence (ECL, Pierce, Rockford, IL), and captured by the ChemiDoc XRS system (Bio-Rad Laboratories, CA). Phosphorylated ERK1/2 or phosphorylated Akt levels were normalized for differences in loading using protein band intensities for total ERK1/2 or AKT. g q The raw SIM images were reconstructed by the OMX SI reconstruction tool available in softWoRx (GE). Channel-speciﬁc OTF ﬁles, channel-speciﬁc K0 angles, Wiener ﬁlter constant (0.001), and bias offset (65) were used during the reconstruction. Wideﬁeld images were generated by averaging phase steps from SIM raw images. To detect the SIM microscope resolution, the bead samples were made similarly to those for STED, but use Vectashield (H-1000, Vectorlabs) as mounting medium instead of Prolong Gold. Slides were sealed with nail polish and imaged. The imaging procedure was similar to the live SIM imaging. FWHM was determined with the same method to STED. After SIM reconstruction, the images were processed in the FIJI software. The rigid image drift during acquisition was corrected with “Correct 3D drift” plugin29. Images were maximum-intensity projected and the linearly contrast was adjusted using Fiji software (National Institutes of Health). For the analysis and display, only those examples that could be tracked and did not bleach more than 20% for consecutive three time frames were selected. Fiji TrackMate was then used to analyze the dynamics of CB1 hotspots in these images30. In TrackMate, the difference of Gaussian (DoG) detector was set with an estimated spot diameter of 80 nm and an appropriate ﬂuorescence intensity as the threshold to detect all the individual CB1 hotspots. The Simple Linear Assignment Problem (LAP) tracker with a linking maximum distance of 200 nm, a gap-closing maximum distance of 200 nm, and a gap-closing maximum frame gap of 2 was used for tracking cells through the time course images. Immunoprecipitation. Following tetracycline induction, CB1-CHO cells were washed with ice-cold PBS and suspended in immunoprecipitation (IP) buffer containing (in mM), 50 Tris-HCl, 120 NaCl, 0.5% Nonidet P-40 and protease cocktail (pH = 7.5). The lysate was sonicated, centrifuged at 13,000 × g for 20 min at 4 °C, and the resulting supernatant was incubated with the rabbit anti-CB1 (CST, 93815) antibody for 20 min at 4 °C. ARTICLE The image ﬂuorescence over time was normalized to initial ﬂuor- escence (average ﬂuorescence value from the 5-s imaging period immediately before photobleaching, deﬁned as 100%) and immediately after bleaching (0%). Analysis of the recovery ﬁtting curves and the half-time recovery were carried out with the Fiji macro (http://imagej.net/Analyze_FRAP_movies_with_a_Jython_script)32. p g Primary neurons were transiently transfected using lipofectamine 2000 (Invitrogen) at DIV 9–12 (2 µg plasmid DNA/20 mm dish). Brieﬂy, neurons were mixed with transfection complexes containing CB1-RFP plasmid DNA and lipofectamine 2000 at a ratio of 1:2 in MEM medium (Thermo Fisher Scientiﬁc, 11090081) for 2 h, then subsequently incubated with the culture medium. One day after transfection, neurons expressed CB1-RFP were imaged under the live 3D-SIM mode. The cells were incubated with Neurobasal medium (without phenol red, Thermo Fisher Scientiﬁc, 12348017) at 37 °C and supplied with 5% CO2 mixed with 95% air during the imaging process. To obtain images with minimized spherical aberration and optimized illumination contrast, immersion oil with different refractive index (RI) were systematically optimized for each individual sample until a symmetrical point spread function was achieved. Usually the oil with RI 1.520 was chosen. To reduced phototoxicity, the illumination laser intensity and the exposure time were carefully adjusted to a minimum value (10% of maximum laser intensity; exposure time, 6–10 ms) and only 1 μm in depth (eight optical sections) were acquired. The time series of 3D-SIM images were acquired six frames with 1 or 3 min acquisition interval. An ultimate focus system was used to maintain the sample Z position, regardless of the mechanical and thermal changes during the acquisition. Western blot. Primary neurons were either treated or not treated with the indi- cated ligands for the times noted, rinsed with ice-cold PBS, and lysed by the addition of 100 μL of ice-cold RIPA lysis buffer (Beyotime) with protease inhibitor cocktail (Roche). Cellular debris was removed by centrifugation at 13,000 × g for 15 min at 4 °C, and the amount of protein was quantiﬁed by BCA protein assay kit (Pierce). Protein samples were separated on 10% SDS-PAGE gels (Bio-Rad) and transferred onto PVDF (polyvinylidene ﬂuoride) membranes (Millipore, Billerica, MA, USA). The membranes were blocked with 5% non-fat dry milk for 2 h at RT and incubated overnight at 4 °C with rabbit anti-phospho-ERK1/2 (1:2000, Cell Signaling), anti-phospho-Ser473-Akt (1:2000, Cell Signaling), anti-ERK1/2 (1:2000, Cell Signaling), and anti-AKT (1:2000, Cell Signaling) antibodies. ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 (7–10 ms per frame) and 1-s interval. No images were acquired during the bleach period. Putative axonal regions transfected with CB1-RFP were chosen and an image stack of 1–2 μm with 0.4-μm interval in in Z axis were acquired. Ultimate focus system was also used to maintain the stability of the microscope. Six pre-bleach images were taken followed by bleaching a 1-μm-diameter circle region for 0.1 ms with a strong laser intensity. Immediately after bleach, post-bleach images were taken until the recovery reached a complete steady state. Images were maximum projected and processed in Fiji. Brieﬂy, FRAP data were quantiﬁed using Fiji software. During image processing, the background level was subtracted from each frame to quantify the recovery of ﬂuorescence. Average ﬂuorescence intensity of the bleached area was corrected for additional bleaching during regular imaging and was normalized to pre- bleach intensity. The image ﬂuorescence over time was normalized to initial ﬂuor- escence (average ﬂuorescence value from the 5-s imaging period immediately before photobleaching, deﬁned as 100%) and immediately after bleaching (0%). Analysis of the recovery ﬁtting curves and the half-time recovery were carried out with the Fiji macro (http://imagej.net/Analyze_FRAP_movies_with_a_Jython_script)32. space of 0.125 μm. Fifteen images were taken with three illumination angles and ﬁve phase shifts for each Z-section. Images were taken in fast 286 MHz mode. space of 0.125 μm. Fifteen images were taken with three illumination angles and ﬁve phase shifts for each Z-section. Images were taken in fast 286 MHz mode. (7–10 ms per frame) and 1-s interval. No images were acquired during the bleach period. Putative axonal regions transfected with CB1-RFP were chosen and an image stack of 1–2 μm with 0.4-μm interval in in Z axis were acquired. Ultimate focus system was also used to maintain the stability of the microscope. Six pre-bleach images were taken followed by bleaching a 1-μm-diameter circle region for 0.1 ms with a strong laser intensity. Immediately after bleach, post-bleach images were taken until the recovery reached a complete steady state. Images were maximum projected and processed in Fiji. Brieﬂy, FRAP data were quantiﬁed using Fiji software. During image processing, the background level was subtracted from each frame to quantify the recovery of ﬂuorescence. Average ﬂuorescence intensity of the bleached area was corrected for additional bleaching during regular imaging and was normalized to pre- bleach intensity. ARTICLE For each ﬁeld of view, a stack of three channels (one per staining) and 10−15 Z stacks with a step size of 1 μm were acquired. Images were opened and processed with Fiji software. The ratio r (number of red spots/ number of cells containing spots) was determined considering a total of 300−600 cells from 6 to 10 different ﬁelds. Methods The acquiring parameters were optimized as follows: logical size, 1024 ×1024 pixels; pixel size, 11 nm; scan speed, 400 Hz; scan direction, unidirectional; line average 4, no frame average; optical zoom factor 10; pinhole 1 airy disc; detector gain, 100%; HyD time gating, 0.3–6 ns for confocal and 1–6 ns for STED; excitation power, 30–40 pW; STED power, 15 mW. Intensity proﬁles across the center of beads were plotted in the FIJI software, and the FWHM (Full width at half maximum) was calculated by ﬁtting with a Gaussian function28. To measure the spatial resolution of the STED imaging microscope, 20 nm ﬂuorescent microspheres (F8887, Thermo Fisher Scientiﬁc) were used to measure the performance of microscope. Beads were diluted 100-fold with ddH2O and then sonicated in order to disperse bead aggregates. 1 μL beads were fully spread on a 0.17-um-thick and PLL-coated coverslip. After about ten minutes with those beads dried, coverslips were mounted on the slide with a thin layer of mounting media (ProLong Gold). The slide was kept at room temperature until the ProLong Gold was fully dried. The microspheres were excited by a 561 nm laser and depleted by a 660 nm laser. The acquiring parameters were optimized as follows: logical size, 1024 ×1024 pixels; pixel size, 11 nm; scan speed, 400 Hz; scan direction, d l l f l f h l Fluorescence labeling of neurons. Cultured neurons were ﬁxed with 4% (w/v) paraformaldehyde (PFA) in phosphate-buffered saline (PBS) for 15 min at 14 day in vitro (DIV 14). After a complete wash with PBS, the samples were then per- meabilized and blocked in the blocking buffer (10% v/v donkey serum, 0.2% v/v Triton X-100 in PBS) for 1 h at room temperature, and subsequently stained with one or two primary antibodies in the incubation buffer (1% donkey serum, 0.1% Triton X-100 in PBS) overnight at 4 °C. The samples were washed in PBS and then Live SIM imaging the structure of CB1 in neurons. SIM images were obtained with the GE DeltaVision OMX microscope, equipped with a 568-nm laser (Coherent), a ×60 objective lens (Olympus, PL Apo N ×60/1.42 oil), and the scientiﬁc CMOS camera (Acquisition pixel size, 82 nm at ×60 objective; PCO edge). 3D-SIM mode was acquired with a ﬁxed 512 × 512 pixel size and an optical section 9 ARTICLE Peptide precursor ions in different brain region protein digests were monitored in the PRM assay by scheduling an inclusion list of each precursor with an isolation window of 1.6 m/z and retention time shift of 2 min. 11. Zhou, R., Han, B., Xia, C. & Zhuang, X. Membrane-associated periodic skeleton is a signaling platform for RTK transactivation in neurons. Science 365, 929–934 (2019). 12. Dudok, B. et al. Cell-speciﬁc STORM super-resolution imaging reveals nanoscale organization of cannabinoid signaling. Nat. Neurosci. 18, 75–86 (2015). 13. Hell, S. W. Microscopy and its focal switch. Nat. methods 6, 24–32 (2009). 14. Zhong, G. S. et al. Developmental mechanism of the periodic membrane skeleton in axons. Elife 3, e04581 (2014). f 15. Heintzmann, R. & Huser, T. Super-resolution structured illumination microscopy. Chem. Rev. 117, 13890–13908 (2017). MS data processing. Mass spectra from the DDA experiment were processed using Proteome Discoverer 2.1 against the Uniprot mouse sequence database. The search parameter included cysteine carbamidomethylation as a ﬁxed modiﬁcation and oxidation of methionine as variable modiﬁcation. Precursor ion mass tolerance was set to 10 ppm and fragment ion tolerance was 0.02 Da. Trypsin was set as the speciﬁc enzyme and two missed cleavages were allowed. The required false dis- covery rate (FDR) was set to 1% at the peptide and protein level. For PRM data analysis, the Skyline software (v3.7.0) was used for targeted peptide quantiﬁcation with settings speciﬁed by the software instruction36. Only b- or y-product ions with m/z values greater than the precursor were selected to quantify the peptides. All the transitions were validated using the mProphet algorithm in Skyline advanced peak picking model that restricts the FDR to <1%. The peptide quantiﬁcation was derived from the sum of the peak areas of 3 −6 product ions for selected peptides. Protein intensity was based on the summed MS responses of one to three unique peptides of the corresponding protein. MS data processing. Mass spectra from the DDA experiment were processed using Proteome Discoverer 2.1 against the Uniprot mouse sequence database. The search parameter included cysteine carbamidomethylation as a ﬁxed modiﬁcation and oxidation of methionine as variable modiﬁcation. Precursor ion mass tolerance was set to 10 ppm and fragment ion tolerance was 0.02 Da. Trypsin was set as the speciﬁc enzyme and two missed cleavages were allowed. The required false dis- covery rate (FDR) was set to 1% at the peptide and protein level. 16. ARTICLE ARTICLE NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 amygdala-dependent synaptic and behavioral functions. J. Neurosci. 33, 10264–10277 (2013). NanoLC-MS/MS analysis. The nanoLC-MS/MS analysis was conducted on an EASY-nLC 1000 connected to Orbitrap Fusion mass spectrometer (Thermo Fisher Scientiﬁc, USA) with a nano-electrospray ionization source. The eluted peptides were separated on an analytical column (200 mm × 75 μm) in-house packed with C18-AQ 3 μm C18 resin (Dr. Maisch, GmbH, Germany) over a 130-min gradient at ﬂow rate of 300 nL/min. For a pooled sample from all six brain region protein digests, a data-dependent (DDA) acquisition method was ﬁrst employed with the following parameters, resolution of 60,000 used for survey scans; the mass range pf 300−1700 m/z; an AGC target value of 4E5; and maximum ion injection time of 50 ms. Up to 12 dynamically chosen and most abundant precursor ions were fragmented. The MS/MS scans were acquired at an Orbitrap resolution of 30,000 (AGC target value 1E5, maximum ion injection time 50 ms). 6. Hua, T. et al. Crystal structure of the human cannabinoid receptor CB1. Cell 167, 750–762.e714 (2016). 7. Hua, T. et al. Crystal structures of agonist-bound human cannabinoid receptor CB1. Nature 547, 468–471 (2017). 8. Xu, K., Zhong, G. S. & Zhuang, X. W. Actin, spectrin, and associated proteins form a periodic cytoskeletal structure in axons. Science 339, 452–456 (2013). 8. Xu, K., Zhong, G. S. & Zhuang, X. W. Actin, spectrin, and associated proteins f i di t k l t l t t i S i 339 452 456 (2013) 9. Liu, Y. et al. Critical role of spectrin in hearing development and deafness. Sci. Adv. 5, eaav7803 (2019). 10. Reddy-Alla, S. et al. Stable positioning of Unc13 restricts synaptic vesicle fusion to deﬁned release sites to promote synchronous neurotransmission. Neuron 95, 1350–1364.e1312 (2017). In order to achieve the accurate quantiﬁcation of selected proteins in the membrane fractions, we developed parallel-reaction-monitoring (PRM) MS assays for all proteins of our interest based on the protein identiﬁcation results from the DDA experiment35. The PRM acquisition method started with a full scan event followed by targeted MS/MS for speciﬁc peptides from the proteins of interest. Major parameters for the MS/MS event in Orbitrap were a resolution of 30,000; an AGC target value of 2E5; and maximum injection time of 100 ms. ARTICLE Sanchez-Blazquez, P., Rodriguez-Munoz, M., Vicente-Sanchez, A. & Garzon, J. Cannabinoid receptors couple to NMDA receptors to reduce the production of NO and the mobilization of zinc induced by glutamate. Antioxid. Redox Signal. 19, 1766–1782 (2013). g 17. Polissidis, A. et al. The cannabinoid CB1 receptor biphasically modulates motor activity and regulates dopamine and glutamate release region dependently. Int. J. Neuropsychoph. 16, 393–403 (2013). y For PRM data analysis, the Skyline software (v3.7.0) was used for targeted peptide quantiﬁcation with settings speciﬁed by the software instruction36. Only b- or y-product ions with m/z values greater than the precursor were selected to quantify the peptides. All the transitions were validated using the mProphet algorithm in Skyline advanced peak picking model that restricts the FDR to <1%. The peptide quantiﬁcation was derived from the sum of the peak areas of 3 −6 product ions for selected peptides. Protein intensity was based on the summed MS responses of one to three unique peptides of the corresponding protein. 18. Bennett, V. & Baines, A. J. Spectrin and ankyrin-based pathways, metazoan inventions for integrating cells into tissues. Physiol. Rev. 81, 1353–1392 (2001). 19. Schallmeiner, E. et al. Sensitive protein detection via triple-binder proximity ligation assays. Nat. Methods 4, 135–137 (2007). 20. Gardner, K. & Bennett, V. Modulation of spectrin-actin assembly by erythrocyte adducin. Nature 328, 359–362 (1987). 21. Blazquez, C. et al. The CB(1) cannabinoid receptor signals striatal neuroprotection via a PI3K/Akt/mTORC1/BDNF pathway. Cell Death Differ. 22, 1618–1629 (2015). Quantiﬁcation and statistical analysis. All image data were ﬁrst processed with Fiji software (National Institutes of Health). Images were resized with the Bicubic interpolation and the brightness and contrast were linearly adjusted. To quantitatively analyze the distribution properties of CB1, segmented lines across the structures were drawn, the intensity proﬁles along the lines were measured and further analyzed in Matlab (MathWorks, Inc.). Individual ﬂuorescence peaks were found and the distance between the neighboring peaks were calculated and pooled together. For distribution pattern analysis, autocorrelation and cross- correlation were performed on the ﬂuorescence intensity proﬁles. The correlation curve was pooled and averaged from many randomly selected lines. All intensity, distance and correlation data were plotted using Graphpad prism (Graphpad Software, Inc.), and all ﬁgure layouts were composed in Illustrator (Adobe Systems, Inc.). 22. Franklin, J. M. & Carrasco, G. A. Cannabinoid receptor agonists upregulate and enhance serotonin 2A (5-HT(2A)) receptor activity via ERK1/2 signaling. Synapse 67, 145–159 (2013). ARTICLE 23. Turu, G. & Hunyady, L. Signal transduction of the CB1 cannabinoid receptor. J. Mol. Endocrinol. 44, 75–85 (2010). 24. Rozenfeld, R. et al. AT1R-CB(1)R heteromerization reveals a new mechanism for the pathogenic properties of angiotensin II. EMBO J. 30, 2350–2363 (2011). 25. Wickert, M. et al. The F238L point mutation in the cannabinoid type 1 receptor enhances basal endocytosis via lipid rafts. Front. Mol. Neurosci. 11, 230 (2018). 26. Perreault, M. L., Hasbi, A., O’Dowd, B. F. & George, S. R. Heteromeric dopamine receptor signaling complexes, emerging neurobiology and disease relevance. Neuropsychopharmacology 39, 156–168 (2014). Results were reported as mean ± s.e.m. Statistical analysis of the data was performed using a Student’s t test, one-way ANOVA. Statistical signiﬁcance was set at p < 0.05. 27. Sungkaworn, T. et al. Single-molecule imaging reveals receptor-G protein interactions at cell surface hot spots. Nature 550, 543–547 (2017). Received: 21 January 2020; Accepted: 15 October 2020; 30. Tinevez, J. Y. et al. TrackMate, an open and extensible platform for single- particle tracking. Methods 115, 80–90 (2017). 31. Tseng, W. C., Jenkins, P. M., Tanaka, M., Mooney, R. & Bennett, V. Giant ankyrin-G stabilizes somatodendritic GABAergic synapses through opposing endocytosis of GABAA receptors. Proc. Natl Acad. Sci. USA 112, 1214–1219 (2015). Mass spectra in brain tissues p Mouse brain tissue preparation. Preparation of mouse brain membrane fractions was performed according to a previous study33. Brieﬂy, six brain regions (olfactory bulb, cerebral cortex, cerebellum, hippocampus, midbrain, and spinal cord) were obtained from 9-week-old C57BL/6 wild-type male mice. Brain regions from three mice were pooled and homogenized in the buffer of 300 mM sucrose, 0.5% bovine serum albumin (BSA), 100 mM EDTA, 30 mM Tris/HCl, pH 7.4 with protease inhibitor (Roche). Crude membrane fractions were isolated from the homogenate by ultra-centrifugation at 160,000 × g at 4 °C for 1 h. The membrane pellet was solubilized in 4% SDS and 100 mM dithiothreitol (DTT) in 100 mM Tris/HCl, pH 7.6, denatured and reduced at 95 °C for 3 min. Protein concentration was deter- mined using the BCA assay. For each brain region, protein sample preparation was conducted in duplicate. Protein digestion. The SDS-assisted digestion of membrane proteins was performed according to methods described previously34. Brieﬂy, 50 μg of protein was diluted in 8 M urea, 50 mM NH4HCO3, and exchanged to the same buffer using the 30 KDa MWCO centrifugal ﬁlter unit (Satorious, Germany) by centrifugation at 13,000×g for 20 min. The following centrifugation steps were performed under the same conditions. Subsequently, 100 μL of 50 mM iodoacetamide in 8 M urea, 50 mM NH4HCO3 was added and incubated at room temperature in darkness for 30 min, then followed by centrifugation. The concentrate was diluted with 200 μL 50 mM NH4HCO3 and centrifuged again, this step was repeated twice. Proteins were digested with trypsin (Promega, Madison, USA) at an enzyme-to-protein ratio of 1:100 (w/w) at 37 °C for 3 h, followed by the addition of trypsin at 1:50 (w/ w) and incubation at 37 °C overnight. After acidiﬁcation, the protein digest was desalted with C18-SepPak columns (Waters, Milford, USA) and lyophilized under vacuum. Fluorescence recovery after photobleaching. FRAP experiments on the culture neuron transfected with CB1-RFP were also performed on the GE DeltaVision OMX microscope with a ×60, 1.47 NA objective. Culture neurons were transfected with CB1-RFP 1 day before imaging. Prior to the experiment, the medium was half changed to Neurobasal without phenol red. The cells were maintained at room temperature with 5% CO2 mixed with 95% air supplied during the whole imaging process. Each FRAP image was taken with a ﬁxed laser intensity and exposure time NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 10 Data availability 28. Nagerl, U. V. & Bonhoeffer, T. Imaging living synapses at the nanoscale by STED microscopy. J. Neurosci. 30, 9341–9346 (2010). y All data are available upon reasonable request. Source data are provided with this paper. All data are available upon reasonable request. Source data are provided with this paper. 29. Parslow, A. C., Cardona, A. & Bryson-Richardson, R. J. Sample drift correction following 4D confocal time-lapse imaging. J. Vis. Exp. https://doi. org/10.3791/51086 (2014). Received: 21 January 2020; Accepted: 15 October 2020; Received: 21 January 2020; Accepted: 15 October 2020; Author contributions Open Access This article is licensed under a Creative Commons Attri- bution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. G.Z. initiated, managed and supervised the project. G.Z. and H.L. conceived and designed the experiments. H.L., J.Y., T.C., S.L., M.D., T.H., performed most of the experiments and data analysis. S.Z., Q.W., F.T., Y.Q., D.D.-G., G.T., Y.Z., W.S., T.W. and Z.-J.L. contributed to data analysis and interpretation. G.Z. wrote the manuscript with contributions from all of the authors. NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 NATURE COMMUNICATIONS \| https://doi.org/10.1038/s41467-020-19510-5 Correspondence and requests for materials should be addressed to G.Z. Reprints and permission information is available at http://www.nature.com/reprints Reprints and permission information is available at http://www.nature.com/reprints Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. Correspondence and requests for materials should be addressed to G.Z. Correspondence and requests for materials should be addressed to G.Z. We thank the Bioimaging Core Facilities of the iHuman Institute and the animal facility of National Center for Protein Science for their support. We thank Dr. Yulong Li from Peking University for providing the GPCR probes. This work was supported by the National Key Research and Development Program of China (2016YFC0905900 (G.Z.), 2017YFC 1001300 (G.Z.)), the National Natural Science Foundation of China (81970878 (G.Z.), 31871036 (T.W.), 31771130 (G.Z.), 81861128023 (G.Z.)), the 2015 Thousand Youth Talents Plan of China (G.Z.) and the Australian Research Council (Grant DE170100546 (T.W.)). Peer review information Nature Communications thanks the anonymous reviewers for their contributions to the peer review of this work. References 1. Herkenham, M. et al. Cannabinoid receptor localization in brain. Proc. Natl Acad. Sci. USA 87, 1932–1936 (1990). 32. Woodruff, J. B. et al. The centrosome is a selective condensate that nucleates microtubules by concentrating tubulin. Cell 169, 1066-+ (2017). 2. Agarwal, N. et al. Cannabinoids mediate analgesia largely via peripheral type 1 cannabinoid receptors in nociceptors. Nat. Neurosci. 10, 870–879 (2007). 2. Agarwal, N. et al. Cannabinoids mediate analgesia largely via peripheral type 1 cannabinoid receptors in nociceptors. Nat. Neurosci. 10, 870–879 (2007). 33. Suski, J. M. et al. Isolation of plasma membrane-associated membranes from rat liver. Nat. Protoc. 9, 312–322 (2014). 34. Wisniewski, J. R., Zougman, A., Nagaraj, N. & Mann, M. Universal sample preparation method for proteome analysis. Nat. Methods 6, 359–362 (2009). 3. Bellocchio, L. et al. Bimodal control of stimulated food intake by the endocannabinoid system. Nat. Neurosci. 13, 281–283 (2010). 35. Gallien, S., Kim, S. Y. & Domon, B. Large-scale targeted proteomics using internal standard triggered-parallel reaction monitoring (IS-PRM). Mol. Cell. Proteom. 14, 1630–1644 (2015). 4. Monory, K. et al. The endocannabinoid system controls key epileptogenic circuits in the hippocampus. Neuron 51, 455–466 (2006). 4. Monory, K. et al. The endocannabinoid system controls key epileptogenic circuits in the hippocampus. Neuron 51, 455–466 (2006). circuits in the hippocampus. Neuron 51, 455–466 (2006). 5. Ruehle, S. et al. Cannabinoid CB1 receptor in dorsal telencephalic glutamatergic neurons, distinctive sufﬁciency for hippocampus-dependent and 36. MacLean, B. et al. Skyline, an open source document editor for creating and analyzing targeted proteomics experiments. Bioinformatics 26, 966–968 (2010). 5. Ruehle, S. et al. Cannabinoid CB1 receptor in dorsal telencephalic glutamatergic neurons, distinctive sufﬁciency for hippocampus-dependent and NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 11 Competing interests p g The authors declare no competing interests. Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41467- 020-19510-5. pp 020-19510-5. NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunications 12 NATURE COMMUNICATIONS \| (2020) 11:5699 \| https://doi.org/10.1038/s41467-020-19510-5 \| www.nature.com/naturecommunicatio
https://openalex.org/W2903095258	https://hal-amu.archives-ouvertes.fr/hal-02002425/file/Frontiers%20Bioeng%20Biotech%202018.pdf	English	null	Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Without Surgical Reinnervation: A Preliminary Study	Frontiers in bioengineering and biotechnology	2,018	cc-by	11,877	Phantom-mobility-based prosthesis control in transhumeral amputees without surgical reinnervation: a preliminary study. Nathanaël Jarrassé, Étienne de Montalivet, Florian Richer, Caroline Nicol, Amelie Touillet, Noel Martinet, Jean Paysant, Jozina B de Graaf To cite this version: Nathanaël Jarrassé, Étienne de Montalivet, Florian Richer, Caroline Nicol, Amelie Touillet, et al.. Phantom-mobility-based prosthesis control in transhumeral amputees without surgical reinnervation: a preliminary study.. Frontiers in Bioengineering and Biotechnology, 2018, 6, 10.3389/fbioe.2018.00164. hal-02002425 Phantom-mobility-based prosthesis control in transhumeral amputees without surgical reinnervation: a preliminary study. Distributed under a Creative Commons Attribution 4.0 International License Keywords: prosthetics, transhumeral amputation, phantom limb, myoelectric control, pattern recognition, voluntary phantom limb mobility Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Without Surgical Reinnervation: A Preliminary Study Nathanaël Jarrassé 1, Etienne de Montalivet 1, Florian Richer 1, Caroline Nicol 2, Amélie Touillet 3, Noël Martinet 3, Jean Paysant 3 and Jozina B. de Graaf 2 1 CNRS, INSERM, Institut des Systèmes Intelligents et de Robotique, ISIR, Sorbonne Université, Paris, France, 2 CNRS, ISM, Aix Marseille Université, Marseille, France, 3 Centre Louis Pierquin, Institut Régional de Médecine Physique et de Réadaptation, UGECAM Nord-Est, Nancy, France Transhumeral amputees face substantial difﬁculties in efﬁciently controlling their prosthetic limb, leading to a high rate of rejection of these devices. Actual myoelectric control approaches make their use slow, sequential and unnatural, especially for these patients with a high level of amputation who need a prosthesis with numerous active degrees of freedom (powered elbow, wrist, and hand). While surgical muscle-reinnervation is becoming a generic solution for amputees to increase their control capabilities over a prosthesis, research is still being conducted on the possibility of using the surface myoelectric patterns speciﬁcally associated to voluntary Phantom Limb Mobilization (PLM), appearing naturally in most upper-limb amputees without requiring speciﬁc surgery. The objective of this study was to evaluate the possibility for transhumeral amputees to use a PLM-based control approach to perform more realistic functional grasping tasks. Two transhumeral amputated participants were asked to repetitively grasp one out of three different objects with an unworn eight-active-DoF prosthetic arm and release it in a dedicated drawer. The prosthesis control was based on phantom limb mobilization and myoelectric pattern recognition techniques, using only two repetitions of each PLM to train the classiﬁcation architecture. The results show that the task could be successfully achieved with rather optimal strategies and joint trajectories, even if the completion time was increased in comparison with the performances obtained by a control group using a simple GUI control, and the control strategies required numerous corrections. While numerous limitations related to robustness of pattern recognition techniques and to the perturbations generated by actual wearing of the prosthesis remain to be solved, these preliminary results encourage further exploration and deeper understanding of the phenomenon of natural residual myoelectric activity related to PLM, since it could possibly be a viable option in some transhumeral amputees to extend their control abilities of functional upper limb prosthetics with multiple active joints without undergoing muscular reinnervation surgery. Edited by: Raviraj Nataraj, Stevens Institute of Technology, United States Reviewed by: Luca Citi, University of Essex, United Kingdom Emanuele Lindo Secco, Liverpool Hope University, United Kingdom Correspondence: Nathanaël Jarrassé jarrasse@isir.upmc.fr Specialty section: This article was submitted to Bionics and Biomimetics, a section of the journal Frontiers in Bioengineering and Biotechnology Received: 01 July 2018 Accepted: 19 October 2018 Published: 29 November 2018 HAL Id: hal-02002425 https://amu.hal.science/hal-02002425v1 Submitted on 31 Jan 2019 L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés. HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. Distributed under a Creative Commons Attribution 4.0 International License ORIGINAL RESEARCH published: 29 November 2018 doi: 10.3389/fbioe.2018.00164 1. INTRODUCTION While there have been improvements of prosthetic solutions over the last years, the situation remains particularly complex for the common case of transhumeral (i.e., above elbow) amputees. The control is diﬃcult to learn, remains non-intuitive, and thus cognitively demanding. Moreover, these prostheses lack functionality and so do not provide the expected assistance in Activities of Daily Life (ADLs) (Biddiss and Chau, 2007). This leads to the development of compensatory strategies involving the rest of the body, causing shoulder, trunk, and contra-lateral limb disorders (Ostlie et al., 2011). The consequence is that transhumeral amputees are more likely to reject their prosthesis than transradial amputees (Wright et al., 1995; Biddiss and Chau, 2007). One of the current major issues, common to all levels of upper-limb amputation, is the growing gap between the available hardware of arm prostheses, becoming more biomimetic with numerous active joints (e.g., polydigital hands), and their still counter-intuitive and sequential control that limit their actual use (Atkins et al., 1996). For instance, the myoelectric control, which is the most common method to command an externally-powered upper limb prosthesis, relies on the use of ElectroMyoGraphic signals (EMG) from two antagonistic muscles of the residual limb (generally the biceps and triceps). An on/oﬀstrategy is applied by thresholding the input signals [amplitude and temporal variations of surface ElectroMyoGrams (sEMG)] that the patient needs to produce with the equipped muscles. Often, each active prosthetic joint that composes the substituting limb is sequentially controlled by the same control inputs. So, despite the potential possibilities oﬀered by the new biomimetic prostheses like whole robotic arms (Resnik et al., 2014) or polydigital hands (Belter and Dollar, 2011), their control remains complex, as it is far from intuitive, and oﬀers few functional Degrees of Freedom (DoF) (Castellini et al., 2014). Most studies on the development of more natural prosthetic control approaches based on PLM decoding (Powell et al., 2014; Atzori et al., 2016; Jarrasse et al., 2017b; Gaudet et al., 2018) have so far been conducted on oﬄine pattern recognition of pre-recorded myoelectric sequences, or either using simple computer interface control of performing simple free motions using a real prosthesis. To our knowledge, no experiments involved the completion of a functional task during which the prosthesis had to be controlled using real time decoding of PLM while the participant controlled the interaction with objects. 1. INTRODUCTION PLM have recently been shown to be controlled as intact limb movements (Raﬃn et al., 2012a,b; Garbarini et al., 2018), with associated muscle activities that vary with the type of executed PLM (Reilly et al., 2006; Raﬃn et al., 2012a; Jarrasse et al., 2017b). This approach has been quite extensively studied for below-elbow amputees whose residual limb usually still contains the muscles that mobilized the ﬁngers before the amputation, and, therefore, provide an adapted measurement site together with relatively strong myoelectric signals. While ﬁrst attempts of adaptation of these approaches to above- elbow amputees date back to the 70s with pioneering work like (Wirta et al., 1978), several studies recently tried to revive such approach using updated classiﬁcation techniques, speciﬁcally of phantom-limb-mobility-related EMG signals (Jarrasse et al., 2017b; Gaudet et al., 2018), even extended to individual ﬁnger movement decoding (Jarrasse et al., 2017a). Also, recent work on the treatment of phantom limb pain with help of virtual reality (e.g., Ortiz-Catalan et al., 2016) illustrates the possibility of decoding these PLM-associated EMG patterns measured on the residual limb of transhumeral amputees. Yet, after transhumeral amputation, PLM-related myoelectric activity is measured over muscle groups of the residual limb which -before amputation- were not naturally related to the missing limb (i.e., hand and wrist), and, therefore, inevitably more complex to decode. So, daily life phantom- based prosthetic control is more challenging, especially without reinforcing PLM-associated sEMG signals through muscular reinnervation surgery (Kuiken et al., 2007). One might, for example, expect an inﬂuence of the fatigue associated to PLM generation and of the remaining mismatch between the actual PLM movements and the ones generated in reaction by the prosthesis (the velocity and range of motion of these PLM being generally limited) (De Graaf et al., 2016), on the number of possible PLM that the patients can execute the one after the other. People with an upper arm amputation represent a signiﬁcant part of the major upper-limb amputees in western countries (for instance 33% in France André and Paysant, 2006 and 45% in the United Kingdom National Amputee Statistical Database, 2009). These patients, most of them young active people, are usually ﬁtted with a functional prosthesis (ﬁnanced by the Social Security) composed of several active joints (hand, wrist and sometimes elbow), allowing them to regain a certain autonomy. Citation: Jarrassé N, de Montalivet E, Richer F, Nicol C, Touillet A, Martinet N, Paysant J and de Graaf JB (2018) Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Without Surgical Reinnervation: A Preliminary Study. Front. Bioeng. Biotechnol. 6:164. doi: 10.3389/fbioe.2018.00164 November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 1 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 1. INTRODUCTION (Right) Photo o FIGURE 1 \| (Left) Global view of the experimental setup during one of the functional tasks of grasping an object (here the foam tennis ball) and releasing it in the dedicated container, with the arm prosthesis controlled through the associated mobilization of the phantom limb. (Right) Photo of the setup being used with P2. 2. MATERIALS AND METHODS availability of the participants during the recording period. The ﬁrst participant (P1), because of an injured plexus brachial leading to limited muscle contraction amplitudes, has never been ﬁtted with a myoelectric prosthesis. On the contrary, the second participant (P2) is daily using a myoelectric prosthesis composed of a motorized Utah© elbow, an active wrist rotator and a polydigital hand (iLimb Ultra, Touch Bionics ©), controlled by two electrodes placed over his biceps and triceps muscle (the control-switching between joints being achieved through a co-contraction of both muscles). Table 1 resumes demographic data. The prosthesis (Figure 1) was composed of an active elbow, a wrist rotator and polydigital hand, attached to a ﬁxed support placed close to the residual limb. A direct control mapping was used to associate a phantom movement of one joint to a movement of the same prosthetic joint. Eight diﬀerent movements of the elbow, wrist (rotation), hand and pinch were used within this experiment. We ﬁrst evaluated the pattern recognition architecture performance in amputees, and then analyzed the grasping task performance through diﬀerent metrics (timing, optimality of sequences, kinematics) in comparison with performance obtained by three non-amputated (control) participants performing the same task with a simple computer interface. The amputated participants were followed-up at the Louis Pierquin Centre of the Regional Institute of Rehabilitation, Nancy, France. Their voluntary mobilization of phantom limb had been explored through a questionnaire and a preliminary evaluation in order to make clear distinctions between residual limb sensations, phantom pain, phantom sensations, and most importantly, between mobility of the residual limb and that of the phantom limb (Touillet et al., 2018). Both participants reported a good feeling and control of their phantom hand, including separate whole hand and pinch (thumb and index) opening and closing, wrist rotation and ﬂexion/extension of the phantom elbow. 1. INTRODUCTION The objective of this study is thus to evaluate the possibility for transhumeral amputees to use a PLM-based control approach to actually perform a set of more realistic functional tasks with an unworn prosthetic arm with numerous active DoF (elbow, wrist rotator, polydigital hand with two diﬀerent types of grasping). Such a task will challenge the precision of control and the fatigue associated to PLM execution (De Graaf et al., 2016), and will oﬀer a ﬁrst evaluation of the “interaction” between “phantom activity” and the prosthetics with tasks in the “real world”. We present in this paper the results of an experiment during which two patients had to voluntarily mobilize their phantom limb to have an arm prosthesis mimicking the phantom movement. To overcome these limitations, pattern-recognition approaches have been developed since the late 60s/70s (Finley and Wirta, 1967; Herberts et al., 1973; Lawrence et al., 1973) aiming a more precise decoding of myoelectric signals in order to improve the recognition of diﬀerent muscle activation patterns and thus to control more types of movements. This requires the use of multiple recording sites, a precise extraction of signal characteristics (not only amplitude) and a multidimensional classiﬁcation architecture. While well established and extensively studied in research institutions, such approaches have only very recently been applied commercially to prosthetics control (e.g., COAPT system, http://www.coaptengineering.com/). One way of feeding pattern recognition myoelectric control is to rely on sEMG activities of the residual limb associated with phantom limb movement (PLM) execution. Indeed, voluntary November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 2 Jarrassé et al. Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees FIGURE 1 \| (Left) Global view of the experimental setup during one of the functional tasks of grasping an object (here the foam tennis ball) and releasing it in the dedicated container, with the arm prosthesis controlled through the associated mobilization of the phantom limb. (Right) Photo of the setup being used with P2. FIGURE 1 \| (Left) Global view of the experimental setup during one of the functional tasks of grasping an object (here the foam t dedicated container, with the arm prosthesis controlled through the associated mobilization of the phantom limb. Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 2.1. Participants Since this is a preliminary study, a limited number of participants were tested. Two participants (in the 35–55 age range) with a unilateral transhumeral amputation of traumatic origin were selected to participate to the study. Three healthy participants (in the 25–27 age range) were recruited as a control group to perform the grasp and release task through a simpliﬁed computer GUI control. This study was carried out in accordance with the recommendations of the Université Paris Descartes ethic committee CERES (N◦IRB 20151900001072), which had approved the protocol. All participants provided written informed consent to participate in the study, and both patients gave written permission for publication of photographs for scientiﬁc and educational purposes. The protocol was performed in accordance with the Declaration of Helsinki. 2.2. Experimental Setup 2.2.1. Upper Limb Prosthetic Platform 2.2. Experimental Setup 2.2.1. Upper Limb Prosthetic Platform 2.2.1. Upper Limb Prosthetic Platform FIGURE 2 \| The arm prosthesis prototype includes a motorized elbow (1), an embedded controller based on a Raspberry Pi 3 (2), an electronic wrist rotator (3), and a Touch Bionics Robolimb (4). FIGURE 2 \| The arm prosthesis prototype includes a motorized elbow (1), an embedded controller based on a Raspberry Pi 3 (2), an electronic wrist rotator (3), and a Touch Bionics Robolimb (4). FIGURE 3 \| View of the residual limbs with the connected six optimal (P1, left), respectively initial twelve (P2, right) pair of electrodes. FIGURE 2 \| The arm prosthesis prototype includes a motorized elbow (1), an embedded controller based on a Raspberry Pi 3 (2), an electronic wrist rotator (3), and a Touch Bionics Robolimb (4). FIGURE 2 \| The arm prosthesis prototype includes a motorized elbow (1), an embedded controller based on a Raspberry Pi 3 (2), an electronic wrist rotator (3), and a Touch Bionics Robolimb (4). FIGURE 3 \| View of the residual limbs with the connected six optimal (P1, left), respectively initial twelve (P2, right) pair of electrodes. Raspberry Pi 3 which drives the elbow and wrist joints through a dedicated position/velocity motor controller (Roboclaw Motor Controller from Ion Motion Control©), and the polydigital hand through a generic serial connection. A dedicated C++ program was developed (running at a frequency of 100 Hz on the Jessie OS© from Debian Linux©) to receive movement instructions (received from a wiﬁsocket connection), to control the active joints accordingly (with respect to a set of predeﬁned parameters like joint velocities), to monitor the kinematic activity and to store recorded data in ﬁles. Raspberry Pi 3 which drives the elbow and wrist joints through a dedicated position/velocity motor controller (Roboclaw Motor Controller from Ion Motion Control©), and the polydigital hand through a generic serial connection. A dedicated C++ program was developed (running at a frequency of 100 Hz on the Jessie OS© from Debian Linux©) to receive movement instructions (received from a wiﬁsocket connection), to control the active joints accordingly (with respect to a set of predeﬁned parameters like joint velocities), to monitor the kinematic activity and to store recorded data in ﬁles. shielded channels at 24-bit resolution was used to record sEMG muscle activities from the participant’s residual limb at a 1 kHz frequency. 2.2.1. Upper Limb Prosthetic Platform Because of the variability in residual limb length and muscle anatomy due to the level of amputation, the scheme of electrodes placement had to be adapted for each participant (see Figure 3). Twelve pairs of sEMG electrodes (Ambu© BlueSensor Ag/AgCl snap bipolar electrodes with a 1.25-cm-diameter circular contact and a 2 cm inter-electrode distance (center point to center point) were initially placed on each participant to measure activity on various parts of the residual biceps, triceps, deltoid and sometimes trapezoidal and pectoralis major muscles. No speciﬁc skin preparation was used before placing the active electrodes on the residual limb. To monitor the kinematic activity of the upper-limb prosthetic, a motion capture dataglove (VMG 30 from Virtual Realities LLC© relying on piezoelectric technology) was placed over the prosthetic hand. Thanks to a dedicated calibration phase (performed once for each patient), the prosthetic ﬁngers activity were recorded at 25 Hz. This glove was also ﬁtted with two Inertial Measurement Units (IMU, 9 degrees-of-freedom) allowing us to track both the elbow ﬂexion and wrist rotation kinematics at a similar frequency. The recorded sEMG signals were then ﬁltered with a [10 Hz ; 400 Hz] third-order bandpass Butterworth ﬁlter and a notch ﬁlter to remove the power line 50 Hz noise (Q factor of 35). A ﬁlter approach exploiting the properties of the Principal Component Analysis (derived from Kvas and Velik, 2008) was then applied to the ﬁrst sEMG recordings for the training of the pattern recognition algorithm. This analysis resulted in the initial selection of an optimal set of 6 pairs of electrodes (maximizing the classiﬁcation results) among the 12 channels initially placed. Once this optimization was performed, the set was used for the whole experimental session. Based on previous individual recordings of intact arm kinematics when mimicking their PLM (De Graaf et al., 2016), the joint velocities of the prosthesis were pre-set as follows: 15°/s for the elbow ﬂexion/extension velocity (slow compared to the natural adopted 105°/s measured in healthy subjects Farthing and Chilibeck, 2003), 40°/s for the wrist rotation and 2.6 s for complete opening or closing the whole hand. 2.2.1. Upper Limb Prosthetic Platform 2.2.1. Upper Limb Prosthetic Platform The upper limb prosthetic used for this experiment (see Figure 2) was designed using modiﬁed commercially available prosthetic parts: a Fillauer© Hosmer E-TWO electric elbow, a conventional electric wrist rotator from Ottobock© (model 10S17), and a Robolimb polydigital hand from Touch Bionics© possessing 6 active joints (one per ﬁnger ﬂexion plus the thumb rotation). This prototype was attached directly to a dedicated support frame made of aluminum proﬁles, which was adjusted to place the prosthetic elbow close to the phantom elbow joint position. The overall prosthetic system was controlled by an embedded The selection of the two amputated participants was based on the level of amputation (only transhumeral amputees), their control ability of a mobile phantom limb (i.e., the possibility to perform several diﬀerent types of phantom movements), the absence of phantom and residual limb pain, and the November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 3 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. TABLE 1 \| Demographic data concerning the 2 amputated participants. Participant Elaps. time Amput. side Dom. ? Amput. type Amput. cause Pain treat. Prosthesis P1 34 years Left No 1/3 arm Left traumatic No None or aesthetic P2 4 years Left No 2/3 arm Left traumatic No Myoelectric M, male; Dom. ?, Dominant limb prior to amputation; Elap. time, Elapsed time since amputation. “Pain Treat.”, pain treatment. “Prosthesis” indicates the type of prosthesis the patients usually wears. FIGURE 2 \| The arm prosthesis prototype includes a motorized elbow (1), an embedded controller based on a Raspberry Pi 3 (2), an electronic wrist rotator (3), and a Touch Bionics Robolimb (4). Raspberry Pi 3 which drives the elbow and wrist joints through FIGURE 3 \| View of the residual limbs with the connected six optimal (P1, left), respectively initial twelve (P2, right) pair of electrodes. TABLE 1 \| Demographic data concerning the 2 amputated participants. Participant Elaps. time Amput. side Dom. ? Amput. type Amput. cause Pain treat. Prosthesis P1 34 years Left No 1/3 arm Left traumatic No None or aesthetic P2 4 years Left No 2/3 arm Left traumatic No Myoelectric M, male; Dom. ?, Dominant limb prior to amputation; Elap. time, Elapsed time since amputation. “Pain Treat.”, pain treatment. “Prosthesis” indicates the type of prosthesis the patients usually wears. TABLE 1 \| Demographic data concerning the 2 amputated participants. 2.3.1. Amputees’ Protocol The participant was comfortably seated in a dedicated chair, ﬁtted with an armrest on the non-amputated side and a head rest. The arm prosthesis, attached to its rigid support, was placed approximatively at the location of the phantom arm of the subject (see Figure 1). Once the electrodes were placed on the subject’s residual limb, the session started with a training phase, was then followed by a phase of preliminary assessment of the classiﬁcation and PLM-based prosthesis control, and ended with the object grasping task. During all three phases, the sEMG signals were recorded and the whole recording session was videotaped and lasted for about 120 min for each participant. The three protocol phases will now be detailed. 2.2.3. Classiﬁcation Architecture A dedicated electrophysiological signal-recording system (Eegosports from ANT-Neuro©, The Netherland) with bipolar As shown on Figure 4, the classiﬁcation of given phantom limb movements cannot only rely on amplitude analyses of November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 4 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. FIGURE 4 \| P1 typical sEMG patterns associated to the voluntary mobilization of the phantom limb recorded by the six selected electrodes when performing successively 8 different phantom limb movements. FIGURE 4 \| P1 typical sEMG patterns associated to the voluntary mobilization of the phantom limb recorded by the six selected electrodes when performing successively 8 different phantom limb movements. sEMG bursts, a method often used for the study of voluntary limb movements with well deﬁned sEMG electrode placement according to SENIAM recommendations (Hermens et al., 2000). Additional features of sEMG signal characterizing its frequency and complexity aspects are needed for optimal classiﬁcation. We used the BCI2000 software suite to develop a global control architecture on a desktop computer running Windows 7 (Intel Core i5-4690K (3.5 GHz) with 16 Go DDR3). BCI2000 is a general-purpose software suite designed for brain-computer interface (BCI) and was used here to run in parallel three principal modules: one acquisition driver to acquire the sEMG data (at a 1 kHz frequency), one Matlab classiﬁcation algorithm script (executed every 128 ms), and one graphical user interface (C++ with Qt) which was also broadcasting in real time the classiﬁcation output to the network, in order to transmit the kinematic instructions to the prosthesis. A generic Linear Discriminant Analysis (LDA) classiﬁer (Englehart et al., 1999) running on Matlab© (relying on the “ﬁt discriminant analysis classiﬁer” function of the Statistics Toolbox) was used to classify the myoelectric activities. The features were computed from the sEMG using a 512-ms-sliding analysis window with a 128-ms- overlap between successive windows. Among the wide variety of features that have been investigated in the literature (Phinyomark et al., 2013), we selected those known to be the most eﬃcient and robust for the classiﬁcation of sEMG with LDA: the root mean square (RMS) value (Oskoei and Hu, 2008), the ﬁrst 4 autoregressive coeﬃcients (AR) (Zardoshti-Kermani et al., 1995; Tkach et al., 2010), the zero crossing and the sample entropy (Richman and Moorman, 2000) of the sEMG were extracted from each channel and used to create the feature vector. 2.2.3. Classiﬁcation Architecture No dimensionality reduction nor post-processing methods was used. sEMG bursts, a method often used for the study of voluntary limb movements with well deﬁned sEMG electrode placement according to SENIAM recommendations (Hermens et al., 2000). Additional features of sEMG signal characterizing its frequency and complexity aspects are needed for optimal classiﬁcation. We used the BCI2000 software suite to develop a global control architecture on a desktop computer running Windows 7 (Intel Core i5-4690K (3.5 GHz) with 16 Go DDR3). BCI2000 is a general-purpose software suite designed for brain-computer interface (BCI) and was used here to run in parallel three principal modules: one acquisition driver to acquire the sEMG data (at a 1 kHz frequency), one Matlab classiﬁcation algorithm script (executed every 128 ms), and one graphical user interface (C++ with Qt) which was also broadcasting in real time the classiﬁcation output to the network, in order to transmit the kinematic instructions to the prosthesis. A generic Linear Discriminant Analysis (LDA) classiﬁer (Englehart et al., 1999) running on Matlab© (relying on the “ﬁt discriminant analysis classiﬁer” function of the Statistics Toolbox) was used to classify the myoelectric activities. The features were computed from the sEMG using a 512-ms-sliding analysis window with a 128-ms- overlap between successive windows. Among the wide variety of features that have been investigated in the literature (Phinyomark et al., 2013), we selected those known to be the most eﬃcient and robust for the classiﬁcation of sEMG with LDA: the root mean square (RMS) value (Oskoei and Hu, 2008), the ﬁrst 4 autoregressive coeﬃcients (AR) (Zardoshti-Kermani et al., 1995; Tkach et al., 2010), the zero crossing and the sample entropy (Richman and Moorman, 2000) of the sEMG were extracted from each channel and used to create the feature vector. No dimensionality reduction nor post-processing methods was used. Only the conﬁdence value of the classiﬁer was used to ﬁlter the algorithm output: if classiﬁcation conﬁdence was below 95%, no movement instruction was sent to the prosthesis. 2.3.2. GUI Control by Healthy Participants 2.3.2. GUI Control by Healthy Participants In order to appreciate the performance obtained by the patients with such a control and setup, which remains very diﬀerent from a natural limb control (especially because it does not allow simultaneous movements), the three healhty participants were asked to perform the exact same experiment but with the prosthesis being controlled through the use of a graphical interface on a computer screen and a mouse (each of the eight movements of the prosthesis being activated by a speciﬁc button on the graphical interface). So, no classiﬁcation of EMG was necessary to control the prothesis. FIGURE 5 \| The objects used for the grasping task. From the left to the right, a cylinder made from Balsa wood from the kit of objects from the SHAP (Light et al., 2002) (diameter 60 mm, weight 30 g), a compliant foam tennis ball (diameter 70 mm, weight 12 g), and a clothespin from the Rolyan Graded Pinch Exerciser kit (model Yellow, 1 lb Pinch Exerciser, weight 20 g). brief initial individual training of the classiﬁer was then used for the whole experiments. 2.3.1.3. Grasping task 2.3.1.3. Grasping task After 10–15 min of passive recovery, the protocol ended by a grasping task performance with the arm prosthesis that included a randomized set of 3 grasping repetitions of 3 diﬀerent objects (shown in Figure 5 through the same control mode. Task duration was deﬁned as the time elapsed since the ﬁrst activation detection by the control architecture (diﬀerent from the “resting state”) until the object impact after falling into the receiving container. The chosen object was placed such that it could be caught by the prosthetic hand (i.e., position within the circle centered on the elbow prosthesis and which radius was equal to the hand/elbow length). The object had to be reached, grasped, detached from its support (velcro ﬁxation requiring approximatively a 3 N force to be detached), brought back to the extended-arm posture and released in a dedicated bucket. Before the start of each trial, the prosthesis was automatically brought back to a standardized initial posture, i.e., elbow fully extended, wrist in pronosupination with the hand palm aligned with the saggital 2.3.1.2. Preliminary assessment of the classiﬁcation and PLM-based prosthesis control The online performance during the preliminary assessment of the classiﬁcation and PLM-based control was analyzed using a Matlab© script to automatically determine within the 5 s interval allocated to each task (i.e., performing one speciﬁc gesture with the prosthesis), the exact starting time of the participant action (when the classiﬁer detected a change from the inactivity with a conﬁdence over 95%) and its end (last time instant of the classiﬁer output detection of an activity until the end of the 5 s time period). The confusion percentages were then calculated over these previously selected times of action, based on the ratio between the time of action and the time during which the correct movement was performed. After a rapid data treatment, the training phase of the classiﬁer was followed by a preliminary assessment of the classiﬁer performance. During this phase, while the online classiﬁcation algorithm was running, the participant was asked to induce given movements of the prosthesis through the phantom limb mobilization. The participant performed a randomized set of PLM (12 repetitions of the 8 selected gestures) during which the movements of the prosthesis reﬂected the PLM that were detected by the classiﬁer. Once the movement instruction provided by the experimenter, the participant had 5 s to perform the task. Within this time period, a failed trial could be followed by a second one. Once a trial was considered as successful, or when it lasted more than 5 s, the experimenter asked the participant to relax before giving the next PLM to perform. In order to avoid incompatible successions of actions, a pseudo-randomized list of 128 for P1 (reduced to 68 for P2, see results) PLM was performed with intermediate resting periods of a few minutes every 20 movements. For the grasping tasks, the prosthetic arm joints were extracted from the worn motion capture glove: the embedded wrist IMU was used to compute both elbow ﬂexion and wrist prono- supination angles. Full elbow extension was deﬁned as a null ﬂexion angle, while the palm parallel to the saggital plane (with the thumb pointing forward) deﬁned the 0°posture for the wrist prono-supination. The hand closing level was averaged by the closing level of the three last ﬁngers (from medium to little). Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 2.3.1.1. Training of the classiﬁer The participant was asked to successively perform once 8 selected phantom movements (8 movements: Elbow Flexion (EF) and Extension (EE), Wrist Pronation (WP) and Supination (WS), hand, resp. Pinch Closing (HC, resp. PC) and Opening (HO, resp. PC) with a few seconds of rest after each movement. This sequence was repeated once to get 2 demonstrations per movement. The experimenter was in charge of verbally asking the subject to execute a given movement and thus determined the rythm of the performance. No instruction was given about the amplitude and the velocity of the gesture, only the need of repeatability was mentioned. It is important to highlight that this November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 5 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. plane, and fully open hand. The ﬁnal/release positions was similar, i.e., elbow fully extended and hand opened but without particular constraint over the wrist orientation. The participants had the opportunity to perform two trials with each object to train before the real grasping task phase started. The object presentation was not randomized and always started with the simplest object to grasp (cylinder) and ﬁnish with the toughest (clothespin). FIGURE 5 \| The objects used for the grasping task. From the left to the right, a cylinder made from Balsa wood from the kit of objects from the SHAP (Light et al., 2002) (diameter 60 mm, weight 30 g), a compliant foam tennis ball (diameter 70 mm, weight 12 g), and a clothespin from the Rolyan Graded Pinch Exerciser kit (model Yellow, 1 lb Pinch Exerciser, weight 20 g). 2.3.1.2. Preliminary assessment of the classiﬁcation and PLM-based prosthesis control The pinch closing level was deﬁned as the closing level between the thumb and the index (which came into contact at around 30% of the pinch closing level because of the associated thumb rotation). 3.2.2. Task Duration The duration of the ﬁrst grasping phase and the total task duration (time to grasp, bring back, and release the object) are shown in Figures 9A,B, for each of the three objects averaged over the two amputated participants using their prosthesis, and over the 3 control participants performing the task by sequential control through the GUI. The ﬁrst phase durations for the patients are globally longer than those obtained for the healthy participants. For the cylinder the diﬀerence is small (average of 11.3 s for the amputees vs. 8.3 s for the controls), but for the ball (26 s vs. 18.7 s) it is longer, and even more for the clothespin (32.5 s vs. 16.8 s), the latter requiring a precise positioning of the ﬁngers through precise manipulation of elbow and wrist. In comparison, the return and release times are shorter (except for the cylinder) and are showing little discrepancies between the two groups (see Figure 9B). Interestingly, the return and release of the ball required an extended time for amputated participants (18 s instead of 7.4 s). Part of the reason is that these participants at the end of the return movement added a wrist rotation before opening the hand (as shown in Figure 7 with the action 8) in order to prevent a possible bouncing of the ball outside of the container whereas the controls did not bother about that. 3.1. Preliminary Evaluation of Classiﬁcation Performance and PLM-based Control 3.1. Preliminary Evaluation of Classiﬁcation Performance and PLM-based Control The confusion matrices for the participants to the preliminary evaluation of online control of the prosthesis are shown in Figure 6, indicating an averaged successful recognition score of November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 6 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. FIGURE 6 \| Confusion matrix of online control of the prosthesis for P1 (12 repetitions of the 8 movements, i.e., 92 movements, performed) and P2 (6 repetitions of the 8 movements, i.e., 48 movements, performed). Confusion matrix color scale is normalized across methods and increases from white to black as a function of increasing classiﬁcation rate. FIGURE 6 \| Confusion matrix of online control of the prosthesis for P1 (12 repetitions of the 8 movements, i.e., 92 movements, performed) and P2 (6 repetitions of the 8 movements, i.e., 48 movements, performed). Confusion matrix color scale is normalized across methods and increases from white to black as a function of increasing classiﬁcation rate. 88.5% for P1 and 86.9% for P2. The amputated participants performed, respectively, 12 and 6 repetitions (since P2 was subject to muscle fatigue, the number of repetitions was dropped by half) of the 8 diﬀerent movements of the prosthesis. For P1, only 92 movements were considered (4 movements among the 96 performed were removed because of some misunderstanding of the participant, invalidating the action). Principal confusions can be observed, respectively, between hand and pinch closing and opening: for P1, 15.7% of the pinch closing actions are considered as hand closing actions, and 8.4% of the hand openings are confused with pinch openings; for P2 rather large confusions exists between hand opening closing (14.9%), and also between hand and pinch opening (15.5 and 14.1%). Interestingly, these pinch/hand confusions were consistent with the sensations reported by the participants who indicated that they had diﬃculties in preventing the enrollment of the three other phantom ﬁngers when performing pinch actions. joints is directly impacted by the classiﬁer output changes, so even a limited (in time) misclassiﬁcation leads to rather important oscillations of the prosthesis (especially at the elbow joint, due to the lever arm eﬀect on the hand position) which perturbed the task and delayed its completion. November 2018 \| Volume 6 \| Article 164 3.2.3. Averaged Kinematic Proﬁles Figure 11 presents the averaged joint kinematic proﬁles of both elbow and wrist joints, normalized in time and averaged between repetitions and participants for the two groups and the three objects. It can be seen that, independently of the durations, a similar elbow path strategy is used by the two groups, with a straight ﬂexion extension for the cylinder, a visible intermediate step for the wrist adjustment for the ball, and a limited ﬂexion for the clothespin which was grasped by the prosthetic ﬁnger tips and not its palm. The wrist path diﬀered between the two groups (along with the variability within groups, more pronounced in amputees), with unnecessary wrist rotation appearing in the grasping of the cylinder by the amputated participants (possibly due to pattern recognition errors), a more segmented path for the clothespin, and, for the two last objects, no return to the initial wrist orientation for control participants. As expected, the trajectories of the control group are smoother than those of the amputated participants, essentially because of the instability of PLM control. The amputees control of the prosthesis was suboptimal compared to the control group performance (11 actions vs. 4 for the cylinder, 35 vs. 14 for the ball and 33 vs. 14 for the clothespin) (Figure 9C). For the two latter objects, even the healthy subjects were not able to fully optimize their sequence of gestures (mean of 14 sequences instead of the optimal values of 6 for the ball and 7 for the clothespin, respectively). 3.2. Grasp and Release Task Performances 3.2.1. Performance Overview Representative scenarios of participant P1 (one of the three repetitions) of the three grasping tasks are shown in Figure 7. The proﬁles of joint kinematics and the associated classiﬁcation output are shown in Figure 8. For the cylinder and ball, the participant P1 was able to perform the task following generic grasping strategies without performing numerous unnecessary and parasitic motor actions with the prosthesis. For the clothespin, he had more trouble stabilizing the elbow at the correct height (with numerous control jumps in the classiﬁer output) to be able to grasp it (Figure 8C). As shown on the joint position proﬁles of Figure 8, the kinematics of the prosthetic Although the total grasp-and-release time is shorter for the controls than for the patients, the relation between the total grasp-and-release time and the needed number of actions to November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 7 Jarrassé et al. Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees FIGURE 7 \| Visualization of representative grasping sequences of P1. FIGURE 7 \| Visualization of representative grasping sequences of P1. complete the task is rather similar. Indeed, Figure 10 shows the equations of the best ﬁtting linear relation is close for the two groups. performing these three tasks, we calculated the averaged number of sequence steps used by the amputees (phantom limb control through classiﬁcation) and the healthy participants (“button control”). The prosthesis control is considered as optimal when the sequence number corresponds to the theoretical minimal number of sequences listed in Table 2. Any increased number therefore can be due to a control error (possibly because of an erroneous classiﬁcation), to a discontinuous movement (for example an elbow ﬂexion discomposed in several ﬂexion submovements), or to an alternative sub-optimal motor strategy. This index actually gives an overview of the global performance of both groups of participants and the control architecture (LDA classiﬁer) since it is more likely to quickly increase with a classiﬁcation architecture rather than with a direct control. 3.2.4. Optimality of Control Sequences Two transhumeral amputated participants have been asked to repetitively grasp three diﬀerent objects with an unworn active eight-DoF prosthetic arm and release them in a dedicated drawer, thanks to a prosthesis control based on phantom limb The optimal sequences of motor actions to perform the grasping- releasing tasks of the three objects are shown in Table 2. In order to evaluate the quality of the prosthesis control in November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 8 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. FIGURE 8 \| Representative proﬁles (associated to the 1–9 action indexes shown on Figure 7) of participant P1 for grasping and releasing the three objects (A: cylinder, B: ball, C: clothespin). Recognized phantom movements (output of the classiﬁcation algorithm) are shown for each task, along with the associated measured kinematic variations of the 4 joints: elbow and wrist angles, and percentages of closing of hand and pinch. The action indexes (from 1 to 9) are related to the similar indices shown on Figure 7. GURE 8 \| Representative proﬁles (associated to the 1–9 action indexes shown on Figure 7) of participant P1 for grasping and releasing the three objects FIGURE 8 \| Representative proﬁles (associated to the 1–9 action indexes shown on Figure 7) of participant P1 for grasping and releasing the three objects (A: cylinder, B: ball, C: clothespin). Recognized phantom movements (output of the classiﬁcation algorithm) are shown for each task, along with the associated measured kinematic variations of the 4 joints: elbow and wrist angles, and percentages of closing of hand and pinch. The action indexes (from 1 to 9) are related to the similar indices shown on Figure 7. over 90% reported in the literature (Al-Timemy et al., 2013; Farina et al., 2014). Yet, ﬁrst, these latter results were obtained on transradial amputees mobilizing their residual hand and wrist muscles. It is noteworthy to recall that the present study was performed with transhumeral amputees, for whom the phantom limb phenomenon is probably related to cortical and/or neuromuscular reorganization after the amputation (Wu and Kaas, 2000; Qi et al., 2004; Gagné et al., 2011), making it more unstable compared to transradial phantom limbs. Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org November 2018 \| Volume 6 \| Article 164 3.2.4. Optimality of Control Sequences FIGURE 9 \| (A) Averaged time (± standard error) to grasp the three different objects for the amputated participants controlling the prosthesis with their phantom limb (blue) as well as for the healthy participants by sequential control through a dedicated GUI (red) (B) Averaged time (± SE) to return and release the three different FIGURE 9 \| (A) Averaged time (± standard error) to grasp the three different objects for the amputated participants controlling the prosthesis with their phantom limb (blue), as well as for the healthy participants by sequential control through a dedicated GUI (red). (B) Averaged time (± SE) to return and release the three different objects for the two groups. (C) Averaged number of actions for completing the 3 “grasp and release” tasks. FIGURE 10 \| Total grasp-and-release time as a function of the number of actions needed to complete the task for all trials and objects and for each patient (in red) and healthy control (in blue). Each symbol represents one trial. The best ﬁtting linear relation and their equations use the same color code. sEMG patterns, reduces cognitive fatigue and ﬁnally improves the robustness of this control approach. Such improvements have indeed been recently reported for the control of PLM of forearm amputees (Powell et al., 2014). This recognition rate of the PLM still allowed the two amputated participants to successfully perform the functional grasp-and-release tasks. And this remains a rather encouraging performance especially when considering the limited training of FIGURE 9 \| (A) Averaged time (± standard error) to grasp the three different objects for the amputated participants controlling the prosthesis with their phantom limb (blue), as well as for the healthy participants by sequential control through a dedicated GUI (red). (B) Averaged time (± SE) to return and release the three different objects for the two groups. (C) Averaged number of actions for completing the 3 “grasp and release” tasks. FIGURE 9 \| (A) Averaged time (± standard error) to grasp the three different objects for the amputated participants controlling the prosthesis with their phantom limb (blue), as well as for the healthy participants by sequential control through a dedicated GUI (red). (B) Averaged time (± SE) to return and release the three different objects for the two groups. (C) Averaged number of actions for completing the 3 “grasp and release” tasks. 3.2.4. Optimality of Control Sequences Indeed, after transradial amputation, depending on the level of amputation (the length of the stump directly conditioning the degree of presence and usability of residual extrinsic hand muscles), the muscles involved in ﬁnger, hand and wrist actions can still be present. In that case, the neuromuscular reorganization is probably less important. Second, none of these two participants mobilize their phantom limb in their daily living activities, so the used PLM were fully unusual and untrained. One might expect that training of PLM-execution probably stabilizes the associated mobilization and myoelectric pattern recognition techniques, with a classiﬁer trained using only two repetitions of each PLM. The participants successfully achieved the tasks, even if the completion times were increased (in comparison with the performances obtained by a control group using a simple GUI control) and the control strategies required numerous corrections. The recognition rate averaged over all movements revealed to be rather high (on the average 86.4% for the two amputated participants) indicating that they were able to control 8 diﬀerent prosthetic movements with a limited set of 6 electrodes placed over their residual upper-arm limb. The main confusions appeared between the hand and pinch actions. However, this confusion did not directly alter the grasp and release performances since even the precise clothespin grasping could be performed with the whole hand instead of a pinch. Although successful, this online recognition and control rate remains slightly lower than the latest classiﬁcation rate of November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 9 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. FIGURE 9 \| (A) Averaged time (± standard error) to grasp the three different objects for the amputated participants controlling the prosthesis with their phantom limb (blue), as well as for the healthy participants by sequential control through a dedicated GUI (red). (B) Averaged time (± SE) to return and release the three different objects for the two groups. (C) Averaged number of actions for completing the 3 “grasp and release” tasks. 3.2.4. Optimality of Control Sequences FIGURE 10 \| Total grasp-and-release time as a function of the number of actions needed to complete the task for all trials and objects and for each patient (in red) and healthy control (in blue). Each symbol represents one trial. The best ﬁtting linear relation and their equations use the same color code. FIGURE 10 \| Total grasp-and-release time as a function of the number of actions needed to complete the task for all trials and objects and for each patient (in red) and healthy control (in blue). Each symbol represents one trial. The best ﬁtting linear relation and their equations use the same color code. sEMG patterns, reduces cognitive fatigue and ﬁnally improves the robustness of this control approach. Such improvements have indeed been recently reported for the control of PLM of forearm amputees (Powell et al., 2014). This recognition rate of the PLM still allowed the two amputated participants to successfully perform the functional grasp-and-release tasks. And this remains a rather encouraging performance especially when considering the limited training of November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 10 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. FIGURE 11 \| Plots of the averaged joint kinematic proﬁles of both elbow and wrist joints, normalized in time and averaged between repetitions and participants for the amputated (blue) and control (red) participants, and the three objects. Standard error is represented by the transparent envelopes around the curves. FIGURE 11 \| Plots of the averaged joint kinematic proﬁles of both elbow and wrist joints, normalized in time and averaged between repetitions and participants for the amputated (blue) and control (red) participants, and the three objects. Standard error is represented by the transparent envelopes around the curves. myoelectric control of active wrist and hand prostheses. Their kinematic strategies (as shown in Figure 11 and action-to-time ratio (as shown in Figure 10 were rather similar but the durations for the two complex grasps (ball and clothespin) were extended in comparison with those obtained by the control participants in their simpliﬁed -but still sequential- control task. When looking into detail the movement sequences of the prosthesis, it appeared that the major task diﬃculty lies in the grasping part of the task, which is not surprising since the grasping is only possible with a precise positioning of the hand. 3.2.4. Optimality of Control Sequences The return and release phase increased in amputees because of an additional wrist reorientation phase at the end before releasing the object. This was probably due to the fact that the amputated participants were more conscientious and wanted to have the object not bouncing out of the bucket, whereas the control participants did not bother about this. TABLE 2 \| Optimal sequences of motor action for completing each grasp and release task. Objects Balsa cylinder Foam ball clothespin Optimal sequences Flex the elbow and stop it at the ﬁnal height Supinate the wrist to orient the palm facing the top Supinate the wrist to orient the palm facing the top Close the hand over the object Flex the elbow and stop it at the ﬁnal height Flex the elbow and stop it at the ﬁnal height Fully extend the elbow Close the hand over the object Correct the wrist orientation to face the clothespin with pinch Open hand to release object Fully extend the elbow Close the pinch over the object Pronate the wrist back to initial orientation Fully extend the elbow Open hand to release object Pronate the wrist back to initial orientation Open hand to release object Optimal number of sequences 4 6 7 Several reasons can be responsible for the longer durations and the higher number of actions needed for completing the task, the most important probably being the control delay induced by the classiﬁcation of the PLM. Precise positioning of the segments of the prosthesis required a precise temporal control of the activation of the corresponding movements, which was clearly diﬀerent between the tasks for the two groups. Indeed, the PLM-based control architecture induced additional delays of up to 512 ms because of the windowing and ﬁltering eﬀect of the classiﬁcation algorithm. This clearly increased the diﬃculty to stop the elbow ﬂexion at the desired height and possibly induced the ﬂexion/extension oscillations observed for the elbow joint, shown in Figure 8 for the clothespin task. November 2018 \| Volume 6 \| Article 164 3.2.4. Optimality of Control Sequences This work demonstrates the eﬀectiveness of a bio- inspired system successfully conjugating the advantages of an underactuated, anthropomorphic hand with a PCA-based control strategy, and opens up promising possibilities for the development of an intuitively controllable hand prosthesis. p y p Despite the above-mentioned diﬃculties, our two patients were able, without any beforehand PLM-training, to control 8 diﬀerent movements of a prosthesis in a more eﬃcient, simple, and dexterous way that conventional (dual-site) myoelectric control can oﬀer. While TMR is now becoming a generic solution for transhumeral amputees to increase their control capabilities over (or simply allow the use of) prosthesis with more than 2 active DoF (i.e., when an active elbow is added to the prosthetic wrist and hand), the preliminary results obtained in this study and other research teams exploiting the natural myoelectric activity related to PLM, pushes toward a reconsideration of the possibility of extracting more control signals without undergoing invasive surgical procedures. Obviously, TMR is providing an interesting way of stabilizing and reinforcing the myoelectric patterns related to PLM. This eases the decoding of myoelectric activities and is possibly one major key to overcome the issue of the perturbations generated by the wearing of the prosthesis. Nonetheless, recent “realistic” use of TMR in transhumeral amputees, controlling a 3-active-DoF-prosthesis through pattern recognition (Hargrove et al., 2017), still required either the use of a 15-electrodes-array placed over the stump (as described in Tkach et al., 2014), or the use of the conventional signal control switch to control the wrist when a direct control with a set of 4 electrodes was used. Therefore, there seems to be a real interest in pushing further our understanding of the phenomenon of natural (i.e., without TMS) residual myoelectric activity related to PLM. This could possibly be a viable option in some transhumeral amputees desiring to control 3 active joints. The obtained results show that amputated participants were able, after a very short appropriation of all task requirements (i.e., 2 repetitions for training, 8 preliminary repetitions of each PLM and two preliminary trials of each grasp-and-release task), to manage the rather complex interaction between their PLM, the associated actions of the prosthesis and, through it, their physical interaction with objects and the environment. The PLM- based control revealed to be rather intuitive. 3.2.4. Optimality of Control Sequences This delay can probably be decreased by the reduction of the sliding window time to 256 ms and the associated overlap to 64 ms, the participants compared to the several days (or weeks) sessions usually required in upper-limb amputees to master a simple (but constraining in terms of muscular contraction amplitudes) the participants compared to the several days (or weeks) sessions usually required in upper-limb amputees to master a simple (but constraining in terms of muscular contraction amplitudes) November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 11 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. by the optimization of the set of features (the sample entropy being computationally time-consuming) and by using compiled programming languages. eliminate the problem of the stump/socket physical connection, (3) electrode implantation (Mastinu et al., 2017) minimizing the issue with skin impedance and movements, and of course (4) more robust architectures of pattern recognition, integrating the stump posture (tracked through IMUs for example) to integrate the actual arm posture in the signal classiﬁcation (Lauretti et al., 2016). p g g g g Another reason for the reduced performance of the amputated participants is that the task was more complex for them (i.e., precise and reproducible mobilization of their phantom limb) than for the control participants (simply acting over a mouse), and, moreover, prolonged phantom limb mobilization induces fatigue (De Graaf et al., 2016), which inﬂuenced the sEMG signals and thus reduced the rate of successful classiﬁcation of the PLM. This might have caused the increase of durations and number of actions for the ball and clothespin objects (which were presented after the cylinder). Finally, the patients reported that they had to concentrate when performing the PLM (specially to prevent from performing “unusual” corrective actions which would have perturbed the classiﬁer) and to perform clear separations (i.e., a time of inactivity) between the diﬀerent PLM to maximize the recognition rate. This also contributed to increasing the task durations. Yet, interestingly, even the durations of the control group are extended when compared to the time it takes to grasp and release the objects with an intact arm (below 5 s). This suggests that, more than the complexity of phantom limb mobilization, the impossibility of performing simultaneous actions is one of the main causes of the prolonged durations needed to perform these grasp-and-release tasks. Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org REFERENCES Finley, F. R., and Wirta, R. W. (1967). Myocoder studies of multiple myopotential response. Arch. Phys. Med. Rehabil. 48:598. Al-Timemy, A. H., Bugmann, G., Escudero, J., and Outram, N. (2013). Classiﬁcation of ﬁnger movements for the dexterous hand prosthesis control with surface electromyography. IEEE J. Biomed. Health Inform. 17, 608–618. doi: 10.1109/JBHI.2013.2249590 Gagné, M., Hétu, S., Reilly, K., and Mercier, C. (2011). The map is not the territory: motor system reorganization in upper limb amputees. Hum. Brain Mapp. 32, 509–519. doi: 10.1002/hbm.21038 Garbarini, F., Bisio, A., Biggio, M., Pia, L., and Bove, M. (2018). Motor sequence learning and intermanual transfer with a phantom limb. Cortex 101, 181–191. doi: 10.1016/j.cortex.2018.01.011 André, J., and Paysant, J. (2006). Les amputés en chiﬀres: épidémiologie. Module de MPR et Appareillage, co. fe mer avril. Available online at: www.cofemer.fr/ UserFiles/File/AP2Amp_Chiﬀres.pdf Gaudet, G., Raison, M., and Achiche, S. (2018). Classiﬁcation of upper limb phantom movements in transhumeral amputees using electromyographic and kinematic features. Eng. Appl. Artif. Intell. 68, 153–164. doi: 10.1016/j.engappai.2017.10.017 Atkins, D. J., Heard, D. C., and Donovan, W. H. (1996). Epidemiologic overview of individuals with upper-limb loss and their reported research priorities. J. Prosthet. Orthot. 8, 2–11. doi: 10.1097/00008526-199600810- 00003 Hargrove, L. J., Miller, L. A., Turner, K., and Kuiken, T. A. (2017). Myoelectric pattern recognition outperforms direct control for transhumeral amputees with targeted muscle reinnervation: a randomized clinical trial. Sci. Rep. 7:13840. doi: 10.1038/s41598-017-14386-w Atzori, M., Gijsberts, A., Castellini, C., Caputo, B., Mittaz Hager, A.-G., Elsig, S., et al. (2016). Clinical parameter eﬀect on the capability to control myoelectric robotic prosthetic hands. J. Rehabil. Res. Dev. 53, 345–358. doi: 10.1682/JRRD.2014.09.0218 He, J., and Zhu, X. (2017). Combining improved gray-level co-occurrence matrix with high density grid for myoelectric control robustness to electrode shift. IEEE Trans. Neural Syst. Rehabil. Eng. 25, 1539–1548. doi: 10.1109/TNSRE.2016.2644264 Belter, J. T., and Dollar, A. M. (2011). “Performance characteristics of anthropomorphic prosthetic hands,” in Rehabilitation Robotics (ICORR), 2011 IEEE International Conference on (Zurich: IEEE), 1–7. Biddiss, E., and Chau, T. (2007). Upper-limb prosthetics: critical factors in device abandonment. Am. J. Phys. Med. Rehabil. 86, 977–987. doi: 10.1097/PHM.0b013e3181587f6c Herberts, P., Almström, C., Kadefors, R., and Lawrence, P. D. (1973). Hand prosthesis control via myoelectric patterns. Acta Orthopaed. Scand. 44, 389– 409. doi: 10.3109/17453677308989075 Castellini, C., Artemiadis, P., Wininger, M., Ajoudani, A., Alimusaj, M., Bicchi, A., et al. (2014). Proceedings of the ﬁrst workshop on peripheral machine interfaces: going beyond traditional surface electromyography. 3.2.4. Optimality of Control Sequences This is conﬁrmed by the fact that the patients never used their PLM in daily life and still were able to control the prosthesis without any beforehand training. Moreover, the patients instinctively tried to correct unplanned actions generated at the prosthesis level with adapted phantom movements (as we do with intact limbs) but unfortunately not known by the classiﬁer. This intrinsic limitation of the pattern recognition techniques that can only recognize speciﬁc and known (trained) PLM, still limits the intuitiveness and naturalness of the approach. j The transhumeral amputees who were recruited in the study had received no prior training with the PLM-based-control approach. Better results, in terms of precision and completion times, can be expected with regular PLM-training. Furthermore, improving the control of the prosthetic joints (especially the elbow involved in numerous and rather long displacements), with exponential velocity and smoothened acceleration proﬁles to restore human-like movement properties (isochrony and minimization of the jerk (Viviani and Flash, 1995) will have a positive eﬀect on the controllability. Bio-inspired approaches minimizing the control dimensionality through mechanical underactuation and models of ﬁnger joint synergies (as proposed by Magenes et al., 2008 or Matrone et al., 2010) could strongly enhance amputees’ abilities in object manipulation tasks. Finally, providing additional sensory feedback to the participant, especially concerning the interaction of the prosthesis with the Obviously, in the current state of the control architecture, the control performance can be expected to be reduced with a worn prosthesis. Indeed, the current limited robustness of pattern recognition techniques is a major obstacle, particularly when they have to be used in realistic scenarios including a worn prosthesis. This will generate additional constraints such as pressure and sweating, along with a mobile residual limb generating non-“phantom limb related” muscle contractions. These factors will aﬀect the sEMG signals and thus decrease the PLM detection rate. Yet, this is a generic problem in the ﬁeld of pattern recognition of electro-physiological signals, and numerous solutions are actually developed that could compensate for the listed issues. Examples of these solutions are (1) robustness to electrodes shift (Muceli et al., 2014; He and Zhu, 2017), (2) use of osseointegration (Ortiz-Catalan et al., 2014) to November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 12 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. NJ, CN, JdG, EdM, and AT participated in the experimental session with the amputated participant. AUTHOR CONTRIBUTIONS NJ, EdM, JdG, CN, AT designed the protocol. EdM, FR, and NJ built the experimental platform, and EdM collected the data with the healthy participants. AT, NM, and JP contacted the amputated participant and organized the experimental session. The authors warmly thank the patients included in this protocol along with the orthopaedic-ﬁtting team, in particular F. Codemard, K. Georges, Y. Corbel, and B. Juncker. We also thank T. Coyle for proofreading the article. FUNDING The study was ﬁnancially supported by the ANR (project PhantoMovCOntrol ANR-15-CE19-0008-02) and the Labex SMART (ANR-11-LABX-65) supported by French state funds managed by the ANR within the Investissements d’Avenir programme under reference ANR-11-IDEX-0004-02. The study was ﬁnancially supported by the ANR (project PhantoMovCOntrol ANR-15-CE19-0008-02) and the Labex SMART (ANR-11-LABX-65) supported by French state funds managed by the ANR within the Investissements d’Avenir programme under reference ANR-11-IDEX-0004-02. 3.2.4. Optimality of Control Sequences NJ, JdG, CN, and AT analyzed the data, and wrote the present report. environment could be a game changer in helping amputated persons in performing such tasks, whatever control technique is used. For instance, artiﬁcially provoking phantom limb referred sensations, as recently tested in (Osborn et al., 2018), could be a relevant technique, particularly when prosthesis control is PLM- based. More amputated participants will be included in future experiments in order to reinforce the ﬁrst results presented here and investigate the inﬂuence of phantom limb mobility training on the control performance. In such way, the often occurring mobility of the phantom limb might become useful instead of something to hide or ignore. REFERENCES “Fusion of m-imu and emg signals for the control of trans-humeral prostheses,” in Biomedical Robotics and Biomechatronics (BioRob), 2016 6th IEEE International Conference on (Singapore: IEEE), 1123–1128. Lawrence, P., Herberts, P., and Kadefors, R. (1973). Experiences with a multifunctional hand prosthesis controlled by myoelectric patterns. Adv. Ext. Control Hum. Extremit. 47–65. Raﬃn, E., Giraux, P., and Reilly, K. T. (2012a). The moving phantom: motor execution or motor imagery? Cortex 48, 746–757. doi: 10.1016/j.cortex.2011.02.003 Raﬃn, E., Mattout, J., Reilly, K. T., and Giraux, P. (2012b). Disentangling motor execution from motor imagery with the phantom limb. Brain 135(Pt 2):582–595. doi: 10.1093/brain/awr337 Light, C. M., Chappell, P. H., and Kyberd, P. J. (2002). Establishing a standardized clinical assessment tool of pathologic and prosthetic hand function: normative data, reliability, and validity. Arch. Phys. Med. Rehabil. 83, 776–783. doi: 10.1053/apmr.2002.32737 Reilly, K. T., Mercier, C., Schieber, M. H., and Sirigu, A. (2006). Persistent hand motor commands in the amputees’ brain. Brain 129, 2211–2223. doi: 10.1093/brain/awl154 Magenes, G., Passaglia, F., and Secco, E. L. (2008). “A new approach of multi-dof prosthetic control,” in Engineering in Medicine and Biology Society, 2008. EMBS 2008. 30th Annual International Conference of the IEEE (Vancoucer, BC: IEEE), 3443–3446. Resnik, L., Klinger, S. L., and Etter, K. (2014). The deka arm: its features, functionality, and evolution during the veterans aﬀairs study to optimize the deka arm. Prosthet. Orthot. Int. 38, 492–504. doi: 10.1177/0309364613506913 Mastinu, E., Doguet, P., Botquin, Y., Håkansson, B., and Ortiz-Catalan, M. (2017). Embedded system for prosthetic control using implanted neuromuscular interfaces accessed via an osseointegrated implant. IEEE Trans. Biomed. Circuits Syst. 11, 867–877. doi: 10.1109/TBCAS.2017.2694710 Richman, J. S., and Moorman, J. R. (2000). Physiological time-series analysis using approximate entropy and sample entropy. Am. J. Physiol. Heart Circul. Physiol. 278, H2039–H2049. doi: 10.1152/ajpheart.2000.278.6.H2039 Matrone, G. C., Cipriani, C., Secco, E. L., Magenes, G., and Carrozza, M. C. (2010). Principal components analysis based control of a multi-dof underactuated prosthetic hand. J. Neuroeng. Rehabil. 7:16. doi: 10.1186/1743-0003-7-16 Tkach, D., Huang, H., and Kuiken, T. (2010). Research study of stability of time-domain features for electromyographic pattern recognition. J. Neuroeng. Rehabil. 7:21. doi: 10.1186/1743-0003-7-21 Tkach, D. C., Young, A. J., Smith, L. H., Rouse, E. J., and Hargrove, L. J. (2014). Real-time and oﬄine performance of pattern recognition myoelectric control using a generic electrode grid with targeted muscle reinnervation patients. IEEE Trans. Neural Syst. Rehabil. Eng. 22, 727–734. REFERENCES Front. Neurorobot. 8:22. doi: 10.3389/fnbot.2014.00022 Hermens, H. J., Freriks, B., Disselhorst-Klug, C., and Rau, G. (2000). Development of recommendations for semg sensors and sensor placement procedures. J. Electromyogr. Kinesiol. 10, 361–374. doi: 10.1016/S1050-6411(00)00027-4 Jarrasse, N., Nicol, C., Richer, F., Touillet, A., Martinet, N., Paysant, J., et al. (2017a). “Voluntary phantom hand and ﬁnger movements in transhumerai amputees could be used to naturally control polydigital prostheses,” in Rehabilitation Robotics (ICORR), 2017 International Conference on (London, UK: IEEE), 1239–1245. De Graaf, J. B., Jarrassé, N., Nicol, C., Touillet, A., Coyle, T., Maynard, L., et al. (2016). Phantom hand and wrist movements in upper limb amputees are slow but naturally controlled movements. Neuroscience 312, 48–57. doi: 10.1016/j.neuroscience.2015.11.007 Jarrasse, N., Nicol, C., Touillet, A., Richer, F., Martinet, N., Paysant, J., et al. (2017b). Classiﬁcation of phantom ﬁnger, hand, wrist, and elbow voluntary gestures in transhumeral amputees with semg. IEEE Trans. Neural Syst. Rehabil. Eng. 25, 71–80. doi: 10.1109/TNSRE.2016.2563222 Englehart, K., Hudgins, B., Parker, P. A., and Stevenson, M. (1999). Classiﬁcation of the myoelectric signal using time-frequency based representations. Med. Eng. Phys. 21, 431–438. doi: 10.1016/S1350-4533(99)00066-1 Farina, D., Jiang, N., Rehbaum, H., Holobar, A., Graimann, B., Dietl, H., et al. (2014). The extraction of neural information from the surface emg for the control of upper-limb prostheses: emerging avenues and challenges. IEEE Trans. Neural Syst. Rehabil. Eng., 22, 797–809. doi: 10.1109/TNSRE.2014.2305111 Kuiken, T. A., Miller, L. A., Lipschutz, R. D., Lock, B. A., Stubbleﬁeld, K., Marasco, P. D., et al. (2007). Targeted reinnervation for enhanced prosthetic arm function in a woman with a proximal amputation: a case study. Lancet 369, 371–380. doi: 10.1016/S0140-6736(07)60193-7 Farthing, J. P., and Chilibeck, P. D. (2003). The eﬀects of eccentric and concentric training at diﬀerent velocities on muscle hypertrophy. Eur. J. Appl. Physiol. 89, 578–586. doi: 10.1007/s00421-003-0842-2 Kvas, G., and Velik, R. (2008). “A ﬁlter approach for myoelectric channel selection,” in 2008 6th IEEE International Conference on Industrial Informatics (Daejeon: IEEE), 1437–1440. November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 13 Phantom-Mobility-Based Prosthesis Control in Transhumeral Amputees Jarrassé et al. Qi, H. X., Stewart Phillips, W., and Kaas, J. H. (2004). Connections of neurons in the lumbar ventral horn of spinal cord are altered after long- standing limb loss in a macaque monkey. Somatosens. Mot. Res. 21, 229–239. doi: 10.1080/08990220400012588 Lauretti, C., Davalli, A., Sacchetti, R., Guglielmelli, E., and Zollo, L. (2016). REFERENCES doi: 10.1109/TNSRE.2014.2302799 Muceli, S., Jiang, N., and Farina, D. (2014). Extracting signals robust to electrode number and shift for online simultaneous and proportional myoelectric control by factorization algorithms. IEEE Trans. Neural Syst. Rehabil. Eng. 22, 623–633. doi: 10.1109/TNSRE.2013.2282898 National Amputee Statistical Database (2009). The Amputee Statistical Database for the United Kingdom 2006/07. Edinburgh: Information Services Division, NHS Scotland. Touillet, A., Peultier-Celli, L., Nicol, C., Jarrassé, N., Loiret, I., Martinet, N., et al. (2018). Characteristics of phantom upper limb mobility encourage phantom-mobility-based prosthesis control. Nat. Sci. Rep. 8:15459. doi: 10.1038/s41598-018-33643-0 Ortiz-Catalan, M., Gudhmundsdóttir, R. A., Kristoﬀersen, M. B., Zepeda- Echavarria, A., Caine-Winterberger, K., Kulbacka-Ortiz, K., et al. (2016). Phantom motor execution facilitated by machine learning and augmented reality as treatment for phantom limb pain: a single group, clinical trial in patients with chronic intractable phantom limb pain. Lancet 388, 2885–2894. doi: 10.1016/S0140-6736(16)31598-7 Viviani, P., and Flash, T. (1995). Minimum-jerk, two-thirds power law, and isochrony: converging approaches to movement planning. J. Exp. Psychol. Hum. Percept. Perform. 21:32. doi: 10.1037/0096-1523.21.1.32 Wirta, R. W., Taylor, D. R., and Finley, F. R. (1978). Pattern-recognition arm prosthesis: a historical perspective—a ﬁnal report. Bull. Prosthet. Res. 10, 8–35. Wright, T. W., Hagen, A. D., and Wood, M. B. (1995). Prosthetic usage in major upper extremity amputations. J. Hand Surg. 20, 619–622. doi: 10.1016/S0363-5023(05)80278-3 Wirta, R. W., Taylor, D. R., and Finley, F. R. (1978). Pattern-recognition arm prosthesis: a historical perspective—a ﬁnal report. Bull. Prosthet. Res. 10, 8–35. Ortiz-Catalan, M., Håkansson, B., and Brånemark, R. (2014). An osseointegrated human-machine gateway for long-term sensory feedback and motor control of artiﬁcial limbs. Sci. Transl. Med. 6:257re6. doi: 10.1126/scitranslmed.3008933 Wright, T. W., Hagen, A. D., and Wood, M. B. (1995). Prosthetic usage in major upper extremity amputations. J. Hand Surg. 20, 619–622. doi: 10.1016/S0363-5023(05)80278-3 Osborn, L. E., Dragomir, A., Betthauser, J. L., Hunt, C. L., Nguyen, H. H., Kaliki, R. R., et al. (2018). Prosthesis with neuromorphic multilayered e-dermis perceives touch and pain. Sci. Robot. 3:eaat3818. doi: 10.1126/scirobotics.aat3818 Wu, C. W., and Kaas, J. H. (2000). Spinal cord atrophy and reorganization of motoneuron connections following long-standing limb loss in primates. Neuron 28, 967–978. doi: 10.1016/S0896-6273(00)00167-7 Oskoei, M. A., and Hu, H. (2008). Support vector machine-based classiﬁcation scheme for myoelectric control applied to upper limb. IEEE Trans. Biomed. Eng. 55, 1956–1965. doi: 10.1109/TBME.2008.919734 Zardoshti-Kermani, M., Wheeler, B., Badie, K., and Hashemi, R. (1995). November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org REFERENCES Emg feature evaluation for movement control of upper extremity prostheses. IEEE Trans. Rehabil. Eng. 3, 324–333. doi: 10.1109/86.481972 Ostlie, K., Franklin, R. J., Skjeldal, O. H., Skrondal, A., and Magnus, P. (2011). Musculoskeletal pain and overuse syndromes in adult acquired major upper-limb amputees. Arch. Phys. Med. Rehabil. 92, 1967–1973. doi: 10.1016/j.apmr.2011.06.026 Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Phinyomark, A., Quaine, F., Charbonnier, S., Serviere, C., Tarpin-Bernard, F., and Laurillau, Y. (2013). EMG feature evaluation for improving myoelectric pattern recognition robustness. Exp. Syst. Appl. 40, 4832–4840. doi: 10.1016/j.eswa.2013.02.023 Copyright © 2018 Jarrassé, de Montalivet, Richer, Nicol, Touillet, Martinet, Paysant and de Graaf. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Powell, M. A., Kaliki, R. R., and Thakor, N. V. (2014). User training for pattern recognition-based myoelectric prostheses: improving phantom limb movement consistency and distinguishability. IEEE Trans. Neural Syst. Rehabil. Eng. 22, 522–532. doi: 10.1109/TNSRE.2013.2279737 November 2018 \| Volume 6 \| Article 164 Frontiers in Bioengineering and Biotechnology \| www.frontiersin.org 14
https://openalex.org/W4362419793	https://figshare.com/articles/journal_contribution/Supplementary_Figure_2_from_Anti-CD20_Antibody_Promotes_Cancer_Escape_via_Enrichment_of_Tumor-Evoked_Regulatory_B_Cells_Expressing_Low_Levels_of_CD20_and_CD137L/22399821/1/files/39845574.pdf	Zulu	null	Supplementary Figure 1 from Anti-CD20 Antibody Promotes Cancer Escape via Enrichment of Tumor-Evoked Regulatory B Cells Expressing Low Levels of CD20 and CD137L	null	2,023	cc-by	162	CD4 CD8 Non-activated Activated B-MOCK B-MDA B-MDA CD20lo a Ki67 b CD20lo CD20hi CD20 c 0 5 10 15 20 25 30 35 40 45 B-MOCK B-MDA B-MDA CD20lo B-MDA CD20hi Proliferation (%) CD4 CD8 * * * * CD4 CD8 Non-activated Activated B-MOCK B-MDA B-MDA CD20lo a Ki67 b CD20lo CD20hi CD20 c 0 5 10 15 20 25 30 35 40 45 B-MOCK B-MDA B-MDA CD20lo B-MDA CD20hi Proliferation (%) CD4 CD8 * * * * CD4 CD8 Non-activated Activated B-MOCK B-MDA B-MDA CD20lo a Ki67 a B-MDA b 0 5 10 15 20 25 30 35 40 45 B-MOCK B-MDA B-MDA CD20lo B-MDA CD20hi Proliferation (%) CD4 CD8 * * * * b CD20lo CD20hi CD20 c 0 5 10 15 20 25 30 35 40 45 B-MOCK B-MDA B-MDA CD20lo B-MDA CD20hi Proliferation (%) CD4 CD8 * * * * b CD20lo CD20hi CD20 c CD20 Suppl. Figure 2 Suppl. Figure 2 Suppl. Figure 2 Suppl. Figure 2
https://openalex.org/W2463922432	https://discovery.ucl.ac.uk/id/eprint/1504657/1/art%253A10.1007%252Fs11302-016-9522-7.pdf	English	null	P2X7R modulation of visually evoked synaptic responses in the retina	Purinergic signalling	2,016	cc-by	13,018	Purinergic Signalling DOI 10.1007/s11302-016-9522-7 Purinergic Signalling DOI 10.1007/s11302-016-9522-7 ORIGINAL ARTICLE P2X7R modulation of visually evoked synaptic responses in the retina Neuromodulation * Thomas E. Salt T.Salt@ucl.ac.uk * Thomas E. Salt T.Salt@ucl.ac.uk 1 Visual Neuroscience, UCL Institute of Ophthalmology, London EC1V 9EL, UK 2 Ocular Biology and Therapeutics, UCL Institute of Ophthalmology, London EC1V 9EL, UK 3 NIHR Biomedical Research Centre in Ophthalmology, London EC1V 9EL, UK 1 Visual Neuroscience, UCL Institute of Ophthalmology, London EC1V 9EL, UK P2X7R modulation of visually evoked synaptic responses in the retina Seetal Chavda1 & Philip J. Luthert2,3 & Thomas E. Salt1 Received: 14 January 2016 /Accepted: 24 June 2016 # The Author(s) 2016. This article is published with open access at Springerlink.com Received: 14 January 2016 /Accepted: 24 June 2016 # The Author(s) 2016. This article is published with open access at Springerlink.com Abstract P2X7Rs are distributed throughout all layers of the retina, and thus, their localisation on various cell types puts into question their specific site(s) of action. Using a dark- adapted, ex vivo mouse retinal whole mount preparation, the present study aimed to characterise the effect of P2X7R acti- vation on light-evoked, excitatory RGC ON-field excitatory post-synaptic potentials (fEPSPs) and on outer retinal electro- retinogram (ERG) responses under comparable conditions. The pharmacologically isolated NMDA receptor-mediated RGC ON-fEPSP was reduced in the presence of BzATP, an effect which was significantly attenuated by A438079 and other selective P2X7R antagonists A804598 or AF27139. In physiological Krebs medium, BzATP induced a significant potentiation of the ERG a-wave, with a concomitant reduction in the b-wave and the power of the oscillatory potentials. Conversely, in the pharmacologically modified Mg2+-free per- fusate, BzATP reduced both the a-wave and b-wave. The ef- fects of BzATP on the ERG components were suppressed by A438079. A role for P2X7R function in visual processing in both the inner and outer retina under physiological conditions remains controversial. The ON-fEPSP was significantly re- duced in the presence of A804598 but not by A438079 or AF27139. Furthermore, A438079 did not have any effect on the ERG components in physiological Krebs but potentiated and reduced the a-wave and b-wave, respectively, when applied to the pharmacologically modified medium. Therefore, activa- tion of P2X7Rs affects the function in the retinal ON pathway. The presence of a high concentration of extracellular ATP would most likely contribute to the modulation of visual transmission in the retina in the pathophysiological microenvironment. Abstract P2X7Rs are distributed throughout all layers of the retina, and thus, their localisation on various cell types puts into question their specific site(s) of action. Using a dark- adapted, ex vivo mouse retinal whole mount preparation, the present study aimed to characterise the effect of P2X7R acti- vation on light-evoked, excitatory RGC ON-field excitatory post-synaptic potentials (fEPSPs) and on outer retinal electro- retinogram (ERG) responses under comparable conditions. Keywords P2X7 receptor . Adenosine 5′-triphosphate . Retina . Neuromodulation Keywords P2X7 receptor . Adenosine 5′-triphosphate . Retina . 2 Ocular Biology and Therapeutics, UCL Institute of Ophthalmology, London EC1V 9EL, UK 3 NIHR Biomedical Research Centre in Ophthalmology, London EC1V 9EL, UK Introduction The P2X7 receptor (P2X7R) is the most diverse receptor sub- type of the P2X family in both structure and function [1]. ATP-gated P2X7Rs have been detected in neurons and glia, in both inner and outer synapse-rich and nuclear layers of the retina. They are functionally expressed by retinal ganglion cells [2–10] and Müller glial cells [11], and their presence on amacrine, horizontal [7, 12, 13] and microglial cells [4, 7] has also been suggested. The upregulation of P2X7R expression and activation by extracellular ATP is associated with the el- evation of intracellular calcium and subsequent photoreceptor and retinal ganglion cell (RGC) death, thus implicating a role for P2X7Rs in retinal degenerations such as retinitis pigmentosa, glaucoma, age-related macular degeneration (AMD) and retinopathies [14, 15]. The P2X7R is relatively insensitive to its extracellular ag- onist, adenosine 5′-triphosphate (ATP), and exhibits a biphasic agonist-evoked response, dependent upon the agonist expo- sure time [16]. Similar to other P2X subtypes, a brief agonist application induces a reversible, transient inward current through a non-selective ion channel, permeable to small cat- ions. Repeated or prolonged agonist application gives rise to a sustained current, which could ultimately lead to cell death, Purinergic Signalling unlimited access to food and water. All procedures were in accordance with the UK Animals (Scientific Procedures) Act 1986 and approved by the UK Home Office and the University College London local ethics committee. assumed to be associated with the opening of a large pore that is permeable to molecules of high molecular weight (<900 Da) [16, 17]. Although P2X7R function in retinal pathophysiology is becoming established, the role of these receptors in synaptic function in the healthy retina has remained largely unex- plored, until now. The specific localisation of P2X7Rs on neu- rons of the vertical rod-mediated pathway suggests that these receptors may have an important role in modulating scotopic visual responses under physiological conditions [7, 13]. Under physiological conditions, in vivo electroretinogram (ERG) studies have shown alterations to both outer and inner retinal function, in response to short-term administration of the P2XR agonist, BzATP, in the rat retina [7, 12, 13]. However, the specific contribution of P2X7Rs to the BzATP- mediated modulation of retinal function remains elusive. A further complexity is that P2X7Rs are distributed throughout all layers of the retina, adding to the uncertainty of the specific site(s) of action for these receptors in modulating neurotrans- mission within the retina. Introduction In line with this notion, it is unclear from the sole use of ERG recordings whether P2X7Rs also directly modulate retinal ganglion cell synaptic function. Response acquisition and recording electrode Transretinal electroretinogram (ERG) and RGC field excitato- ry post-synaptic potential (fEPSP) recordings were made via Krebs medium-filled extracellular recording electrodes (2– 8 MΩ), pulled from borosilicate capillaries (1.2 × 0.69 mm, Clark Electromedical Instruments). Responses were recorded with an Axoprobe-1A amplifier (Axon Instruments), digitised (5 kHz) via a CED1401 interface and stored on a computer with Spike2 software (Cambridge Electronic Design Ltd., Cambridge, UK). Mains-related noise (50 Hz) was eliminated with a Humbug noise reduction system (Quest Scientific). Preparation of the ex vivo mouse retinal whole mount All procedures were carried out in low-light conditions and red light was used for illumination, unless otherwise stated. Animals were killed by cervical dislocation and decapitation. The eyes were rapidly removed and placed into an ice-cold, oxygenated sucrose Krebs medium containing the following (mM): sucrose 202, KCl 2, KH2PO4 1.25, MgSO4 10, CaCl2 0.5, NaHCO3 26 and glucose 10. The eyes were placed in a Petri dish containing sucrose Krebs medium, under a dissec- tion microscope, and the optic nerve and extra-ocular muscle tissue were removed from the sclera. A cut was made at the ora serrata, and the cornea and iris were removed, followed by the removal of the lens and vitreous. The open eye-cup, consisting of the intact retina, retinal pigment epithelium and sclera, was cut into quadrants to allow it to lie flat, as a whole mount preparation, with the integrity of the retinal circuitry maintained and attached to the retinal pigment epithelium. A large body of evidence suggests a role for neuronal P2X7Rs in the modulation of neurotransmitter release and subsequent regulation of synaptic function [17]. P2X7R-me- diated neurotransmitter release has been detected in the cere- bellum [18], cerebral cortex and midbrain preparations [19–21], hippocampus [22] and neuromuscular junction [23]. Indeed, similar mechanisms may be at play within the retina, and the expression of P2X7Rs on Müller glia and mi- croglia also implicates a contribution to neuromodulation by non-neuronal sources to visual processing, by the activation of these receptors. Perfusion and recording chamber The retinal whole mount, ganglion cell layer side up, was transferred to a blacked-out interface recording chamber. The retina was held partially submerged and superfused with oxygenated Krebs medium containing the following (mM): NaCl 124, KCl 2, KH2PO4 1.25, MgSO4 1, CaCl2 2, NaHCO3 26 and glucose 10, at 0.4 ml/min and 36 ± 0.2 °C. The whole mount was left to recover for approximately 60 min prior to experimentation. The aim of the present study was to investigate the effect of P2X7R activation on visually evoked synaptic responses at the level of the photoreceptor, the ON bipolar cell, and RGC. We have performed this using a dark-adapted ex vivo mouse ret- inal whole mount preparation to record extracellular, light- evoked retinal responses and with the use of selective P2X7R compounds. Our findings suggest that functional P2X7Rs are present throughout the visual pathway in the ret- ina and that their activation leads to modulation of visual responses. Animals To obtain transretinal ERG responses, the recording elec- trode was manually lowered onto the surface of the retinal whole mount, without breaking into the retinal layers. For RGC fEPSP acquisition, a MS314 electronic micromanipula- tor (Marzhauser-Wetzler) was used to further lower the elec- trode vertically, from the surface of the whole mount in 5– Adult (4–12 weeks) C57BL/6 mice (Harlan Laboratories, UK) were housed on a 12-h light/dark cycle, in the Biological Resources Unit, Institute of Ophthalmology, University College London. The animals were given Purinergic Signalling 10 μm steps, at a rate of 1 μm/step, every 10–15 s until break- through to the ganglion cell layer, where recordings were ob- tained. Figure 1 illustrates the positions of the recording elec- trodes for acquisition of ERG and fEPSP responses. guide set 2 cm above the preparation. ERG responses were stimulated with a 10-ms duration flash, repeated every 3 s. RGC fEPSPs were stimulated with a 1-s duration flash, repeated every 3 s. For all experiments, the light stimulus was set to an intensity at which sub-maximal responses were elicited, typical- ly within the range of −1.47 to −0.96 log cd s/m2. Visual stimulation and characterisation of responses Upon stimulation, the transretinal ERG exhibited a charac- teristic waveform, composed of the negative a-wave, a prom- inent and large-amplitude positive b-wave and oscillatory po- tentials superimposed on the leading edge of the b-wave, qual- itatively similar to rodent rod ERGs obtained in vivo [24]. For the RGC fEPSP experiments, the onset and termination of the light stimulus produced negative RGC ON- and OFF-fEPSPs, respectively, similar to those described previously [25, 26]. The retinal whole mount was stimulated with a full-field, green light-emitting diode (LED; λ = 562 nm) via a fibre-optic light g g ( ) p g LIGHT ON OFF Fig. 1 Placement of recording electrodes for acquisition of transretinal electroretinogram and retinal ganglion cell fEPSP responses. Schematic illustrates retinal circuitry and the respective positions of the recording electrode for the electrophysiological experiments. The retinal whole mount was placed in the bath, ganglion cell layer side up. The full-field light flash stimulus was placed above the slice, and the arrow shows the direction in which the light travelled. Top trace, Typical transretinal electroretinogram response recorded by lowering the electrode onto the surface of the retinal whole mount, without breaking into the layers. The response was repeatedly elicited by a single 10-ms duration flash (green arrow), with a 3-s interval. Bottom trace, Retinal ganglion cell fEPSPs were recorded by placing the electrode in the ganglion cell layer. The green bar represents the duration of the stimulus (1 s), which was repeated at 3-s intervals. Upon stimulus onset, an ON retinal ganglion cell fEPSP was generated. Stimulus offset produced an OFF retinal ganglion cell fEPSP. ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, IPL inner plexiform layer, GCL ganglion cell layer, M microglia, cBC cone bipolar cell, rBC rod bipolar cell, AII AII amacrine cell LIGHT ON OFF LIGHT ON OFF Fig. 1 Placement of recording electrodes for acquisition of transretinal electroretinogram and retinal ganglion cell fEPSP responses. Schematic illustrates retinal circuitry and the respective positions of the recording l d f h l h i l i l i Th i l h l Pharmacological investigations Pharmacological compounds were added into the bathing me- dium, which continuously superfused the retinal preparation through the course of each experiment. All ERG experiments were carried out in either physiological, low Mg2+, or a phar- macologically modified Krebs medium (hereafter referred to as PMSTN) that was composed of the following (mM): NaCl 124, KCl 2, KH2PO4 1.25, CaCl2 2, NaHCO3 26, glucose 10, NBQX 0.01, picrotoxin 0.1, strychnine 0.005 and tetrodotoxin 0.0001. AMPA, GABAA/C and glycine receptors and sodium channels were blocked with NBQX, picrotoxin, strychnine and tetrodotoxin, respectively. The effect of P2X7R activation on synaptic transmission at rod photoreceptor-ON bipolar cell synapses was investigated under three different conditions: (1) physiological Krebs medium, (2) low Mg2+ Krebs medium and (3) PMSTN (as with the RGC experiments). Firstly, nor- mal Krebs medium was used to explore the effect of P2X7R activation on the ERG under physiological conditions. PMSTN medium was used as it allowed comparison of the effect of P2X7R activation on the ERG and the ON-RGC fEPSP under similar conditions. Additionally, it also enabled us to identify whether an upstream mechanism in the outer retina could be influencing the effect further downstream at the level of the retinal ganglion cells. The use of the low Mg2+ perfusate enabled further characterisation of the role of P2X7Rs in the retina, as their activation is considered to be enhanced in the absence of divalent cations. All RGC fEPSP experiments were undertaken using the PMSTN medium, in order prevent action potential firing and to isolate and reveal the NMDA-mediated component of the RGC ON-fEPSPs. Fig. 1 Placement of recording electrodes for acquisition of transretinal electroretinogram and retinal ganglion cell fEPSP responses. Schematic illustrates retinal circuitry and the respective positions of the recording electrode for the electrophysiological experiments. The retinal whole mount was placed in the bath, ganglion cell layer side up. The full-field light flash stimulus was placed above the slice, and the arrow shows the direction in which the light travelled. Top trace, Typical transretinal electroretinogram response recorded by lowering the electrode onto the surface of the retinal whole mount, without breaking into the layers. The response was repeatedly elicited by a single 10-ms duration flash (green arrow), with a 3-s interval. Bottom trace, Retinal ganglion cell fEPSPs were recorded by placing the electrode in the ganglion cell layer. Pharmacological investigations The green bar represents the duration of the stimulus (1 s), which was repeated at 3-s intervals. Upon stimulus onset, an ON retinal ganglion cell fEPSP was generated. Stimulus offset produced an OFF retinal ganglion cell fEPSP. ONL outer nuclear layer, OPL outer plexiform layer, INL inner nuclear layer, IPL inner plexiform layer, GCL ganglion cell layer, M microglia, cBC cone bipolar cell, rBC rod bipolar cell, AII AII amacrine cell Differences in protocols for investigating P2X7R function in ERG and RGC fEPSP experiments are highlighted where necessary. The P2X7R agonist, BzATP, was applied for 10 min. For the concentration-response investigation of BzATP on RGC ON-fEPSPs, no more than two applications of BzATP were used on one preparation, regardless of con- centration. The same protocol was followed for the AII amacrine cell Purinergic Signalling signal by high-frequency noise. A secondary low-pass filter (<15 Hz) was applied to isolate a smoothed ERG and eliminate the oscillatory potentials. Figure 2 shows no overall distortion in the power spectra or waveforms of the low-frequency ERG components (a-wave and b-wave), following two orders of low-pass filtering. The smoothed waveform ensured accurate measurement of the a-wave slope and b-wave amplitude with- out contamination from the oscillatory potentials. concentration-response investigation of adenosine on RGC fEPSP responses. To test the effect of BzATP in the presence of the P2X7R antagonist A438079 on both ERG and RGC fEPSP responses, BzATP (300 μM) was added to the PMSTN medium for 10 min, followed by a 30–40-min washout period. A438079 was then added to the medium and allowed to superfuse the retina until a stable response was reached (∼20–30 min). BzATP was then co-applied with the antagonist-containing me- dium. In separate experiments, the effect of the P2X7R antago- nists A804598 and AF27139 (both prepared in DMSO) on the BzATP-mediated effect on RGC fEPSPs was tested using the same protocol as for A438079. As a control, DMSO (0.1 %) was added to the PMSTN medium in these experiments. The slope of the a-wave, which reflects photoreceptor ac- tivity [27], was measured between the onset and peak of the a- wave. The b-wave peak amplitude was measured from the peak of the a-wave to half the recovery phase of the b-wave and is known to reflect ON bipolar cell function [28]. Sweeps were analysed by averaging these parameters over 60 s, throughout the time course of the experiment. Digital filtering and analysis of ERG components The ERG traces were analysed offline using Spike2 software (Cambridge Electronic Design, UK). Waveforms were initially low-pass-filtered (<250 Hz) to reduce contamination of the 10 20 30 0 0 50 100 150 200 Power (µV.sec)2 Frequency (Hz) 200 ms 40 μV < 250 Hz < 250 Hz + < 15 Hz < 250 Hz + > 60 Hz Filter frequencies Low to mid-frequency a-wave and b-wave components dominate the power spectrum High frequency oscillatory potentials A B Fig. 2 Digital filtering of transretinal electroretinogram responses. a Dark-adapted electroretinogram (ERG) power spectra (300 s averages) under various filter settings. The ERG was low-pass-filtered (<250 Hz; dashed traces), with the prominent a-wave and b-wave components within the low to mid frequency range. Further low-pass filtering (<15 Hz; dotted traces) eliminated the high-frequency component (oscillatory potentials). A high-pass filter (>60 Hz; solid trace) was applied after the first filter in order to isolate the high-frequency oscillatory potentials, which exhibited a peak frequency of approximately 100 Hz. b Waveform averages illustrate the effect of the filter settings on the shape of the ERG. Application of two orders of low- pass filter did not distort the waveform keeping the a-wave and b-wave intact. Application of the high-pass filter eliminated the a-wave and b- wave isolating the oscillatory potentials for further analysis. Arrows represent onset of flash stimulus (10 ms) 10 20 30 0 0 50 100 150 200 Power (µV.sec)2 Frequency (Hz) Low to mid-frequency a-wave and b-wave components dominate the power spectrum High frequency oscillatory potentials A A Low to mid-frequency a-wave and b-wave components dominate the power spectrum 200 ms 40 μV < 250 Hz < 250 Hz + < 15 Hz < 250 Hz + > 60 Hz Filter frequencies B B Fig. 2 Digital filtering of transretinal electroretinogram responses. a Dark-adapted electroretinogram (ERG) power spectra (300 s averages) under various filter settings. The ERG was low-pass-filtered (<250 Hz; dashed traces), with the prominent a-wave and b-wave components within the low to mid frequency range. Further low-pass filtering (<15 Hz; dotted traces) eliminated the high-frequency component (oscillatory potentials). A high-pass filter (>60 Hz; solid trace) was applied after the first filter in order to isolate the high-frequency oscillatory potentials, which exhibited a peak frequency of approximately 100 Hz. b Waveform averages illustrate the effect of the filter settings on the shape of the ERG. Pharmacological investigations Drug effects on both response components were measured by calculating the mean ± SEM percentage change in the measured parame- ters (slope or amplitude), relative to control values. The im- plicit time of the b-wave, in particular, can provide a measure of photoreceptor sensitivity, whereas possible contamination Digital filtering and analysis of ERG components Application of two orders of low- pass filter did not distort the waveform keeping the a-wave and b-wave intact. Application of the high-pass filter eliminated the a-wave and b- wave isolating the oscillatory potentials for further analysis. Arrows represent onset of flash stimulus (10 ms) Purinergic Signalling of the a-wave by the onset of the b-wave must also be consid- ered during analysis of the kinetics of these ERG components [29]. The latencies were calculated from stimulus onset to the peak amplitudes of the a-wave or b-wave, and values are shown as mean ± SEM implicit time. medium.AsshowninFig.3,BzATPwasappliedtotheperfusate for10min,andtheagonist-inducedmodulationoftheON-fEPSP wascomparedacrossconcentrationsrangingfrom10to300μM. There was no significant effect on the ON-fEPSP with low mi- cromolar concentrations of BzATP. However, a concentration- relatedreductionintheON-fEPSPpeakamplitudewasmeasured at higher micromolar concentrations. BzATP at 300 μM, the highest concentration tested, elicited a significant reduction in the ON-fEPSP to 78.1 ± 3 % of control (P < 0.05, n = 21) (Fig. 3a, b). BzATP elicited its effect within approximately 2– 3 min of reaching the retina and was sustained throughout the duration of application (Fig. 3b(ii)). This effect was partially re- coverable on washout to 92.4 ± 2 % of control (P < 0.05). Oscillatory potentials were analysed in the frequency do- main. These high-frequency wavelets are believed to reflect inner retinal network function, particularly the activity of amacrine and ganglion cells [30]. A secondary high-pass filter (>60 Hz) was applied to isolate the oscillatory potentials from the first filtered waveform, as shown in Fig. 2. The oscillatory potentials exhibited a peak frequency of ∼90–110 Hz, which is comparable to that of dark-adapted mouse ERGs in vivo [31]. A fast Fourier transform was used to derive the power spectrum of the oscillatory potentials, from stimulus onset to the half-width of the b-wave recovery phase (∼240 ms). To analyse drug effects, average spectra were generated 120 s before, during and after drug application, and the mean ± SEM percentage change in peak power was compared with control values. ( ) The effect of BzATP on the ON-fEPSP was assessed in the presence of the selective, competitive P2X7R antagonists A438079, A804598 or AF27139. Compared to the more clas- sically used P2X7R antagonists, A438079 and A804598 are generally considered amongst the most potent suppressors of mouse P2X7R function, of those commercially available [32, 33]. Analysis of RGC ON-fEPSPs RGC ON-fEPSP traces were also analysed offline using Spike2 software (Cambridge Electronic Design, UK). Responses were visualised as 3-s ‘sweeps’ and were analysed by measuring the average amplitude of the ON-fEPSPs over 60 s, throughout the time course of the experiment. The effect of pharmacological agents on the ON responses was measured by calculating the mean ± SEM percentage change in fEPSP amplitude, compared with control values. Experiments were separate for each antagonist and were carried out in PMSTN medium. The effect of each antagonist on the BzATP-induced changes in the ON- fEPSP was directly compared to the preceding applica- tion of BzATP in the absence of the antagonist. A438079 (10 μM), A804598 (30 μM) and AF27139 (30 μM) all partially attenuated the BzATP-mediated reduction of the ON-fEPSP to 68.7 ± 2 %, 55.1 ± 11 % and 75.2 ± 7 % of control, when compared to the max- imum effect of BzATP (P < 0.05 for all) (Fig. 3c). Digital filtering and analysis of ERG components AF27139 (Lundbeck, DK) is not yet commercially avail- able, and A804598 has not been widely tested, but both have shown to potently block P2X7R-mediated cytokine release in cell culture preparations across species [33]. The less novel A438079 was chosen for this study as it has demonstrated potent suppression of P2X7R function in models of chronic pain, microglial neuroinflammation and neurodegenerative disorders in the spinal cord and brain [34–38]. The use of A438079 in ex vivo retinal whole mount preparations has not previously been reported, although the compound has proven potent in suppressing swelling-induced damage to iso- lated retinal ganglion cells [9]. Statistical analysis Data were imported from Spike2 to Excel (Microsoft, USA) for analysis and generating graphs. Statistical testing was car- ried out using GraphPad Prism (v.6 for Windows, CA, USA). For comparisons between baseline and drug effect on the re- sponses, statistical significance was established using the Wilcoxon matched pairs test. Comparisons between treatment groups were carried out using the Mann-Whitney test. For all tests, statistical significance was observed if P < 0.05. To explore whether the release of endogenous ATP acting on P2X7Rs during visual stimulation affects ON-centre retinal gan- glion cell activity, the effects of A438079, A804598 and AF27139, when applied alone, were investigated (Fig. 3d). The ON-fEPSP was reduced in the presence of A804598 to 85.0 ± 15 % of control (P < 0.05, n = 7) but was not significantly affected by A438079 or AF27139 at the concentrations tested. Results Suppression of the retinal ganglion cell ON-fEPSP by BzATP and reduction of the effect by the selective and competitive P2X7 receptor antagonists A438079, A804598 and AF27139 Adenosine potentiates the retinal ganglion cell ON-fEPSP ns not significant; P < 0.05, compared to control % Control ON fEPSP amplitude 50 60 70 80 90 100 110 * ns * -30 -20 -10 0 10 -10 0 10 20 30 Time (min) % Change ON fEPSP amplitude * A B b c a % Control ON fEPSP amplitude 50 60 70 80 90 100 110 * * ns * A -30 -20 -10 0 10 -10 0 10 20 30 Time (min) % Change ON fEPSP amplitude * B b c a B A 0 20 40 60 80 100 Control A438079 (10 μM) A804598 (30 μM) AF27139 (30 μM) % BzATP-evoked effect * * * + BzATP (300 μM) C D % Control ON fEPSP amplitude 0 20 40 60 80 100 Control A438079 (10 μM) A804598 (30 μM) AF27139 (30 μM) * ns ns D D C + BzATP (300 μM) Fig. 3 The effect of BzATP on the light-evoked ON-fEPSP. Responses are pharmacologically isolated, NMDAR-mediated. BzATP was applied (10 min) at concentrations of 10 (n = 2), 30 (n = 6), 100 (n = 6) and 300 μM (n = 21). Values are mean ± SEM percentage of pre-treatment control fEPSP peak amplitude. a BzATP elicited a concentration-related reduction of the ON-fEPSP. Washout of BzATP (300 μM) induced recovery of the ON-fEPSP. b Time course plot shows the effect of BzATP (300 μM) on the ON-fEPSP. Plotted values are mean ± SEM percentage change in fEPSP peak amplitude, relative to pre-treatment control. Representative traces (120 s averages) illustrate the effect of BzATP (300 μM) on the ON-fEPSP. a Control, b BzATP (300 μM), c wash. Scale bars = 400 ms, 20 μV. c Actions of selective P2X7R antagonists on the BzATP-induced suppression of the ON-fEPSP. The selective P2X7R antagonists, A438079 (n = 6), A804598 (n = 7) or AF27139 (n = 6), significantly attenuated the BzATP (300 μM)- mediated reduction of the ON-fEPSP. All three compounds exhibited similar potency in blocking P2X7R function at the concentrations tested. Note that in the presence of all antagonists, a large residual effect on the ON-fEPSP persisted with BzATP application. d Direct actions of selective P2X7R antagonists on the NMDAR-mediated ON- fEPSP. The ON-fEPSP was significantly reduced in the presence of A804598 (n = 7) but not by A438079 (n = 6) or AF27139 (n = 6) at the concentrations tested. Adenosine potentiates the retinal ganglion cell ON-fEPSP ns not significant; P < 0.05, compared to control adenosine on the ON-fEPSP are due to the activation of different purinergic receptors. suggested that the actions of BzATP in the hippocampus are due to the effect of its breakdown product Bz- adenosine through the subsequent activation of A1 re- ceptors [39]. Adenosine receptors are known to be expressed in the retina. Therefore, it was investigated whether the effect of BzATP on the RGC fEPSPs was mediated by the activation of adenosine receptors. Adenosine potentiates the retinal ganglion cell ON-fEPSP TheeffectofP2X7RactivationonsynapticresponsesofRGCwas investigated. All experiments were carried out in PMSTN Extracellular ATP is rapidly degraded into its constituent adenosine by ectonucleotidases. It has previously been Purinergic Signalling % Control ON fEPSP amplitude 50 60 70 80 90 100 110 * ns * 0 20 40 60 80 100 Control A438079 (10 μM) A804598 (30 μM) AF27139 (30 μM) % BzATP-evoked effect * * * + BzATP (300 μM) -30 -20 -10 0 10 -10 0 10 20 30 Time (min) % Change ON fEPSP amplitude * A B C b c a % Control ON fEPSP amplitude 0 20 40 60 80 100 Control A438079 (10 μM) A804598 (30 μM) AF27139 (30 μM) * ns ns D Fig. 3 The effect of BzATP on the light-evoked ON-fEPSP. Responses are pharmacologically isolated, NMDAR-mediated. BzATP was applied (10 min) at concentrations of 10 (n = 2), 30 (n = 6), 100 (n = 6) and 300 μM (n = 21). Values are mean ± SEM percentage of pre-treatment control fEPSP peak amplitude. a BzATP elicited a concentration-related reduction of the ON-fEPSP. Washout of BzATP (300 μM) induced recovery of the ON-fEPSP. b Time course plot shows the effect of BzATP (300 μM) on the ON-fEPSP. Plotted values are mean ± SEM percentage change in fEPSP peak amplitude, relative to pre-treatment control. Representative traces (120 s averages) illustrate the effect of BzATP (300 μM) on the ON-fEPSP. a Control, b BzATP (300 μM), c wash. Scale bars = 400 ms, 20 μV. c Actions of selective P2X7R antagonists on the BzATP-induced suppression of the ON-fEPSP. The selective P2X7R antagonists, A438079 (n = 6), A804598 (n = 7) or AF27139 (n = 6), significantly attenuated the BzATP (300 μM)- mediated reduction of the ON-fEPSP. All three compounds exhibited similar potency in blocking P2X7R function at the concentrations tested. Note that in the presence of all antagonists, a large residual effect on the ON-fEPSP persisted with BzATP application. d Direct actions of selective P2X7R antagonists on the NMDAR-mediated ON- fEPSP. The ON-fEPSP was significantly reduced in the presence of A804598 (n = 7) but not by A438079 (n = 6) or AF27139 (n = 6) at the concentrations tested. P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission b Time course plot shows the effect of adenosine (300 μM; green) and vehicle (yellow) on the ON-fEPSP. Plotted values are mean ± SEM percentage change in fEPSP peak amplitude relative to pre-treatment control. c Representative traces (120 s averages) illustrate the effect of adenosine (300 μM) on the ON-fEPSPs -20 -10 0 10 20 -5 0 5 10 15 20 25 Time (min) % Change ON fEPSP amplitude B ADO 300 μM, 12 min Wash ON-fEPSP. ns not significant; P < 0.05, compared to control. b Time course plot shows the effect of adenosine (300 μM; green) and vehicle (yellow) on the ON-fEPSP. Plotted values are mean ± SEM percentage change in fEPSP peak amplitude relative to pre-treatment control. c Representative traces (120 s averages) illustrate the effect of adenosine (300 μM) on the ON-fEPSPs -20 -10 0 10 20 -5 0 5 10 15 20 25 Time (min) % Change ON fEPSP amplitude B 50 60 70 80 90 100 110 ns * ns A % Control ON fEPSP amplitude B A Time (min) Time (min) ADO 300 μM, 12 min C Control 20 μV 400 ms ADO 300 μM, 5 min C ADO 300 μM, 12 min Control Wash Fig. 4 The effect of adenosine on the NMDAR-mediated RGC ON- fEPSP. Adenosine (ADO) was applied (10 min) at concentrations of 100 and 300 μM. n = 8 for both concentrations tested. a Values are mean ± SEM percentage of pre-treatment control fEPSP peak amplitude, taken ∼5–7 min after the start of adenosine application. The ON-fEPSP was reversibly and significantly potentiated by adenosine, in a concentration-related manner. Vehicle application elicited no effect on the ON-fEPSP. ns not significant; P < 0.05, compared to control. b Time course plot shows the effect of adenosine (300 μM; green) and vehicle (yellow) on the ON-fEPSP. Plotted values are mean ± SEM percentage change in fEPSP peak amplitude relative to pre-treatment control. c Representative traces (120 s averages) illustrate the effect of adenosine (300 μM) on the ON-fEPSPs A concomitant reduction of the ERG b-wave ampli- tude to 80.6 ± 4 % of control (P < 0.05, n = 13) was also measured in the presence of BzATP (Fig. 5a(ii)). The changes in the b-wave within 1–2 min of application appeared to reach maximum reduction in amplitude im- mediately prior to washout. P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission Considering the direct out- put of the photoreceptors onto the ON bipolar cells along the vertical pathway, it would be expected that a potentiation in the a-wave slope would lead to a subse- quent increase in the b-wave amplitude. However, this does not seem to be the case here, and is consistent with the in vivo findings of the effect of BzATP on the rat ERG [13]. The b-wave amplitude recovered to baseline within approximately 10 min following wash- out of BzATP (97.3 ± 3 % of control, P > 0.05). BzATP application also induced a potentiation in the implicit time of the b-wave (Fig. 5a(ii)). The differing effect of BzATP on the kinetics of the a-wave and b-wave, as has been previously demonstrated in vivo [13], suggests that there are independent photoreceptor and post- photoreceptor sites of action stimulated by the P2X7R agonist. response components, the P2X7R agonist was applied in nor- mal physiological medium that did not contain magnesium ions. Under these conditions, BzATP induced a reduction in the a-wave slope to 90.7 ± 3 % of control (P < 0.05, n = 6), which was typically recoverable 5 min post-wash to 97.2 ± 3 % of control (P > 0.05) (Fig. 5b(i)). Furthermore, BzATP increased the implicit time of the a-wave maximum amplitude (Fig. 5b(i)). The BzATP-mediated reduction in the a-wave slope was contrary to that seen in the presence of extracellular magnesium, suggesting that the effect of P2X7R activation is reversed under these conditions. In the absence of extracellular magnesium, BzATP sig- nificantly reduced the b-wave amplitude to 89.5 ± 1 % of control (P < 0.05, n = 6) (Fig. 5b(ii)). The temporal profile of the b-wave amplitude was comparable to that of the a- wave slope during the early phase of BzATP application, where an exponential reduction in both components was apparent. However, the differing recovery periods of the a- and b-waves strongly support the notion that BzATP elic- ited independent effects on photoreceptor and post- photoreceptor components. Additionally, BzATP did not elicit a significant effect on the implicit time of the b-wave maximum amplitude (Fig. 5b(ii)). The suppressive effect of BzATP on the b-wave amplitude in the presence and absence of extracellular magnesium was similar in magni- tude. P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission The effect of P2X7R activation on synaptic transmission at rod photoreceptor-ON bipolar cell synapses was investigated un- der three different conditions. Dark-adapted retinal whole mounts were used to record visually evoked ERG responses. In physiological Krebs medium, a 10-min application of BzATP induced a marked potentiation of the a-wave slope to 149.0 ± 14 % of control (P < 0.05, n = 13) (Fig. 5a(i)). The BzATP-mediated changes in the a-wave were relatively rapid in onset, reaching a sustained maximum effect within approx- imately 5–6 min after the start of drug application. Following washout of BzATP, the a-wave slope recovered to 108.2 ± 3 % of control (P > 0.05). BzATP also significantly increased the implicit time of the a-wave maximum amplitude compared to pre-treatment control (P < 0.05) (Fig. 5a(i)). The ON-fEPSP was significantly potentiated by aden- osine, exhibiting a dose-related effect. For the highest concentration tested (300 μM), maximum potentiation was achieved within 3–5 min of adenosine exposure, to 109.5 ± 2 % of control (P < 0.05, n = 8) (Fig. 4). The effect on the ON-fEPSP peak amplitude was not sustained but decayed prior to washout. A transient re- duction in the ON-fEPSP was also observed before re- covering fully, to 99.8 ± 2 % of control approximately 10 min later (P > 0.05), which is illustrated by the rep- resentative traces (Fig. 4c). Vehicle application had no significant effect on the ON-fEPSP (Fig. 4). These re- sults suggest that the differential effects of BzATP and Purinergic Signalling 50 60 70 80 90 100 110 ns * ns -20 -10 0 10 20 -5 0 5 10 15 20 25 Time (min) % Change ON fEPSP amplitude A B C Control 20 μV 400 ms % Control ON fEPSP amplitude ADO 300 μM, 5 min ADO 300 μM, 12 min Wash Fig. 4 The effect of adenosine on the NMDAR-mediated RGC ON- fEPSP. Adenosine (ADO) was applied (10 min) at concentrations of 100 and 300 μM. n = 8 for both concentrations tested. a Values are mean ± SEM percentage of pre-treatment control fEPSP peak amplitude, taken ∼5–7 min after the start of adenosine application. The ON-fEPSP was reversibly and significantly potentiated by adenosine, in a concentration-related manner. Vehicle application elicited no effect on the ON-fEPSP. ns not significant; P < 0.05, compared to control. P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission The differences in the changes to the b-wave The effect of BzATP-mediated activation of P2X7Rs is considerably enhanced in the absence of extracellular divalent cations, particularly magnesium ions (Mg2+) [16, 40]. To fur- ther characterise the BzATP-mediated effect on the ERG Purinergic Signalling % Change b-wave amplitude A Physiological Krebs’ B Low Mg2+ C Low Mg2+ + PIC + STY + TTX + NBQX Time (min) % Change a-wave slope -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 74.4 ± 2 BzATP 80.6 ± 3 ns Wash 77.6 ± 3 ns -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 95.4 ± 3 BzATP 101.8 ± 2 * Wash 101.2 ± 2 * * -60 -40 -20 0 20 40 60 80 -5 0 5 10 15 20 25 I. T. (ms): Control 39.8 ± 2 BzATP 43.9 ± 2 * Wash 41.0 ± 2 * * -80 -60 -40 -20 0 20 40 -5 0 5 10 15 20 25 I. T. (ms) Control 29.7 ± 2 BzATP 32.5 ± 2 * Wash 31.0 ± 2 * * -80 -60 -40 -20 0 20 40 -5 0 5 10 15 20 25 I. T. (ms) Control 61.1 ± 7 BzATP 68.5 ± 7 * Wash 60.8 ± 7 ns * -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 104.9 ± 3 BzATP 109.8 ± 3 * Wash 108.3 ± 3 * * i ii iii i ii iii i ii iii Control BzATP Wash Control BzATP Wash Control BzATP Wash g. 5 The effect of BzATP on the dark-adapted ERG when recorded (n = 6), BzATP reduced the a-wave slope (i) and also caused a reduc A Physiological Krebs’ B Low Mg2+ C Low Mg2+ + PIC + STY + TTX + NBQX % Change a-wave slope -60 -40 -20 0 20 40 60 80 -5 0 5 10 15 20 25 I. T. (ms): Control 39.8 ± 2 BzATP 43.9 ± 2 * Wash 41.0 ± 2 * * -80 -60 -40 -20 0 20 40 -5 0 5 10 15 20 25 I. T. P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission (ms) Control 29.7 ± 2 BzATP 32.5 ± 2 * Wash 31.0 ± 2 * * -80 -60 -40 -20 0 20 40 -5 0 5 10 15 20 25 I. T. (ms) Control 61.1 ± 7 BzATP 68.5 ± 7 * Wash 60.8 ± 7 ns * i i i % Change b-wave amplitude Time (min) -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 74.4 ± 2 BzATP 80.6 ± 3 ns Wash 77.6 ± 3 ns * -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 95.4 ± 3 BzATP 101.8 ± 2 * Wash 101.2 ± 2 * * -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 104.9 ± 3 BzATP 109.8 ± 3 * Wash 108.3 ± 3 * * ii ii ii 5 -60 -40 -20 0 20 -5 0 5 10 15 20 25 I. T. (ms) Control 104.9 ± 3 BzATP 109.8 ± 3 * Wash 108.3 ± 3 * * ii iii iii Control BzATP Wash Control BzATP Wash iii iii Control BzATP Wash Control BzATP Wash iii Control BzATP Wash iii Control BzATP Wash iii Control BzATP Wash Control BzATP Wash Control BzATP Wash Control BzATP Wash Fig. 5 The effect of BzATP on the dark-adapted ERG when recorded under different conditions. Time course plots show the effect of BzATP (300 μM; 10 min) on the a-wave slope and b-wave amplitude (bars). Values are mean ± SEM percentage change and insets show mean ± SEM implicit time values. For all, ns not significant; P < 0.05, compared to pre-treatment control. a In physiological Krebs medium (n = 13), BzATP induced a sustained potentiation of the a-wave slope (i), which reached maximal effect within approximately 5 min of application. BzATP concomitantly reduced the b-wave amplitude (ii) over the course of application, an effect which was recoverable following washout. b In the absence of extracellular magnesium ions (n = 6), BzATP reduced the a-wave slope (i) and also caused a reduction in the b-wave amplitude (ii) over the course of application, an effect which was not fully recoverable following washout. c In the absence of extracellular magnesium ions, and in the presence of NBQX, picrotoxin, strychnine and TTX (n = 7), BzATP reduced the a-wave slope (i). P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission a In physiological Krebs medium (n = 6), the BzATP-mediated effects on the a-wave slope (i) and b-wave amplitude (ii) were markedly reduced in the presence of A438079 (10 μM) compared to control. b In the absence of extracellular magnesium ions, and in the presence of NBQX, picrotoxin, strychnine and TTX (n = 7), A438079 almost completely suppressed the BzATP-mediated potentiation of the ERG a-wave compared to control. b A438079-induced attenuation of the BzATP-mediated reduction in the b- wave was also observed. P < 0.05, compared to pre-treatment control 0 20 40 60 80 100 Control A438079 + BzATP * % BzATP-evoked effect on a wave 0 20 40 60 80 100 Control A438079 + BzATP % BzATP-evoked effect on a-wave * i i A Physiological Krebs’ B Low Mg2++PIC+STY+TTX+NBQX 0 20 40 60 80 100 Control A438079 + BzATP * % BzATP-evoked effect on a wave i B Low Mg2++PIC+STY+TTX+NBQX 0 20 40 60 80 100 Control A438079 + BzATP % BzATP-evoked effect on a-wave * i A Physiological Krebs’ B 0 20 40 60 80 100 Control A438079 + BzATP * on a wave 0 20 40 60 80 100 Control A438079 + BzATP * on b-wave i B Low Mg2++PIC+STY+TTX+NBQX ii extracellular magnesium ions, and in the presence of NBQX, picrotoxin, strychnine and TTX (n = 7), A438079 almost completely suppressed the BzATP-mediated potentiation of the ERG a-wave compared to control. b A438079-induced attenuation of the BzATP-mediated reduction in the b- wave was also observed. P < 0.05, compared to pre-treatment control A 0 20 40 60 80 100 Control A438079 + BzATP % BzATP-evoked effect on b-wave ii 0 20 40 60 80 100 Control A438079 + BzATP * % BzATP-evoked effect on b-wave ii ii extracellular magnesium ions, and in the presence of NBQX, picrotoxin, strychnine and TTX (n = 7), A438079 almost completely suppressed the BzATP-mediated potentiation of the ERG a-wave compared to control. b A438079-induced attenuation of the BzATP-mediated reduction in the b- wave was also observed. P < 0.05, compared to pre-treatment control Fig. 6 Actions of the selective P2X7R antagonist A438079 on the BzATP-mediated effects on the ERG components. Values are mean ± SEM percentage of control (max. BzATP effect, i.e. 100 %). P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission The b- wave amplitude was markedly suppressed (ii) and was not fully recoverable following washout. Note the relatively similar effect of BzATP on the b-wave, compared to that shown in b(ii). Representative traces (a(iii), b(iii), c(iii)) illustrate the effect of BzATP on the ERG under the respective conditions. Scale bars = 100 ms, 20 μV amplitude during the post-wash phase under both condi- tions imply that magnesium was required for full recovery following BzATP application. was also significantly delayed in the presence of BzATP, re- covering on washout (Fig. 5c(i)). Furthermore, there was a significant difference in the BzATP-mediated effect on the a- wave slope compared to that observed in the low magnesium solution (P < 0.05). In separate experiments, the effects of BzATP on the ERG a-wave and b-wave components were investigated in the pres- ence of the antagonists NBQX, picrotoxin, strychnine and tetrodotoxin and the absence of magnesium ions (PMSTN medium). As shown in Fig. 5c(i), there was a significant re- duction in the a-wave slope to 70.1 ± 8 % of pre-treatment control after 5 min (P < 0.05, n = 7), with partial recovery fol- lowing washout (76.3 ± 9 % of control, P < 0.05). The a-wave The effect of BzATP on the b-wave amplitude, in the PMSTN bathing medium, followed a similar temporal profile to that of the a-wave slope, with a significant reduction to 66.6 ± 10 % of control (P < 0.05, n = 7) (Fig. 5c(ii)) with little re- covery following washout of BzATP. The b-wave implicit time was also significantly increased with BzATP Purinergic Signalling 0 20 40 60 80 100 Control A438079 + BzATP % BzATP-evoked effect on a wave 0 20 40 60 80 100 Control A438079 + BzATP % BzATP-evoked effect on a-wave * 0 20 40 60 80 100 Control A438079 + BzATP % BzATP-evoked effect on b-wave * 0 20 40 60 80 100 Control A438079 + BzATP * % BzATP-evoked effect on b-wave i i A Physiological Krebs’ B Low Mg2++PIC+STY+TTX+NBQX ii ii Fig. 6 Actions of the selective P2X7R antagonist A438079 on the BzATP-mediated effects on the ERG components. Values are mean ± SEM percentage of control (max. BzATP effect, i.e. 100 %). P2X7R activation modulates rod photoreceptor-ON bipolar cell synaptic transmission a In physiological Krebs medium (n = 6), the BzATP-mediated effects on the a-wave slope (i) and b-wave amplitude (ii) were markedly reduced in the presence of A438079 (10 μM) compared to control. b In the absence of (Fig. 5c(ii)). Under these conditions, BzATP induced effects on the b-wave amplitude (P > 0.05) which were comparable to those seen in the absence of magnesium alone. effect; P < 0.05) (Fig. 6b(ii)). Thus, the A438079-mediated attenuation in the effect of BzATP on these components of the ERG confirms the involvement of P2X7R activation. The effect of direct application of the P2X7 receptor antagonist A438079 on the dark-adapted electroretinogram a-wave and b-wave components In order to further delineate P2X7R involvement, the effect of BzATP on the ERG components was assessed in the presence of the selective, competitive P2X7R antagonist, A438079 (10 μM). The effect of A438079 on the BzATP-induced changes in the a- and b-waves was directly compared to the preceding application of BzATP in the same experiment. When tested in physiological Krebs medium, A438079 sig- nificantly reduced the potentiating effect of BzATP on the a- wave slope to 57.3 ± 18 % of the effect of BzATP when ap- plied alone (P < 0.05, n = 6) (Fig. 6a(i)). Similarly, the BzATP- mediated suppression of the b-wave was markedly reduced in the presence of the antagonist to 41.0 ± 30 % of control (P < 0.05, n = 6) (Fig. 6a(ii)). Next, the effect of the P2X7R antagonist on the action of BzATP was tested in PMSTN medium. As shown in Fig. 7b(i), A438079 abolished the BzATP-mediated reduction of the a-wave slope to 4 ± 16 % of the effect of BzATP when applied alone (P < 0.05, n = 7). Furthermore, A438079 reduced the effect of BzATP on the b- wave amplitude to 58.3 ± 16 % of control (maximum BzATP To explore whether the release of endogenous ATP act- ing on P2X7Rs during visual stimulation affects outer retinal processing, the effect of A438079 on the ERG when applied alone was investigated. In physiological Krebs medium, A438079 (10 μM) alone did not induce an effect on the a-wave (108.9 ± 10 % of control, P > 0.05, n = 9) (Fig. 7a(i)). Similarly, the b-wave also remained unaffected by the application of the antagonist (93.3 ± 22 %, P > 0.05, n = 9) (Fig. 7a(ii)). Conversely, in the pharmacologically modified perfusion medium, the a-wave slope was significantly potentiated in the pres- ence of A438079 (10 μM) alone, to 125.7 ± 12 % of pre-treatment control (P < 0.05, n = 10). A438079 also markedly reduced the b-wave amplitude to 79.1 ± 15 % of control (P < 0.05, n = 10). The implicit times for both a- and b-waves were increased in the presence of the antagonist. These results suggest that P2X7R activation may modulate retinal function under baseline conditions Purinergic Signalling but that this depends on the particular ionic conditions of the microenvironment. The effect of direct application of the P2X7 receptor antagonist A438079 on the dark-adapted electroretinogram a-wave and b-wave components Discussion P2X7Rs modulate the ON pathway in the outer and inner ti 0 20 40 60 80 100 120 Control A438079 0 20 40 60 80 100 Control A438079 * 0 20 40 60 80 100 120 140 Control A438079 * % Control a-wave slope % Control b-wave amplitude % Control a-wave slope 0 20 40 60 80 100 120 Control A438079 % Control b-wave amplitude ns ns i i A Physiological Krebs’ B Low Mg2++PIC+STY+TTX+NBQX ii ii Fig. 7 The direct effect of the selective P2X7R antagonist A438079 on ERG components under different physiological conditions. Values are mean ± SEM percentage of pre- treatment control; n = 10. A438079 (10 μM) was superfused for 20–30 min until the response was stable. a In physiological Krebs medium, application of A438079 alone had no overall effect on the a-wave (i) or b-wave (ii). b In separate experiments using the pharmacologically modified Krebs medium, A438079 induced a marked potentiation in the a- wave (i) compared to control (P < 0.05). The b-wave (ii) was significantly reduced in the presence of A438079 (P < 0.05) 0 20 40 60 80 100 120 Control A438079 % Control a-wave slope ns i A Physiological Krebs’ 0 20 40 60 80 100 120 140 Control A438079 * % Control a-wave slope i B Low Mg2++PIC+STY+TTX+NBQX B A 0 20 40 60 80 100 120 Control A438079 % Control b-wave amplitude ns ii 0 20 40 60 80 100 Control A438079 * % Control b-wave amplitude ii ii but that this depends on the particular ionic conditions of the microenvironment. P2X7Rs modulate the ON pathway in the outer and inner retina ERG oscillatory potentials are suppressed by P2X7R activation The present study has demonstrated that neuronal function in both the outer and inner retina of the mouse under dark- adapted conditions is considerably altered by P2X7R activa- tion. An overview of the key findings are summarised in Table 1. Furthermore, P2X7R modulation of visual responses appears to occur through independent mechanisms, at differ- ent levels of the rod pathway. These findings are not unex- pected as P2X7Rs are expressed on photoreceptors; horizon- tal, amacrine and retinal ganglion cells; and on non-neuronal cells including microglia, retinal pigment epithelial cells and vascular endothelial cells within the retina. As of yet, P2X7Rs have not been localised to bipolar cells. The effect of P2X7R activation on the ERG oscillatory poten- tials was also explored, using physiological Krebs medium. The high-frequency oscillatory potentials were digitally ex- tracted from the raw ERG waveform (see Fig. 2). Changes in the peak power output of the oscillatory potentials were assessed with their frequency spectra before, during and after application of the pharmacological compound. As seen with the a-wave and b-wave components recorded in physiological Krebs medium, A438079 alone had no significant effect on the peak power of the oscillatory potentials (124.6 ± 25 % of control, P > 0.05, n = 6). Modulation of outer retinal function BzATP significantly reduced the peak power to 58.8 ± 8 % of control (P < 0.05, n = 13) (Fig. 8a, b). Following washout of BzATP, the peak power of the oscillatory potentials recovered to 97.9 ± 10 % of control (P > 0.05). Next, the effect of A438079 was tested on the BzATP-induced suppression of the oscillatory potentials. A438079 significantly reduced the effect of BzATP on the oscillatory potential peak power to 68.2 ± 8 % of control (P < 0.05, n = 6) (Fig. 8c). These results suggest that BzATP considerably suppressed the activity of inner retinal networks, and further imply that P2X7R modula- tion in the inner retina is at least partially independent of that in the outer retina. P2X7R-mediated alterations to synaptic transmission from rod photoreceptors to rod-ON bipolar cells were assessed using the dark-adapted ERG. Under physiological conditions, P2X7R activation by its agonist, BzATP, potentiated the a- wave, with a concomitant reduction in the b-wave. This find- ing is consistent with previous work in the rat in vivo [13]. BzATP also delayed the time to peak of the a-wave, which could indicate a reduction in photosensitivity of the rods [41]. However, in the study by Puthussery et al. [13], computational extraction of the sensitivity parameter of the a-wave leading Purinergic Signalling 0 20 40 60 80 100 120 Control BzATP Wash * ns % Control peak power A % BzATP-evoked effect 0 20 40 60 80 100 Control A438079 + BzATP * C 0 50 100 150 200 250 300 0 1 2 3 Frequency (Hz) Power (µV.sec)2 20 µV 100 ms BzATP Wash Control 0 20 40 60 80 100 120 Control BzATP Wash * ns % Control peak power A % BzATP-evoked effect 0 20 40 60 80 100 Control A438079 + BzATP * C B Fig. 8 The effect of BzATP on ERG oscillatory potentials. Experiments were done in physiological Krebs medium. a Mean ± SEM percentage of control peak power; n = 7. BzATP (300 μM; 10 min) caused a reversible and significant reduction in the peak power output of the oscillatory potentials compared to control (P < 0.05), indicating modulation of inner retinal function. b(i) Representative oscillatory potentials, before, during and post-BzATP application. Black, control; blue, BzATP; grey, wash. Under control conditions, the oscillatory potentials exhibited a peak frequency ∼107 Hz, typical of dark-adapted ERG responses. Modulation of outer retinal function In the presence of BzATP, no overall shift in frequency was detected. b(ii) Power Spectra (120 s) illustrate the BzATP-mediated suppression of the oscillatory potentials and recovery on washout. Black control; blue, BzATP; grey, wash. c A438079 (10 μM) significantly reduced the BzATP-evoked suppression of the oscillatory potentials compared to control. P < 0.05, compared to pre-treatment control 0 50 100 150 200 250 300 0 1 2 3 Frequency (Hz) Power (µV.sec)2 20 µV 100 ms BzATP Wash Control B B B A Frequency (Hz) fect of BzATP on ERG oscillatory potentials. Experiments Fig. 8 The effect of BzATP on ERG oscillatory potentials. Experiments were done in physiological Krebs medium. a Mean ± SEM percentage of control peak power; n = 7. BzATP (300 μM; 10 min) caused a reversible and significant reduction in the peak power output of the oscillatory potentials compared to control (P < 0.05), indicating modulation of inner retinal function. b(i) Representative oscillatory potentials, before, during and post-BzATP application. Black, control; blue, BzATP; grey, wash. Under control conditions, the oscillatory potentials exhibited a peak frequency ∼107 Hz, typical of dark-adapted ERG responses. In the presence of BzATP, no overall shift in frequency was detected. b(ii) Power Spectra (120 s) illustrate the BzATP-mediated suppression of the oscillatory potentials and recovery on washout. Black control; blue, BzATP; grey, wash. c A438079 (10 μM) significantly reduced the BzATP-evoked suppression of the oscillatory potentials compared to control. P < 0.05, compared to pre-treatment control edge indicated no change in photoreceptor sensitivity as a result of P2X7R activation. Thus, the change in implicit time of the a-wave in this investigation could instead be attributed to a possible contamination of the BzATP-induced effect on the b-wave. terminal induced by P2X7R activation would subsequently potentiate the a-wave during photostimulation, due to a great- er change in the photoreceptor dark current, as has been sug- gested previously [13]. Concurrently, facilitation of the pho- toreceptor dark current would increase the rate of glutamate release in the dark, thus eliciting a smaller light-induced depolarisation of the ON BCs. Stable calcium influx through L-type channels is imperative for the sustained release of glu- tamate from rod terminals in darkness [42]. A P2X7R- The opposing effects of BzATP on the photoreceptor and bipolar cell response components could be explained by a mechanism associated with a change in the dark current of the photoreceptors. Modulation of outer retinal function Depolarisation of the photoreceptor Table 1 Summary of the effect of BzATP on the RGC ON-fEPSP and ERG components in the absence and presence of P2X7R antagonists under the physiological conditions tested. All values are expressed as mean ± SEM. See text for n and P values. The effect of BzATP alone is compared to pre-treatment control. The effect of BzATP in the presence of an antagonist is compared to the maximum BzATP-mediated effect Physiological Krebs Modified Krebs (PMSTN) Low Mg2+ RGC ON-fEPSP – BzATP: 78.1 ± 3 % – BzATP + antagonist (% of max. BzATP effect) A438079: 68.7 ± 2 % A804598: 55.1 ± 11 % AF27139: 75.2 ± 7 % ERG a-wave BzATP: 149.0 ± 14 % BzATP: 70.1 ± 8 % BzATP: 90.7 ± 3 % BzATP + A438079: 57.3 ± 18 % BzATP + A438079: 4 ± 16 % – ERG b-wave BzATP: 80.6 ± 4 % BzATP: 66.6 ± 10 % BzATP: 89.5 ± 1 % BzATP + A438079: 41.0 ± 30 % BzATP + A438079: 58.3 ± 16 % – ERG OPs BzATP: 58.8 ± 8 % – – BzATP + A438079: 68.2 ± 8 % OPs oscillatory potentials Low Mg2+ Purinergic Signalling modulation of the ON bipolar cell response under both condi- tions tested. The presence of a residual effect of BzATP even with the addition of a selective P2X7R antagonist supports the notion that a consistent or secondary stress-related release of large amounts of endogenous ATP, possibly by non-neuronal cells, could act to influence retinal synaptic signalling. dependent rise in intracellular calcium has previously shown to directly stimulate vesicular glutamate release from hippo- campal and cortical presynaptic terminals [20–22]. The pres- ent findings suggest that P2X7R activation may modulate ERG responses through a similar calcium-dependent mecha- nism, although direct P2X7R modulation of ribbon synapses has not yet been shown. To explore whether P2X7Rs have a physiological role in outer retinal function upon visual stimulation, we investigated the effects of A438079 alone on the ERG components. When recorded in the physiological Krebs medium, blockade of P2X7Rs had no overall effect on the a-wave or b-wave. This contradicts previous findings whereby the b-wave was signif- icantly potentiated in a P2X7R knockout model [7]. However, both ERG components were markedly suppressed when the antagonist was applied in the modified perfusion medium. Modulation of outer retinal function A sensitised or stressed cellular microenvironment can initiate acute changes to glutamate and ATP homeostasis and signal- ling. Such deviations from the physiological norm may in- crease the likelihood of ATP release in the outer retina, at a sufficient concentration to activate P2X7Rs and subsequently affect the processing of the visual signal. P2X7R activity is altered and most likely enhanced in the absence of extracellular magnesium [16, 40]. In the present study, removal of magnesium ions from the perfusate reversed the effect of P2X7R activation on the a-wave, instead inducing a marked reduction in the photoreceptor response. This sup- ports the concept of a bi-directional function of P2X receptors in modulating synaptic activity dependent upon the physio- logical context, as has been consistently evident in central synapses, particularly in the hippocampus [43]. Without ex- tracellular magnesium, the b-wave was suppressed by BzATP, to a similar magnitude, when compared to the effect in the presence of the divalent cation. Contrary to the observed effect in physiological perfusate, the BzATP-mediated suppression of the b-wave was not reversible in the absence of magnesium. The differential temporal profiles of the effect of BzATP on the a- and b-waves highlight independent sites of modulation by P2X7Rs in both the inner and outer retina. These results strongly suggest a direct role for P2X7Rs in photoreceptors in modulating outer retinal processing and are supported by studies that have shown P2X7R immunoreactiv- ity at ribbon synapses within the rod spherule, in the rodent retina [7, 12, 13]. However, P2X7R modulation of photore- ceptor function through alternative mechanisms cannot be ruled out. The P2X7R expression profile in the healthy rabbit [9] and human retina [36, 45] has shown no evidence of im- munoreactivity in photoreceptors and may instead support the concept that P2X7Rs may modulate photoreceptor function through its expression in retinal pigment epithelial cells in the absence of pathology. Indeed, RPE cells are sources of ATP, which can be released in response to hyperosmotic stress [46], through various mechanisms including ATP-triggered secondary ATP release, CFTR-dependent ATP efflux as well as vesicular ATP secretion, as demonstrated in human retinal cells [47]. This autostimulatory ATP release may be driven through P2X7R-associated panx-1 hemichannels, as shown in the bovine retina [48]. Modulation of outer retinal function Furthermore, RPE cells have dem- onstrated the presence of ectonucleotidases [46], which may rapidly degrade ATP into its constituents, in order to activate other P2 and adenosine receptors in cells of the outer retina, and therefore influence synaptic signalling through alternative mechanisms. In the present study, the suppressive effect of P2X7R acti- vation on the photoreceptor response was enhanced in the presence of blockers of GABAA/C, glycine and AMPA/KA receptors and of voltage-gated sodium channels. It has previ- ously been suggested that P2X7R-associated modulation of the mouse ERG may be due to alterations in horizontal and amacrine cell feedback [7]. However, the underlying mecha- nism of modulation of the photoreceptor response may not be associated with the release of inhibitory inputs to the rod ter- minals under dark-adapted conditions [44]. Indeed, the study by Vessey and Fletcher [7] found that the ERG a-wave was unaltered, whereas the b-wave was markedly potentiated in a P2X7R-knockout mouse model. In the present investigation, there was no overall difference in the BzATP-induced reduc- tion in the b-wave, between responses recorded in the absence of magnesium alone, and with the addition of pharmacologi- cal blockers of inhibitory neurotransmission. This indicates that P2X7R-mediated modulation of the ON BCs is not due to an effect on inhibitory feedback at the ON BC terminal in the inner retina. Since BCs do not themselves express P2X7Rs, these results could reflect a possible modulation of both ERG components by non-neuronal cells during P2X7R stimulation. Modulation of inner retinal function The involvement of P2X7R-mediated modulation of the a- wave was confirmed with the selective antagonist A438079, which markedly suppressed and abolished the effect of P2X7R activation on the photoreceptor response, when tested in phys- iological Krebs and the pharmacologically modified medium, respectively. Similarly, A438079 partially blocked P2X7R The present study has demonstrated that P2X7R activation distinctly influences the activity of inner retinal networks, as shown by the changes observed in the ERG oscillatory poten- tials. A clear effect of P2X7R activation on the ON pathway was apparent, as BzATP suppressed the ON-fEPSP in a Purinergic Signalling concentration-related manner. The involvement of P2X7Rs in modulating the ON-fEPSP was supported by the partially at- tenuated effect of BzATP in the presence of the selective P2X7R antagonists A438079, A804598 or AF27139. With each antagonist, a residual effect of BzATP on the ON- fEPSP remained, which could be explained by a lack of po- tency of these compounds for blocking P2X7R-mediated ef- fects in the inner retinal layers. monkey retina [52], although evidence of its expression and function in the mouse is uncertain. In the rodent, P2X3Rs have been detected in the inner plexiform layer [53]. On activation, both P2X1Rs and P2X3Rs rapidly desensitise [1], and thus their kinetic profiles do not correlate with the BzATP-mediated ef- fects observed in the present study. Selective expression of P2X2Rs in starburst amacrine cells within the OFF vertical path- way has been demonstrated in the mouse retina [54, 55]. Although P2X2Rs display slowly desensitising properties simi- lar to P2X7Rs, it is uncertain whether they would directly affect neuronal responses within the ON pathway. As with the outer retina, the lack of potency of these com- pounds in suppressing the BzATP-mediated reduction in ON- centre RGC responses could be due to the stimulated release of endogenous ATP within the inner retinal layers. ATP is present in high concentrations within all metabolically active cells and so all can function as sources of ATP, which may then act to compete with the antagonist at the receptor. Alternatively, P2X7Rs are known to drive ATP release which can then act cell-autonomously. Indeed, P2X7R-mediated, non-vesicular release of ATP has been demonstrated in retinal ganglion cells [9]. P2X7Rs are expressed on amacrine, gangli- on and microglial cells in the inner mouse retina and are all therefore potential sources of ATP-mediated ATP release. Modulation of inner retinal function Metabotropic P2Y receptors may also modulate visual re- sponses in the retina as they have been suggested to play a role in glial-neuronal cross-talk [48]. Indeed both functional and im- munohistochemical expression of P2Y1Rs on Müller cells [11, 48, 56, 57] and P2Y4R on rod bipolar and amacrine cells [58] have been demonstrated in the retina. Considering that ADP and UTP are markedly more potent agonists at these receptors, re- spectively [59], their direct contribution to the BzATP-mediated effect on visual responses in the present study is unlikely. However, it is important to note that the remarkable diversity of purinergic receptors in such an intimate retinal microenviron- ment opens the possibility of their contribution to the overall glial-neuronal cross-talk in the processing of visual information, through the secondary release of endogenous purinergic ligands. In the absence of pathology or inflammation, the expression of functional P2X7Rs has only been demonstrated in Müller cells of the human retina but not in those of other mammalian species [49]. However, it has been shown in the rat retina that light-evoked ATP release from ganglion cells increases Müller cell Ca2+ transients through activation of glial P2YRs. This in turn mediates the release of glial ATP, which is rapidly hydro- lysed into adenosine, in turn hyperpolarising ganglion cells through the activation of A1 receptors [50]. It has been pro- posed that the BzATP-mediated suppression of excitatory re- sponses in the rodent hippocampus and brainstem was attribut- ed to its breakdown product Bz-adenosine and subsequent ac- tivation of adenosine receptors [39, 51]. Adenosine facilitated the ON-fEPSP for a short period before the effect diminished, leading to a temporary suppression of the ON-fEPSP before recovery on washout. This may be due to a number of factors, which include receptor desensitisation or clearance of adeno- sine from the extracellular environment via nucleoside trans- porters expressed in Müller cells. In the present study, the pos- sible contribution of A1 receptors to the BzATP-mediated re- duction of the ON-fEPSP was ruled out. Adenosine produced a dose-related potentiation in the ON-fEPSP, an effect which followed a differential temporal profile to that seen for the effect of BzATP. This suggests that adenosine facilitates possibly through A3 receptors, whereas P2X7Rs appear to reduce neu- ronal activity within the ON pathway. Funding This study was funded in part by Lundbeck Research. Funding This study was funded in part by Lundbeck Research. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appro- priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Modulation of inner retinal function The present findings are in accord with previous findings [8, 10] on P2X7R function in the retina and extend these find- ings to show that activation of these receptors can affect func- tion at different sites in the retinal ON pathway. Thus, activa- tion of these receptors under pathophysiological conditions by high concentrations of ATP would likely affect function at all stages of visual transmission in the retina. Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest. 1. North RA (2002) Molecular physiology of P2X receptors. Physiol Rev 82:1013–1067 2. Brändle U, Kohler K, Wheeler-Schilling TH (1998) Expression of the P2X7-receptor subunit in neurons of the rat retina. Mol Brain Res 62:106–109 2. Brändle U, Kohler K, Wheeler-Schilling TH (1998) Expression of the P2X7-receptor subunit in neurons of the rat retina. Mol Brain Res 62:106–109 References Zhang K, Yao G, Gao Y et al (2007) Frequency spectrum and amplitude analysis of dark- and light-adapted oscillatory potentials in albino mouse, rat and rabbit. Doc Ophthalmol 115:85–93 9. Xia J, Lim JC, Lu W et al (2012) Neurons respond directly to mechanical deformation with pannexin-mediated ATP release and autostimulation of P2X7 receptors. J Physiol 10:2285–2304 32. Donnelly-Roberts DL, Namovic MT, Han P, Jarvis MF (2009) Mammalian P2X7 receptor pharmacology: comparison of recom- binant mouse, rat and human P2X7 receptors. Br J Pharmacol 157: 1203–1214 10. Zhang X, Zhang M, Laties AM, Mitchell CH (2005) Stimulation of P2X7 receptors elevates Ca2+ and kills retinal ganglion cells. Invest Ophthalmol Vis Sci 46:2183–2191 11. Pannicke T, Fischer W, Biedermann B et al (2000) P2X7 receptors in Müller glial cells from the human retina. J Neurosci 20:5965– 5972 33. Donnelly-Roberts DL, Namovic MT, Surber B et al (2009) [3H]A- 804598 ([3H]2-cyano-1-[(1S)-1-phenylethyl]-3-quinolin-5- ylguanidine) is a novel, potent, and selective antagonist radioligand for P2X7 receptors. Neuropharmacology 56:223–229 12. Puthussery T, Fletcher EL (2004) Synaptic localization of P2X7 receptors in the rat retina. J Comp Neurol 472:13–23 34. Donnelly-Roberts DL, Jarvis MF (2007) Discovery of P2X7 receptor-selective antagonists offers new insights into P2X7 recep- tor function and indicates a role in chronic pain states. Br J Pharmacol 151:571–579 13. Puthussery T, Yee P, Vingrys AJ, Fletcher EL (2006) Evidence for the involvement of purinergic P2X receptors in outer retinal pro- cessing. Eur J Neurosci 24:7–19 14. Fletcher EL (2010) Mechanisms of photoreceptor death during ret- inal degeneration. Optom Vis Sci 87:269–275 35. McGaraughty S, Chu KL, Namovic MT et al (2007) P2X7-related modulation of pathological nociception in rats. Neuroscience 146: 1817–1828 15. Mitchell CH, Lu W, Hu H et al (2009) The P2X(7) receptor in retinal ganglion cells: a neuronal model of pressure-induced dam- age and protection by a shifting purinergic balance. Purinergic Signal 5:241–249 36. Teixeira JM, Oliveira MCG, Parada CA, Tambeli CH (2010) Peripheral mechanisms underlying the essential role of P2X7 re- ceptors in the development of inflammatory hyperalgesia. Eur J Pharmacol 644:55–60 16. Surprenant A, Rassendren F, Kawashima E et al (1996) The cyto- lytic P2Z receptor for extracellular ATP identified as a P2X receptor (P2X7). Science 272:735–738 37. References Although BzATP is a potent agonist of P2X7Rs, it is not selective for the receptor and can bind to P2X1, P2X2 and P2X3 receptors with similar potency [17], and therefore, the possibility remains that these receptor subtypes may also be activated in the presence of the agonist. P2X1R immunoreactiv- ity has been localised to the inner retinal layers of the cat and 1. North RA (2002) Molecular physiology of P2X receptors. Physiol Rev 82:1013–1067 2. Brändle U, Kohler K, Wheeler-Schilling TH (1998) Expression of the P2X7-receptor subunit in neurons of the rat retina. Mol Brain Res 62:106–109 Purinergic Signalling 25. Massey C, Miller RF (1990) Receptors of ganglion cells in rabbit retina. J Neurophysiol 63:16–30 3. Franke H, Klimke K, Brinckmann U et al (2005) P2X(7) receptor- mRNA and -protein in the mouse retina; changes during retinal degeneration in BALBCrds mice. Neurochem Int 47:235–242 26. Mittman S, Taylor WR, Copenhagen DR (1990) Concomitant acti- vation of two types of glutamate receptor mediates excitation of salamander retinal ganglion cells. J Physiol 428:175–197 4. Innocenti B, Pfeiffer S, Zrenner E et al (2004) ATP-induced non- neuronal cell permeabilization in the rat inner retina. J Neurosci 24: 8577–8583 27. Hood DC, Birch DG (1990) A quantitative measure of the electrical activity of human rod photoreceptors using electroretinography. Vis Neurosci 5:379–387 5. Ishii K, Kaneda M, Li H et al (2003) Neuron-specific distribution of P2X7 purinergic receptors in the monkey retina. J Comp Neurol 459:267–277 28. Stockton RA, Slaughter MM (1989) B-wave of the electroretino- gram. A reflection of ON bipolar cell activity. J Gen Physiol 93: 101–122 6. Sugiyama T, Oku H, Komori A, Ikeda T (2006) Effect of P2X7 receptor activation on the retinal blood velocity of diabetic rabbits. Arch Ophthalmol 124:1143–1149 29. Hood DC, Birch DG (1992) A computational model of the ampli- tude and implicit time of the b-wave of the human ERG. Vis Neurosci 8:107–126 7. Vessey KA, Fletcher EL (2012) Rod and cone pathway signalling is altered in the P2X7 receptor knock out mouse. PLoS One 7:e29990 8. Wheeler-Schilling TH, Marquordt K, Kohler K et al (2001) Identification of purinergic receptors in retinal ganglion cells. Brain Res Mol Brain Res 92:177–180 30. Wachtmeister L (1998) Oscillatory potentials in the retina: what do they reveal. Prog Retin Eye Res 17:485–521 31. References Marcellino D, Suárez-Boomgaard D, Sánchez-Reina MD et al (2010) On the role of P2X(7) receptors in dopamine nerve cell degeneration in a rat model of Parkinson’s disease: studies with the P2X(7) receptor antagonist A-438079. J Neural Transm 117: 681–687 17. Sperlágh B, Vizi ES, Wirkner K, Illes P (2006) P2X7 receptors in the nervous system. Prog Neurobiol 78:327–346 18. León D, Sánchez-Nogueiro J, Marín-García P, Miras-Portugal MAT (2008) Glutamate release and synapsin-I phosphorylation induced by P2X7 receptors activation in cerebellar granule neurons. Neurochem Int 52:1148–1159 38. Clark AK, Staniland AA, Marchand F et al (2010) P2X7-dependent release of interleukin-1beta and nociception in the spinal cord fol- lowing lipopolysaccharide. J Neurosci 30:573–582 19. Lundy PM, Hamilton MG, Mi L et al (2002) Stimulation of Ca2+ influx through ATP receptors on rat brain synaptosomes: identifi- cation of functional P2X7 receptor subtypes. Br J Pharmacol 135: 1616–1626 39. Kukley M, Stausberg P, Adelmann G et al (2004) Ecto- nucleotidases and nucleoside transporters mediate activation of adenosine receptors on hippocampal mossy fibers by P2X7 recep- tor agonist 2′-3′-O-(4-benzoylbenzoyl)-ATP. J Neurosci 24:7128– 7139 20. Marín-García P, Sánchez-Nogueiro J, Gómez-Villafuertes R et al (2008) Synaptic terminals from mice midbrain exhibit functional P2X7 receptor. Neuroscience 151:361–373 40. Virginio C, Church D, North R, Surprenant A (1997) Effects of divalent cations, protons and calmidazolium at the rat P2X7 recep- tor. Neuropharm 36:1285–1294 21. Alloisio S, Cervetto C, Passalacqua M et al (2008) Functional ev- idence for presynaptic P2X7 receptors in adult rat cerebrocortical nerve terminals. FEBS Lett 582:3948–3953 41. Hood DC, Birch DG (1997) Assessing abnormal rod photoreceptor activity with the a-wave of the electroretinogram: applications and methods. Doc Ophthalmol 92:253–267 22. Sperlágh B, Köfalvi A, Deuchars J et al (2002) Involvement of P2X7 receptors in the regulation of neurotransmitter release in the rat hippocampus. J Neurochem 81:1196–1211 42. Okawa H, Sampath AP (2007) Optimization of single-photon re- sponse transmission at the rod-to-rod bipolar synapse. Physiology (Bethesda) 22:279–286 23. Deuchars S, Atkinson L, Brooke RE et al (2001) Neuronal P2X7 receptors are targeted to presynaptic terminals in the central and peripheral nervous systems. J Neurosci 21:7143–7152 43. Pankratov Y, Lalo U, Krishtal OA, Verkhratsky A (2009) P2X receptors and synaptic plasticity. Neuroscience 158:137–148 24. Robson JG, Maeda H, Saszik SM, Frishman LJ (2004) In vivo studies of signaling in rod pathways of the mouse using the elec- troretinogram. Vision Res 44:3253–3268 44. 3. Franke H, Klimke K, Brinckmann U et al (2005) P2X(7) receptor- mRNA and -protein in the mouse retina; changes during retinal degeneration in BALBCrds mice. Neurochem Int 47:235–242 References Bloomfield S, Xin D (2000) Surround inhibition of mammalian AII amacrine cells is generated in the proximal retina. J Physiol 523(Pt 3):771–783 Purinergic Signalling 45. Niyadurupola N, Sidaway P, Ma N et al (2013) P2X7 receptor activation mediates retinal ganglion cell death in a human retina model of ischemic neurodegeneration. Invest Ophthalmol Vis Sci 54:2163–2170 52. Yazulla S, Studholme KM (2004) Vanilloid receptor like 1 (VRL1) immunoreactivity in mammalian retina: colocalization with so- matostatin and purinergic P2X1 receptors. J Comp Neurol 474: 407–418 53. Puthussery T, Fletcher EL (2007) Neuronal expression of P2X3 purinoceptors in the rat retina. Neuroscience 146:403–414 46. Eldred JA, Sanderson J, Wormstone M et al (2003) Stress-induced ATP release from and growth modulation of human lens and retinal pigment epithelial cells. Biochem Soc Trans 31:1213–1215 54. Kaneda M, Ishii T, Hosoya T (2008) Pathway-dependent modula- tion by P2-purinoceptors in the mouse retina. Eur J Neurosci 28: 128–136 47. Reigada D, Mitchell CH (2005) Release of ATP from retinal pig- ment epithelial cells involves both CFTR and vesicular transport. Am J Physiol Cell Physiol 288(1):C132–C140 55. Kaneda M, Ito K, Shigematsu Y, Shimoda Y (2010) The OFF- pathway dominance of P2X(2)-purinoceptors is formed without visual experience. Neurosci Res 66:86–91 48. Reigada D, Lu W, Zhang M, Mitchell CH (2008) Elevated pressure triggers a physiological release of ATP from the retina: possible role for pannexin hemichannels. Neuroscience 157:396–404 56. Bringmann A, Pannicke T, Weick M et al (2002) Activation of P2Y receptors stimulates potassium and cation currents in acutely isolat- ed human Müller (glial) cells. Glia 37:139–152 49. Wurm A, Pannicke T, Iandiev I et al (2011) Purinergic signaling involved in Müller cell function in the mammalian retina. Prog Retin Eye Res 30:324–342 57. Ward MM, Fletcher EL (2009) Subsets of retinal neurons and glia express P2Y1 receptors. Neuroscience 160:555–566 50. Newman E (2005) Calcium increases in retinal glial cells evoked by light-induced neuronal activity. J Neurosci 25:5502–5510 58. Ward MM, Puthussery T, Fletcher EL (2008) Localization and pos- sible function of P2Y(4) receptors in the rodent retina. Neuroscience 155:1262–1274 51. Tautenhahn M, Leichsenring A, Servettini I et al (2012) Purinergic modulation of the excitatory synaptic input onto rat striatal neurons. Neuropharmacology 62:1756–1766 59. Burnstock G (2009) Purines and purinoceptors: molecular biology overview. Curr Top Membr 54:1–27
https://openalex.org/W4225329783	https://zenodo.org/record/6516874/files/V2I4-474.pdf	Russian	null	ЖАЗОДАН ОЗОД ҚИЛИШ ТУРИ СИФАТИДА ШАРТЛИ ҲУКМ ҚИЛИШ	Zenodo (CERN European Organization for Nuclear Research)	2,022	cc-by	4,705	MAQOLA TARIXI Qabul qilindi: 20-aprel 2022 Ma’qullandi: 25- aprel 2022 Chop etildi: 30- aprel 2022 Ушбу мақолада жазодан озод қилиш институти тушунчаси, келиб чиқиш тарихи, ҳуқуқий аҳамияти, турлари ҳамда қўлланилиш доираси ҳақида фикр юритилади. Бундан ташқари, тадқиқот ишида ривожланган хорижий давлатлар қонунчилигида жазодан озод қилишнинг ўзига хос томонилари ва миллиц қонунчилигимизга имплементация қилиш масалари ёритилган. 1 Сергеевский Н.Д. Русское уголовное право: Пособие к лекциям. – Пг., 1904. – 420-б.; Таганцев Н.С. Русское уголовное право. Часть Общая: Декции. – СПб., 1902. Т.2. – 1392–1404-б.; Познышев С.В. Основные начала науки уголовного права. – М., 1912. – 651-б. 2 Дриль Д.А. Уголовное право. – СПб., 1909. – 549- б.; Фойницкий И.Я. Учение о наказании в связи с тюрьмоведением. – СПб., 1889. – 160-б. жазони ҳам, янги жазони ҳам ўташига тўғри келади. жазони ҳам, янги жазони ҳам ўташига тўғри келади. Шартли ҳукм қилишнинг ҳуқуқий табиати тўғрисидаги масала узоқ йиллардан бери олимлар ўртасида қизғин баҳс-мунозараларга сабаб бўлиб келади. Шартли ҳукм қилишнинг ҳуқуқий табиати хусусидаги у ёки бу нуқтаи назарларнинг вужудга келиш тарихини ўрганиш республикамиз ҳуқуқида тегишли нормалар 1917 йилдан кейин мустаҳкамланган, деган хулосага келиш имконини беради. Мазкур нормани жиноят қонунчилигига киритишнинг зарурлиги ва мақсадга мувофиқлиги тўғрисидаги масала россиялик криминалистлар томонидан ҳам қизғин муҳокама қилинади. Аммо мазкур институтнинг ҳуқуқий табиатини тушунишга нисбатан ягона ёндашув олимлар орасида мавжуд эмас эди. Унга афв этиш каби жазодан озод қилиш сифатида; жиноят содир этган шахснинг айбини енгиллаштирувчи ҳолатлар мавжуд бўлган ҳолда суднинг жазо ҳажмини камайтириш олий ҳуқуқи сифатида; жазоловчи моҳияти жазонинг амалда ижро этилиши хавфи сақланиб қолувчи мустақил, алоҳида (маънавий-ахлоқий хусусиятга эга бўлган) жазо сифатида ёндашишни таклиф қиладилар. ХХ аср бошида Бельгия, Франция, Швейцария ва бошқа бир қанча мамлакатлар қонунчилигида жазони бекор қилиш имконияти назарда тутилади. Бу ҳолда қисқа муддатга озодликдан маҳрум қилиш тариқасидаги жазога ҳукм қилинган шахс қонунда белгиланган муддат мобайнида янги жиноий қилмишлар содир этмаслик шарти билан жазони ўтамайди. Янги жиноят содир этган тақдирда маҳкум илгари тайинланган Айни пайтда Н.Д.Сергеевский, Н.С.Таганцев, С.В.Познишев ва айрим Volume 2 Issue 4, April 2022 ISSN 2181-2020 ISSN 2181-2020 Volume 2 Issue 4, April 2022 Page 474 L OF ACADEMIC RESEARCH my Research Support Center www.in-academy.uz Шартли ҳукм қилишни қисмларга ажратишнинг баҳслилигига келтирилган концепция муаллифининг ўзи эътиборни қаратади ва бу асосдан холи эмас. Бундай ажратиш муайян шартли ҳукм қилиш ва шартли ҳукм қилинган шахснинг хулқ-атвори устидан назорат алоҳида-алоҳида мавжуд бўлиши мумкин, деб тахмин қилиш имконини беради. Бироқ назорат маҳкум ўтиши лозим бўлган синовнинг ажралмас қисми. Ўз навбатида синов шахснинг хулқ-атворига қўйиладиган муайян талабларни ўз ичига олади ва шартли ҳукм қилишни қўлланиш шарти сифатида амал қилади. Бундан ташқари, шахсга жиноят-ҳуқуқий таъсир чораларининг иккита алоҳида тури бир пайтнинг ўзида қўлланиши мумкинлигини тахмин қилиш ҳам тўғри бўлмас эди. 3 Ружников А.Н. Условное осуждение и отсрочка исполнения приговора. – М., 1994. – 32-б. 5 Ўзбекистон Республикаси Олий суди Пленумининг “Судлар томонидан жиноят учун жазо тайинлаш амалиёти тўғрисида” 2006 йил 3 февралдаги 1-сон Қарори. 36-банд // Ўзбекистон Республикаси Олий суди Ахборотномаси 2006 й., 1-сон. 4 Каримов И.А. Мамлакатимизда демократик ислоҳотларни янада чуқурлаштириш ва фуқаролик жамиятини ривожлантириш концепцияси // Халқ сўзи. – Тошкент, 2010. 13 ноябрь. бошқа олимлар шартли ҳукм қилиш институтини жиноят содир этишда айбдор шахсларнинг барчаси қилмишнинг оғирлик даражасига мувофиқ жазога тортилишини талаб қилувчи жазоловчи олий судловнинг асосий вазифасига мутлақо зид1, деб ҳисоблайдилар. Шартли ҳукм қилиш тўғрисидаги нормаларни жиноят қонунига киритиш тарафдорлари уларда қисқа муддатларга озодликдан маҳрум қилишнинг жиноятчига янада кучлироқ ахлоқ тузатувчи таъсир кўрсатишга қодир бўлган муқобилини кўрганлар2. Бироқ уларнинг ғоялари ўша даврдаги қонунчиликда рўёбга чиқарилмаган. Шартли ҳукм қилишнинг ҳуқуқий табиати тўғрисида сўз юритганда, А.Н.Ружниковнинг шартли ҳукм қилишда ўзгача ҳуқуқий табиатга эга бўлган чора – айбдорнинг хулқ- атвори устидан назоратни ажратиш мумкин, деган фикрига алоҳида тўхталиб ўтиш, бизнингча, ўринли бўлади. Муаллиф фикрига кўра, ушбу назорат жиноий жавобгарликни амалга оширишнинг мустақил чораси ёки маҳкумни синашдан иборат бўлган алоҳида жиноят-ҳуқуқий таъсир чораси ҳисобланади, шартли ҳукм қилишнинг ўзи эса жазони ўташдан озод қилишнинг турларидан биридир3. Шундай қилиб, шартли ҳукм қилишнинг ҳуқуқий табиатини аниқлашнинг таклиф қилинган вариантлари таҳлили шартли ҳукм қилиш жазодан озод қилишнинг турларидан биридир, деган хулосага келиш имконини беради. Шу туфайли ҳам шартли ҳукм қилишни қўлланиш ва бекор қилиш асосларини тартибга солувчи моддалар Ўзбекистон Республикаси ЖКнинг “Жазодан озод қилишнинг турлари” деб номланган XIII бобида жамланса, бизнингча, мақсадга мувофиқ бўлар эди. Сўнгги йилларда шартли ҳукмлар чиқариш амалиётида юз берган ўзгаришларни маълум даражада шарти равишда тайинлашга йўл қўйиладиган жазо турлари рўйхати кенгайгани билан изоҳлаш мумкин. Ўзбекистон Республикаси ЖК 72-моддасининг биринчи қисмига мувофиқ, суд ISSN 2181-2020 Volume 2 Issue 4, April 2022 озодликдан маҳрум қилиш, интизомий қисмга жўнатиш, хизмат бўйича чеклаш ёки ахлоқ тузатиш ишлари тариқасидаги жазоларни шартли равишда тайинлаш тўғрисида қарор чиқариши мумкин. 1959 йилги ЖКга мувофиқ фақат ахлоқ тузатиш ишлари ва озодликдан маҳрум қилишга шартли ҳукм қилишга йўл қўйилар эди. Озодликдан маҳрум қилиш тариқасидаги жазони тайинлашда ҳукмни ижро этиш кечиктирилиши мумкин эди. Шартли ҳукм қўлланилганда, бир йилдан уч йилгача бўлган синов муддати белгиланади ва бу муддат ҳукм чиқарилган кундан бошлаб ҳисобланади. қонун синов муддати давомийлилигини тайинланган жазо миқдори билан боғламайди. Бу масала суд томонидан ҳар сафар содир этилган жиноятнинг хусусияти ва судланувчининг шахсидан келиб чиқиб ҳал этилади6. Сўнгги йилларда шартли ҳукм қилинаётган шахслар сонининг кўпайиб бораётганлиги жиноят қонунчилиги сиёсатининг инсонпарварлашуви талабларига тўла жавоб беради. Ушбу сиёсатни амалга ошириш йўналишларидан бири озодликдан маҳрум қилишга муқобил чораларни ёки маҳкумни жамиятдан ажратиш билан боғлиқ бўлмаган чораларни янада кенгроқ қўллаш ҳисобланади. Зотан, ижтимоий хавфи унча катта бўлмаган ишлар бўйича одамларни қамаш “давлатга жуда қимматга тушади, қолаверса, бундай жазо тури маҳкумларни тарбиялаш ва қайта тарбиялаш вазифасини ҳал қилмайди”4. Сўнгги йилларда шартли ҳукм қилинаётган шахслар сонининг кўпайиб бораётганлиги жиноят қонунчилиги сиёсатининг инсонпарварлашуви талабларига тўла жавоб беради. Ушбу сиёсатни амалга ошириш йўналишларидан бири озодликдан маҳрум қилишга муқобил чораларни ёки маҳкумни жамиятдан ажратиш билан боғлиқ бўлмаган чораларни янада кенгроқ қўллаш ҳисобланади. 6 Ўша ерда. – 34-банд. бошқа олимлар шартли ҳукм қилиш институтини жиноят содир этишда айбдор шахсларнинг барчаси қилмишнинг оғирлик даражасига мувофиқ жазога тортилишини талаб қилувчи жазоловчи олий судловнинг асосий вазифасига мутлақо зид1, деб ҳисоблайдилар. Зотан, ижтимоий хавфи унча катта бўлмаган ишлар бўйича одамларни қамаш “давлатга жуда қимматга тушади, қолаверса, бундай жазо тури маҳкумларни тарбиялаш ва қайта тарбиялаш вазифасини ҳал қилмайди”4. Айни пайтда шуни қайд этиш лозимки, жазо шартли тайинланган шахсларни ва унча оғир бўлмаган жиноятни қайта содир этганлик учун ҳукм қилинган шахсларни тавсифловчи кўрсаткичларга Ўзбекистон Республикаси ЖК 72-моддаси саккизинчи қисмининг шартли ҳукм қилинган шахс синов муддати давомида Шартли ҳукм қилинган шахсларнинг хулқ-атвори устидан назорат ички ишлар органлари, ҳарбий хизматчиларга нисбатан эса – улар хизматни ўтаётган жойдаги ҳарбий қисмлар (муассасалар) қўмондонлиги томонидан амалга оширилади (Ўзбекистон Республикаси ЖИК 15- моддасининг биринчи қисми). Шартли ҳукм қилинган шахсларнинг ҳисобини олиб бориш, уларнинг ҳулқ-атвори устидан назорат олиб боришни ички ишлар идоралари ҲООБ жазони ижро этиш инспекциялари ташкил этади. 6 Ўша ерда. – 34-банд. Volume 2 Issue 4, April 2022 ISSN 2181-2020 Page 476 Бундай назорат шартли ҳукм қилинганларнинг хулқ-атворини кузатиш, уларга ишга жойлашиш ва турар жой билан таъминлашда ёрдам кўрсатиш, улар қонун талабларига риоя қилишлари ва суд уларга юклаган вазифаларни бажаришларини текширишдан иборат. Унинг вазифаси – шартли ҳукм қилинганлар синов муддати мобайнида ҳуқуқий кўрсатмаларни бузишларининг олдини олиш. инспекциялари томонидан амалга оширилади. Бунда жазони ижро этиш инспекцияси томонидан:  маҳкумлар билан суҳбат ўтказилиб, жазо ўташ шарт-шароитлари тушунтирилади;  ишлаш ва ўқиш жойлари маъмуриятларидан маҳкумларга нисбатан тавсияномалар сўралади, керак бўлган пайтда маъмурият аъзолари билан учрашув ўтказилади;  маҳкумлар томонидан жамоат тартибини бузиш ҳоллари содир этилганлигини аниєлаш мақсадида ҳар ойда ички ишлар бўлимининг маълумот картотекаси билан танишилади; Шартли ҳукм қилинганлар устидан назоратни амалга ошириш тартиби Ўзбекистон Республикаси ЖИК 32-бобида белгиланган. Бундан ташқари, шартли ҳукм қилинганларни назорат қилиш масалалари Ўзбекистон Республикаси Ички ишлар вазирлигининг 2001 йил 298-сонли буйруғи билан тасдиқланган “Ахлоқ тузатиш ишлари, муайян ҳуқуқдан маҳрум қилиш тариқасидаги жазоларни ижро этиш ва шартли ҳукм қилинганларни назорат қилиш тартибига оид йўриқнома”да атрофлича тартибга солинган. Хусусан, ушбу норматив-ҳуқуқий ҳужжатда шартли ҳукм қилиш тўғрисидаги суд ҳукмларини ижрога қабул қилишни расмийлаштириш, маҳкумлар ахлоқи устидан назорат қилиш, шартли ҳукмни бекор қилиш ҳамда ҳукмда тайинланган жазони ижро этиш тўғрисида судга тақдимнома киритиш, шартли ҳукм қилинганларнинг қидируви, шартли ҳукм қилинганларни ҳисобдан чиқариш тартиби белгиланган.  суд томонидан белгиланган мажбуриятларнинг бажарилиши (ҳар чоракда камида бир маротаба) текширилади, уларни бажармай юрган шахсларга нисбатан тегишли чоралар кўрилади. Натижаси ҳақида маълумотнома тузилиб, назорат йиғма жилдига қўшиб борилади;  суд томонидан рўйхатдан ўтиб туриш мажбурияти юклатилган маҳкумларни рўйхатдан ўтказиш амалга оширилади. 8 Усмоналиев М. Жиноят ҳуқуқи. Умумий қисм. Олий ўқув юртлари учун дарслик. – Т.: “Янги аср авлоди”, 2005. – 600-б. 7 Ўзбекистон Республикаси Олий суди Пленумининг “Судлар томонидан жиноят учун жазо тайинлаш амалиёти тўғрисида” 2006 йил 3 февралдаги 1-сон Қарори. 35-банд // Ўзбекистон Республикаси Олий суди Ахборотномаси 2006 й., 1-сон. бошқа олимлар шартли ҳукм қилиш институтини жиноят содир этишда айбдор шахсларнинг барчаси қилмишнинг оғирлик даражасига мувофиқ жазога тортилишини талаб қилувчи жазоловчи олий судловнинг асосий вазифасига мутлақо зид1, деб ҳисоблайдилар. Жазони ижро этиш инспекцияси маҳкум томонидан қуйидаги ҳуқуқбузарликлар содир этилганда, судга тақдимнома киритади: а) синов муддати давомида жамоат тартибини ёки меҳнат интизомини бузганлиги учун маъмурий ёки интизомий жазо қўлланган бўлса; б) шартли ҳукм қилинган муайян муддат мобайнида ишга ёки ўқишга кирмаган, етказилган зарарни бартараф этмаган, ички ишлар органининг рухсатисиз турар жойини ўзгартирган, рўйхатдан ўтиш учун келмаётган ҳамда ҳукмда кўрсатилган Юқорида айтиб ўтилганидек, синов муддати мобайнида шартли ҳукм қилинган шахсларнинг хулқ-атвори устидан назорат ички ишлар идоралари ҲООБ жазони ижро этиш ISSN 2181-2020 Page 477 бошқа мажбуриятларни бажармаётган бўлса; Бундай ҳуқуқбузарликлар орасида майда безорилик, жамоат тартибига таҳдид соладиган материалларни тайёрлаш ёки тарқатиш мақсадида сақлаш, ўқотар қуроллардан белгиланган тартибни бузган ҳолда отиш, хонаки усулда ҳозирланадиган ўткир спиртли ичимликларни сотиш мақсадида тайёрлаш ёки сотиш, жамоат жойларида спиртли ичимликлар ичиш, вояга етмаган шахсни ғайриижтимоий хатти-ҳаракатларга жалб қилиш, порнографик мазмундаги материалларни тарқатиш, фоҳишалик билан шуғулланиш, қимор ва таваккалчиликка асосланган бошқа ўйинлар, маиший шовқинга қарши кураш талабларини бузиш бор. Суднинг шартли ҳукм қилиш тўғрисидаги қарори ўз кучида қолиши учун маҳкум белгиланган синов бажариши, технология интизомига, меҳнат муҳофазаси, техника хавфсизлиги ва ишлаб чиқариш санитарияси талабларига риоя қилиши, иш берувчининг мол-мулкини авайлаб асраши шарт. Ходимнинг меҳнат вазифалари ички тартиб қоидаларида, интизом тўғрисидаги устав ва низомларда, корхонада қабул қилинадиган локал ҳужжатларда (жамоа шартномаларида, йўриқномалар ва ҳоказоларда), меҳнат шартномасида аниқ белгилаб қўйилади. Ходимга меҳнат интизомини бузганлиги учун иш берувчи қуйидаги интизомий жазо чораларини қўллашга ҳақли: 1) ҳайфсан; 2) ўртача ойлик иш ҳақининг ўттиз фоизидан ортиқ бўлмаган миқдорда жарима. Ички меҳнат тартиби қоидаларида ходимга ўртача ойлик иш ҳақининг эллик фоизидан ортиқ бўлмаган миқдорда жарима солиш ҳоллари ҳам назарда тутилиши мумкин; 3) меҳнат шартномасини бекор қилиш. Ўзбекистон Республикаси Меҳнат кодексида назарда тутилмаган интизомий жазо чораларини қўлланиш тақиқланади (Ўзбекистон Республикаси Меҳнат кодексининг 181-моддаси). Суд шартли ҳукм қилиш тўғрисида қарор чиқарар экан, маҳкумга бажармаслик шартли ҳукмнинг бекор қилинишига олиб келиши мумкин бўлган мажбуриятлар EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center UIF = 8 1 \| SJIF = 5 685 www in academy uz белгиланган барча ҳуқуқлардан фойдаланадилар. бошқа мажбуриятларни бажармаётган бўлса; мобайнида етказилган зарарни бартараф қилиш; ишга ёки ўқишга кириш; яшаш жойи, иш ёки ўқиш жойи ўзгариб қолса, бу ҳақда шартли ҳукм қилинган шахснинг хулқи устидан назорат олиб борувчи органга хабар бериб туриш; вақти-вақти билан келиб бу органларда рўйхатдан ўтиб туриш; муайян жойларда бўлмаслик, муайян вақтда яшаш жойида бўлиш; алкоголизм, гиёҳвандлик, заҳарвандлик ёки таносил касаллигидан даволаниш курсини ўташ ва ҳ.к. в) шартли ҳукм қилинган шахснинг назоратдан бўйин товлаш мақсадида яширинганига етарли асос бўлса. Ўзбекистон Республикаси ЖК 72-моддасининг олтинчи қисмига мувофиқ, агар шартли ҳукм қилинган шахс синов муддати давомида суд томонидан унга юклатилган мажбуриятларини бажармаса ёхуд жамоат тартиби ёки меҳнат интизомини бузганлиги учун унга маъмурий ёки интизомий таъсир чораси қўлланилган бўлса, суд бундай шахснинг хулқи устидан назорат олиб борувчи орган тақдимномасига биноан жазонинг шартлилигини бекор қилиб, ҳукмда белгиланган жазони ижро этиш тўғрисида қарор қабул қилишга ҳақли. Шу билан бирга, содир этилган ҳуқуқбузарликнинг хусусияти ва муайян ҳолатларини, маҳкумнинг шахси ва синов муддати давомидаги хулқ- атворига доир маълумотларни ҳисобга олиб, суд шартли ҳукмни аслича қолдириш тўғрисида қарор қабул қилиши мумкин7. Шартли ҳукм қилинган шахсга нисбатан юклатилган мажбуриятларнинг аҳамияти шундаки, синов муддати тугаганидан кейин шахсга нисбатан тайинланган жазони ижро этиш ёки шахсни жазодан озод қилиш масаласини ҳал қилишда айнан ана шу унга юклатилган мажбуриятларни қай даражада бажарган ёки бажармаганлиги асос бўлади. Демак, шахсга нисбатан ЖК 72- моддасининг 1-қисмида кўрсатилган жазоларни тайинлаб, сўнг шартли ҳукм қилишнинг ўзи ҳали шахсни жазодан озод қилиш дегани эмас. Суд шахсни шартли ҳукм қилиб, синов муддатини белгилайди ва муайян шартларни шахс зиммасига юклайди8. Шартли ҳукмни бекор қилиш тўғрисидаги тақдимнома етказилган зарарни бартараф қилиш учун суд томонидан белгиланган муддат тамом бўлиши билан ўн кун ичида юборилади. Шартли ҳукм қилинган шахсларнинг устидан назорат қонуний кучга кирган суд ҳукмига мувофиқ амалга оширилади. Шартли ҳукм қилинганлар Ўзбекистон Республикаси қонунларида ва суд ҳукмида кўзда тутилган чеклаш ва мажбуриятлардан ташқари Ўзбекистон Республикаси фуқаролари учун қонунларда Суд шартли ҳукм қилинган шахсга, башарти, бунга асослар мавжуд бўлса, қуйидаги мажбуриятларни юклаши мумкин: муайян вақт Volume 2 Issue 4, April 2022 ISSN 2181-2020 Page 478 EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center UIF = 8.1 \| SJIF = 5.685 www.in-academy.uz белгиланган барча ҳуқуқлардан фойдаланадилар. Жамоат тартибини бузиш Маъмурий жавобгарлик тўғрисидаги кодексда назарда тутилган ҳуқуқбузарликлар орасида энг оғири ҳисобланади. Шу туфайли ҳам улар содир этилиши шартли ҳукм бекор қилинишига олиб келиши мумкин бўлган ҳуқуқбузарликлар сифатида танланган. Жамоат тартибига тажовуз қилувчи ҳуқуқбузарликлар Ўзбекистон Республикаси Маъмурий жавобгарлик тўғрисидаги кодексининг XV бобида жамланган. белгиланган барча ҳуқуқлардан фойдаланадилар. Ўзбекистон Республикаси ЖК 72-моддасининг бешинчи қисмига мувофиқ, ҳукм қилинган шахснинг хулқи устидан назорат олиб борувчи органнинг тақдими билан суд синов муддати давомида унга юклатилган мажбуриятларнинг ҳаммасини ёки бир қисмини олиб ташлаши ёки унинг зиммасига янги мажбуриятлар юклаши ҳам мумкин. Шартли ҳукм қилинган шахс яшаш жойини ташлаб кетган бўлса, жазони ижро этиш инспекцияси ходимлари томонидан унинг қаердалигини аниқлаш чоралари кўрилади. Бу мақсадда: а) маҳкумнинг қариндош ва танишлари унинг қаерда бўлиши мумкинлигини аниқлаш мақсадида суриштирилади; б) манзиллар маълумотхонасидан ва бошқа тегишли давлат органларидан ҳамда уй жойларида жойлашган ҳудуддаги ички ишлар бўлимига сўровлар юборилади; д) тиббий муассасаларда маҳкумлар тўғрисида бирор маълумот бор-йўқлиги текширилади; е) ички ишлар бўлимларида ушланган, ҳибсга олинганлар, судланганлар ҳамда қидирувда бўлганлар рўйхати кўриб чиқилади; м) маҳкумларни аниқлаш мақсадида кўзда тутилган бошқа чоралар ҳам қўлланилади; з) тўпланган ҳужжатлар алоҳида йиғма жилдда тахланиб, назорат йиғма жилдига қўшилади; и) агарда бир ой муддатда қидирилаётган шахсни аниқлашнинг иложи бўлмаса, жазони ижро этиш инспекциясида тўпланган ҳужжатлар билан хулосани ва суд ҳукмининг нусхасини унга нисбатан қидирув эълон қилишни сўраб маҳкумнинг турар жойи бўйича ички ишлар бўлими бошлиғига топширади. Маҳкумнинг назоратдан бўйин товлаш мақсадида яширингани учун етарли асос бўлса, шартли ҳукмни бекор қилиш тўғрисида судга тақдимнома киритилади. Шартли ҳукм ўта оғир жинояти учун ҳукм қилинганларга, шунингдек илгари қасддан содир этган жинояти учун озодликдан маҳрум этиш жазосига ҳукм қилинган шахсларга нисбатан белгиланган барча ҳуқуқлардан фойдаланадилар. ф р Жамоат тартибини бузиш Маъмурий жавобгарлик тўғрисидаги кодексда назарда тутилган ҳуқуқбузарликлар орасида энг оғири ҳисобланади. Шу туфайли ҳам улар содир этилиши шартли ҳукм бекор қилинишига олиб келиши мумкин бўлган ҳуқуқбузарликлар сифатида танланган. Жамоат тартибига тажовуз қилувчи ҳуқуқбузарликлар Ўзбекистон Республикаси Маъмурий жавобгарлик тўғрисидаги кодексининг XV бобида жамланган. Бундай ҳуқуқбузарликлар орасида майда безорилик, жамоат тартибига таҳдид соладиган материалларни тайёрлаш ёки тарқатиш мақсадида сақлаш, ўқотар қуроллардан белгиланган тартибни бузган ҳолда отиш, хонаки усулда ҳозирланадиган ўткир спиртли ичимликларни сотиш мақсадида тайёрлаш ёки сотиш, жамоат жойларида спиртли ичимликлар ичиш, вояга етмаган шахсни ғайриижтимоий хатти-ҳаракатларга жалб қилиш, порнографик мазмундаги материалларни тарқатиш, фоҳишалик билан шуғулланиш, қимор ва таваккалчиликка асосланган бошқа ўйинлар, маиший шовқинга қарши кураш талабларини бузиш бор. Ходимга меҳнат интизомини бузганлиги учун иш берувчи қуйидаги интизомий жазо чораларини қўллашга ҳақли: 1) ҳайфсан; 2) ўртача ойлик иш ҳақининг ўттиз фоизидан ортиқ бўлмаган миқдорда жарима. Ички меҳнат тартиби қоидаларида ходимга ўртача ойлик иш ҳақининг эллик фоизидан ортиқ бўлмаган миқдорда жарима солиш ҳоллари ҳам назарда тутилиши мумкин; 3) меҳнат шартномасини бекор қилиш. Ўзбекистон Республикаси Меҳнат кодексида назарда тутилмаган интизомий жазо чораларини қўлланиш тақиқланади (Ўзбекистон Республикаси Меҳнат кодексининг 181-моддаси). ҳ д дд ) Суд шартли ҳукм қилиш тўғрисида қарор чиқарар экан, маҳкумга бажармаслик шартли ҳукмнинг бекор қилинишига олиб келиши мумкин бўлган мажбуриятлар юклаши мумкин. Бироқ айни пайтда суд маҳкумга бирон-бир мажбурият юкламаслиги ҳам мумкин. Бу ҳолда шартли ҳукм қилинган шахснинг хулқ- атвори устидан назоратни амалга ошириш шартли ҳукм қилинган шахснинг жамоат тартиби ва меҳнат интизомига риоя қилишидан ҳамда Суднинг шартли ҳукм қилиш тўғрисидаги қарори ўз кучида қолиши учун маҳкум белгиланган синов муддати мобайнида меҳнат интизомига ҳам риоя қилиши шарт. Ўзбекистон Республикаси Меҳнат кодексининг 176- моддасига мувофиқ ходим ўз меҳнат вазифаларини ҳалол, виждонан бажариши, меҳнат интизомига риоя қилиши, иш берувчининг қонуний фармойишларини ўз вақтида ва аниқ Volume 2 Issue 4, April 2022 ISSN 2181-2020 ISSN 2181-2020 Volume 2 Issue 4, April 2022 Page 479 EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center UIF = 8.1 \| SJIF = 5.685 www.in-academy.uz жазони ижро этиш инспекцияси ходимлари билан суҳбатлашишдан иборат бўлади. Бунда мазкур шахсга шартли ҳукмни бекор қилиш учун асос сифатида, у янги жиноят содир этган ҳоллардан ташқари, жамоат тартиби ёки меҳнат интизомини бузганлиги учун унга маъмурий ёки интизомий таъсир чораси қўлланилгани амал қилади. 10 Золотарев В. Условное осуждение с отбыванием в колонии // Рос. юстиция. 1999. №3. – 38-б. 9 Ўзбекистон Республикаси Олий суди Пленумининг “Судлар томонидан жиноят учун жазо тайинлаш амалиёти тўғрисида” 2006 йил 3 февралдаги 1-сон Қарори. 36-банд // Ўзбекистон Республикаси Олий суди Ахборотномаси 2006 й., 1-сон. www.in-academy.uz жазони ижро этиш инспекцияси ходимлари билан суҳбатлашишдан иборат бўлади. Бунда мазкур шахсга шартли ҳукмни бекор қилиш учун асос сифатида, у янги жиноят содир этган ҳоллардан ташқари, жамоат тартиби ёки меҳнат интизомини бузганлиги учун унга маъмурий ёки интизомий таъсир чораси қўлланилгани амал қилади. д) тиббий муассасаларда маҳкумлар тўғрисида бирор маълумот бор-йўқлиги текширилади; е) ички ишлар бўлимларида ушланган, ҳибсга олинганлар, судланганлар ҳамда қидирувда бўлганлар рўйхати кўриб чиқилади; м) маҳкумларни аниқлаш мақсадида кўзда тутилган бошқа чоралар ҳам қўлланилади; Ўзбекистон Республикаси ЖК 72-моддасининг бешинчи қисмига мувофиқ, ҳукм қилинган шахснинг хулқи устидан назорат олиб борувчи органнинг тақдими билан суд синов муддати давомида унга юклатилган мажбуриятларнинг ҳаммасини ёки бир қисмини олиб ташлаши ёки унинг зиммасига янги мажбуриятлар юклаши ҳам мумкин. з) тўпланган ҳужжатлар алоҳида йиғма жилдда тахланиб, назорат йиғма жилдига қўшилади; и) агарда бир ой муддатда қидирилаётган шахсни аниқлашнинг иложи бўлмаса, жазони ижро этиш инспекциясида тўпланган ҳужжатлар билан хулосани ва суд ҳукмининг нусхасини унга нисбатан қидирув эълон қилишни сўраб маҳкумнинг турар жойи бўйича ички ишлар бўлими бошлиғига топширади. Маҳкумнинг назоратдан бўйин товлаш мақсадида яширингани учун етарли асос бўлса, шартли ҳукмни бекор қилиш тўғрисида судга тақдимнома киритилади. Шартли ҳукм қилинган шахс яшаш жойини ташлаб кетган бўлса, жазони ижро этиш инспекцияси ходимлари томонидан унинг қаердалигини аниқлаш чоралари кўрилади. Бу мақсадда: а) маҳкумнинг қариндош ва танишлари унинг қаерда бўлиши мумкинлигини аниқлаш мақсадида суриштирилади; Шартли ҳукм ўта оғир жинояти учун ҳукм қилинганларга, шунингдек илгари қасддан содир этган жинояти учун озодликдан маҳрум этиш жазосига ҳукм қилинган шахсларга нисбатан қўлланилмайди, ўн саккиз ёшга тўлмаган шахслар, биринчи ва иккинчи гуруҳ ногиронлари, аёллар, шунингдек олтмиш ёшдан ошган шахслар бундан мустасно (Ўзбекистон Республикаси ЖК 72-моддасининг еттинчи қисми). б) манзиллар маълумотхонасидан ва бошқа тегишли давлат органларидан ҳамда уй дафтаридан маҳкумнинг турар жойидаги рўйхатдан чиқмагани аниқланади; в) ўқиш ёки иш жойидаги кадрлар бўлимидан маҳкумнинг қаерда бўлиши мумкинлиги тўғрисидаги маълумотларга эга бўлган шахслар орқали аниқлаш чоралари кўрилади; Айни пайтда, шуни назарда тутиш керакки, Ўзбекистон Республикаси ЖКнинг 72-моддасида шартли ҳукмни бекор қилиш учун асос г) яширинган шахсни аниқлаш мақсадида унинг қариндошлари турар Volume 2 Issue 4, April 2022 ISSN 2181-2020 ISSN 2181-2020 Page 480 Foydalanilgan adabiyotlar: 1.1. Мирзиёев Ш.М. Эркин ва фаровон, демократик Ўзбекистон давлатини мард ва олижаноб халқимиз билан бирга қурамиз// Халқ сўзи.-2016.-15 дек. 1.1. Мирзиёев Ш.М. Эркин ва фаровон, демократик Ўзбекистон давлатини мард ва олижаноб халқимиз билан бирга қурамиз// Халқ сўзи.-2016.-15 дек. 1.2. Мирзиёев Ш.М. Буюк келажагимизни мард ва олижаноб халқимиз билан бирга қурамиз. Тoшкeнт: Ўзбeкиcтaн, 2017. 1.2. Мирзиёев Ш.М. Буюк келажагимизни мард ва олижаноб халқимиз билан бирга қурамиз. Тoшкeнт: Ўзбeкиcтaн, 2017. EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center UIF = 8.1 \| SJIF = 5.685 www.in-academy.uz Volume 2 Issue 4, April 2022 ISSN 2181-2020 Page 481 EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center UIF = 8.1 \| SJIF = 5.685 www.in-academy.uz адолат принципини бузади, балки уларнинг қилмишларидан жабрланганларнинг ҳуқуқларини камситади. Жиноятчиларга инсонийлик кўрсатиш оқилона чегарада бўлиши лозим. Бироқ Ўзбекистон Республикаси ЖПК суд қарорини назорат тартибида қайта кўришда нафақат ишда йб ё балки маҳкумнинг шахси тўғрисидаги маълумотларни янги очилган ҳолатлар ёки янги ҳолатлар қаторига киритмайди. Бизнингча, жиноят- процессуал қонунчиликдаги бу камчилик ЖПКга тегишли қўшимча киритиш йўли билан бартараф этилиши лозим. адолат принципини бузади, балки уларнинг қилмишларидан жабрланганларнинг ҳуқуқларини камситади. Жиноятчиларга инсонийлик кўрсатиш оқилона чегарада бўлиши лозим. Бироқ Ўзбекистон Республикаси ЖПК суд қарорини назорат тартибида қайта кўришда нафақат ишда айбдорнинг аҳволини ёмонлаштирувчи томонга бурилиш ясашни тақиқлайди, EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center UIF = 8.1 \| SJIF = 5.685 www.in-academy.uz озодликдан шартли маҳрум қилишни ҳам ўтайди10. бўладиган ҳолатлар доираси қатъий белгиланган. Шу сабабли, агар шартли ҳукм қилинган шахснинг ўша иш бўйича ҳукм чиқарилгунга қадар содир этилган бошқа жиноятда ҳам айбдорлиги аниқланса, ушбу ҳолат жазонинг шартлилигини бекор қилиш учун асос бўла олмайди. Бундай ҳолда Ўзбекистон Республикаси ЖКнинг 59-моддаси (“Бир неча жиноят содир этганлик учун жазо тайинлаш”) қоидалари қўлланилмайди (чунки шартли ҳукм юридик табиатига кўра жиноий жазодан озод қилишнинг бир тури ҳисобланади) ва биринчи ҳамда иккинчи иш бўйича чиқарилган ҳар қайси ҳукм алоҳида ижро қилинади9. Ўз-ўзидан равшанки, бундай ҳолларда шартли ҳукм қилишнинг моҳиятига путур етади ва одиллик принципи бузилади. Шартли ҳукм қилишда шахсга жазони ижро этиш инспекцияси ходимлари томонидан амалга ошириладиган ахлоқ тузатиш тусидаги айрим чоралар ёрдамида мустақил равишда ахлоқан тузалиш имконияти берилади. Бундан фарқли ўлароқ озодликдан маҳрум қилиш жазосини амалда ўташда маҳкумга уни жамиятдан ажратиш шароитларида жазони ижро этиш муассасаси маъмурияти томонидан ахлоқ тузатиш ва тарбиявий таъсир сурункали кўрсатилади. Бундан ташқари, жазони ижро этиш муассасалари шартли ҳукм қилинганлар устидан назоратни амалга ошириш ваколатига эга эмас. Бу функция ички ишлар органларининг жазони ижро этиш инспекциялари зиммасига юкланган. Бинобарин, улар ушбу функцияни озодликдан маҳрум қилиш жойларидаги шахсларга нисбатан бажармайдилар. Натижада мазкур шахс жазо шартли тайинланган жиноят учун амалда оқланади, чунки шартли ҳукм қилиш билан боғлиқ бўлган салбий оқибатлар озодликдан маҳрум қилишга хос бўлган инсон ҳуқуқларининг янада жиддийроқ чекланиши билан қамраб олинади. Ўзбекистон Республикаси Олий суди Пленуми жиноятлар қандай кетма- кетликда содир этилишига эътиборни қаратмайди. Шартли жазо тайинланган жиноят вақтга кўра бошқа жиноятдан олдин содир этилган бўлиши ҳам, кейин содир этилган бўлиши ҳам мумкин. Жиноятлардан бири бошқа жиноят учун шартли ҳукм қилиш тўғрисидаги қарор чиқарилгунга қадар содир этилгани ҳал қилувчи аҳамиятга эга бўлади. Олий суд Пленуми қарорининг юқорида зикр этилган қоидасига асосланган ҳолда суд реал жазо тайинлаш ва шартли жазо тайинлаш тўғрисидаги ҳукмларни алоҳида- алоҳида ижро этиш ҳақида қарор қабул қилади. Бунинг натижасида маҳкум, масалан, озодликдан маҳрум қилиш жазосини жазони ижро этиш муассасасида ўтиш билан бир вақтда Бизнингча, бундай ҳолларда шартли ҳукм қилиш тўғрисидаги қарор процессуал тартибда бекор қилинса, мақсадга мувофиқ бўлади. Жиноятлар содир этган шахсларга ҳаддан ташқари инсоний ёндашув нафақат ижтимоий 9 Ўзбекистон Республикаси Олий суди Пленумининг “Судлар томонидан жиноят учун жазо тайинлаш амалиёти тўғрисида” 2006 йил 3 февралдаги 1-сон Қарори. 36-банд // Ўзбекистон Республикаси Олий суди Ахборотномаси 2006 й., 1-сон. EURASIAN JOURNAL OF ACADEMIC RESEARCH 11. Ўзбекистон Республикаси Президенти Ш.М.Мирзиёевнинг 2018-йил 14-майда “Жиноят ва жиноят-процессуал қонунчилиги тизимини тубдан такомиллаштириш чора-тадбирлари тўғрисида”ги 3723-сонли қарори. 11. Ўзбекистон Республикаси Президенти Ш.М.Мирзиёевнинг 2018-йил 14-майда “Жиноят ва жиноят-процессуал қонунчилиги тизимини тубдан такомиллаштириш чора-тадбирлари тўғрисида”ги 3723-сонли қарори. 12. Ўзбекистон Республикаси Президентининг 2017-йил 7-февралдаги “Ўзбекистон Республикасини янада ривожлантириш бўйича харакатлар стратегияси тўғрисида”ги ПФ-4947-сонли Фармони. 12. Ўзбекистон Республикаси Президентининг 2017-йил 7-февралдаги “Ўзбекистон Республикасини янада ривожлантириш бўйича харакатлар стратегияси тўғрисида”ги ПФ-4947-сонли Фармони. 13. СССР Олий Суди Пленумининг 1961-йил 4-мартдаги “Шартли ҳукмни қўллаш бўйича суд амалиёти тўғрисида”ги қарорининг 2-банди. 14. Ўзбекистон Республикаси ИИВ нинг 2013-йил 29-июлдаги Ўзбекистон Республикаси ИИВ озодликдан маҳрум қилиш туридаги жазони ижро этиш муассасаларининг ички тартиб қоидаларини тасдиқлаш тўгрисидаги 2495-сонли Буйруғи. 15. “Strafgesetzbuch in der Fassung der Bekantnachtung vom 13. November 1998 (BGBI. I S. 3322), das zuletzt durch Artikel 1 des Gesetzes vom 3. Marz 2020 (BGBI. I S. 431) geandert worden ist). (56. Artikel). 15. “Strafgesetzbuch in der Fassung der Bekantnachtung vom 13. November 1998 (BGBI. I S. 3322), das zuletzt durch Artikel 1 des Gesetzes vom 3. Marz 2020 (BGBI. I S. 431) geandert worden ist). (56. Artikel). 16. Ўзбекистон Республикаси ИИВ нинг “Муайян ҳуқуқдан маҳрум қилиш, ахлоқ тузатиш ишлари ва озодликдан чеклаш тариқасидаги жазоларнинг ижросини ташкил этиш ҳамда шартли ҳукм қилинган шахслар устидан назоратни амалга ошириш тартиби тўғрисида”ги 2017-йил 27-июлдаги 157-сон йўриқномани тасдиқлаш ҳақидаги 2017-йил 28-августдаги 2922-сонли Буйруғи. 16. Ўзбекистон Республикаси ИИВ нинг “Муайян ҳуқуқдан маҳрум қилиш, ахлоқ тузатиш ишлари ва озодликдан чеклаш тариқасидаги жазоларнинг ижросини ташкил этиш ҳамда шартли ҳукм қилинган шахслар устидан назоратни амалга ошириш тартиби тўғрисида”ги 2017-йил 27-июлдаги 157-сон йўриқномани тасдиқлаш ҳақидаги 2017-йил 28-августдаги 2922-сонли Буйруғи. Норматив-ҳуқуқий ҳужжатлар: 1. Ўзбекистон Республикасининг Конститутсияси. – Т.: “Ўзбекистон”, 2017-йил. збекистон Республикасининг Конститутсияси. – Т.: “Ўзбекистон”, 2017-йил. у р 3. Ўзбекистон Республикасининг Жиноят-процессуал кодекси. Расмий нашр. – Т.:“Адолат”, 2019. 3. Ўзбекистон Республикасининг Жиноят-процессуал кодекси. Расмий нашр. – Т.:“Адолат”, 2019. 4. Ўзбекистон Республикасининг Жиноят-ижроия кодекси. T.: “Адолат” 2017-йил. 5. Ўзбекистон Республикаси Олий Суди Пленуми томонидан 2006-йил 3-февралда қабул қилинган “Судлар томонидан жиноят учун жазо тайинлаш амалиёти тўғрисида”ги 1-сонли қарори. 6. Ўзбекистон Республикаси Олий Суди Пленумининг 2000 – йил 15-сентябрдаги Вояга етмаганларнинг жиноятлари ҳақидаги ишлар бўйича суд амалиёти тўғрисидаги 21- сонли Қарори. 6. Ўзбекистон Республикаси Олий Суди Пленумининг 2000 – йил 15-сентябрдаги Вояга етмаганларнинг жиноятлари ҳақидаги ишлар бўйича суд амалиёти тўғрисидаги 21- сонли Қарори. 7. Ўзбекистон Республикаси Олий Суди Пленумининг 2007-йил 14-ноябрдаги “Судлар томонидан амнистия актларини қўллашнинг айрим масалалари тўғрисида”ги 17-сонли Қарори. 7. Ўзбекистон Республикаси Олий Суди Пленумининг 2007-йил 14-ноябрдаги “Судлар томонидан амнистия актларини қўллашнинг айрим масалалари тўғрисида”ги 17-сонли Қарори. 8. Ўзбекистон Республикаси Олий Суди Пленумининг, 2004-йил 23-май “Суд ҳукми тўғрисида”ги 07-сонли Қарори. 8. Ўзбекистон Республикаси Олий Суди Пленумининг, 2004-йил 23-май “Суд ҳукми тўғрисида”ги 07-сонли Қарори. 8. Ўзбекистон Республикаси Олий Суди Пленумининг, 2004-йил 23-май “Суд ҳукми тўғрисида”ги 07-сонли Қарори. 9. Ўзбекистон Республикаси Олий Суди Пленумининг, 2009-йил 10-апрелдаги “Суд ҳужжатларини бажаришдан бўйин товлаш ва уларнинг ижро этилишига тўсқинлик қилиш учун жиноий жавобгарликка доир қонунларни қўллашнинг айрим масалалари тўғрисида”ги 7-сонли Қарори. 9. Ўзбекистон Республикаси Олий Суди Пленумининг, 2009-йил 10-апрелдаги “Суд ҳужжатларини бажаришдан бўйин товлаш ва уларнинг ижро этилишига тўсқинлик қилиш учун жиноий жавобгарликка доир қонунларни қўллашнинг айрим масалалари тўғрисида”ги 7-сонли Қарори. 10. Ўзбекистон Республикаси Олий суди Пленумининг 2007 йил 14 ноябрдаги 15-сонли “Умрбод озодликдан маҳрум қилиш тариқасидаги жазони қўллашнинг айрим масалалари тўғрисида”ги қарори 8-бандининг иккинчи ва учинчи хатбошилари. 10. Ўзбекистон Республикаси Олий суди Пленумининг 2007 йил 14 ноябрдаги 15-сонли “Умрбод озодликдан маҳрум қилиш тариқасидаги жазони қўллашнинг айрим масалалари тўғрисида”ги қарори 8-бандининг иккинчи ва учинчи хатбошилари. Volume 2 Issue 4, April 2022 ISSN 2181-2020 Page 482 Volume 2 Issue 4, April 2022 ISSN 2181-2020 Page 482 Дарслик ва ўқув қўлланмалар: 1. М.Х.Рустамбаев. Ўзбекистон Республикаси Жиноят ҳуқуқи курси. Том 2. “Жазо ҳақида таълимот”. Дарслик. 2-нашр, тўлдирилган ва қайта ишланган. – Т.: “Комплех Принт”. Ўзбекистон Республикаси Миллий гвардияси Ҳарбий-техника институти, 2018- йил 205-бет. 1. М.Х.Рустамбаев. Ўзбекистон Республикаси Жиноят ҳуқуқи курси. Том 2. “Жазо ҳақида таълимот”. Дарслик. 2-нашр, тўлдирилган ва қайта ишланган. – Т.: “Комплех Принт”. Ўзбекистон Республикаси Миллий гвардияси Ҳарбий-техника институти, 2018- йил 205-бет. 2. Рустамбоев М.Ҳ. Ўзбекистон Республикаси Жиноят кодексига шарҳлар. Умумий қисм. – Т.: “Адолат”, 2016. 2. Рустамбоев М.Ҳ. Ўзбекистон Республикаси Жиноят кодексига шарҳлар. Умумий қисм. – Т.: “Адолат”, 2016. 3. Рустамбоев М.Ҳ. Ўзбекистон Республикаси Жиноят ҳуқуқи курси. “Жазо тўғрисида таълимот”. Умумий қисм. – Т.: “ИЛМ ЗИЁ”, 2011-й. 4. Тургунбоев Э.О. Жиноят ҳуқуқи. (Умумий қисм) (Албомлар схемаси) – Т.: “Мунис десигн гроуп” 2013-йил. 99-бет. 5. Ўзбекистон Республикаси Жиноят кодекси, илмий-амалий шарҳлар Ўзбекистон Республикаси ИИВ Академияси. – Т.: “Адолат”. 1997-йил. 6. Жиноят ҳуқуқи (дарслик). Ўзбекистон Республикаси ИИВ Академияси, Т.: 2012-йил. 295-б. 7. А.С.Якубов, Р.Кабулов. Ўзбекистон Республикасининг Жиноят ҳуқуқи (чизмалар албоми). Ўзбекистон Республикаси ИИВ Академияси, Т.: 2014-йил. 123-124-бетлар. 8. Jung, Heike, Heribert Ostendorf Wolfgang Schild u. a.: STGB: Kommentar zum Strafgesetztbuch: BAND 3: && 80-145 d. Reihe Alternativkommentar. 1983 j. Комиссаров. В. С. “Уголовный право” РФ, - Москва, 2012 года. – С. 58. 10. Уголовный кодекс РФ с комментариями и изменениями. (Обшая часть УК РФ) - Москва, 2019-2020 года. Volume 2 Issue 4, April 2022 ISSN 2181-2020 Volume 2 Issue 4, April 2022 Page 483 ISSN 2181-2020 EURASIAN JOURNAL OF ACADEMIC RESEARCH Innovative Academy Research Support Center 11. Спиридонов, Анастас Павлович Угаловнои право Российский Федерации правовая глобализация влияние и соотношение, 2006 года. Даврий нашр ва журналлар: 1. Ўзбекистон Республикаси ИИВ Академиясининг Ахборотномаси. – Тошкент, 2014. - № 4. 2. Ўзбекистон Республикаси Олий суди ахборотномаси. – Тошкент, 2015. - № 1. Il i d bi tl l l 2. Ўзбекистон Республикаси Олий суди ахборотномаси. – Тошкент, 2015. - № 1. Барсукова С.Г. Условно-досрочное освобождение от отбывания наказания по российскому законодательству. Автореф. дисс. ... канд. юр. наук. – Самара, 2000. – С.11– 12. 2. Малин П.М. Уголовно-правовые и уголовно исполнительные проблемы условно- досрочного освобождения от отбывания наказания в виде лишения свободы. Атореф. дисс. … канд. юр. наук. – Рязань, 2000. – С.8.,13.; 3. Евтушенко И.И. Условно-досрочное освобождение осужденных к лишению свободы и их ресоциализация: Монография. –Волгоград, 2005. С. 42.; 4. Носенко А. М. Условно-досрочное освобождение от лишения свободы. Автореф. дисс. ... канд. юр. наук. – Киев, 1993. Volume 2 Issue 4, April 2022 ISSN 2181-2020 Page 484 Volume 2 Issue 4, April 2022 ISSN 2181-2020 Volume 2 Issue 4, April 2022 Page 484
https://openalex.org/W1910208674	https://hau.collections.crest.ac.uk/15100/1/Rob%20Graham%20development%20of%20a%20real%20time.pdf	English	null	Development of a Real-Time qPCR Assay for Quantification of Covert Baculovirus Infections in a Major African Crop Pest	Insects	2,015	cc-by	6,904	insects ISSN 2075-4450 www.mdpi.com/journal/insects/ OPEN ACCESS insects ISSN 2075-4450 www.mdpi.com/journal/insects/ OPEN ACCESS Insects 2015, 6, 746-759; doi:10.3390/insects6030746 insects ISSN 2075-4450 www.mdpi.com/journal/insects/ Article Development of a Real-Time qPCR Assay for Quantification of Covert Baculovirus Infections in a Major African Crop Pest Robert I. Graham 1,†,, Yamini Tummala 1, Glenn Rhodes 2, Jenny S. Cory 3, Alan Shirras 4, David Grzywacz 5 and Kenneth Wilson 1 1 Lancaster Environment Centre, Lancaster University, Lancaster LA1 4YQ, UK; E-Mails: yaminitummala@gmail.com (Y.T.); ken.wilson@lancaster.ac.uk (K.W.) 2 Lake Ecosystems Group, Centre for Ecology and Hydrology, Bailrigg, Lancaster LA1 4AP, UK; E-Mail: glenn@ceh.ac.uk 3 Department of Biological Sciences, Simon Fraser University, Burnaby, BC V5A 1S6, Canada; E-Mail: jsc21@sfu.ca 4 Biomedical and Life Sciences, Lancaster University, Lancaster LA1 4YQ, UK; E-Mail: a.shirras@lancaster.ac.uk 5 Natural Resources Institute, University of Greenwich, Chatham Maritime, Kent ME4 4TB, UK; E-Mail: d.grzywacz@greenwich.ac.uk † Current address: Crop and Environment Sciences, Harper Adams University, Edgmond, Shropshire TF10 8NB, UK. Author to whom correspondence should be addressed; E-Mail: rgraham@harper-adams.ac.uk. Academic Editor: Brian T. Forschler Received: 23 April 2015 / Accepted: 19 August 2015 / Published: 25 August 2015 Abstract: Many pathogens and parasites are present in host individuals and populations without any obvious signs of disease. This is particularly true for baculoviruses infecting lepidopteran hosts, where studies have shown that covert persistent viral infections are almost ubiquitous in many species. To date, the infection intensity of covert viruses has rarely been quantified. In this study, we investigated the dynamics of a covert baculovirus infection within the lepidopteran crop pest Spodoptera exempta A real time quantitative OPEN ACCESS Insects 2015, 6, 746-759; doi:10.3390/insects6030746 1. Introduction Diseases can be an important factor in the population dynamics of insect pests of agricultural importance as well as being useful model systems for exploring host pathogen interactions. Among the most important insect crop pests are the species of Lepidoptera such as Spodoptera spp. and Heliothis/Helicoverpa spp. which are globally important crop pests whose conventional control is challenging due to their ability to develop resistance to chemical pesticides. The viral insect diseases caused by Baculoviruses, double-stranded DNA virus [1] are among the most studied and have been developed for both classical biological control and as biological pesticides. These viruses are characterised by having the infectious virions embedded in proteinaceous occlusion bodies (OB) adapted for persistence in the environment and classical horizontal transmission between susceptible hosts [2]. With advances in molecular biology, it has become evident that for many insect viral pathogens (including baculoviruses) the ability to cause persistent non-lethal covert infections within their hosts that are capable of being transmitted from parent to offspring is an important characteristic in the ecology of the virus [3–14]. Persistent infections are believed to exist in one of two forms: either as a latent infection in which the virus is non-replicating and transcriptional activity is minimal; or as an infection in which the virus is transcriptionally active and replication occurs at a low level, as widely accepted to occur in baculovirus systems studied to date [5–8]. The precise mechanism by which covert viruses persist in cells of infected insects without causing cell death remains unclear. Studies have shown that when larvae are subjected to periods of stress (such as poor diet, crowding, or the presence of other pathogens), covert infections may be triggered into lethal overt infections that produce OBs for subsequent horizontal transmission (e.g., [4,15]). However, the various protocols for triggering covert viruses have proved inherently unreliable and unrepeatable indicating that our fundamental knowledge of this phenomenon is deficient and that we have much to learn about the dynamics and behaviour of non-persistent infections and their role in the ecology of lepidopteran populations, especially those of economic and agricultural importance. The larval stage of the African armyworm moth, Spodoptera exempta (Lepidoptera: Noctuidae), is a devastating cyclical and migratory crop pest of maize, wheat, sorghum, and other staple crops in a large part of sub-Saharan Africa [16]. High-density outbreaks of armyworm caterpillars typically exceed densities of 100–200 larvae per m2. Insects 2015, 6 Insects 2015, 6 peaking in early larval instars and being lowest in adults and final-instar larvae. Adult dissections indicated that, contrary to expectation, viral load aggregation was highest in the head, wings and legs, and lowest in the thorax and abdomen. The data presented here have broad implications relating to our understanding of transmission patterns of baculoviruses and the role of covert infections in host-pathogen dynamics. Keywords: Spodoptera exempta; baculovirus; nucleopolyhedrovirus; covert infections; TaqMan real-time qPCR Keywords: Spodoptera exempta; baculovirus; nucleopolyhedrovirus; covert infections; TaqMan real-time qPCR Robert I. Graham 1,†,, Yamini Tummala 1, Glenn Rhodes 2, Jenny S. Cory 3, Alan Shirras 4, David Grzywacz 5 and Kenneth Wilson 1 1 Lancaster Environment Centre, Lancaster University, Lancaster LA1 4YQ, UK; E-Mails: yaminitummala@gmail.com (Y.T.); ken.wilson@lancaster.ac.uk (K.W.) 1 Lancaster Environment Centre, Lancaster University, Lancaster LA1 4YQ, UK; E-Mails: yaminitummala@gmail.com (Y.T.); ken.wilson@lancaster.ac.uk (K.W.) 2 Lake Ecosystems Group, Centre for Ecology and Hydrology, Bailrigg, Lancaster LA1 4AP, UK; E-Mail: glenn@ceh.ac.uk 2 Lake Ecosystems Group, Centre for Ecology and Hydrology, Bailrigg, Lancaster LA1 4AP, UK; E-Mail: glenn@ceh.ac.uk 3 Department of Biological Sciences, Simon Fraser University, Burnaby, BC V5A 1S6, Canada; E-Mail: jsc21@sfu.ca 3 Department of Biological Sciences, Simon Fraser University, Burnaby, BC V5A 1S6, Canada; E-Mail: jsc21@sfu.ca 4 Biomedical and Life Sciences, Lancaster University, Lancaster LA1 4YQ, UK; E-Mail: a.shirras@lancaster.ac.uk 5 Natural Resources Institute, University of Greenwich, Chatham Maritime, Kent ME4 4TB, UK; E-Mail: d.grzywacz@greenwich.ac.uk 5 Natural Resources Institute, University of Greenwich, Chatham Maritime, Kent ME4 4TB, UK; E-Mail: d.grzywacz@greenwich.ac.uk † Current address: Crop and Environment Sciences, Harper Adams University, Edgmond, Shropshire TF10 8NB, UK. Author to whom correspondence should be addressed; E-Mail: rgraham@harper-adams.ac.uk. Academic Editor: Brian T. Forschler * Author to whom correspondence should be addressed; E-Mail: rgraham@harper-adam Received: 23 April 2015 / Accepted: 19 August 2015 / Published: 25 August 2015 Abstract: Many pathogens and parasites are present in host individuals and populations without any obvious signs of disease. This is particularly true for baculoviruses infecting lepidopteran hosts, where studies have shown that covert persistent viral infections are almost ubiquitous in many species. To date, the infection intensity of covert viruses has rarely been quantified. In this study, we investigated the dynamics of a covert baculovirus infection within the lepidopteran crop pest Spodoptera exempta. A real-time quantitative polymerase chain reaction (qPCR) procedure using a 5' nuclease hydrolysis (TaqMan) probe was developed for specific detection and quantification of Spodoptera exempta nucleopolyhedrovirus (SpexNPV). The qPCR assay indicated that covert baculovirus dynamics varied considerably over the course of the host life-cycle, with infection load 747 2.2. Overt Baculovirus Provenance Spodoptera exempta nucleopolyhedrovirus (SpexNPV) was collected from a single larval cadaver in central Tanzania in 2008 and isolated using standard centrifugation techniques, as previously described [23]. The concentration of SpexNPV occlusion bodies (OBs) of each virus preparation was measured using a Neubauer Improved haemocytometer, with replicated samples taken twice at two dilutions. Viral DNA was extracted as described in Graham et al. [24]. Briefly, OBs were lysed by addition of 0.5 M Na2CO3, 0.1% SDS and incubated for 2 h at 37 °C with proteinase K (200 mg/mL). DNA was purified by phenol/chloroform extraction, dialysed in 1× TE buffer, and stored at 4 °C until required. Insects 2015, 6 748 Insects 2015, 6 S. exempta nucleopolyhedrovirus (SpexNPV) are prevalent in natural populations [17] and SpexNPV has been known to cause up to 98% mortality in some natural populations, though infection levels are highly variable and generally much lower than this [16,21]. While it has been shown that covert infections occur in SpexNPV and it is hypothesised that they play an important role in the seasonal population cycle [8], very little is known about the population dynamics of the covert form of this baculovirus or how the switch from covert to overt infection is initiated. One hypothesis that has been proposed is that covert infections need to reach a certain threshold intensity or viral load before they can be triggered into overt infections. Greater insight into this process could be a key to better understanding of the dynamics of SpexNPV and thus to better prediction of pest outbreaks and to an understanding of how to manipulate the host pathogen system to reduce pest outbreaks and crop damage. Here, we report on the design and implementation of a real-time qPCR technique allowing the assessment of covert baculovirus infections within both larval and adult forms of the host moth. Specifically, the qPCR assay was used to: (i) quantify the dynamics of covert SpexNPV infections over the course of host development; (ii) determine the principal body parts harbouring covert infections; and (iii) track the replication dynamics of an orally-induced overt infection. Assessing these three parameters allowed us to quantify the infection intensity (viral load) dynamics of asymptomatic covert infections, evaluate hypotheses of covert to overt triggering and thereby significantly furthering our knowledge of host-pathogen interactions within this major agricultural crop pest. 2.1. Culture of S. exempta 2.1. Culture of S. exempta The Spodoptera exempta laboratory culture was collected from central Tanzania in January 2011, and maintained on standard culture protocol on a wheatgerm diet [8,22] at 24 °C under a 12 h light/dark cycle. Insects used for experiments were at least two generations under laboratory conditions. 1. Introduction The adult moth is highly migratory, and most armyworm outbreaks are characterized as single-generation eruptions [16]. S. exempta harbours a baculovirus exhibiting both overt horizontal [17,18] and covert vertical [8,19,20] transmission strategies. Overt infections of Insects 2015, 6 Insects 2015, 6 (Qiagen Ltd., Crawley, UK), according to the manufacturer’s instructions. DNA quantification was undertaken using a Nanodrop2000 Spectrophotometer (ThermoScientific, Willimington, DE, USA). DNA quality was assessed by testing the 260/280 nm ratio readings during the quantification process, and amplifying the S. exempta mitochondrial cytochrome oxidase I (COI) gene, using universal primers LCO-1490 and HCO-1298 [25]. (Qiagen Ltd., Crawley, UK), according to the manufacturer’s instructions. DNA quantification was undertaken using a Nanodrop2000 Spectrophotometer (ThermoScientific, Willimington, DE, USA). DNA quality was assessed by testing the 260/280 nm ratio readings during the quantification process, and amplifying the S. exempta mitochondrial cytochrome oxidase I (COI) gene, using universal primers LCO-1490 and HCO-1298 [25]. 2.4. Real-Time Quantitative PCR (qPCR) Conceptual considerations and nomenclature for real-time qPCR were carried out and reported here in accordance with the MIQE guidelines [26]. Primer Express software (v3.0; Applied Biosystems, Warrington, UK) was used to design primers and the hydrolysis (Taqman, Thermo Fisher Scientific, Warrington, UK) probe specific to the SpexNPV polyhedrin gene in the Tanzania isolate (Accession number JX488468). The forward primer (P1) sequence 5'-CCCGTGTACGTAGGAAACAACA-3', reverse primer (P2) 5'-CAACCGCCGCCCTTCT-3' and hydrolysis probe (TaqMan) 5'-6FAM-CGAGTA CCGCATCAGCCTGGCC-TAMRA-3' amplified a 62 bp region of the SpexNPV polyhedrin gene. To check for target specificity in silico, the viral polyhedrin sequence and the primer sequences were compared with sequences published on NCBI GenBank database, using the online basic local alignment search tool (BLAST). A perfect match along the entire length of the primers was found only for the polyhedrin target gene sequence. Template DNA (5 µL) was used in 25 µL reactions containing 12.5 µL 2× TaqMan Universal PCR Master Mix (P/N 4304437, Life Technologies, UK) and 0.4 µM each of primers P1 and P2 (P/N 4304972, Life Technologies, UK) and hydrolysis probe (P/N 450003, Life Technologies, UK, UK). Reactions were run on an ABI Prism 7000 SDS machine (Applied Biosystems) in triplicate, using a thermal cycling program consisting of an initial denaturation step at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. For each qPCR assay, a standard curve was constructed using 10-fold serial dilutions of viral genomes, originally isolated from pure NPV OBs (range, 5–5 × 106 viral genomes); and each dilution was processed in triplicate on the same 96-well PCR plate with the samples. SpexNPV genome size was determined from a complete viral-genome sequence (genome size of 129.5 Kbp; Escasa and Cory, unpublished data [27]). Negative controls (“no template controls” NTC; water instead of template DNA) were included in all reactions. Data were analysed using Sequence Detection Software (v1.2.3) 7000 (Applied Biosystems, Warrington, UK) and SpexNPV viral loads were reported as number of viral genomes per µg of total DNA. Only standard curves in which the regression coefficients of determination (R2) exceeded 0.990 were considered sufficiently accurate for determination of persistent virus levels. PCR efficiency was calculated using the equation, Efficiency = −1 + 10(−1/slope) [28]. 2.3. Total Genomic DNA Extraction from Insects All insects used in this study were sourced from the laboratory culture where there was no history of viral contamination or spontaneous outbreaks. However, to remove any potential contaminants from our samples (such as surface pathogens, including NPV OBs), insects used for DNA extraction were surfaced-sterilised using 10% hypochlorite solution, and then washed twice in 70% ethanol. Total genomic DNA was extracted from whole armyworm adults and larvae using the DNA/RNA Allprep Kit 749 2.7. Baculovirus Dynamics Following SpexNPV Ingestion Newly-moulted 4th-instar larvae (L4) were allowed to ingest a diet-plug (n = 27 larvae per treatment) treated with either distilled H2O (control) or 1 × 104 SpexNPV OBs. After 24 h, larvae that had eaten all the diet-plug and virus were transferred to individual polypots (30 mL) containing artificial diet. Larvae were kept at 24 °C and checked every 24 h thereafter. A subset of larvae from both treatment groups was sacrificed at each 24 h time-point to monitor baculovirus dynamics. 3.1. Analytical Sensitivity of the qPCR Assay Serial 10-fold dilutions of purified SpexNPV DNA were used to produce a qPCR standard curve and determine the cut-off value for the assay. The assay was shown to be reproducible (r2 > 0.990) and amplified polyhedrin across a seven log serial dilution range with an efficiency of 99.5% (Figure S1, supplementary material). The most reliable and repeatable limit of detection (LOD) was equivalent to 5 viral genome copies of SpexNPV (approximate Cq of 39). 2.6. Localisation of SpexNPV Infection within Body Regions of Adult Moths Twenty-four adult moths were randomly selected from our laboratory culture and dissected into five body parts (abdomen, thorax, head, legs, and wings) under a Leica MZ7.5 stereo microscope (Leica, Milton Keynes, UK). Total genomic DNA was isolated from each individual body region, and qPCR undertaken to investigate viral load within specific body parts (groups of three individuals were pooled in order to provide sufficient DNA for quantitation). 2.5. Covert Virus Dynamics during S. exempta Development Three pairs of 2nd generation laboratory adults were used to establish a time-course experiment to investigate covert-virus dynamics over host development. Mated females laid eggs for 2–3 days and were then frozen for subsequent analysis. Eggs were placed in 30 mL polypots containing artificial diet. When neonates emerged, approximately 30 insects per pair were individually reared in 30 mL diet pots at 24 °C. Four individuals per pair were sacrificed after 4, 6, 8, 10, and 12 days post hatch, and stored at Insects 2015, 6 750 Insects 2015, 6 −20 °C in 100% ethanol. Total DNA from insects was extracted as above, and TaqMan qPCR used to investigate SpexNPV load. Head capsule widths were recorded as an additional function of larval age. 2.6. Localisation of SpexNPV Infection within Body Regions of Adult Moths 3.2. Covert SpexNPV Dynamics during Host Developmental Cycle To determine whether SpexNPV viral load changed over the course of host developmental cycle, a time-course experiment was established in which S. exempta larvae were sacrificed at specific time-points from the egg stage through to adult maturity (Figure 1a). All insects used in this study were asymptomatic for SpexNPV infection, with no overt viral mortality observed during the experiment. One larva with a very high covert viral load was excluded from the analysis due to undue leverage, however removal of this outlier did not qualitatively alter the results and improved the model fit. Covert viral load was found to vary widely over the course of S. exempta development (up to 104-fold). The highest viral copy number was observed at L2 and L3, with 3349 ± 881 copies/µg DNA and 8180 ± 6953 copies/µg DNA, respectively (n = 12, each). Adults (86.5 ± 25.4 copies/µg DNA; n = 6) and L5 larvae (65.0 ± 13.7 copies/µg DNA; n = 24) produced the lowest viral readings. The viral load was higher in egg and neonate stages then there was a decline in viral load as the larva grew post hatching with a linear decline in viral load from 1st instar neonates to 5th instar larvae (Linear regression: viral load = 2.27 – 5.79 × instar; R2 = 0.64; F1,59 = 106.1, p < 0.0001; Figure 1b), and a quadratic relationship between viral load and 751 3.3. Localisation of Baculovirus in Adult Moths To determine whether the covert baculovirus was systemic or localised in particular body tissues, adult moths were dissected into five regions (abdomen, thorax, head, legs, and wings). There was a highly significant difference between the body regions in the amount of SpexNPV detected (ANOVA: F4,34 = 8.03, p = 0.0001; R2 = 0.49; Figure 2). Post hoc Fisher’s least significant difference tests indicated that covert viral loads were significantly higher in the head, legs and wings (log10(mean) ± S.E. = −3.780 ± 0.140) than in the abdomen and thorax (−2.084 ± 0.239). Figure 2. Covert SpexNPV DNA levels in different body regions of adult S. exempta moths. Means ± S.E. are shown. Different letters above the error bars indicate means that are significantly different according to Fisher’s least significant difference test. Figure 2. Covert SpexNPV DNA levels in different body regions of adult S. exempta moths. Means ± S.E. are shown. Different letters above the error bars indicate means that are significantly different according to Fisher’s least significant difference test. Insects 2015, 6 Insects 2015, 6 Insects 2015, 6 capsule width (viral load = 2.23 – 5.20 × head-capsule + 1.50 × head-capsu p < 0.0001; Figure 1c). e 1. Covert SpexNPV DNA as measured in asymptomatic insects sampled neonates, and 4, 6, 8, 10 and 12 days post egg hatch. SpexNPV load as a fu e (days post-hatch); (B) larval instar (1st to 5th); and (C) head capsule wid the curve shown is the smoothing spline through the raw data. In (B,C) sion lines are shown. larval head-capsule width (viral load = 2.23 – 5.20 × head-capsule + 1.50 × head-capsule2; R2 = 0.61; F2,56 = 43.3, p < 0.0001; Figure 1c). Figure 1. Covert SpexNPV DNA as measured in asymptomatic insects sampled as adults, eggs, neonates, and 4, 6, 8, 10 and 12 days post egg hatch. SpexNPV load as a function of (A) age (days post hatch); (B) larval instar (1st to 5th); and (C) head capsule width (mm) Figure 1. Covert SpexNPV DNA as measured in asymptomatic insects sampled as adults, eggs, neonates, and 4, 6, 8, 10 and 12 days post egg hatch. SpexNPV load as a function of (A) age (days post-hatch); (B) larval instar (1st to 5th); and (C) head capsule width (mm). In (A) the curve shown is the smoothing spline through the raw data. In (B,C) the linear regression lines are shown. Insects 2015, 6 Insects 2015, 6 752 3.4. Baculovirus Dynamics Following Oral Ingestion Using qPCR to track an orally-induced overt infection allowed us to monitor the progress of infection, and give an insight into the dynamics and magnitude of overt baculoviral disease in comparison to covert infections. To quantify the dynamics of viral replication during an overt infection, early L4 larvae were fed 1 × 104 SpexNPV OBs. Baculovirus dynamics were monitored throughout the infection process and compared to control larvae that were not inoculated with virus (Figure 3). Initial background covert readings at day 0 (before treatment) were 589 ± 83 (mean ± S.E.) SpexNPV copies/µg DNA. Throughout larval development, control larvae had on average 4.1 ± 1.9 × 103 copies/µg DNA, with viral load being a quadratic function of time post-infection, declining slightly towards the end of the larval instar (Linear regression: Viral load = 3.17 − 0.70 × Time – 2.55 × Time2; R2 = 0.41; F2,24 = 8.17, p = 0.002). In contrast, in larvae inoculated with SpexNPV viral load increased asymptotically between one and three days post infection to reach an average of 1.3 ± 0.8 × 1010 copies/µg DNA (Viral load = 8.27 + 15.32 × Time – 9.01 × Time2; R2 = 0.88; F2,23 = 87.35, p < 0.0001), with viral load increasing asymptotically with larval age. 753 Insects 2015, 6 Insects 2015, 6 Figure 3. Dynamics of SpexNPV genome replication following ingestion of 104 OBs (treated) and dH2O (control). The lines are the fitted Gompertz curves (see text for details). Figure 3. Dynamics of SpexNPV genome replication following ingestion of 104 OBs (treated) and dH2O (control). The lines are the fitted Gompertz curves (see text for details). Arguably, a better way to describe the SpexNPV dynamics of infected larvae is to fit a Gompertz function to the data: f(x) = a × exp(–b × exp(−k × x)). When this non-linear growth function was fitted to the infection data, the parameter estimates (+ S.E.) for a, b and k were as follows: Control larvae: a = 3.213 + 0.135, b = 0.149 + 0.156, k = 2.849 + 0.843; SpexNPV-treated larvae: a = 10.286 + 0.325, b = 1.963 + 0.371, k = 1.144 + 0.193. Only the asymptote parameter, a, was significantly different from zero for the model describing SpexNPV dynamics in control larvae, whereas all three model parameters were statistically significant for the dynamics in virus-treated larvae, consistent with viral replication in these insects. 3.4. Baculovirus Dynamics Following Oral Ingestion All SpexNPV-treated larvae had died by the final time-point, remaining largely asymptomatic until 5 days post infection. 4. Discussion The aim of this study was to produce and validate a reliable qPCR methodology to quantify the dynamics of covert baculovirus infections within the lepidopteran host, Spodoptera exempta. To achieve this, we developed a specific real-time qPCR detection assay using a hydrolysis (TaqMan) probe, for fast and successful quantitation of SpexNPV. Covert infections are widespread in insect systems, and baculovirus persistent infections have long been thought to be prevalent in lepidopterans (for example, [17]). Their existence has been used to explain the widely observed phenomena of the appearance of occurrence of overt infection in healthy populations [29] and the finding of heterologous viral progeny after oral-infection with baculoviruses [4,30–34]. Traditional methods for NPV detection include phase-contrast microscopy, homogenization and staining of insect tissues. This requires specialized skills, is not very sensitive or species specific, and can only be used to detect virus in samples Insects 2015, 6 Insects 2015, 6 754 with fairly high viral loads, such as infected larvae. Molecular methods now present an opportunity to develop much more sensitive tools for virus detection, such as this new qPCR method. In our host-development study, covert viral load appeared to peak in early instar larvae (0, 4 and 6 days post hatch), and then decrease markedly with larval age, specifically after 8 days post hatch, by which time the larvae had entered larval-stage L4. In an assessment of the closely related Spodoptera exigua host-NPV system, Murillo et al. [9] used a SYBRgreen qPCR assay. They reported high covert viral loads at L1 and L3, with lower viral levels at later instars, pupae and adults, concluding that the host may be suppressing infection during these later life-stages. It should be noted that in our study we have looked at “number of viral genomes per µg DNA” whereas Murillo et al. [9] looked at pg DNA. This should be noted when comparing different published studies. In common with many other Spodoptera species, S. exempta is a phase-polyphenic species [22], and it is at L4 stage that the cuticular colour differentiation becomes most obvious. It is possible that this cuticular colour change (from green to a black melanic form) and the accumulation of melanin and other immune-associated factors [35,36] could be driving the reduction in viral loads; thus density-dependent prophylaxis (DDP) may directly impact upon covert virus dynamics, with dark melanic forms of larvae having lower viral loads. Indeed, Vilaplana et al. 4. Discussion [20] showed that in larva from parents reared under solitary conditions, and so which did not show DDP, the viral infection was more prevalent than those from parents reared gregariously. This is consistent with a finding in the field, where covert viral loads were lower in populations of very high larval density [21]. The insects in our current study were reared individually (and therefore DDP per se would not be a factor), but individual cuticular melanisation remains a consideration. When S. exempta larvae were subjected to a superinfection via ingestion of SpexNPV OBs, rapid replication of SpexNPV occurred within 24–48 h post infection (the number of virus genomes increased approximately 106-fold). This indicated rapid virus replication, despite larvae remaining largely asymptomatic for a further 72 h. At this stage of our investigations, it is not clear whether the viral replication was due to the ingested virus only, replication of covert virus only, or a combination of the two. It is possible that the covert NPV infection was triggered into rapid replication somehow by the challenge of ingested NPV [4,30–34]. Further investigation whereby S. exempta larvae are fed heterologous virus (one that is not detected by this qPCR assay) could resolve the nature of covert virus dynamics when challenged by a competing pathogen. Vilaplana et al. [8] demonstrated that persistent infections of SpexNPV were common in S. exempta field populations of larvae collected from a small geographical area in northern Tanzania. The detection of virus in adult moths is considered particularly important due to the fact that adults cannot be directly infected by baculoviruses, nor are they thought to support extensive (if any) viral replication [37]. Therefore, detection here of SpexNPV in the adult life stage validates a key assumption of the hypothesis of vertical transmission of the virus. Interestingly, our results on the location of virus in covertly infected adult body parts are perhaps contrary to what one might expect: the virus was found in all the body parts tested and not restricted to specific locations or organs. In a viral species adapted to vertical transmission, one would expect high viral loads within the reproductive organs (found in the host abdomen) but in this study, the highest loads were found not in the abdomen but in the wings, head and legs. Insects 2015, 6 755 Insects 2015, 6 Insects 2015, 6 artefact of DNA isolation, with perhaps different body parts having more PCR inhibitory chemicals for which we did not test, although the protocol was undertaken to ensure these would be kept to a minimum. In future experiments, diluting samples 10-, 100- or even 1000-fold before qPCR could enable us to test for possible contaminants. In addition, despite the specificity and sensitivity of our SpexNPV polyhedrin primers, we cannot completely rule out the possibility that highly similar and not yet discovered viruses also could be recognized. As discussed by Krokene et al. [38], targeting additional genes in the virus genome of interest would make the qPCR method more robust and reduce the chance of detecting highly similar viruses. In a wider context, baculoviruses present an interesting model system to study due to the presence of what appears to be two transmission strategies (horizontally transmitted, highly pathogenic; and vertically transmitted, seemingly asymptomatic). In the armyworm system, it may be the specific dynamics of the host populations that selects for the occurrence of these two different viral transmission strategies. Armyworm populations alternate during the year between many outbreaks with high populations of gregarious armyworms in the rainy seasons when food is abundant and very low density solitary populations in the dry season when food is scarce. Persistent covert asymptomatic virus strains are most likely the superior competitors (as compared to the highly pathogenic horizontally-transmitted variants) during the dry season (approximately 6 months of the year in Tanzania) when host population densities are extremely low and dispersed [39]. However, when conditions change and high density host populations appear and become suitable for horizontal transmission of the virus, a persistent virus that can reactivate into an overt-infection will become most competitive. To date, modelling approaches suggest that persistent infections could create a variety of host dynamics: from host-pathogen cycles to endemic persistent interactions, whereby the healthy individuals are lost from the host system [40–43]. The high prevalence of covert virus infection found in field populations of S. exempta (>97%, [21]) suggests that the populations studied to date are close to such an endemic fixation in which virus-free insects have been lost. At present it is not known how long covert infections can persist in a population, although assays on field-collected S. exempta show that they can persist for at least seven generations maintaining 100% infection prevalence [8]. 4. Discussion As yet, we are unsure why this may be, but possibly it is due to virus accumulation in areas of high chitin concentration, or areas of lower haemolymph levels (resulting in the virus being exposed to a reduced immune challenge). However, at this stage we cannot rule out the possibility that the differences are due to an Insects 2015, 6 Thus we might expect baculovirus variants that produce covert infections to be adapted in some way to vertical transmission, such as by showing reduced pathogenicity. There is some recent evidence from Spodoptera exigua NPV that supports this supposition [44], and that vertical-adapted variants may be produced (i.e., variants that only undertake covert transmission). To date we cannot be sure of the precise nature of the covert infection in S. exempta, other than to note that the 62 bp amplified region detected by this qPCR assay is identical to overt-SpexNPV sequence data (Genbank accession number JX488468), and differs over the same region from polyhedrin in Spodoptera frugiperda at only 2 positions (see accession number KC845532). More research on the genetics of covert infections is required to determine the relationship of covert and overt variants, as the qPCR amplicon is too short to attain any meaningful phylogenetic data. However, a previous study investigating SpexNPV covert infections using the lef-8 gene sequence, Vilaplana et al. [8] did observe minor nucleotide variation between covert and overt infections. Thus, it is possible that covert infections of SpexNPV may be restricted to a few closely related variants; or several very different genotypes, as observed in overt SpexNPV populations [18]. Elucidating this issue would help us to understand the Insects 2015, 6 756 Insects 2015, 6 potential for wider use of the current qPCR assay or modifications of it for allelic discrimination assays or the use of high resolution melting (HRM) analysis. Acknowledgments This work was supported by a grant from the Sustainable Agriculture Research for International Development (SARID) programme, funded by the United Kingdom’s Biotechnology and Biological Sciences Research Council (BBSRC) and the Department for International Development (DfID). The views expressed are not necessarily those of DfID or BBSRC. We are extremely grateful for the support and assistance of the Tanzanian Ministry of Agriculture, Food and Cooperatives, to Mr Wilfred Mushobozi for field assistance, and to Phill Nott for technical expertise. 5. Conclusions In conclusion, we present a qPCR assay that allows the investigation of covert baculovirus dynamics within a lepidopteran host, in this case the African armyworm, S. exempta. From an applied perspective, the existence of covert and sub-lethal infections in pest populations may open up interesting new options for the implementation of biological pest control programs, given that covert infection can be transmitted to their offspring, which in turn may be more likely to succumb to patent NPV disease following application of viral biopesticide [20]. As a result, the role of covert infections could be an important consideration when managing biological pest control programs. Author Contributions Kenneth Wilson, Jenny S. Cory, David Grzywacz and Alan Shirras conceived the original research idea; Robert I. Graham and Kenneth Wilson designed the current experiments; Robert I. Graham and Yamini Tummala performed the experiments; Kenneth Wilson, Robert I. Graham, Yamini Tummala and Glenn Rhodes analysed the data; Robert I. Graham, Kenneth Wilson, Glenn Rhodes, Jenny S. Cory, Alan Shirras and David Grzywacz wrote the paper. Conflicts of Interest The authors declare no conflict of interest. The funding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. 1. Jehle, J.A.; Blissard, G.W.; Bonning, B.C.; Cory, J.S.; Herniou, E.A.; Rohrmann, G.F.; Theilmann, D.A.; Thiem, S.M.; Vlak, J.M. On the classification and nomenclature of baculoviruses: A proposal for revision. Arch. Virol. 2006, 151, 1257–1266. 2. Dwyer, G.; Dushoff, J.; Elkinton, J.S.; Levin, S.A. Pathogen-driven outbreaks in forest defoliators revisited: Building models from experimental data. Am. Nat. 2000, 156, 105–120. 3. Kukan, B. Vertical transmission of nucleopolyhedrovirus in insects. J. Invertebr. Pathol. 1999, 74, 103–111. issard, G.W.; Bonning, B.C.; Cory, J.S.; Herniou, E.A.; Rohrmann, G.F.; Theilmann, D.A.; References 1. Jehle, J.A.; Blissard, G.W.; Bonning, B.C.; Cory, J.S.; Herniou, E.A.; Rohrmann, G.F.; Theilmann, D.A.; Thiem, S.M.; Vlak, J.M. On the classification and nomenclature of baculoviruses: A proposal for revision. Arch. Virol. 2006, 151, 1257–1266. 2. Dwyer, G.; Dushoff, J.; Elkinton, J.S.; Levin, S.A. Pathogen-driven outbreaks in forest defoliators revisited: Building models from experimental data. Am. Nat. 2000, 156, 105–120. 3. Kukan, B. Vertical transmission of nucleopolyhedrovirus in insects. J. Invertebr. Pathol. 1999, 74, 103–111. 3. Kukan, B. Vertical transmission of nucleopolyhedrovirus in insects. J. Invertebr. Pathol. 1999, 74, 103–111. Insects 2015, 6 757 Insects 2015, 6 4. Hughes, D.S.; Possee, R.D.; King, L.A. Activation and detection of a latent baculovirus resembling Mamestra brassicae nuclear polyhedrosis virus in M. brassica insects. Virology 1993, 194, 608–615. 4. Hughes, D.S.; Possee, R.D.; King, L.A. Activation and detection of a latent baculovirus resembling Mamestra brassicae nuclear polyhedrosis virus in M. brassica insects. Virology 1993, 194, 608–615. 5. Hughes, D.S.; Possee, R.D.; King, L.A. Evidence for the presence of a low-level, persistent baculovirus infection of Mamestra brassicae insects. J. Gen. Virol. 1997, 78, 1801–1805. 6. Burden, J.P.; Griffiths, C.M.; Cory, J.S.; Smith, P.; Sait, S.M. Vertical transmission of sublethal granulovirus infection in the Indian meal moth, Plodia interpunctella. Mol. Ecol. 2002, 11, 547–555. 7. Burden, J.P.; Nixon, C.P.; Hodgkinson, A.E.; Possee, R.D.; Sait, S.M.; King, L.A.; Hails, R.S. Covert infections as a mechanism for long-term persistence of baculoviruses. Ecol. Lett. 2003, 6, 524–531. 8. Vilaplana, L.; Wilson, K.; Redman, E.M.; Cory, J.S. Pathogen persistence in migratory insects: High levels of vertically-transmitted virus infection in field populations of the African armyworm. Evol. Ecol. 2010, 24, 147–160. 9. Murillo, R.; Hussey, M.S.; Possee, R.D. Evidence for covert baculovirus infections in a Spodoptera exigua laboratory culture. J. Gen. Virol. 2011, 92, 1061–1070. 10. Kemp, E.M.; Woodward, D.T.; Cory, J.S. Detection of single and mixed covert baculovirus infections in eastern spruce budworm, Choristoneura fumiferana populations. J. Invertebr. Pathol. 2011, 107, 202–205. 11. Virto, C.; Zarate, C.A.; Lopez-Ferber, M.; Murillo, R.; Caballero, P.; Williams, T. Gender-mediated differences in vertical transmission of a nucleopolyhedrovirus. PLoS ONE 2013, 8, doi:10.1371/ journal.pone.0070932 12. Virto, C.; Navarro, D.; Tellez, M.M.; Herrero, S.; Williams, T.; Murillo, R.; Caballero, P. Natural populations of Spodoptera exigua are infected by multiple viruses that are transmitted to their offspring. J. Invertebr. Pathol. 2014, 122, 22–27. 13. Insects 2015, 6 758 Insects 2015, 6 Insects 2015, 6 20. Vilaplana, L.; Redman, E.M.; Wilson, K.; Cory, J.S. Density-related variation in vertical transmission of a virus in the African armyworm. Oecologia 2008, 155, 237–246. 21. Graham, R.I.; Grzywacz, D.; Mushobozi, W.L.; Wilson, K. Wolbachia in a major African crop pest increases susceptibility to viral disease rather than protects. Ecol. Lett. 2012, 15, 993–1000. 22. Reeson, A.F.; Wilson, K.; Gunn, A.; Hails, R.S.; Goulson, D. Baculovirus resistance in the noctuid Spodoptera exempta is phenotypically plastic and responds to population density. Proc. R Soc. Lond. B. 1998, 265, 1787–1791. 23. Hunter-Fujita, F.R.; Entwhistle, P.F.; Evans, H.F.; Crook, N.E. In Insect Viruses and Pest Management; Wiley and Sons: Chichester, UK, 1998. 24. Graham, R.I.; Tyne, W.I.; Possee, R.D.; Sait, S.M.; Hails, R.S. Genetically variable nucleopolyhedroviruses isolated from spatially separate populations of the winter moth Operophtera brumata (Lepidoptera: Geometridae) in Orkney. J. Invertebr. Pathol. 2004, 87, 29–38. 25. Folmer, O.; Black, M.; Hoeh, W.; Lutz, R.; Vrijenhoek, R. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol. Mar. Biol. Biotech. 1994, 3, 294–299. 26. Bustin, S.A.; Benes, V.; Garson, J.A.; Hellemans, J.; Huggett, J.; Kubista, M.; Mueller, R.; Nolan, T.; Pfaffl, M.M.; Shipley, G.L.; et al. The MIQE guidelines: Minimum information for publication of quantitative Real-Time PCR experiments. Clin. Chem. 2009, 55, 611–622. 27. Escasa, S.; Cory, J.S. Algoma University, Sault Set Marie, Ontario, Canada; Unpublished data, 2008. 28. Nybo, K. qPCR efficiency calculations. BioTechniques 2011, 51, 401–402. , ; y, g y, , , ; p , 28. Nybo, K. qPCR efficiency calculations. BioTechniques 2011, 51, 401–402. 28. Nybo, K. qPCR efficiency calculations. BioTechniques 2011, 51, 401–402. 29. Fuxa, J.R.; Sun, J.Z.; Weidner, E.H.; LaMotte, L.R. Stressors and rearing diseases of Trichoplusia ni: Evidence of vertical transmission of NPV and CPV. J. Invertebr. Pathol. 1999, 74, 149–155. 30. Fuxa, J.R.; Richter, A.R. Virulence and multigeneration passage of a nuclear polyhedrosis virus selected for an increased rate of vertical transmission. Biol. Control 1992, 2, 171–175. 31. Fuxa, J.R.; Weidner, E.H.; Richter, A.R. Polyhedra without virions in a vertically transmitted nuclear polyhedrosis-virus. J. Invertebr. Pathol. 1992, 60, 53–58. 32. Cooper, D.; Cory, J.S.; Theilmann, D.A.; Myers, J.H. Nucleopolyhedroviruses of forest and western tent caterpillars: Cross-infectivity and evidence for activation of latent virus in high-density field populations. Ecol. Entomol. 2003, 28, 41–50. 33. Longworth, J.F.; Cunningham, J.C. 37. Federici, B.A. Baculovirus pathogenesis. In The Baculoviruses; Miller, L.K., Ed.; Plenum Press: New York, NY, USA, 1997. References Jakubowska, A.K.; D’Angiolo, M.; Gonzalez-Martinez, R.M.; Millan-Leiva, A.; Carballo, A.; Murillo, R.; Caballero, P.; Herrero, S. Simultaneous occurrence of covert infections with small RNA viruses in the lepidopteran Spodoptera exigua. J. Invertebr. Pathol. 2014, 121, 56–63. 14. Xu, P.; Liu, Y.; Graham, R.I.; Wilson, K.; Wu, K. Densovirus is a mutualistic symbiont of a global crop pest (Helicoverpa armigera) and protects against a baculovirus and Bt biopesticide. PLoS Pathog. 2014, 10, doi:10.1371/journal.ppat.1004490. 15. Fuxa, J.R.; Richter, A.R.; Ameen, A.O.; Hammock, B.D. Vertical transmission of TnSNPV, TnCPV, AcMNPV, and possibly recombinant NPV in Trichoplusia ni. J. Invertebr. Pathol. 2002, 79, 44–50. 16. Rose, D.J.W.; Dewhurst, C.F.; Page, W.W. The African Armyworm Handbook, 2nd Ed.; NRI: Chatham, UK, 2000. 17. Brown, E.S.; Swaine, G. Virus disease of African armyworm Spodoptera exempta (Wlk). Bull. Entomol. Res. 1965, 56, 95–116. 18. Redman, E.M.; Wilson, K.; Grzywacz, D.; Cory, J.S. High levels of genetic diversity in Spodoptera exempta NPV from Tanzania. J. Invertebr. Pathol. 2010, 105, 190–193. 19. Swaine, G. Generation to generation passage of nuclear polyhedral virus of Spodoptera exempta (Wlk). Nature 1966, 210, 1053–1054. Insects 2015, 6 Activation of occult nuclear-polyhedrosis viruses by foreign nuclear polyhedra. J. Invertebr. Pathol. 1968, 10, 361–367. 34. Kouassi, L.N.G.; Tsuda, K.; Goto, C.; Mukawa, S.; Sakamaki, Y.; Kusigemati, K.; Nakamura, K. Prevalence of latent virus in Spodoptera. litura (Fabricius) (Lepidoptera: Noctuidae) and its activation by a heterologous virus. Appl. Entomol. Zool. 2009, 44, 95–102. 35. Wilson, K.; Cotter, S.C.; Reeson, A.F.; Pell, J.K. Melanism and disease resistance in insects. Ecol. Lett. 2001, 4, 637–649. 36. Cotter, S.C.; Myatt, J.P.; Benskin, C.M.H.; Wilson, K. Selection for cuticular melanism reveals immune function and life-history trade-offs in Spodoptera littoralis. J. Evol. Biol. 2008, 21, 1744–1754. 37. Federici, B.A. Baculovirus pathogenesis. In The Baculoviruses; Miller, L.K., Ed.; Plenum Press: New York, NY, USA, 1997. Insects 2015, 6 759 Insects 2015, 6 38. Krokene, P.; Heldal, I.; Fossdal, C.G. Quantifying Neodiprion sertifer nucleopolyhedrovirus DNA from insects, foliage and forest litter using the quantitative real-time polymerase chain reaction. Agric. For. Entomol. 2013, 15, 120–125. 39. Dewhurst, C.F.; Page, W.W.; Rose, D.J.W. The relationship between outbreaks, rainfall and low density populations of the African armyworm, Spodoptera exempta, in Kenya. Entomol. Exp. Appl. 2001, 98, 285–294. 40. Boots, M.; Greenman, J.; Ross, D.; Norman, R.; Hails, R.; Sait, S. The population dynamical implications of covert infections in host-microparasite interactions. J. Anim. Ecol. 2003, 72, 1064–1072. 41. Jones, E.O.; White, A.; Boots, M. Interference and the persistence of vertically transmitted parasites. J. Theor. Biol. 2007, 246, 10–17. 42. Sorrell, I.; White, A.; Pedersen, A.B.; Hails, R.S.; Boots, M. The evolution of covert, silent infection as a parasite strategy. Proc. R Soc. Lond. B 2009, 276, 2217–2226. 43. Bonsall, M.B.; Sait, S.M.; Hails, R.S. Invasion and dynamics of covert infection strategies in structured insect-pathogen populations. J. Anim. Ecol. 2005, 74, 464–474. 44. Cabodevilla, O.; Villar, E.; Virto, C.; Murillo, R.; Williams, T.; Caballero, P. Intra- and intergenerational persistence of an insect Nucleopolyhedrovirus: Adverse effects of sublethal disease on host development, reproduction, and susceptibility to superinfection. Appl. Environ. Microbiol. 2011, 77, 2954–2960. © 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/). © 2015 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/). © 2015 by the authors; licensee MDPI, Basel, Switzerland. Insects 2015, 6 This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W3019570233	https://digitalcommons.uri.edu/cgi/viewcontent.cgi?article=2167&context=oa_diss	English	null	DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR RING OPENING POLYMERIZATION OF CYCLIC LACTONES	null	2,020	cc-by	86,170	DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR RING OPENING POLYMERIZATION OF CYCLIC LACTONES RING OPENING POLYMERIZATION OF CYCLIC LACTONES Nayanthara Dharmaratne U i it f Rh d I l d t dh t @ il Follow this and additional works at: https://digitalcommons.uri.edu/oa_diss Terms of Use All rights reserved under copyright. Recommended Citation Recommended Citation Recommended Citation Recommended Citation Dharmaratne, Nayanthara, "DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR RING OPENING POLYMERIZATION OF CYCLIC LACTONES" (2020). Open Access Dissertations. Paper 1152. https://digitalcommons.uri.edu/oa_diss/1152 University of Rhode Island University of Rhode Island DigitalCommons@URI DigitalCommons@URI Open Access Dissertations 2020 DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR RING OPENING POLYMERIZATION OF CYCLIC LACTONES RING OPENING POLYMERIZATION OF CYCLIC LACTONES Nayanthara Dharmaratne University of Rhode Island, nayantaradharmaratne@gmail.com Open Access Dissertations Recommended Citation Recommended Citation Dharmaratne, Nayanthara, "DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR RING OPENING POLYMERIZATION OF CYCLIC LACTONES" (2020). Open Access Dissertations. Paper 1152. https://digitalcommons.uri.edu/oa_diss/1152 This Dissertation is brought to you by the University of Rhode Island. It has been accepted for inclusion in Open Access Dissertations by an authorized administrator of DigitalCommons@URI. For more information, please contact digitalcommons-group@uri.edu. For permission to reuse copyrighted content, contact the author directly. DOCTOR OF PHILOSOPHY DISSERTATION OF NAYANTHARA U. DHARMARATNE DEVELOPMENT OF NOVEL (THIO)UREA ORGANOCATALYSTS FOR RING OPENING POLYMERIZATION OF CYCLIC LACTONES. BY NAYANTHARA U. DHARMARATNE ABSTRACT Amidst multiple catalysts for ROP, H-bonding organocatalysts stand out in the precise level of reaction control they are able to render during ROP. The H-bonding class of organocatalysts are thought to effect ROP via dual activation of both monomer and chain end. The only drawback of these class of catalysts is the low activity shown despite their high selectivity towards ROP. A new class of H-bond donating ureas was developed for the ring-opening polymerization (ROP) of lactone monomers to address this problem of low activity. The most active of these new catalysts, a tris-urea H-bond donor, is among the most active organocatalysts known for ROP, yet it retains the high selectivity of H-bond mediated organocatalysts. The urea cocatalyst, along with an H-bond accepting base, exhibits characteristics of a “living” ROP, is highly active, in one case, accelerating a reaction from days to minutes, and remains active at low catalyst loadings. To understand the structure function relationship of the multi H-bonding (thio)ureas, a series of conformationally flexible bis(thio)urea H-bond donors plus base cocatalyst were applied to the ring-opening polymerization of lactones. The rate of the ROP displays a strong dependence upon the length and identity of the tether, where a circa five methylene- unit long tether exhibits the fastest ROP. The developing urea class of H-bond donors and the discovery of Triclocarban as a commericially available efficient catalysts for ROP, facilitates the solvent-free ROP of lactones at ambient and elevated temperatures, displaying enhanced rates and control versus other known organocatalysts for ROP under solvent-free conditions. One-pot block copolymerizations of lactide and valerolactone, which had previously been inaccessible in solution phase organocatalytic ROP, can be achieved under these reaction conditions, and one-pot triblock copolymers are also synthesized. For the ROP of lactide, however, thioureas remain the more effective H-bond donating class. For all (thio)urea catalysts under solvent-free conditions and in solution, the more active catalysts are generally more controlled. copolymerizations of lactide and valerolactone, which had previously been inaccessible in solution phase organocatalytic ROP, can be achieved under these reaction conditions, and one-pot triblock copolymers are also synthesized. For the ROP of lactide, however, thioureas remain the more effective H-bond donating class. For all (thio)urea catalysts under solvent-free conditions and in solution, the more active catalysts are generally more controlled. ABSTRACT With the discovery of multi H-bonding (thio)urea that is highly efficient and effective towards ROP, a novel bis-(thio)urea catalysts was developed as an enantio-selective chiral catalysts. This catalysts not only shows high stereoselectivity at mild conditions, but also show faster rates for the polymerization of rac-LA forming stereoblock PLA with precise control in molecular weight and enhanced thermal properties. ACKNOWLEDGEMENT I’d like to thank my parents, husband and the rest of my family for their constant support and motivation throughout my education career. I would like to thank Prof. Matthew Kiesewetter, for his endless encouragemet, continuous flow of projects that kept me busy and productive, his guidance and most of all for believing in me even more than I believed in my self. A big thank you to my committee members for their support in making this process smooth. I would also like to thank the past and present members in the Kiesewetter lab for making the work environment a place I look forward coming in to everyday, and for the continuous support. Thank you to the rest of the chemistry graduate community and the Department of Chemistry and Graduate school at the University of Rhode Island for all the support. -Nayanthara U. Dharmaratne iv iv PREFACE This dissertation is written in Manuscript Format. issertation is written in Manuscript Format. Chapter 1 : A review article that links the latest mechanistic advancements in H-bonding (thio)urea catalyst systems, in carrying out ROPs of cyclic lactones under have been presented. (Org. Biomol. Chem., 2019,17, 3305-3313) Chapter 2 : A combined experimental and computational investigation of hydrophobicity trend for oxygen containing functional groups commonly found in monomers and polymers have been presented. Analytical experiments for this study was conducted by Prof. Rob Mathers. Synthesis of oligomers and polymers for the study were conducted by me. (Macromolecules. 2018, 51(21), 8461-8468) Chapter 3: A study which reports the development of multi H-bond donating catalyst systems to perform ROP has been reported. Polymerization of lactide with these catalysts and studies on small molecule transformations were performed by me. (ACS Macro Lett. 2016, 5 (8), 982-986.). Chapter 4: Structure function relationship, by varying tether lengths of bis-(thio)urea catalysts has been reported. The relationship between acidity of thioureas and its activity in ROP of lactide was studied by me. (Macromolecules. 2019, 52(23), 9232-9237). v v Chapter 5: A study where a commercially available urea – triclocarban was discovered for its efficient catalytic activity in ROP of cyclic lactones in non-polar and polar solvents has been reported. (ACS Macro Lett. 2017, 6 (4), 421-425) Chapter 6: Application of (thio)urea catalyst systems under solvent-free conditions has been reported. Polymerizations of lactide under solvent free conditiond were conducted by me that revealed the activity of these catalysts at high temperatures. (Macromolecules 2017, 50 (22), 8948-8954) Chapter 7: The study reports the activity of (thio)urea catalysts at high temperatures and reveals different factors such as solvent, identity of co-catalysts that determine the activity of these catalysts at high temperature. Studies with lactide were conducted by me. (Macromolecules. 2018, 51(24), 10121-10126). Chapter 8: The study reports the development of a novel chiral bis-thiourea catalysts for enantioselective ROP of rac-LA. Preliminary catalysts screening was conducted by Oleg Kazakov. The manuscript is in progress. vi TABLE OF CONTENTS ABSTRACT………………………………………………………………………….......ii ACKNOWLEDGEMENTS…………………………………………………………….iv PREFACE………………………………………………………………………………..v TABLE OF CONTENTS……………………………………………………………….vii LIST OF TABLES……………………………………………………………………..viii LIST OF SCHEMES………………………………..……………………………….......x LIST OF FIGURES………………………………………………………………...…...xi CHAPTER 1……………………………………………………………………………...1 CHAPTER 2…………………………………………………………………………….30 CHAPTER 3………………………………………………………………………….....69 CHAPTER 4…………………………………………………………………………........................126 CHAPTER 5…………………………………………………………………………...175 CHAPTER 6…………………………………………………………………………...219 CHAPTER 7…………………………………………………………………………...264 CHAPTER 8…………………………………………………………………………...317 vii vii Table Table Page Table 2.1. Summation of hydrophobicity trends for small molecules and polymers…...55 Table 3.1. MTBD and 1-S, 2-S or 3-S catalyzed ROP of VL and CL ……….…….…...89 Table 3.2. 1-O, 2-O or 3-O and MTBD cocatalyzed ROP of lactones. ………….……...90 Table 3.3. Transesterification of Ethyl Acetate………………………………………….91 Table 3.4. Low 3-O/MTBD Cocatalyst Loadings in the ROP of VL……………………92 Table 3.5. Solvent Screen in the 3-O/MTBD Cocatalyzed ROP of VL……….…………93 Table 3.6. Post-polymerization Transesterification in 3-O/MTBD Cocatalyzed ROP….94 Table 4.1. Bis(thio)urea plus MTBD cocatalyzed ROP of VL.…………………..……..149 Table 4.2. Bis(thio)urea and MTBD cocatalyzed ROP of VL ………………..….……..150 Table 4.3. Bis(thio)urea Plus MTBD cocatalyzed ROP of VL in Acetone-d6 and Solvent- free conditions ………………………………..…………………………….151 Table 4.4. ROP of VL or CL cocatalyzed by MTBD plus Bisureas with heteroatom- containing tethers.……………………………………………...…………...152 Table 4.5. ROP of VL cocatalyzed by MTBD plus Bisthioureas with heteroatom- containing tethers…………………………………………………………...153 Table 4.6. Bis(thio)urea and Me6TREN cocatalyzed ROP of L-LA.…………………..154 Table 4.7. Mono(thio)urea and Me6TREN cocatalyzed ROP of L-LA ………………..155 Table 5.1. MTBD and TCC Catalyzed ROP of VL and CL ………………….…….….199 Table 5.2. Urea or Thiourea Plus MTBD Cocatalyzed ROP of VL in Acetone ………200 Table 5.3. Triclocarban Plus BEMP Cocatalyzed ROP of VL and CL ………………….201 viii Table 5.4. Chain Length Variation for the TCC or di-CC plus MTBD cocatalyzed ROP of VL ……………………………………………………………………..…...202 Table 5.5. Solvent Screen of TCC/MTBD cocatalyzed ROP of VL…………………...202 Table 5.6. Chain Length Variation for the TCC/MTBD cocatalyzed ROP of VL in acetone- d6. ……………………………………………………………………………203 Table 6.1. TCC plus base cocatalyzed ROP of VL…………………………………….243 Table 6.2. H-bond donor plus base cocatalyzed ROP of VL…………………………..244 Table 6.3. TCC plus base cocatalyzed ROP of macrolactones………………………...245 Table 6.4. H-bond Mediated Solvent-free ROP of L-LA………………………………246 Table 6.5. Solvent free ROP of VL with TCC/BEMP…………………………………247 Table 6.6. TCC plus MTBD or BEMP cocatalyzed ROP of CL……………….……...247 Table 6.7. Base Screen in the 2-S Mediated ROP of L-LA………………………….…248 Table 7.1. Activation Parameters for H-bond Donor/MTBD Cocatalyzed ROP of CL..284 Table 7.2. Thermal Decomposition of H-bond Donors with and without MTBD……..285 Table 7.3. Mn and Mw/Mn for polymerizations conducted at 80°C…………………….286 Table 8.1. Polymerization of rac-LA at room temperature …………………………....349 Table 8.2. Polymerization of rac-LA at -15o C………………………………………...349 ix LIST OF SCHEMES Scheme Page Scheme 1.1. H-bonding versus Imidate Mediated ROP of Lactones…………………...25 Scheme 1.2. Proposed activation pathway of covalently linked bifunctional thiourea in ROP of LA………………………………………………………………..25 Scheme 1.3. Bifunctional activation of monomer and initiator/chain end by TBD….…26 Scheme 1.4. Equilibrium between neutral versus imidate TCC with base……………..26 Scheme 1.5. ROP of epoxides and thiourea mediated conversion of alkoxide to alcohol and thioimdate for the ROP of lactones…………………………………..27 Scheme 2.1. Comparison of sulfur-based functional groups with ether and ketone analogues………………………………………………………………….56 Scheme 3.1. Highly active and highly selective H-bond donor 3-O…………………….95 Scheme 3.2. Proposed mechanism for 3-O/MTBD catalyzed ROP……………………..95 Scheme 4.1. Neutral H-bond versus imidate mediated ROP of VL…………………...155 Scheme 5.1. Proposed mechanism for TCC/base cocatalyzed ROP…………………..203 Scheme 6.1. Mechanism for the urea or thiourea plus base cocatalyzed ROP………...249 Scheme 8.1. Mechanism of catalysis…………………………………………………...350 x LIST OF FIGURES Figures Page Figure 1.1. Strength of cocatalyst binding is predictive of catalytic activity…………...28 Figure 1.2. Representative (thio)ureas in red and proposed activated-TU mode activation for multi-donors…………………………………………………………….28 Figure 1.3. Monomer activation by activated-3-O and intramolecular deactivation of 3-S. ………………………………………………………………………………29 Figure 1.4. The thiourea/base mediated ROP of lactones slows in polar solvent while urea/base mediated ROP remain active……………………………………..29 Figure 2.1. UV/Vis spectroscopy data for absorbance (lmax) of Nile Red in molecules with oxygen-containing functional groups……………………………………….56 Figure 2.2. Computational octanol-water partition coefficients (LogPoct) for cyclic and tricyclic monomers………………………………………………………….57 Figure 2.3. Correlation between experimental HPLC retention times and computational LogPoct values normalized by surface area………………………………….58 Figure 2.4. UV/Vis spectroscopy data for solvatochromatic behavior of Nile Red comparing a) cyclic carbonates, esters, and anhydrides with the corresponding acyclic molecule and b) evaluation of functional groups with and without vinyl groups. Lower lmax values indicate more hydrophobicity…………………..59 Figure 2.5. Comparison of hydrophobicity values for a) ethers and acetals and b) acetals, esters, and carbonates versus number of monomer units……………………60 xi Figure 2.6. UV/Vis spectroscopy data for solvatochromatic behavior of Nile Red comparing oligomers of PVL and PCL……………………………………..61 Figure 2.7. Electronic effect of benzene reduced the gap between carbonate and anhydride functional groups……………………………………………………………62 Figure 2.8. Comparison of hydrophobicity values for polyesters with different C:O ratios. ...…………………………………………………………………………….63 Figure 2.9. Contact angle of static drop on Teflon coated septa and Restek butyl rubber septa for liquid carbonates and anhydrides………………………………….64 Figure 2.10. TGA data and first derivative for PCL……………………………………..65 Figure 2.11. TGA data and first derivative for PVL……………………………………..66 Figure 2.12. Differential scanning calorimetry data for PCL……………………………67 Figure 2.13. Differential scanning calorimetry data for PVL…………………………...68 Figure 3.1. Base and (thio)urea cocatalysts evaluated for ROP……………………….96 Figure 3.2. Mn versus conversion for the 2-S/MTBD catalyzed ROP of VL………….96 Figure 3.3. First order evolution of [VL] versus time for the 2-S/MTBD catalyzed ROP of VL………………………………………………………………………97 Figure 3.4. Mn versus [VL]o/[I]o for the 2-S/MTBD catalyzed ROP of VL…………..97 Figure 3.5. GPC traces of the polymer resulting from the 2-S/MTBD cocatalyzed ROP and subsequent chain extension of VL (from 1-pyrenebutanol……………98 Figure 3.6. Observed rate constant versus [MTBD] in the 2-S/MTBD catalyzed ROP of VL…………………………………………………………………………99 Figure 3.7. Observed rate constant versus [MTBD] in the 2-S/MTBD catalyzed ROP of CL…………………………………………………………………………99 xii xii Figure 3.8. Mn vs conversion of VL for the 3-O/MTBD catalyzed ROP of VL………100 Figure 3.9. First order evolution of [VL] vs time for the 3-O/MTBD catalyzed ROP of VL………………………………………………………………………..100 Figure 3.10. LIST OF FIGURES Mn vs conversion for the 3-O/MTBD catalyzed ROP of CL……………..101 Figure 3.11. First order evolution of [CL] vs time for the 3-O/MTBD catalyzed ROP of CL………………………………………………………………………..101 Figure 3.12. First order evolution of [CL] and [VL] vs time for the 3-O/MTBD catalyzed copolymerization of CL………………………………………………….102 Figure 3.13. GPC traces of the polymer resulting from the 3-O/MTBD cocatalyzed ROP and subsequent chain extension of CL from 1-pyrenebutanol………….102 Figure 3.14. Observed rate constant (kobs, min-1) vs [3-O] in the 3-O/MTBD catalyzed ROP of VL……………………………………………………………….103 Figure 3.15. DFT B3LYP//6-31G geometry optimized structures of 3-S……………104 Figure 3.16. DFT B3LYP//6-31G geometry optimized structures of 3-O……………104 Figure 3.17. MALDI-TOF of the PLA resulting from ROP of L-LA………………….105 Figure 3.18. 1H NMR spectra of 1-O, 2-O, and 3-O………………………………….106 Figure 3.19. 1H NMR (300 MHz, DMSO-d6) of 2-O…………………………………..107 Figure 3.20. 13C NMR (75 MHz, DMSO-d6) of 2-O……………………………………108 Figure 3.21.1H NMR (300 MHz, DMSO-d6) of 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane………………………………………………...109 Figure 3.22. 1H NMR (300 MHz, DMSO-d6) of 2-OS………………………………….110 Figure 3.23. 13C NMR (75 MHz, acetone-d6) of 2-OS………………………………….111 Figure 3.24. 1H NMR (300 MHz, acetone-d6) of 3-O…………………………………..112 xiii Figure 3.25. 13C NMR (75 MHz, acetone-d6) of 3-O…………………………………..113 Figure 3.26. 1H NMR (300 MHz, acetone-d6) of 3-S…………………………………..114 Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S…………………………………...115 Figure 4.1. Mono(thio)urea, bis(thio)urea donors evaluated and proposed activated- thiourea mode activation for bis-(thio)urea H-bond donors………………156 Figure 4.2. Mn and Mw/Mn versus conversion for the H-bond donor plus MTBD cocatalyzed ROP of VL using 2-S5 and 2-O5………………………….…156 Figure 4.3. Proposed activated (thio)urea anion mechanism for the bisurea plus MTBD mediated ROP of VL……………………………………………………...157 Figure 4.4. 1H NMR spectra of 2-O5 plus MTBD in acetone-d6……………………..157 Figure 4.5. First order evolution of VL versus time for the 2-O5/MTBD catalyzed ring- opening polymerization of VL……………………………………………158 Figure 4.6. Mn and Mw/Mn versus conversion for 2-O5 catalyst…………………..….158 Figure 4.7. Mn and Mw/Mn versus conversion for 2-O5-O catalyst…………………...159 Figure 4.8. Mn and Mw/Mn versus conversion for 2-S5, Mn versus conversion for 2-O5 catalyst……………………………………………………………………159 Figure 4.9. 1 H NMR spectra of 2-O5-O and 2-S5-O with and without Me6TREN in acetone- d6………………………………………………………………..160 Figure 4.10. NMR spectrum of (2-O2)……………………………………………….161 Figure 4.11. NMR spectrum of (2-O4)……………………………………………….162 Figure 4.12. NMR spectrum of (2-O5)……………………………………………….163 Figure 4.13. NMR spectrum of (2-O6)……………………………………………….164 Figure 4.14. NMR spectrum of (2-O12)…………………………………………..….165 xiv Figure 4.15. NMR spectrum of (2-O5-N)…………………………………………….166 Figure 4.16. NMR spectrum of (2-O5-O)…………………………………………….167 Figure 4.17. NMR spectrum of (2-O3-diMe)…………………………………………168 Figure 4.18. NMR spectrum of (2-S4)………………………………………………...169 Figure 4.19. NMR spectrum of (2-S5)………………………………………………...170 Figure 4.20. LIST OF FIGURES NMR spectrum of (2-S6)………………………………………………...171 Figure 4.21. NMR spectrum of (2-S12)…………………………………………….....172 Figure 4.22. NMR spectrum of (2-S5-N)……………………………………………...173 Figure 4.23. NMR spectrum of (2-S5-O)…………………………...………………...174 Figure 5.1. Base and (thio)urea cocatalysts evaluated for ROP………………………204 Figure 5.2. First order evolution of VL vs time for the TCC/MTBD catalyzed ring- opening polymerization of VL……………………………………………204 Figure 5.3. Mn and Mw/Mn vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL……………………………………………………..205 Figure 5.4. GPC traces of the polymers resulting from the chain extension experiment of VL………………………………………………………………………...205 Figure 5.5. First order evolution of CL vs time for the TCC/MTBD catalyzed ring- opening polymerization of CL……………………………………………206 Figure 5.6. Mn and Mw/Mn catalyzed ring-opening polymerization of CL……......…..206 Figure 5.7. Approach to equilibrium evolution of [VL] vs time for the TCC/MTBD catalyzed ring-opening polymerization of VL……………………………207 Figure 5.8. Mn and Mw/Mn vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL……………………………………………………..207 xv Figure 5.9. GPC traces of the polymers resulting from the chain extension of PVL in acetone………………………………………………………………….208 Figure 5.10. Mn and Mw/Mn vs conversion for the TCC/BEMP catalyzed ring-opening polymerization of VL………………………………………………….209 Figure 5.11. GPC traces of the polymers resulting from the chain extension experiment of VL……………………………………………………...209 Figure 5.12. First order evolution of [L-LA] vs time for the TCC/Me6TREN catalyzed ring-opening polymerization…………………………………………..210 Figure 5.13. Mn and Mw/Mn vs conversion for the TCC/Me6TREN catalyzed ring- opening polymerization of L-LA………………………………………211 Figure 5.14. MALDI-TOF of the PLLA resulting from TCC/Me6TREN cocatalyzed ROP of L-lactide……………………………………………………...212 Figure 5.15. Titration binding curve for the CL/1-O binding in benzene-d6………213 Figure 5.16. Titration binding curve for the CL/1-S binding in benzene-d6……….213 Figure 5.17. Methine region of the methyl-decoupled 1H NMR spectrum of PLLA obtained via TCC/Me6TREN cocatalyzed ROP of L-LA…………….214 Figure 5.18. First order evolution of [VL] vs time for the di-CC/MTBD catalyzed ROP of VL……………………………………………………………215 Figure 5.19. NMR spectrum of mono-CC………………………………………...216 Figure 5.20. NMR spectrum of di-CC………………………………………….....217 Figure 5.21. 1H NMR spectra of TCC plus base in acetone-d6…………………….218 Figure 6.1. First order evolution of [monomer] versus time and Mn and Mw/Mn versus conversion for the TCC/MTBD co catalyzed ROP of VL…………..250 xvi Figure 6.2. RI and UV GPC traces of the ROP initiated from pyrenebutanol for PVL (TCC/BEMP) and PLA (2-S/PMDTA)………………………….................251 Figure 6.3. First order evolution of [monomer] versus time; and Mn and Mw/Mn versus conversion for the TCC/BEMP co catalyzed ROP of VL…………………..252 Figure 6.4. LIST OF FIGURES First order evolution of [monomer] versus time and Mn and Mw/Mn versus conversion for the TCC/BEMP co catalyzed ROP of CL………………….253 Figure 6.5. Percent conversion to polymer of VL solutions of MTBD, BEMP and TCC plus benzyl alcohol at -10°C and room temperature……………………….254 Figure 6.6. Solvent-free ROP allows for the direct-from-monomer creation of a negative mold………………………………………………………………………..255 Figure 6.7. 400 MHz 1H NOESY in acetone-d6 of TCC/MTBD and TCC/BEMP)…..256 Figure 6.8. Mn and Mw/Mn vs conversion for the TCC/BEMP co catalyzed ROP of EB………………………………………………………………………….257 Figure 6.9. First order evolution of [LA] vs time, and Mn and Mw/Mn vs conversion...258 Figure 6.10. First order evolution of [LA] vs time for the 2-S/PMDTA cocatalyzed ROP LA………………………………………………………………………..259 Figure 6.11. First order evolution of [VL] and [CL] vs time for the one-pot copolymerization catalyzed by TCC/BEMP…………………………….260 Figure 6.12. First order evolution of [monomer] vs time for the copolymerization of EB………………………………………………………………………..261 Figure 6.13. NMR of poly(VL-co-CL-co-EB)…………….…………………………..262 Figure 6.14. First order evolution of [monomer] vs time for the copolymerization of VL and L-LA…………………………………………………………………263 xvii Figure 6.15. Percentage conversion of EB vs time for the copolymerization of: VL, CL and EB…………………………………………………………………263 Figure 7.1. H-bonding and imidate mediated ROP of CL………….………………287 Figure 7.2. Eyring plots for the ROP of CL from benzyl alcohol in toluene catalyzed by TCC/MTBD, and 2-O/MTBD…………………………………………288 Figure 7.3. Eyring plot for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene and MIBK solvent……………………………………………289 Figure 7.4. Eyring plot for the solvent-free ROP of LA from benzyl alcohol catalyzed by 2-S/PMDETA……………………………………………………………289 Figure 7.5. Mn and Mw/Mn versus conversion plot for the 3-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………………….290 Figure 7.6. Temperature drop ROP of CL……………………………………………291 Figure 7.7. 1H NMR spectra of TCC/MTBD cocatalyst in the presence of varying [CL] in CDCl3…………………………………………………………………….292 Figure 7.8. Eyring plot constructed from the linear portion of the observed first order rate constants for the 2-S/MTBD co catalyzed ROP of CL in toluene………..293 Figure 7.9. Eyring plot constructed from the linear portion of the observed first order rate constants for the 1-O/MTBD co catalyzed ROP of CL in toluene………..293 Figure 7.10. Eyring plot constructed from the linear portion of the observed first order rate constants for the TCC/MTBD co catalyzed ROP of CL in toluene….294 Figure 7.11. Eyring plot constructed from the linear portion of the observed first order rate constants for the 1-S/MTBD co catalyzed ROP of CL in toluene……294 xviii Figure 7.12. LIST OF FIGURES Eyring plot constructed from the linear portion of the observed first order rate constants for the 3-S/MTBD co catalyzed ROP of CL in toluene…...295 Figure 7.13. Eyring plot constructed from the linear portion of the observed first order rate constants for the 2-O/MTBD co catalyzed ROP of CL in toluene…..295 Figure 7.14. Eyring plot constructed from the linear portion of the observed first order rate constants for the 3-O/MTBD co catalyzed ROP of CL in toluene…..296 Figure 7.15. Eyring plot constructed from the linear portion of the observed first order rate constants for the TCC/DBU co catalyzed ROP of CL in toluene…..296 Figure 7.16. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene…..297 Figure 7.17. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene……....297 Figure 7.18. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD cocatalyzed ROP of CL from benzyl alcoholat 1M, 2M and 3M concentrations in toluene…………………………………………………298 Figure 7.19. Eyring plot constructed from the observed first order rate constants for the 2- S/PMDETA cocatalyzed ROP of L-LA from benzyl alcohol in methyl isobutyl ketone……………………………………………………………299 Figure 7.20. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone………………………………………………….………………….300 xix Figure 7.21. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone……………………………………………………………………..300 Figure 7.22. Eyring plot constructed from the observed first order rate constants for the 2- S/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone……………………………………………………………………..301 Figure 7.23. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone……………………………………………………………….…….301 Figure 7.24. Eyring plot constructed from the observed first order rate constants for the 3- O/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone……………………………………………………………………..302 Figure 7.25. Eyring plot constructed from the observed first order rate constants for the 1- O/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone……………………………………………………………………..302 Figure 7.26. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in benzene…..303 Figure 7.27. LIST OF FIGURES First order evolution of [CL] versus time for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….303 Figure 7.28. First order evolution of [CL] versus time for the 1-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….304 Figure 7.29. First order evolution of [CL] versus time for the 1-S/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C……………………304 xx xx Figure 7.30. First order evolution of [CL] versus time for the 2-S/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 80°C…………………...…305 Figure 7.31. First order evolution of [CL] versus time for the 3-S/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….305 Figure 7.32. First order evolution of [CL] versus time for the 2-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….306 Figure 7.33. First order evolution of [CL] versus time for the 3-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….306 Figure 7.34. First order evolution of [CL] versus time for the TCC/DBU cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….307 Figure 7.35. First order evolution of [CL] versus time for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol toluene at 110°C……………………….307 Figure 7.36. First order evolution of [CL] versus time for the 2-O/MTBD ROP of CL from benzyl alcohol in toluene at 110°C…………………………………308 Figure 7.37. First order evolution of [CL] versus time for the 1-S/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C…………………….308 Figure 7.38. First order evolution of [CL] versus time for the 1-S/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C……………………………309 Figure 7.39. First order evolution of [CL] versus time for the 1-S/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone at 110°C……310 Figure 7.40. First order evolution of [CL] versus time for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone at 110°C……310 xxi Figure 7.41. First order evolution of [CL] versus time for the 2-S/MTBD cocatalyzed ROP of CL from benzyl alcohol methyl isobutyl ketone at 60°C..………311 Figure 7.42. First order evolution of [CL] versus time for the 2-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone at 110°C……..311 Figure 7.43. LIST OF FIGURES First order evolution of [CL] versus time for the 1-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone at 110°C……..312 Figure 7.44. First order evolution of [CL] versus time for the 3-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in methyl isobutyl ketone at 110°C……312 Figure 7.45. First order evolution of [CL] versus time for the TCC/MTBD cocatalyzed ROP of CL from benzyl alcohol in benzene at 110°C……………………313 Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80°C……………………314 Figure 7.47. Mn and Mw/Mn versus conversion plot for 1-S at 110°C…………………...314 Figure 7.48. Mn and Mw/Mn versus conversion plot for 1-S at 110°C………………….315 Figure 7.49. Mn and Mw/Mn Versus Conversion plot for 2-O at 110°C………………….315 Figure 7.50. Mn and Mw/Mn Versus Conversion plot for 2-O at 90°C………………….316 Figure 7.51. Mn and Mw/Mn versus conversion plot for 2-O at 90°C……………………316 Figure 8.1. Microstructures of PLA formed by the polymerization of rac-LA…………350 Figure 8.2. Mechanisms of stereocontrolled ROP………………………………………351 Figure 8.3. Thioureas and bases used in this study………………………………………351 Figure 8.4. First order plots for the polymerization of D-LA, rac-LA and L-LA with TU1 and TU8 with Me6TREN…………………………………………………352 Figure 8.5. First order plots for the polymerization of D-LA, rac-LA and L-LA respectively with TU1/ BEMP…………………………………………………………...352 Figure 8.6. RI and UV GPC traces of the PLA initiated by 1-pyrenebutanol…..…..….353 xxii xxii Figure 8.7. Mn and Mw/Mn catalyzed ring-opening polymerization of rac-LA………353 Figure 8.8. MALDI-TOF of the PLA resulting from TU1/Me6TREN cocatalyzed ROP of rac-LA………………………………..……………………………………..354 Figure 8.9. Homonuclear decoupled 1H NMR spectrum of the methine region of PLA...354 Figure 8.10. DSC thermograms of PLA…………………………………………………355 Figure 8.11. DSC thermograms of PLA…………………………………………………355 Figure 8.12. HPLC chromatograms of L-LA, D-LA. rac-LA and the unreacted monomer at 47 % monomer conversion……………………………………………….356 Figure 8.13. Downfield portion of 1 H NMR spectra of TU1-TU4 with and without Me6TREN in CH2Cl2……………………………………………………..357 Figure 8.14. NMR spectrum of TU1…………………………………………..……….358 Figure 8.15. NMR spectrum of TU2…………………………………………..……….359 Figure 8.16. NMR spectrum of TU3…………………………………………..……….360 Figure 8.17. NMR spectrum of TU5…………………………………………..……….361 Figure 8.18. NMR spectrum of TU6…………………………………………..……….362 Figure 8.19. NMR spectrum of TU7…………………………………………..……….363 Figure 8.20. NMR spectrum of TU9…………………………………………..……….364 …353 xxiii MANUSCRIPT – I Submitted for publication to Organic & Biomolecular Chemistry INTRODUCTION Organocatalysis for polymer synthesis has come to be synonymous with the construction of precisely tailored materials through the ring-opening polymerization (ROP) of esters, carbonates and other cyclic monomers.1–10 While organocatalysts have gained a beachhead in the synthesis of other polymers,11 a host of organic systems for transesterification polymerization have been developed. 1,2,12–17 The purview of organocatalysts for polymerization are ‘living’ ROP. A living ROP is a type of chain growth polymerization characterized by the lack of chain transfer and termination events – a kinetic definition.18 Organocatalysis for polymer synthesis has come to be synonymous with the construction of precisely tailored materials through the ring-opening polymerization (ROP) of esters, carbonates and other cyclic monomers.1–10 While organocatalysts have gained a beachhead Organocatalysis for polymer synthesis has come to be synonymous with the construction of precisely tailored materials through the ring-opening polymerization (ROP) of esters, carbonates and other cyclic monomers.1–10 While organocatalysts have gained a beachhead in the synthesis of other polymers,11 a host of organic systems for transesterification polymerization have been developed. 1,2,12–17 The purview of organocatalysts for polymerization are ‘living’ ROP. A living ROP is a type of chain growth polymerization characterized by the lack of chain transfer and termination events – a kinetic definition.18 A controlled, ‘living’, polymerization is one that features predictable molecular weights (Mn) and molecular weight distributions close to unity (dispersity= Ð = Mw/Mn) and are capable of yielding polymers with well-defined architectures.1,2,12 The selectivity of catalysts towards ROP versus non-enchainment reactions is vital to minimizing the molecular weight distribution.1,2,12,19,20 Additionally, functional group tolerance,1,2,21–24 activity of catalysts under a wide range of temperatures 11,25–29 and pressure, 30,31 and a variety of solvents and solvent-free conditions32 facilitate the implementation of diverse reaction conditions which facilitates advanced polymer design. A controlled, ‘living’, polymerization is one that features predictable molecular weights (Mn) and molecular weight distributions close to unity (dispersity= Ð = Mw/Mn) and are capable of yielding polymers with well-defined architectures.1,2,12 The selectivity of catalysts towards ROP versus non-enchainment reactions is vital to minimizing the molecular weight distribution.1,2,12,19,20 Additionally, functional group tolerance,1,2,21–24 activity of catalysts under a wide range of temperatures 11,25–29 and pressure, 30,31 and a variety of solvents and solvent-free conditions32 facilitate the implementation of diverse reaction conditions which facilitates advanced polymer design. Among the various catalysts for ROP, H-bonding organocatalysts stand out in the precise level of reaction control they are able to render during ROP. ABSTRACT Among the various catalysts for ROP, H-bonding organocatalysts stand out in the precise level of reaction control they are able to render during ROP. The H-bonding class of organocatalysts are thought to effect ROP via dual activation of both monomer and chain end. (Thio)urea mediated ROP has experienced a renaissance as a new polymerization mechanism – mediated by imidate or thioimidate species – facilitates new modes of reactivity and new synthetic abilities. Indeed, the urea class of H-bond donors have shown to be more active than their corresponding thioureas. The imidate mechanism remains highly active in polar solvents and exhibits remarkable control – and ‘living’ behavior - under solvent-free conditions, and a broad range of temperatures is accessible. The advancements in synthetic abilities have all evolved through a greater understanding of reaction mechanism. Through the continued synergistic advances of catalysis and material, the (thio)urea class of catalyst can find use in a host of potential applications, research and industrial environments. 2 INTRODUCTION The (thio)urea H-bonding class of organocatalysts are thought to effect ROP via dual activation of both monomer and chain end, Scheme 1.1.1,2,12,33 In this approach, a typical catalyst system consisting of a thiourea (TU) and base cocatalyst can render high functional group tolerance and yield polymers with predictable molecular weights and narrow Mw/Mn.1,2,4,34–36 Despite the high 3 3 selectivity shown by this class of catalyst, the major disadvantage had been the slow rates for ROP.34,35,37 The development of advanced catalyst systems continues apace, this shortcoming has largely been mitigated. Indeed, (thio)urea H-bond mediated ROP has experienced a renaissance as a new polymerization mechanism – mediated by imidate or thioimidate species – facilitates new synthetic abilities and new modes of reactivity, Scheme 1. It should be noted that there are many structural manifestations of H-bond mediated catalysts,11,38–41 but this review is narrowly focused on the evolution of the (thio)urea/base cocatalyst system as it pertains to the ROP of δ-valerolactone (VL), e- caprolactone (CL) and lactide (LA), in particular. INTRODUCTION 4 4 THIO)UREA H-BOND MEDIATED RING-OPENING POLYMERIZATION The naissance of H-bond mediated ROP occurred in 2005 when the Takemoto thiourea (Figure 1) was applied for the polymerization of LA.4 This unimolecular, bifunctional catalyst consists of an H-bond donating moiety and an H-bond accepting moiety that can activate monomer and initiator/chain end, respectively (Scheme 2), yielding, in addition to typical ‘living’ behavior, highly selective ROP with minimum broadening of Mw/Mn even at monomer conversions ³ 95%.4 However, the reaction time is protracted (2 days), and ROP is most effective in non-H-bonding solvents.4 This study cemented common themes among H-bond mediated catalysts for ROP: a 3,5-bistrifluoromethyl aryl group for its electron withdrawing abilities and a cyclohexyl group, which is not required versus other alkyl groups for catalysis.4 Amazingly, this study also revealed similar catalytic activity to the bifunctional Takemoto catalyst when bimolecular catalysts were employed; the thiourea 1-S plus N,N-dimethylcyclohexylamine (NCyMe2) cocatalyzed ROP of LA demonstrated that covalently tethering the H-bond donor and acceptor is not essential.4,8 A base screen conducted using 1-S and commercially available bases revealed (-)-sparteine to exhibit the highest activity, achieving 95% conversion of LA in 2 h (25-fold faster than the parent system), producing PLA with minimal epimerization and narrow Mw/Mn.8 Thiourea plus alkylamine base cocatalysts are limited to the ROP of LA.42 F H b d di t d ROP t i b t l t i d ith TU f th A base screen conducted using 1-S and commercially available bases revealed (-)-sparteine to exhibit the highest activity, achieving 95% conversion of LA in 2 h (25-fold faster than the parent system), producing PLA with minimal epimerization and narrow Mw/Mn.8 Thiourea plus alkylamine base cocatalysts are limited to the ROP of LA.42 For H-bond mediated ROP, strong organic base cocatalysts are required with TUs for the ROP of lactones other than lactide.35 The guanidine base N-methyl-1,5,7- triazabicyclo[4.4.0]dec-5-ene (MTBD) and amidine base 1,8-diazabicyclo[5.4.0]- undec- 7-ene (DBU) are only active for ROP of VL and CL from alcoholic initiators in the presence of 1-S. INTRODUCTION Under typical reaction conditions (2M monomer, 5 mol% cocatalysts), 5 5 the MTBD or DBU plus 1-S cocatalyzed ROP of VL ([M]o/[I]o = 100) reached full conversion in ~4 h, and the polymerization of CL was much slower (full conversion in 2- 5 days).42 Although slower than other catalyst systems, these cocatalysts are highly controlled, leading to polymers with narrow molecular weight distributions (Mw/Mn ≤ 1.08), predictable molecular weights up to [M]o/[I]o = 200 with good end group fidelity.42 The selectivity of these catalysts for monomer versus polymer could be ascribed to the high affinity of thiourea for s-cis esters (lactones) in contrast with negligible binding to s-trans esters (i.e. polymer backbone).1,2,42 In general, thioureas featuring aryl rings with strong electron withdrawing groups result in faster rates, but the trend is not robust.43 Further, enhanced H-bonding to base cocatalyst will attenuate catalytic activity.16,35 Mechanistic studies on the thiourea/alkylamine base mediated ROP of LA informed the development of advanced catalyst systems for ROP. Kinetic studies on the ROP of LA cocatalyzed by 1-S and certain alkylamine bases (i.e. not all cocatalyst combinations) revealed second order dependence on [1-S]o; a mechanistic account was proposed.44 As a direct result, the bisthiourea 2-S was synthesized and applied with base cocatalysts for the ROP of lactide, which resulted in enhanced rates (k2-S/k1-S ~12).37 Unexpectedly, the application of 2-S (plus base cocatalyst) results in rate accelerated ROP versus 1-S for all base cocatalysts and monomers examined, regardless of kinetics for the analogous 1-S system. The 2-S plus base cocatalyzed ROPs of LA and lactones exhibit similar rate equations (Rate = kobs[M]; kobs = [2-S + base]o[initiator]o) which suggests that 2-S is acting as a discrete catalyst (one bisthiourea per base per monomer in the transition state).34,37 Buttressed by computational studies37 and indirect evidence,34 an activated-TU mechanism was proposed, whereby the ‘extra’ thiourea stabilizes the catalytic thiourea via H-bond 6 6 activation, Figure 2. The ROP of cyclic lactones in the presence of 2-S and base proceeded with lower catalyst loadings and enhanced rates compared to monothiourea 1-S, yet selectivity and control are retained.37 Since 2-S proved to be superior to 1-S in all comparisons, an obvious question becomes, what about a tristhiourea? INTRODUCTION However, the tristhiourea 3-S is markedly inactive for ROP.34 This observation was attributed to intramolecular H-bonding between all three thiourea moieties, generating a C3 symmetric structure, rendering all thioureas inaccessible for catalysis.34 Computational studies suggested that contracting the length of the H-bond donor moieties by changing C=S to C=O would break the C3 symmetry and result in the generation of a ‘frustrated’ system that cannot form a completed, intramolecular H-bonded network, thereby liberating a urea moiety for monomer activation (Figure 3). The prediction proved prophetic, and the trisurea 3-O proved to be the gateway, at least for our group, to the incredibly active imidate mediated polymerizations. The ROP of VL with 3-O/MTBD achieved full conversion 25 times faster than with 2- S/MTBD, producing PVL in 3 min (Mn = 7.5 kDa, Mw/Mn = 1.07). The ROP of CL with 3-O/MTBD was slower but was completed in 30 min compared to 10 h or 45 h with 2- S/MTBD or 1-S/MTBD, respectively. These transformations were not only more rapid but proceeded with high control, exhibiting the characteristics of a ‘living’ ROP. A comparative study conducted for the ROP of CL (2M from benzyl alcohol, M/I = 50) with the highly-active base 1,5,7-triazabicyclodec-5-ene (TBD), a go-to commercially available organocatalyst for ROP,45 versus 3-O/MTBD displayed the superior ROP abilities of the nascent urea H-bond donors: 3-O/MTBD (33 mM) 26 min, 97% conversion, Mw/Mn = 1.05; TBD (33 mM) 140 min, 93% conversion, Mw/Mn = 1.37.34 The marked success of 3- 7 7 7 O mediated ROP suggested that other urea H-bond donors would be active as well. Indeed, the monourea 1-O and bisurea 2-O were more active than the analogous thiourea H-bond donors when applied with a base cocatalyst for ROP.34 A commercially available monourea H-bond donor, triclocarban (TCC), exhibits catalytic rates and selectivities for all lactone monomers that rival that of 3-O. IMIDATE MEDIATED RING-OPENING POLYMERIZATION IMIDATE MEDIATED RING-OPENING POLYMERIZATION Consideration of the enchainment mechanism of the highly-active organocatalyst TBD provides a point of comparison for the enchainment mechanism of the nascent urea/base mediated ROP. TBD is highly active for a wide range of monomers; however, TBD- catalyzed ROP have been observed to lack selectivity and control especially at high monomer conversions.1,2,7,42 Mechanistically, the lowest energy enchainment pathway has been computationally and experimentally suggested to be H-bonding, where TBD acts as a bifunctional molecule activating both monomer and chain end (Scheme 3).1,2,49,50 The mode of activity displayed by TBD serves a as an analogy to the advance made by Waymouth and coworkers whereby a thiourea is treated with a strong base to form a thioimidate species, which is highly-active for ROP, Scheme 3.35,51 The treatment of a thiourea with strong bases like sodium and potassium methoxides form the thioimidate salt and an alcohol which can be used as catalyst/initiator systems for the ROP of lactones. The thioimidate (anionic thiourea) can function both as an H-bond donor and acceptor similar to TBD (Scheme 3).51 When the ROP of LA ([LA]o/[NaOCH3]o = 200) was conducted using 1-10 equivalents of thiourea to NaOCH3, monomer conversion >90% was achieved in ≤ 6 min, where faster rates were seen with lower amounts of thiourea.51 However, a molar excess of thiourea to base was vital to minimize the molecular weight distribution of the PLA (Mw/Mn = 1.55 to 1.18). The identity of the alkoxide counterion was shown to influence the selectivity of ROP, and slower rates but enhanced selectivity were observed with K+ versus Na+.51 The ROPs with thiourea/alkoxides showed characteristics of ‘living’ polymerizations. The transformations were controlled and highly selective compared to ROPs mediated by alkoxides alone, producing highly isotactic PLA with predictable Consideration of the enchainment mechanism of the highly-active organocatalyst TBD provides a point of comparison for the enchainment mechanism of the nascent urea/base mediated ROP. INTRODUCTION For reasons that are not entirely clear, the 2- S/alkylamine system remains the more active and controlled system for the ROP of lactide.11,32 The lower solubility of urea versus thiourea H-bond donors had restricted their application as catalysts, but almost all urea cocatalysts examined are fully soluble in the presence of base and/or monomer.34,46 The initial reports of urea plus base cocatalyst mediated ROP showed that stronger organic bases yielded more active ROP.34,47 This result stands in contrast to that of thiourea mediated ROP where catalytic activity is related to the binding between the cocatalysts (see above);44,48 this may have been the first indication that a different mechanism of enchainment was operative. The initial reports also disclosed that urea H-bond donors remain active in polar solvent, Figure 4. This result was particularly surprising giving the large suppression of rate in polar solvent (e.g. THF) displayed by thiourea H-bond donors.4,8,34,47 Again, it was becoming apparent that a new mechanism was engendering abilities that were historically out of reach. 8 8 IMIDATE MEDIATED RING-OPENING POLYMERIZATION TBD is highly active for a wide range of monomers; however, TBD- catalyzed ROP have been observed to lack selectivity and control especially at high monomer conversions.1,2,7,42 Mechanistically, the lowest energy enchainment pathway has been computationally and experimentally suggested to be H-bonding, where TBD acts as a bifunctional molecule activating both monomer and chain end (Scheme 3).1,2,49,50 The mode of activity displayed by TBD serves a as an analogy to the advance made by Waymouth and coworkers whereby a thiourea is treated with a strong base to form a thioimidate species, which is highly-active for ROP, Scheme 3.35,51 The treatment of a thiourea with strong bases like sodium and potassium methoxides form the thioimidate salt and an alcohol which can be used as catalyst/initiator systems for the ROP of lactones. The thioimidate (anionic thiourea) can function both as an H-bond donor and acceptor similar to TBD (Scheme 3).51 When the ROP of LA ([LA]o/[NaOCH3]o = 200) was conducted using 1-10 equivalents of thiourea to NaOCH3, monomer conversion >90% was achieved in ≤ 6 min, where faster rates were seen with lower amounts of thiourea.51 However, a molar excess of thiourea to base was vital to minimize the molecular weight distribution of the PLA (Mw/Mn = 1.55 to 1.18). The identity of the alkoxide counterion was shown to influence the selectivity of ROP, and slower rates but enhanced selectivity were observed with K+ versus Na+.51 The ROPs with thiourea/alkoxides showed characteristics of ‘living’ polymerizations. The transformations were controlled and highly selective compared to ROPs mediated by alkoxides alone producing highly isotactic PLA with predictable Consideration of the enchainment mechanism of the highly-active organocatalyst TBD provides a point of comparison for the enchainment mechanism of the nascent urea/base mediated ROP. TBD is highly active for a wide range of monomers; however, TBD- catalyzed ROP have been observed to lack selectivity and control especially at high monomer conversions.1,2,7,42 Mechanistically, the lowest energy enchainment pathway has 9 9 molecular weights and minimal epimerization. The adaptability of this system was shown by its efficacy in ROP of VL and CL. Computational and mechanistic studies indicate that the active catalyst species is characterized by the metal ion complexed to S, and a mode of enchainment was proposed, Scheme 3. IMIDATE MEDIATED RING-OPENING POLYMERIZATION The larger association constant for the binding of TU-K+/HOtBU to VL (24±4 M-1) versus ethyl acetate (5±2 M-1) indicates that the selectivity of the ROP is rendered by the different binding of the anionic adduct to the cyclic lactone versus the open chain ester.51 The treatment of a urea H-bond donor with a strong base form a urea anion which is incredibly active and controlled in the ROP of lactones. One method of generating the urea anion (imidate) is to employ a strong inorganic base, alkoxide (e.g. KOCH3) or hydride (e.g. KH). In the latter method, an ex situ alcohol initiator can be introduced. Just as neutral urea catalysts were previously shown to be much more active than their thiourea counterparts in performing ROP, Figure 4,34,47 the urea anions are much faster than the corresponding thiourea anions.35 The slowest imidate was not only 25 times faster than the analogous thioimidate, but also exhibited enhanced selectivity.35,51 Kinetic studies indicated first order behavior in [monomer] and[initiator]o and inverse first order dependence on urea when [alkoxide]o ≤ [urea]o. This was suggested to be a result of reversible neutral urea:imidate dimer formation which could inhibit catalytic activity.35,52,53 This study also revealed a correlation between the pKa of the urea or thiourea and its activity, where ureas with lower acidity form more active (basic) imidates/thioimidates.35 Hence, ureas featuring more or stronger electron withdrawing groups produce urea anions that are less active for ROP, and an imidate is more active than its analogous thioimidate due to the increased acidity of thioureas versus ureas.34,35,43,47,54 The high selectivity and 10 versatility of imidates, unlike TBD, was attributed to the ability to fine tune the basicity and H-bond donating ability by changing the substituent groups on the phenyl ring.35,51 An N-methylated monofunctional urea exhibited slower rates and decreased selectivity compared to ureas featuring two N-H donors, suggesting a bifunctional mode of activation (c.f. TBD) is preferred.35 These hyperactive imidate mediated H-bonding catalysts were reported to be faster and more selective than other organocatalysts, resembling some metal- containing catalysts in their activity.1,2,14,34,35,51 When ureas or thioureas are subjected to strong organic bases, an equilibrium is established between neutral H-bond mediated ROP and the more active imidate mechanism. Our group studied the mechanism of TCC/base mediated ROP, and a simple 1H NMR experiment of TCC with and without base cocatalyst proved highly diagnostic. IMIDATE MEDIATED RING-OPENING POLYMERIZATION Imidate formation is indicated by an upfield shift of TCC resonances in the presence of base, and cocatalyst H-bonding is indicated by the downfield shift of TCC resonances in the presence of base.47 The equilibrium between neutral urea and imidate species (Scheme 4) shifts more towards imidate in the presence of stronger bases (BEMP-H+ pKaMeCN = 27.6 > MTBD-H+ pKaMeCN = 25.4 > DBU-H+ pKaMeCN = 24.3) and upon the application of polar solvent (which presumably stabilizes the charged catalyst species).47,54 More imidate character is associated with faster rates of ROP.35,47,54,55 However, once the H- bonding/imidate equilibrium is shifted mostly to imidate, catalytic activity will diminish if more acidic (thio)ureas or stronger bases are applied. This is also attributable to the reduced basicity of the resulting (thio)imidate; non-linear Hammett behavior has been observed.35,43,51 The very progress of the reaction was shown to influence the nature of the active catalyst because, during an ROP, the highly polar monomer is converted to less polar 11 polymer. Hence the H-bonding/imidate equilibrium (Scheme 4) was shown to shift towards neutral catalysts late in the ROP.11 This may constitute an advantage of applying organic (versus alkoxides or hydrides) bases whose reactivity becomes attenuated late in the ROP, thereby increasing reaction control.11 12 NEW REACTIONS AND ABILITIES The development of new catalytic abilities – the imidate mechanism of enchainment – has provided new synthetic opportunities. For example, (thio)imidate mediated ROP are operative under solvent-free conditions.32 The polar lactone monomer is ironically a poor solvent for H-bond mediate ROP of lactones; the monomer interrupts cocatalyst H-bonding and severely attenuates reactivity. However, in solvent-free conditions, the urea plus base cocatalyst system is highly active for ROP. These conditions even allow for the synthesis of block copolymers that are inaccessible in solution conditions.32 New opportunities in additive manufacturing can be envisaged. Imidate mediated enchainment allowed for the production of high molecular weight poly(γ- butyrolactone)s (PγBLs) via selective ROP of “nonpolymerizable” γ-butyrolactone (γBL) at -40oC.28,29 The utility of commercially available phosphazene super bases and (thio)ureas facilitated the formation of linear PγBL initiated by the alcohol species. These species display among the highest activity for the organocatalytic ROP of γBL.28 The ROP of γBL with alkoxide/urea catalysts show high activity even at -20oC. Although this system produces linear polymers, careful manipulation of monomer/catalyst/initiator was required to ensure initiation from the alcohol (versus monomer).29 Those ROP using less acidic (thio)ureas displayed greater catalytic activity, presumably due to the generation of a more basic (thio)urea anion. The utility of the imidate/neutral H-bonding duality of (thio)ureas were further demonstrated in a study where a sequential one pot copolymerization of epoxides and LA was reported.32,56 One pot synthesis of polyether-polylactide copolymers has been successful only in a few cases.56 The strong base required for the ROP of cyclic ethers can 13 lead to deleterious epimerization of LA and transesterification of PLA.56 The 1-S H-bond donor in the presence of tetrabutyl ammonium fluoride (TBAF) is effective for the copolymerization of glycidyl phenyl ether and LA, yielding polymers with predictable molecular weights and narrow dispersities (Mw/Mn = 1.13 – 1.19).56 The proposed mechanism proceeds by an anionic initiation of the epoxide by TBAF; the addition of 1-S allows the conversion of the incipient alkoxide to the 1-S thioimidate, which is competent for the controlled ROP of LA, Scheme 5. NEW REACTIONS AND ABILITIES Hence, the mechanistic duality of the 1-S system directly facilitates the one pot copolymerization of epoxide and LA.14,56–58 lead to deleterious epimerization of LA and transesterification of PLA.56 The 1-S H-bond donor in the presence of tetrabutyl ammonium fluoride (TBAF) is effective for the copolymerization of glycidyl phenyl ether and LA, yielding polymers with predictable molecular weights and narrow dispersities (Mw/Mn = 1.13 – 1.19).56 The proposed mechanism proceeds by an anionic initiation of the epoxide by TBAF; the addition of 1-S allows the conversion of the incipient alkoxide to the 1-S thioimidate, which is competent for the controlled ROP of LA, Scheme 5. Hence, the mechanistic duality of the 1-S system directly facilitates the one pot copolymerization of epoxide and LA.14,56–58 14 CONCLUSION Since the application of the Takemoto thiourea for ROP and the discovery that covalent tethering the H-bond donor and base cocatalyst is not essential, the field of (thio)urea mediated ROP has advanced in spurts to among the more active and controlled systems for the enchainment of cyclic monomers. (Thio)urea catalysts were conventionally known to follow a dual H-bonding mechanism with assistance of organic bases; however, the differing activity of these base cocatalysts and the ability to manipulate the H-bond donating ability by changing the acidity of (thio)ureas provided substantial evidence for a second mechanism. In the presence of weaker organic bases, (thio)ureas promote ROP via a neutral H-bonding mechanism, whereas with stronger bases they proceed via an imidate H-bonding mechanism which may exhibit dual H-bonding activity like TBD. In most cases, the recently developed urea class of H-bond donors were shown to be more active than their corresponding thioureas. The imidate mechanism remains highly active in polar solvents and exhibits remarkable control under solvent-free conditions, and high temperature applications are accessible.11 It should be emphasized that the advancements in synthetic abilities have all evolved through a greater understanding of reaction mechanism. We expect that the enhanced utility – greater range of solvents, temperatures and substrates – will expose weakness and strengths of the nascent catalysts which will precipitate further advances, perhaps via mechanistic study. New substrates with new demands for selectivity remain to be studied. Through the continued synergistic advances of catalysis and material, the H-bonding class of catalyst can find use in a host of potential applications, research and industrial environments. 15 LIST OF REFERENCES (1) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R. M. Organocatalysis: Opportunities and Challenges for Polymer Synthesis. Macromolecules 2010, 43 (5), 2093–2107. (2) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Organocatalytic Ring-Opening Polymerization. Chem. Rev. 2007, 107 (12), 5813–5840. (3) Reichmanis, E.; Wooley, K. L.; Lodge, T. P. Research in Macromolecular Science: Challenges and Opportunities for the Next Decade. Macromolecules 2009, 42 (2). (4) Dove, A. P.; Pratt, R. C.; Lohmeijer, B. G. G.; Waymouth, R. M.; Hedrick, J. L. Thiourea-Based Bifunctional Organocatalysis: Supramolecular Recognition for Living Polymerization. J. Am. Chem. Soc. 2005, 127 (40), 13798–13799. (5) McKinlay, C. J.; Waymouth, R. M.; Wender, P. A. Cell-Penetrating, Guanidinium- Rich Oligophosphoesters: Effective and Versatile Molecular Transporters for Drug and Probe Delivery. J. Am. Chem. Soc. 2016, 138 (10), 3510–3517. (6) Pratt, R. C.; Nederberg, F.; Waymouth, R. M.; Hedrick, J. L. Tagging Alcohols with Cyclic Carbonate: A Versatile Equivalent of (Meth)Acrylate for Ring- Opening Polymerization. Chem. Commun. 2008, 114–116. (7) Sanders, D. P.; Fukushima, K.; Coady, D. J.; Nelson, A.; Fujiwara, M.; Yasumoto, M.; Hedrick, J. L. A Simple and Efficient Synthesis of Functionalized Cyclic 16 Carbonate Monomers Using a Versatile Pentafluorophenyl Ester Intermediate. J Carbonate Monomers Using a Versatile Pentafluorophenyl Ester Intermediate. J. Am. Chem. Soc. 2010, 132 (42), 14724–14726. Am. Chem. Soc. 2010, 132 (42), 14724–14726. (8) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. a.; Lundberg, P. N. P.; Dove, A. P.; Li, H.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L. Exploration, Optimization, and Application of Supramolecular Thiourea−Amine Catalysts for the Synthesis of Lactide (Co)Polymers. Macromolecules 2006, 39 (23), 7863–7871. (9) Zhang, L.; Nederberg, F.; Messman, J. M.; Pratt, R. C.; Hedrick, J. L.; Wade, C. G. Organocatalytic Stereoselective Ring-Opening Polymerization of Lactide with Dimeric Phosphazene Bases. J. Am. Chem. Soc. 2007, 129 (42), 12610–12611. (10) Nomura, N.; Ishii, R.; Akakura, M.; Aoi, K. Stereoselective Ring-Opening Polymerization of Racemic Lactide Using Aluminum-Achiral Ligand Complexes: Exploration of a Chain-End Control Mechanism. J. Am. Chem. Soc. 2002, 124 (21), 5938–5939. (11) Coderre, D. N.; Fastnacht, K. V.; Wright, T. J.; Dharmaratne, N. U.; Kiesewetter, M. K. H-Bonding Organocatalysts for Ring-Opening Polymerization at Elevated Temperatures. Macromolecules 2018, 51, 10121–10126. (12) Dubois, P.; Coulembier, O.; Raquez, J.-M. Handbook of Ring-Opening Polymerization; John Wiley & Sons, 2009. (13) Hillmyer, M. A.; Tolman, W. B. LIST OF REFERENCES Aliphatic Polyester Block Polymers: Renewable, Degradable, and Sustainable. Acc. Chem. Res. 2014, 47 (8), 2390–2396. (14) Guillaume, S. M.; Kirillov, E.; Sarazin, Y.; Carpentier, J. F. Beyond Stereoselectivity, Switchable Catalysis: Some of the Last Frontier Challenges in Ring-Opening Polymerization of Cyclic Esters. Chem. - A Eur. J. 2015, 21 (22), (13) Hillmyer, M. A.; Tolman, W. B. Aliphatic Polyester Block Polymers: Renewable, Degradable, and Sustainable. Acc. Chem. Res. 2014, 47 (8), 2390–2396. (14) Guillaume, S. M.; Kirillov, E.; Sarazin, Y.; Carpentier, J. F. Beyond Stereoselectivity, Switchable Catalysis: Some of the Last Frontier Challenges in Ring-Opening Polymerization of Cyclic Esters. Chem. - A Eur. J. 2015, 21 (22), 17 7988–8003. 7988–8003. (15) Bakkali-Hassani, C.; Rieger, E.; Vignolle, J.; Wurm, F. R.; Carlotti, S.; Taton, D. The Organocatalytic Ring-Opening Polymerization of: N -Tosyl Aziridines by an N-Heterocyclic Carbene. Chem. Commun. 2016, 52 (62), 9719–9722. (16) Thomas, C.; Bibal, B. Hydrogen-Bonding Organocatalysts for Ring-Opening Polymerization. Green Chem. 2014, 16 (4), 1687. (17) Organic Catalysis for Polymerization; Dove, A. P., Sardon, H., Naumann, S., Eds.; Royal Society of Chemistry: London, 2019. (18) Odian, G. Ring-Opening Polymerization. In Principles of Polymerization; New York, 1991; pp 532–603. (19) Penczek, S.; Duda, A.; Kowalski, A.; Libiszowski, J.; Majerska, K.; Biela, T. On the Mechanism of Polymerization of Cyclic Esters Induced by Tin(II) Octoate. Macromol. Symp. 2000, 157, 61–70. (20) Naumann, S.; Wang, D. Dual Catalysis Based on N-Heterocyclic Olefins for the Copolymerization of Lactones: High Performance and Tunable Selectivity. Macromolecules 2016, 49 (23), 8869–8878. (21) Pascual, A.; Sardon, H.; Veloso, A.; Ruiperez, F.; Mecerreyes, D. Organocatalyzed Synthesis of Aliphatic Polyesters from Ethylene Brassylate: A Cheap and Renewable Macrolactone. ACS Macro Lett. 2014, 3 (9), 849–853. (22) Bannin, T. J.; Kiesewetter, M. K. Poly(Thioester) by Organocatalytic Ring- Opening Polymerization. Macromolecules 2015, 48 (16), 5481–5486. (23) Winnacker, M.; Sag, J. Sustainable Terpene-Based Polyamides: Via Anionic Polymerization of a Pinene-Derived Lactam. Chem. Commun. 2018, 54 (7), 841– 18 844. 844. (24) Nachtergael, A.; Coulembier, O.; Dubois, P.; Helvenstein, M.; Duez, P.; Blankert, B.; Mespouille, L. Organocatalysis Paradigm Revisited : Are Metal-Free Catalysts Really Harmless? Biomacromolecules 2015, 16, 507–514. (25) Mezzasalma, L.; Dove, A. P.; Coulembier, O. Organocatalytic Ring-Opening Polymerization of L-Lactide in Bulk: A Long Standing Challenge. Eur. Polym. J. 2017, 95 (May), 628–634. (26) Danko, M.; Mosnázek, J. Ring-Opening Polymerization of γ-Butyrolactone and Its Derivatives: A Review. Polimery/Polymers 2017, 62 (4), 272–282. (27) Hong, M.; Chen, E. Y. X. Completely Recyclable Biopolymers with Linear and Cyclic Topologies via Ring-Opening Polymerization of γ-Butyrolactone. Nat. Chem. 2016, 8 (1), 42–49. (28) Zhang, C. J.; Hu, L. F.; Wu, H. L.; Cao, X. H.; Zhang, X. H. Dual Organocatalysts for Highly Active and Selective Synthesis of Linear Poly(γ-Butyrolactone)s with High Molecular Weights. Macromolecules 2018, 51 (21), 8705–8711. (29) Lin, L; Han, D; Qin, J.; Wang, S.; Xiao, M; Sun, L; Meng, Y. Nonstrained γ- Butyrolactone to High-Molecular-Weight Poly(γ-Butyrolactone): Facile Bulk Polymerization Using Economical Ureas/Alkoxides. Macromolecules 2018, 51 (22), 9317–9322. (30) Yamashita, K.; Yamamoto, K.; Kadokawa, J. Acid-Catalyzed Ring-Opening Polymerization of γ-Butyrolactone under High-Pressure Conditions. Chem. Lett. 2014, 43 (2), 213–215. 19 (31) A. J, K.; M. A, U.; M. O, E.; R. U., O. Emerging Issues in the Mechanisms of High Pressure Free Radical Ethylene Polymerization: A Review. Am. J. Polym. Sci. 2012, 2 (5), 91–101. (32) Pothupitiya, J. U.; Dharmaratne, N. U.; Jouaneh, T. M; Fastnacht, K. V.; Coderre, D. N.; Kiesewetter, M. K. H-Bonding Organocatalysts for the Living, Solvent-Free Ring-Opening Polymerization of Lactones: Toward an All-Lactones, All- Conditions Approach. Macromolecules 2017, 50 (22), 8948–8954. Conditions Approach. Macromolecules 2017, 50 (22), 8948–8954. (33) Zhu, N.; Liu, Y.; Liu, J.; Ling, J.; Hu, X.; Huang, W.; Feng, W.; Guo, K. Organocatalyzed Chemoselective Ring-Opening Polymerizations. Sci. Rep. 2018, 8 (1), 4–11. (34) Fastnacht, K. V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; Kiesewetter, E. T.; Kiesewetter, M. K. Bis- and Tris-Urea H-Bond Donors for Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst. ACS Macro Lett. 2016, 5 (8), 982–986. (35) Lin, B.; Waymouth, R. M. Urea Anions: Simple, Fast, and Selective Catalysts for Ring-Opening Polymerizations. J. Am. Chem. Soc. 2017, 139 (4), 1645–1652. (36) Thomas, C.; Peruch, F.; Bibal, B. Ring-Opening Polymerization of Lactones Using Supramolecular Organocatalysts under Simple Conditions. RSC Adv. 2012, 2 (33), 12851–12856. (37) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. (44) Kazakov, O. I; Kiesewetter, M. K. Cocatalyst Binding Effects in Organocatalytic 844. T.; Kiesewetter, M. K. Rate Accelerated Organocatalytic Ring-Opening Polymerization of l-Lactide via the Application of a Bis(Thiourea) H-Bond Donating Cocatalyst. Macromolecules 2015, 48 (17), 6127–6131. 20 (38) Fastnacht, K. V.; Datta, P. P.; Kiesewetter, M. K. Bifunctional and Supramolecular Organocatalysts for Polymerization. In Organic Catalysis for Polymerization; Dove, A. P., Sardon, H., Naumann, S., Eds.; Royal Society of Chemistry: London, 2019; pp 87–120. (39) Liu, J.; Chen, C.; Li, Z.; Wu, W.; Zhi, X.; Zhang, Q.; Wu, H.; Wang, X.; Cui, S.; Guo, K. A Squaramide and Tertiary Amine: An Excellent Hydrogen-Bonding Pair Organocatalyst for Living Polymerization. Polym. Chem. 2015, 6 (20), 3754–3757. (40) Liu, J.; Xu, J.; Li, Z.; Xu, S.; Wang, X.; Wang, H.; Guo, T.; Gao, Y.; Zhang, L.; Guo, K. Squaramide and Amine Binary H-Bond Organocatalysis in Polymerizations of Cyclic Carbonates, Lactones, and Lactides. Polym. Chem. 2017, 8, 7054–7068. (41) Rostami, A.; Sadeh, E.; Ahmadi, S. Exploration of Tertiary Aminosquaramide Bifunctional Organocatalyst in Controlled/Living Ring-Opening Polymerization of l-Lactide. J. Polym. Sci. Part A Polym. Chem. 2017, 55, 2483–2493. (42) Lohmeijer, B. G. G., R. C. Pratt, F. Leibfarth; Logan, J. W.; Long, D. A; Dove, A. P., Nedergerg, J; Choi, C; Wade, C; Waymouth, R. M; Hedrick, J. L. Guanidine and Amidine Organocatalysts for Ring-Opening Polymerization of Cyclic Esters. Macromolecules 2006, 39 (25), 8574–8583. (43) Pothupitiya, J. U.; Hewawasam, R. S.; Kiesewetter, M. K. Urea and Thiourea H- Bond Donating Catalysts for Ring-Opening Polymerization: Mechanistic Insights via (Non)Linear Free Energy Relationships. Macromolecules 2018, No. 51, 3203– 3211. (44) Kazakov, O. I; Kiesewetter, M. K. Cocatalyst Binding Effects in Organocatalytic 21 Ring-Opening Polymerization of l -Lactide. Macromolecules 2015, 48 (17), 6121– 6126. Ring-Opening Polymerization of l -Lactide. Macromolecules 2015, 48 (17), 6121– 6126. (45) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Waymouth, R. M.; Hedrick, J. L. Triazabicyclodecene: A Simple Bifunctional Organocatalyst for Acyl Transfer and Ring-Opening Polymerization of Cyclic Esters. J. Am. Chem. Soc. 2006, 128 (14), 4556–4557. (46) Curran, D. P.; Kuo, L. H. Acceleration of a Dipolar Claisen Rearrangement by Hydrogen Bonding to a Soluble Diaryl^urea. Tetrahedron Lett. 1995, 36 (37), 6647–6650. (47) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, M. K. Triclocarban: Commercial Antibacterial and Highly Effective H-Bond Donating Catalyst for Ring-Opening Polymerization. ACS Macro Lett. 2017, 6 (4), 421–425. (48) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. (58) Teator, A. J.; Lastovickova, D. N.; Bielawski, C. W. Switchable Polymerization Catalysts. Chem. Rev. 2016, 116 (4), 1969–1992. Synthesis of α,ω-Difunctionalized Poly(Ethylene Oxide)s and Poly(Ethylene Oxide)-b-Poly(ε-Caprolactone) Block Copolymers. J. Am. Chem. Soc. 2009, 131 (9), 3201–3209. 844. K. Cooperative Hydrogen-Bond Pairing in Organocatalytic Ring-Opening Polymerization. Macromolecules 2014, 47 (21), 7463–7468. (49) Simón, L.; Goodman, J. M. The Mechanism of TBD-Catalyzed Ring-Opening Polymerization of Cyclic Esters. J. Org. Chem. 2007, 72 (25), 9656–9662. (50) Chuma, A.; Horn, H. W.; Swope, W. C.; Pratt, R. C.; Zhang, L.; Lohmeijer, B. G. G.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L.; Rice, J. E. The Reaction Mechanism for the Organocatalytic Ring-Opening Polymerization of l-Lactide Using a Guanidine-Based Catalyst: Hydrogen-Bonded or Covalently Bound? J. Am. Chem. Soc. 2008, 130 (21), 6749–6754. 22 (51) Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Fast and Selective Ring- Opening Polymerizations by Alkoxides and Thioureas. Nat. Chem. 2016, 8, 1047– 1053. (52) Reddy, L. S.; Basavoju, S.; Vangala, V. R.; Nangia, A. Hydrogen Bonding in Crystal Structures of N,N′-Bis(3-Pyridyl)Urea. Why Is the N-H⋯O Tape Synthon Absent in Diaryl Ureas with Electron-Withdrawing Groups? Cryst. Growth Des. 2006, 6 (1), 161–173. (53) Obrzud, M.; Rospenk, M.; Koll, A. Self-Aggregation Mechanisms of N-Alkyl Derivatives of Urea and Thiourea. Phys. Chem. Chem. Phys. 2014, 16 (7), 3209– 3219. (54) Lin, B.; Waymouth, R. M. Organic Ring-Opening Polymerization Catalysts: Reactivity Control by Balancing Acidity. Macromolecules 2018, 51 (8), 2932– 2938. (55) Lemaire, C. F.; Aerts, J. J.; Voccia, S.; Libert, L. C.; Mercier, F.; Goblet, D.; Plenevaux, A. R.; Luxen, A. J. Fast Production of Highly Reactive No-Carrier- Added [18F] Fluoride for the Labeling of Radiopharmaceuticals. Angew. Chemie - Int. Ed. 2010, 49 (18), 3161–3164. (56) Liu, Y.; Wang, X.; Li, Z.; Wei, F.; Zhu, H.; Dong, H.; Chen, S.; Sun, H.; Yang, K.; Guo, K. A Switch from Anionic to Bifunctional H-Bonding Catalyzed Ring- Opening Polymerizations towards Polyether-Polyester Diblock Copolymers. Polym. Chem. 2018, 9 (2), 154–159. (57) Raynaud, J.; Absalon, C.; Gnanou, Y.; Taton, D. N-Heterocyclic Carbene-Induced Zwitterionic Ring-Opening Polymerization of Ethylene Oxide and Direct 23 Synthesis of α,ω-Difunctionalized Poly(Ethylene Oxide)s and Poly(Ethylene Oxide)-b-Poly(ε-Caprolactone) Block Copolymers. J. Am. Chem. Soc. 2009, 131 (9), 3201–3209. (58) Teator, A. J.; Lastovickova, D. N.; Bielawski, C. W. Switchable Polymerization Catalysts. Chem. Rev. 2016, 116 (4), 1969–1992. 24 Scheme 1.1. H-bonding versus Imidate Mediated ROP of Lactones Scheme 1.1. H-bonding versus Imidate Mediated ROP of Lactones Scheme 1.2. Proposed activation pathway of covalently linked bifunctional thiourea in ROP of LA Scheme 1.2. Proposed activation pathway of covalently linked bifunctional thiourea in ROP of LA Scheme 1.2. MANUSCRIPT – II Published in ACS Macromolecules 844. Proposed activation pathway of covalently linked bifunctional thiourea in ROP of LA 25 Scheme 1.3. (upper) Bifunctional activation of monomer and initiator/chain end by TBD (lower) Formation of imidate catalyst and suggested activation modality. Scheme 1.3. (upper) Bifunctional activation of monomer and initiator/chain end by TBD (lower) Formation of imidate catalyst and suggested activation modality. Scheme 1.3. (upper) Bifunctional activation of monomer and initiator/chain end by TBD (lower) Formation of imidate catalyst and suggested activation modality. Scheme 1.4. Equilibrium between neutral versus imidate TCC with base Scheme 1.4. Equilibrium between neutral versus imidate TCC with base 26 Scheme 1.5. ROP of epoxides and thiourea mediated conversion of alkoxide to alcohol and thioimdate for the ROP of lactones. Scheme 1.5. ROP of epoxides and thiourea mediated conversion of alkoxide to alcohol and thioimdate for the ROP of lactones. Scheme 1.5. ROP of epoxides and thiourea mediated conversion of alkoxide to alcohol and thioimdate for the ROP of lactones. Scheme 1.5. ROP of epoxides and thiourea mediated conversion of alkoxide to alcohol and thioimdate for the ROP of lactones. 27 Figure 1.1. Strength of cocatalyst binding is predictive of catalytic activity. Figure 1.1. Strength of cocatalyst binding is predictive of catalytic activity. Figure 1.2. Representative (thio)ureas in red and proposed activated-TU mode activation for multi-donors. Figure 1.2. Representative (thio)ureas in red and proposed activated-TU mode activation for multi-donors. 28 Figure 1.3. Monomer activation by activated-3-O and intramolecular deactivation of 3-S. Figure 1.3. Monomer activation by activated-3-O and intramolecular deactivation of 3-S. Figure 1.3. Monomer activation by activated-3-O and intramolecular deactivation of 3-S. Figure 1.4. The thiourea/base mediated ROP of lactones slows in polar solvent while urea/base mediated ROP remain active. Figure 1.4. The thiourea/base mediated ROP of lactones slows in polar solvent while urea/base mediated ROP remain active. Figure 1.4. The thiourea/base mediated ROP of lactones slows in polar solvent while urea/base mediated ROP remain active. 29 titative Measurements of Polymer Hydrophobicity Based on Functional Group Nayanthara U. Dharmaratne,1 Terra Marie M. Jouaneh,1 Matthew K. Kiesewetter.,1 and Robert T. Mathers2 1Department of Chemistry, University of Rhode Island, Kingston, RI 02881 2Department of Chemistry, The Pennsylvania State University, New Kensington, Pennsylvania 15068. ABSTRACT A combined experimental and computational investigation revealed a hydrophobicity trend for oxygen containing functional groups commonly encountered in monomers and polymers. Based on solvatochromatic dye experiments, HPLC retention times, and theoretical LogP values, the 3-oxygen atom in carbonates results in more hydrophobicity than other permutations like anhydrides. Another trend emerged for functional groups with 2-oxygen atoms (acetals > esters). Overall, when comparing aliphatic polymers with similarly sized monomers, hydrophobicity decreased as follows: carbonates > acetals > esters > anhydrides. These trends have important implications for degradation, conductivity, and many other applications. 31 INTRODUCTION Many scientists who study polymers from the perspective of surface chemistry,1 biochemistry,2 biology,3 or physics4 routinely use words like “hydrophobic” and “hydrophilic.”5 These valuable concepts provide a necessary framework for describing physical properties of various polymers and nanostructures.6 While such constructs can facilitate a nice qualitative comparison of similar functional groups, like a series of lactones with different ring sizes or acrylates with various alkyl substituents, using the term “hydrophobic” has limitations when ranking dissimilar polymers. To some extent, ambiguity arises due to the instinctual perspective of counting carbon atoms in functional groups or calculating a carbon:oxygen (C:O) ratio for monomers. Considering a number of oxygen-containing monomers units with ketones, ethers, esters, acetals, anhydrides, and carbonates may have similar C:O ratios, structural diversity creates a dilemma when evaluating homopolymers and copolymers. For instance, while poly(tetrahydrofuran) (PTHF) is viewed as “hydrophobic” relative to poly(N- isopropylacrylamide) (PNIPAM),7 other researchers view PTHF as “hydrophilic” relative to polycaprolactone (PCL).8 Since the C:O ratio for PTHF (4:1) is larger than PCL (3:1) but less than PNIPAM, this perspective illustrates the confusion when counting atoms in a monomer unit. In addition, various PCL derivatives have been designated as “hydrophobic,”9 “moderately hydrophobic,”10 and “highly hydrophobic.”11 To further compound the uncertainty with homopolymers, copolymers with carbonate units are more hydrophobic than esters in some cases but less hydrophobic than esters in other copolymers.12 32 Because terms like hydrophobic have potential for ambiguity when ranking polymers with oxygen-containing functional groups, quantitative metrics offer a potential pathway to reduce misperception. As such, a number of computational strategies, like octanol-water partition coefficients (LogPoct)13,14 and Molecular Dynamics (MD) simulations of water contact angles15 and solubility parameters16 greatly assist efforts to understand hydrophobicity. These computational approaches complement experimental methods based on contact angle measurements,17 swelling experiments,7,18 liquid chromatography,19 and solvatochromatic dye experiments.20 Because terms like hydrophobic have potential for ambiguity when ranking polymers with oxygen-containing functional groups, quantitative metrics offer a potential pathway to reduce misperception. As such, a number of computational strategies, like octanol-water partition coefficients (LogPoct)13,14 and Molecular Dynamics (MD) simulations of water contact angles15 and solubility parameters16 greatly assist efforts to understand hydrophobicity. INTRODUCTION These computational approaches complement experimental methods based on contact angle measurements,17 swelling experiments,7,18 liquid chromatography,19 and solvatochromatic dye experiments.20 Recently, combined experimental and computational LogPoct values have demonstrated the ability to discern hydrophobicity of anhydride monomer libraries,21 polyacrylate homopolymers,22 star polymers,23 post-polymer modification,18 polyester electrolytes,24 solubility of B vitamins in melt polymerizations,25 and diblock copolymers for crystallization driven self-assembly (CDSA).26 As the structural diversity of petroleum27-30 and renewable31-38 monomers continues to expand, understanding and predicting hydrophobicity for various permutations of oxygen atoms leads to challenges. Consequently, an opportunity exists to rank common functional groups, like ethers,39 carbonates,27 esters,25 anhydrides,21 and acetals.40-42 We hypothesize that functional groups with oxygen atoms exert a major influence on hydrophobicity and suggest that using the C:O ratio or size of alkyl substituents oversimplifies assessment when comparing structurally diverse monomers. For instance, will functional groups with the same number of oxygen atoms, like esters and acetals, have different amounts of hydrophobicity? How about carbonates and anhydrides, which each have three oxygen atoms? 33 With these questions in mind, we have begun to investigate the connection between oxygen-containing functional groups, like acetal, ester, ether, carbonate and anhydrides, and hydrophobicity of monomers and polymers. To answer these questions, a combination of experimental and computational methods were investigated for acyclic model compounds, cyclic and tricyclic monomers, as well as some selected polymer examples. 34 EXPERIMENTAL SECTION Acetic anhydride (Acros Organics, 99+ %, AA), acrolein diethyl acetal (Sigma-Aldrich, 96 %, ADA), butyl acrylate (Sigma-Aldrich, > 99%, BA), diethyl ether (Fisher Scientific, laboratory grade, DE), dimethyl carbonate (Acros Organics, 99 %, DMC), dimethoxymethane (Sigma Aldrich, 99 %, DMM), 2,2-dimethoxy-2-phenylacetophenone (Sigma Aldrich, 99 %, DMPA), ethyl acetate (EMD Chemicals, > 99.5 %, EtAc), methyl succinic anhydride (Sigma Aldrich, 98 %, MSA), propylene carbonate (Sigma Aldrich, 99.7 %, PC), 3-pentanone (Acros Organics, 98 %, 3P), and gamma-valerolactone (Sigma Aldrich, 99 %, GVL) were used as received. δ-Valerolactone (Acros Organics, 99 %, VL) and ε-caprolactone (Sigma Aldrich, 97 %, CL) were distilled from calcium hydride under high vacuum. UV/Vis spectroscopy measurements conducted with a Shimadzu UV-1650PC spectrophotometer at ambient temperature. Samples analyzed in semi-micro UV-cuvettes (path length 10.0 mm) (Brand GMBH, Germany). The max values of Nile Red taken from maximum height of absorbance. Reported values are an average of 5 experiments. Typical standard deviation was ± 0.5 nm. For consistency, Nile Red concentrations were adjusted to achieve absorbance values of <1. Octanol-water partition coefficients were calculated by Materials Studio 7.0 and extracted using the ALogP function on the QSAR menu. Connolly surface area (SA) calculated with a 1.40 Å probe after Forcite Geometry Optimization with a Smart algorithm and COMPASS II forcefield. The maximum number of iterations during minimization varied depending on molecular weight but were typically 500 for monomers and 10,000-20,000 35 for oligomers. Convergence tolerance for the Smart algorithm included a 1.0e-4 kcal/mol energy convergence, 0.005 kcal/mol/Å force convergence, and a 5.0e-5 Å displacement convergence. Contact angle measurements were conducted at ambient temperature with a custom-built goniometer using a 22 gauge needle.43 Drops (~10 µL) of organic liquid were placed on Restek Thermolite® rubber septa (10 mm diameter) and Teflon coated silicon septa (10 mm diameter). Reported contact angle values were an average of 6-8 measurements. Contact angle measurements were conducted at ambient temperature with a custom-built goniometer using a 22 gauge needle.43 Drops (~10 µL) of organic liquid were placed on Restek Thermolite® rubber septa (10 mm diameter) and Teflon coated silicon septa (10 mm diameter). Reported contact angle values were an average of 6-8 measurements. HPLC analysis was investigated with a Shimadzu HPLC system equipped with two pumps (20AT), an autosampler (SIL-20), degasser, column oven (20HT), and a Waters UV/Vis detector. EXPERIMENTAL SECTION Acetonitrile (0.15 mL/min) was eluted through a Phenomenex C18 column (3 mm x 100 mm, 3 µm particles) at 30 °C. Samples (1 mg/mL) were prepared in acetonitrile and injected (1 µL) via autosampler. The UV detector was set at 254 nm. HPLC analysis was investigated with a Shimadzu HPLC system equipped with two pumps (20AT), an autosampler (SIL-20), degasser, column oven (20HT), and a Waters UV/Vis detector. Acetonitrile (0.15 mL/min) was eluted through a Phenomenex C18 column (3 mm x 100 mm, 3 µm particles) at 30 °C. Samples (1 mg/mL) were prepared in acetonitrile and injected (1 µL) via autosampler. The UV detector was set at 254 nm. Photopolymerization of neat butyl acrylate was conducted at ambient temperature. Initially, filtering butyl acrylate (BA) through neutral alumina removed inhibitor. DMPA was crushed to a fine powder with a spatula. Then, BA (2.4 g) and DMPA (0.5 wt. %) were added to a cuvette, capped with a septum, and vortexed until homogeneous. After bubbling the solution with nitrogen for 3 min, the capped cuvette was immediately irradiated at 365 nm (7.7 ± 0.2 mW/cm2) for 1 min. After the polymerization, Nile Red (~ 0.05 mg) powder was added. The solution stirred for several minutes before taking a UV/Vis spectrum. Example oligomerization of VL and CL. To a 20 mL vial, CL (2.000 g, 17.5 mmol) and benzyl alcohol (0.1890 g, 1.75 mmol) were added, and the contents were stirred to mix. To a second 7 ml vial, TBD (0.0240 g, 0.175 mmol) and acetone (2.0 mL) were added and 36 agitated until homogenous. The contents of the second vial were transferred to the first vial via Pasteur pipette, and the contents were agitated to mix. The reaction was quenched in 1 h by the addition of benzoic acid (0.0430 g, 0.345 mmol). Polymer was purified using a silica column where 100 % dichloromethane was used as the mobile phase. PCL was removed of volatiles under high vacuum prior to characterization. Yield 97%, Mn = 1200 g/mol determined by NMR. agitated until homogenous. The contents of the second vial were transferred to the first vial via Pasteur pipette, and the contents were agitated to mix. The reaction was quenched in 1 h by the addition of benzoic acid (0.0430 g, 0.345 mmol). Polymer was purified using a silica column where 100 % dichloromethane was used as the mobile phase. EXPERIMENTAL SECTION PCL was removed of volatiles under high vacuum prior to characterization. Yield 97%, Mn = 1200 g/mol determined by NMR. Synthesis of Trimethylene Carbonate (TMC) Synthesis of Trimethylene Carbonate (TMC) TMC was synthesized as reported.44 Ethyl chloroformate (25 ml, 0.264 mol) was added in a single portion to a solution of 1,3-propanediol (9.5 ml, 0.132 mol) dissolved in 500 mL of THF in a Schlenk flask under N2 atmosphere. The contents were cooled using an ice bath and triethylamine (37 ml, 0.264 mol) was added dropwise over a period of 30 minutes. After 2 h, a precipitate formed, was filtered off and the filtrate was concentrated under vacuum to yield a yellow oil. Diethyl ether (200 mL) was added to the remaining oil, and the contents were stored in a refrigerator overnight to crystallize TMC. The crystals were further purified by recrystallizing in diethyl ether. Yield 52 %. Oligomerization of TMC to produce poly(trimethylene carbonate) (PTMC) was conducted by adding benzyl alcohol (0.424 g, 3.92 mmol) to a solution of TMC (2.000 g, 19.6 mmol) in 7 mL of CHCl3 in a 20 mL vial. In a second 7 mL vial, a solution of TBD (0.014 g, 0.098 mmol) in CHCl3 (1 mL) was prepared. The TBD solution was then added to the 20 mL vial via Pasteur pipette, and the contents were shaken to mix. The reaction was quenched in 15 min using benzoic acid (0.024 g, 0.195 mmol) to yield 98% PTMC, Mn (NMR) = 650 g/mol, Mn (GPC) = 1100 g/mol. 37 RESULTS AND DISCUSSION In order to rank hydrophobicity of common functional groups, a series of acyclic model compounds (Figure 2.1) of similar size were investigated by UV/Vis spectroscopy. Accordingly, analyzing lmax values for a solvatochromatic dye in these solvents quantified the relationship between hydrophobicity and functional groups. For example, an inverse relationship exists between lmax values for Nile Red (NR) and hydrophobicity. In order to rank hydrophobicity of common functional groups, a series of acyclic model compounds (Figure 2.1) of similar size were investigated by UV/Vis spectroscopy. Accordingly, analyzing lmax values for a solvatochromatic dye in these solvents quantified the relationship between hydrophobicity and functional groups. For example, an inverse relationship exists between lmax values for Nile Red (NR) and hydrophobicity. In Figure 2.1, NR confirmed the 3 oxygen atoms in dimethyl carbonate (DMC) provided a more hydrophobic environment than the 3 oxygen atoms in acetic anhydride (AA). Surprisingly, DMC is also more hydrophobic than functional groups with fewer oxygen atoms, like ethyl acetate (EtAc) and 3-pentanone (3P). These results indicate that simply counting the number of oxygen atoms in a functional does not convey the whole picture. Likewise, dimethoxymethane, which has an acetal functional group with 2 oxygen atoms, is more hydrophobic than diethyl ether (DE) and EtAc. In Figure 2.1, NR confirmed the 3 oxygen atoms in dimethyl carbonate (DMC) provided a more hydrophobic environment than the 3 oxygen atoms in acetic anhydride (AA). Surprisingly, DMC is also more hydrophobic than functional groups with fewer oxygen atoms, like ethyl acetate (EtAc) and 3-pentanone (3P). These results indicate that simply counting the number of oxygen atoms in a functional does not convey the whole picture. Likewise, dimethoxymethane, which has an acetal functional group with 2 oxygen atoms, is more hydrophobic than diethyl ether (DE) and EtAc. To put these observations in context, consider that carbon tetrachloride has more electronegative atoms than dichloromethane, but is more hydrophobic due to symmetry and lack of a molecular dipole.45 Interestingly, a similar situation occurs when examining the 2 oxygen atoms in 1,4-dioxane. Due to symmetry, 1,4-dioxane has a dielectric constant (e = 2.209) much closer to cyclohexane (e = 2.023) compared to other molecules with same number of oxygen and carbon atoms, like ethyl acetate (e = 6.02).46 So compared to 3P, the location and symmetry of extra oxygen atoms in DMC actually attenuate the polarity of the carbonyl. RESULTS AND DISCUSSION 38 To focus the experimental perspective in Figure 2.1 on monomers, select examples from polycondensation reactions and ring opening polymerizations (ROP) have been emphasized. In Figure 2.2, a series of acetals, lactones,47 anhydrides,21,48,49 and carbonates50 were ranked by octanol-water partition coefficients (LogPoct) (Equation 1) and normalized by surface area (SA). As such, positive LogPoct/SA values indicate a preference for dissolving in the 1-octanol layer. In contrast, hydrophilic molecules exhibit negative LogPoct/SA values and prefer to partition into the water layer. Conveniently, LogPoct calculations provide an avenue to compare liquid monomers with solid bicyclic monomers. In contrast, dissolving Nile Red at ambient temperature is optimal for the liquids used in Figure 2.1. LogPoct = log([soluteoctanol layer]/[solutewater layer]) (1) LogPoct = log([soluteoctanol layer]/[solutewater layer]) (1) (1) The combination of solvatochromatic experiments (Figure 2.1) and computational LogPoct values (Figure 2.2) offer compelling evidence of a hydrophobicity trend amongst acetals, carbonates, esters, and anhydrides. Based on these examples, carbonates, like EC, TMC, and PC, are always more hydrophobic than their corresponding anhydrides. In fact, this large difference between EC and SA occurs without matching the C:O ratio. In contrast, the gap between acetals/esters is smaller than carbonates/anhydrides. As such, comparing cyclic molecules with similar C:O ratios, as shown in Figure 2.2a and 2.2b, provided the most accurate picture and indicated the hydrophobicity of acetals > esters. In Figure 2.3, HPLC analysis reinforced the concept of a hydrophobicity trend and demonstrated a correlation between experimental retention time and computational 39 LogPoct/SA values. As such, the 5-membered acetal (i.e. 1,3-benzodioxole) was more hydrophobic than the corresponding ester and anhydride. In addition to mirroring the trends of 5-membered rings in Figure 2.2, this data had reasonable agreement with other molecules, like benzaldehyde and styrene oxide. Unexpectedly, a caveat emerged when attaching two large groups, like benzyl groups, to a carbonate or ester. As such, these large groups overwhelmed the influence of a single functional group and caused HPLC retention times to converge. Additionally, electronic effects (Figure 2.7) on the functional group were observed when comparing benzyl and phenyl groups. Moving from small molecule trends in Figures 2.1-2.3 towards polymers, a number of unanswered questions exist. As such, this investigation clarifies a number of critical questions: What occurs to hydrophobicity upon the ring-opening of cyclic monomers (Figure 2.4a)? How much influence would vinyl groups exert on hydrophobicity (Figure 2.4b)? RESULTS AND DISCUSSION How relevant is the C:O ratio in determining hydrophobicity for polyesters (Figure 2.8), polyacetals (Figure 2.5a) and polyethers? Do the trends from Figures 2.1-2.3 apply to polymers over a range of oligomer sizes (Figure 2.5)? Would the influence of endgroups alter hydrophobicity trends (Figure 2.5b)? To investigate what happens during a polymerization, Figure 2.4a considers ring-opening polymerization (ROP) while Figure 2.4b explores whether the addition of an alkene will compete with the influence of esters and acetals. Figure 2.4a revealed cyclic monomers, like PC, GVL, VL, and MSA, exhibit less hydrophobicity than the ring-opened model compounds. Surprisingly, this observation was most evident for carbonates like DMC. Contact angle measurements (Figure 2.9) confirm the polarity difference between AA and MSA is smaller than the difference between PC and DMC. These remarks about Figure 40 2.4a are important because cyclic monomers would underestimate hydrophobicity of corresponding polymers obtained by ROP. For example, cyclic VL (lmax = 546.0 nm) is less hydrophobic than an oligomer of polyvalerolactone (DP =5, lmax = 542.7 nm). Although cyclic structures resulted in significantly less hydrophobicity than linear analogues, the effect of adding an alkene (Figure 2.4b) to DMM and EtAc is less pronounced. Similarly, switching the alkyl length in ethyl acrylate (EA) to a butyl group has minimal effect on the dominant influence on the ester group. As a result, photoinitiated polymerization of neat butyl acrylate (BA) (lmax = 525.1 nm) with DMPA produced polybutylacrylate with a similar Nile Red absorbance (lmax = 523.1 nm). A comparison of BA with an acetal containing identical numbers of carbon and oxygen atoms, like ADA, indicates the acetal is more hydrophobic than the ester. The intuitive process of examining the C:O ratio has some relevance for qualitatively assessing hydrophobicity of polymers with similar functional groups. For example, the hydrophobicity trend for polyesters (Figure 2.8) like PCL, PVL, and PGBL clearly follows the different C:O ratios in the repeat unit. However, the situation becomes more complex when entertaining structurally diverse monomers or comparing different functional groups. For instance, Figure 2.5a highlights an example where the C:O ratio fails to account for the differences between polyacetals and polyethers. Remarkably, in Figure 2.5a, the acetal units in POM contain a minimal number of carbon atoms (C:O =1:1) but exhibit more hydrophobicity than PEO (C:O =2:1), polyoxetane (C:O =3:1), and slightly more than polytetrahydrofuran (PTHF) (C:O =4:1). RESULTS AND DISCUSSION As such, to approximate the hydrophobicity of the acetal in POM, a polyether would need a C:O ratio ≥ 4:1. Furthermore, the computational results in Figure 2.5a correlate to experimental 41 contact angle measurements observed for POM (q = 69°51, q = 67°52), PEO (qadv = 34-45°)43 and analogues like poly(ethylene glycol) (PEG).53 While some polyethers, like PEO, are widely recognized as hydrophilic, Figure 2.5a represents the first systematic comparison as a function of oligomer length. To understand if small molecule trends from Figures 2.1-2.3 apply to a range of oligomer sizes, a few ring-opening polymerizations (ROP) were investigated. In Figure 2.5b, a series of oligomers made with carbonate, acetal, and ester-based monomers from Figure 2.2 were constructed. Then, computational LogPoct/SA values for these oligomers were compared. Overall, the functional group trends for polymers, like poly(PC), poly(GVL), and poly(PA), had good agreement with the monomers. In Figure 2.5, OH chain ends exhibit a strong influence on hydrophobicity when the degree of polymerization (DP) is below ~8 monomer units. To compare these theoretical results with experimental values, a series of oligomers ranging from DP 2 to DP 10 were synthesized via ROP of VL and CL. Based on analysis by GPC, thermogravimetric analysis (TGA) (Figures 2.10 and 2.11), differential scanning calorimetry (DSC) (Figures 2.12 and 2.13), the oligomer lengths of DP 2, DP 5, and DP 10 represent average values. In Figure 2.6, solvatochromatic experiments with these oligomers indicated chain end influence on hydrophobicity diminished as molecular weight increases from DP 5 and DP 10. These experimental results in Figure 2.6 correlate well with computational results in Figure 2.5. 42 CONCLUSION Computational LogPoct values, Nile Red (NR) solvatochromatic experiments, HPLC retention times, and contact angle measurements confirmed hydrophobicity depends on number of oxygen atoms and presence of alkyl groups or aromatic rings. As such, among functional groups with 3-oxygen atoms, aliphatic carbonates are more hydrophobic than anhydrides. Based on model compounds, electronic effect of benzene rings (Figure 2.7) narrows the hydrophobicity gap between carbonates and anhydrides. Although the difference in hydrophobicity for carbonates and anhydrides in Figure 2.2 is substantial, the gap between acetals and ester was smaller. When comparing similar C:O ratios, acetals were more hydrophobic than esters. These observations are summarized in Table 2.1. As a general trend, symmetry of oxygen atoms increases hydrophobicity. For instance, Figure 2.1 highlights how DMC is more hydrophobic than EtAc and 3P. Interestingly, sulfur-based analogues in Scheme 2.1 also demonstrate hydrophobicity trends that depend on symmetry of the oxygen atoms. As such, symmetry of two oxygen atoms in phenylmethyl sulfone results in more hydrophobicity compared to the single oxygen atom in the corresponding sulfoxide. This perspective on hydrophobicity helps explain conductivity trends for polymers containing sulfoxide, sulfone, and sulfide groups.54 Considering LogPoct values span 5-6 orders of magnitude, this metric assesses the concept of hydrophobicity on a larger continuum compared to certain types of experimental techniques, like contact angle measurements. As a result, LogPoct values may offer 43 valuable insight in regards to understanding experimental studies on polymer degradability of polyesters,55 polyacetals56 and polycarbonates.42 Furthermore, the nuances of copolymers often create confusion. For example, opposite degradation trends are observed for poly(TMC-co-LA) and poly(LA-co-DHA) copolymers.12 So, why do esters provide more hydrolytic degradation than carbonates in one case and less in another? The answer lies in recognizing a hydrophobicity trend: TMC > lactic acid (LA) > dihydroxyacetone (DHA). 44 (5) Chandler, D. Interfaces and the driving force of hydrophobic assembly. Nature 2005, 437, 640. (11) Jahani, H.; Jalilian, F. A.; Wu, C.-Y.; Kaviani, S.; Soleimani, M.; Abassi, N.; Ou, K.- L.; Hosseinkhani, H. Controlled surface morphology and hydrophilicity of L.; Hosseinkhani, H. Controlled surface morphology and hydrophilicity of LIST OF REFERENCES (1) Xin, B. W.; Hao, J. C. Reversibly switchable wettability. Chemical Society Reviews 2010, 39, 769. (2) Stangl, M.; Hemmelmann, M.; Allmeroth, M.; Zentel, R.; Schneider, D. A Minimal Hydrophobicity Is Needed To Employ Amphiphilic p(HPMA)-co-p(LMA) Random Copolymers in Membrane Research. Biochemistry 2014, 53, 1410. (3) Meneksedag-Erol, D.; Kc, R. B.; Tang, T.; Uludağ, H. A Delicate Balance When Substituting a Small Hydrophobe onto Low Molecular Weight Polyethylenimine to Improve Its Nucleic Acid Delivery Efficiency. ACS Appl. Mat. Inter. 2015, 7, 24822. (3) Meneksedag-Erol, D.; Kc, R. B.; Tang, T.; Uludağ, H. A Delicate Balance When Substituting a Small Hydrophobe onto Low Molecular Weight Polyethylenimine to Improve Its Nucleic Acid Delivery Efficiency. ACS Appl. Mat. Inter. 2015, 7, 24822. (4) Robert, C. H.; Pettitt, B. M. Reconciling the understanding of ‘hydrophobicity’ with physics-based models of proteins. Journal of Physics: Condensed Matter 2016, 28, 083003. 45 (6) Figg, C. A.; Carmean, R. N.; Bentz, K. C.; Mukherjee, S.; Savin, D. A.; Sumerlin, B. S. Tuning Hydrophobicity To Program Block Copolymer Assemblies from the Inside Out. Macromolecules 2017, 50, 935. (7) Guan, Y.; Ding, X.; Zhang, W.; Wan, G.; Peng, Y. Polytetrahydrofuran Amphiphilic Networks, 5. Synthesis and Swelling Behavior of Thermosensitive Poly(N- isopropylacrylamide)-l-polytetrahydrofuran Networks. Macromol. Chem. Phys. 2002, 203, 900. 6) Figg, C. A.; Carmean, R. N.; Bentz, K. C.; Mukherjee, S.; Savin, D. A.; Sumer (6) Figg, C. A.; Carmean, R. N.; Bentz, K. C.; Mukherjee, S.; Savin, D. A.; Sumerlin, B. S. Tuning Hydrophobicity To Program Block Copolymer Assemblies from the Inside Out. Macromolecules 2017, 50, 935. (7) Guan, Y.; Ding, X.; Zhang, W.; Wan, G.; Peng, Y. Polytetrahydrofuran Amphiphilic Networks, 5. Synthesis and Swelling Behavior of Thermosensitive Poly(N- isopropylacrylamide)-l-polytetrahydrofuran Networks. Macromol. Chem. Phys. 2002, 203, 900. (8) Vaikkath, D.; Anitha, R.; Sumathy, B.; Nair, P. D. A simple and effective method for making multipotent/multilineage scaffolds with hydrophilic nature without any postmodification/treatment. Colloids and Surfaces B: Biointerfaces 2016, 141, 112. (9) Cai, L.; Lu, J.; Sheen, V.; Wang, S. Lubricated Biodegradable Polymer Networks for Regulating Nerve Cell Behavior and Fabricating Nerve Conduits with a Compositional Gradient. Biomacromolecules 2012, 13, 358. (10) Tse, K.-H.; Sun, M.; Mantovani, C.; Terenghi, G.; Downes, S.; Kingham, P. J. In vitro evaluation of polyester-based scaffolds seeded with adipose derived stem cells for peripheral nerve regeneration. J. Biomed. Mat. Res.: Part A 2010, 95A, 701. (11) Jahani, H.; Jalilian, F. A.; Wu, C.-Y.; Kaviani, S.; Soleimani, M.; Abassi, N. (16) Belmares, M.; Blanco, M.; Goddard, W. A.; Ross, R. B.; Caldwell, G.; Chou, S.-H.; Pham, J.; Olofson, P. M.; Thomas, C. Hildebrand and Hansen solubility parameters from Molecular Dynamics with applications to electronic nose polymer sensors. Journal of Computational Chemistry 2004, 25, 1814. LIST OF REFERENCES (10) Tse, K.-H.; Sun, M.; Mantovani, C.; Terenghi, G.; Downes, S.; Kingham, P. J. In vitro evaluation of polyester-based scaffolds seeded with adipose derived stem cells for peripheral nerve regeneration. J. Biomed. Mat. Res.: Part A 2010, 95A, 701. 46 polycaprolactone toward selective differentiation of mesenchymal stem cells to neural like cells. J. Biomed. Mat. Res.: Part A 2015, 103, 1875. polycaprolactone toward selective differentiation of mesenchymal stem cells to neural like cells. J. Biomed. Mat. Res.: Part A 2015, 103, 1875. (12) Weiser, J. R.; Zawaneh, P. N.; Putnam, D. Poly(carbonate-ester)s of Dihydroxyacetone and Lactic Acid as Potential Biomaterials. Biomacromolecules 2011, 12, 977. (13) Petrauskas, A. A.; Kolovanov, E. A. ACD/Log P method description. Perspectives in Drug Discovery and Design 2000, 19, 99. (14) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Deliv. Rev. 2001, 46, 3. (14) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Deliv. Rev. 2001, 46, 3. (15) Liu, H.; Li, Y.; Krause, W. E.; Rojas, O. J.; Pasquinelli, M. A. The Soft-Confined Method for Creating Molecular Models of Amorphous Polymer Surfaces. J. Phys. Chem. B 2012, 116, 1570. (16) Belmares, M.; Blanco, M.; Goddard, W. A.; Ross, R. B.; Caldwell, G.; Chou, S.-H.; Pham, J.; Olofson, P. M.; Thomas, C. Hildebrand and Hansen solubility parameters from Molecular Dynamics with applications to electronic nose polymer sensors. Journal of Computational Chemistry 2004, 25, 1814. (16) Belmares, M.; Blanco, M.; Goddard, W. A.; Ross, R. B.; Caldwell, G.; Chou, S.-H.; Pham, J.; Olofson, P. M.; Thomas, C. Hildebrand and Hansen solubility parameters from Molecular Dynamics with applications to electronic nose polymer sensors. Journal of Computational Chemistry 2004, 25, 1814. 47 (17) Gao, L.; McCarthy, T. J. Teflon is Hydrophilic. Comments on Definitions of Hydrophobic, Shear versus Tensile Hydrophobicity, and Wettability Characterization. Langmuir 2008, 24, 9183. (18) Waggel, J.; Mathers, R. T. Post polymer modification of polyethylenimine with citrate esters: selectivity and hydrophobicity. RSC Advances 2016, 6, 62884. (19) Som, A.; Reuter, A.; Tew, G. N. Protein Transduction Domain Mimics: The Role of Aromatic Functionality. Angew. Chem. Int. Ed. 2012, 51, 980. (20) Jessop, P. G.; Jessop, D. A.; Fu, D.; Phan, L. (23) Unverferth, M.; Meier, M. A. R. Tuning the polarity of ADMET derived star-shaped polymers via thiol-ene chemistry. Polymer 2014, 55, 5571. LIST OF REFERENCES Solvatochromic parameters for solvents of interest in green chemistry. Green Chem. 2012, 14, 1245. (21) Dakshinamoorthy, D.; Weinstock, A. K.; Damodaran, K.; Iwig, D. F.; Mathers, R. T. Diglycerol-Based Polyesters: Melt Polymerization with Hydrophobic Anhydrides. ChemSusChem 2014, 7, 2923. (22) Magenau, A. J. D.; Richards, J. A.; Pasquinelli, M. A.; Savin, D. A.; Mathers, R. T. Systematic Insights from Medicinal Chemistry To Discern the Nature of Polymer Hydrophobicity. Macromolecules 2015, 48, 7230. (23) Unverferth, M.; Meier, M. A. R. Tuning the polarity of ADMET derived star-shaped polymers via thiol-ene chemistry. Polymer 2014, 55, 5571. (23) Unverferth, M.; Meier, M. A. R. Tuning the polarity of ADMET derived star-shaped polymers via thiol-ene chemistry. Polymer 2014, 55, 5571. (23) Unverferth, M.; Meier, M. A. R. Tuning the polarity of ADMET derived star-shaped polymers via thiol-ene chemistry. Polymer 2014, 55, 5571. 48 (24) Yildirim, E.; Dakshinamoorthy, D.; Peretic, M. J.; Pasquinelli, M. A.; Mathers, R. T. Synthetic Design of Polyester Electrolytes Guided by Hydrophobicity Calculations. Macromolecules 2016, 49, 7868. (25) Soxman, A. G.; DeLuca, J. M.; Kinlough, K. M.; Iwig, D. F.; Mathers, R. T. Functionalization of polyesters with multiple B vitamins. Journal of Polymer Science Part A: Polymer Chemistry 2017, 55, 3308. (24) Yildirim, E.; Dakshinamoorthy, D.; Peretic, M. J.; Pasquinelli, M. A.; Mathers, R. T. Synthetic Design of Polyester Electrolytes Guided by Hydrophobicity Calculations. Macromolecules 2016, 49, 7868. (25) Soxman, A. G.; DeLuca, J. M.; Kinlough, K. M.; Iwig, D. F.; Mathers, R. T. Functionalization of polyesters with multiple B vitamins. Journal of Polymer Science Part A: Polymer Chemistry 2017, 55, 3308. (26) Inam, M.; Cambridge, G.; Pitto-Barry, A.; Laker, Z. P. L.; Wilson, N. R.; Mathers, R. T.; Dove, A. P.; O'Reilly, R. K. 1D vs. 2D shape selectivity in the crystallization-driven self-assembly of polylactide block copolymers. Chemical Science 2017, 8, 4223. (27) Tempelaar, S.; Mespouille, L.; Coulembier, O.; Dubois, P.; Dove, A. P. Synthesis and post-polymerisation modifications of aliphatic poly(carbonate)s prepared by ring-opening polymerisation. Chem. Soc. Rev. 2013, 42, 1312. (28) Pesko, D. M.; Jung, Y.; Hasan, A. L.; Webb, M. A.; Coates, G. W.; Miller Iii, T. F.; Balsara, N. P. Effect of monomer structure on ionic conductivity in a systematic set of polyester electrolytes. Solid State Ionics 2016, 289, 118. (29) Noy, J.-M.; Koldevitz, M.; Roth, P. J. Thiol-reactive functional poly(meth)acrylates: multicomponent monomer synthesis, RAFT (co)polymerization and highly efficient thiol- para-fluoro postpolymerization modification. Polym. Chem. (36) Zhu, Y.; Romain, C.; Williams, C. K. Sustainable polymers from renewable resources. Nature 2016, 540, 354. LIST OF REFERENCES 2015, 6, 436. 49 (30) Vancoillie, G.; Frank, D.; Hoogenboom, R. Thermoresponsive poly(oligo ethylene glycol acrylates). Progress in Polymer Science 2014, 39, 1074. (31) Green Polymer Chemistry: Biocatalysis and Biomaterials; Cheng, H. N.; Gross, R. A., Eds.; ACS Symposium Series 1043, American Chemical Society: Washington DC, 2010. (32) Mathers, R. T.; Meier, M. A. R. In Green Polymerization Methods: Renewable Starting Materials, Catalysis and Waste Reduction; Mathers, R. T., Meier, M. A. R., Eds.; Wiley-VCH: Weinheim, 2011. (33) Wilbon, P. A.; Chu, F.; Tang, C. Progress in Renewable Polymers from Natural Terpenes, Terpenoids, and Rosin. Macromol. Rapid Commun. 2013, 34, 8. (34) Coates, G. W.; Hillmyer, M. A. Polymers from renewable resources. Macromolecules 2009, 42, 7987. (35) Schroder, K.; Matyjaszewski, K.; Noonan, K. J. T.; Mathers, R. T. Towards sustainable polymer chemistry with homogeneous metal-based catalysts. Green Chem. 2014, 16 1673 (36) Zhu, Y.; Romain, C.; Williams, C. K. Sustainable polymers from renewable resources. Nature 2016, 540, 354. (36) Zhu, Y.; Romain, C.; Williams, C. K. Sustainable polymers from renewable resources. Nature 2016, 540, 354. 50 (37) Winkler, M.; Romain, C.; Meier, M. A. R.; Williams, C. K. Renewable polycarbonates and polyesters from 1,4-cyclohexadiene. Green Chemistry 2015, 17, 300. (38) Yang, G. Z.; Rohde, B. J.; Tesefay, H.; Robertson, M. L. Biorenewable Epoxy Resins Derived from Plant-Based Phenolic Acids. ACS Sustain. Chem. Eng. 2016, 4, 6524. (37) (39) Janvier, M.; Hollande, L.; Jaufurally, A. S.; Pernes, M.; Ménard, R.; Grimaldi, M.; Beaugrand, J.; Balaguer, P.; Ducrot, P.-H.; Allais, F. Syringaresinol: A Renewable and Safer Alternative to Bisphenol A for Epoxy-Amine Resins. ChemSusChem 2017, 10, 738. (40) Pemba, A. G.; Flores, J. A.; Miller, S. A. Acetal metathesis polymerization (AMP): A method for synthesizing biorenewable polyacetals. Green Chemistry 2013, 15, 325. (41) Miller, S. A. Sustainable Polymers: Opportunities for the Next Decade. ACS Macro Letters 2013, 2, 550. (42) Ortmann, P.; Heckler, I.; Mecking, S. Physical properties and hydrolytic degradability of polyethylene-like polyacetals and polycarbonates. Green Chem. 2014, 16, 1816. (43) Quirk, R. P.; Mathers, R. T.; Cregger, T.; Foster, M. D. Anionic synthesis of block copolymer brushes grafted from a 1,1-diphenylethylene monolayer. Macromolecules 2002, 35, 9964. (43) Quirk, R. P.; Mathers, R. T.; Cregger, T.; Foster, M. D. Anionic synthesis of block copolymer brushes grafted from a 1,1-diphenylethylene monolayer. Macromolecules 2002, 35, 9964. 51 (44) Jiménez-Pardo, I.; van der Ven, L. G. J.; van Benthem, R. A. T. LIST OF REFERENCES M.; Esteves, A. C. C.; de With, G. Effect of a Set of Acids and Polymerization Conditions on the Architecture of Polycarbonates Obtained via Ring Opening Polymerization. Journal of Polymer Science Part A: Polymer Chemistry 2017, 55, 1502. (45) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic Chemistry; University Science Books, 2006. (46) Haynes, W. M. Handbook of Chemistry and Physics; 93rd ed.; CRC Press: Boca Raton, 2012. (47) Azadi, P.; Carrasquillo-Flores, R.; Pagan-Torres, Y. J.; Gurbuz, E. I.; Farnood, R.; Dumesic, J. A. Catalytic conversion of biomass using solvents derived from lignin. Green Chem. 2012, 14, 1573. (48) Dakshinamoorthy, D.; Lewis, S. P.; Cavazza, M. P.; Hoover, A. M.; Iwig, D. F.; Damodaran, K.; Mathers, R. T. Streamlining the conversion of biomass to polyesters: bicyclic monomers with continuous flow. Green Chem. 2014, 16, 1774. (49) Jezowski, S. R.; Monaco, S.; Yennawar, H. P.; Wonderling, N. M.; Mathers, R. T.; Schatschneider, B. Unusual physical behaviour and polymorphic phase transitions in crystalline bicyclic anhydrides. CrystEngComm 2017, 19, 276. (49) Jezowski, S. R.; Monaco, S.; Yennawar, H. P.; Wonderling, N. M.; Mathers, R. T.; Schatschneider, B. Unusual physical behaviour and polymorphic phase transitions in crystalline bicyclic anhydrides. CrystEngComm 2017, 19, 276. 52 (50) Helou, M.; Miserque, O.; Brusson, J.-M.; Carpentier, J.-F.; Guillaume, S. M. Poly(trimethylene carbonate) from Biometals-Based Initiators/Catalysts: Highly Efficient Immortal Ring-Opening Polymerization Processes. Adv. Syn. Catal. 2009, 351, 1312. (50) Helou, M.; Miserque, O.; Brusson, J.-M.; Carpentier, J.-F.; Guillaume, S. M. Poly(trimethylene carbonate) from Biometals-Based Initiators/Catalysts: Highly Efficient Immortal Ring-Opening Polymerization Processes. Adv. Syn. Catal. 2009, 351, 1312. (50) (51) Yousef, S.; Visco, A.; Galtieri, G.; Njuguna, J. Flexural, impact, rheological and physical characterizations of POM reinforced by carbon nanotubes and paraffin oil. Polymers for Advanced Technologies 2016, 27, 1338. (52) Bannwarth, M. B.; Klein, R.; Kurch, S.; Frey, H.; Landfester, K.; Wurm, F. R. Processing and adjusting the hydrophilicity of poly(oxymethylene) (co)polymers: nanoparticle preparation and film formation. Polymer Chemistry 2016, 7, 184. (52) Bannwarth, M. B.; Klein, R.; Kurch, S.; Frey, H.; Landfester, K.; Wurm, F. R. Processing and adjusting the hydrophilicity of poly(oxymethylene) (co)polymers: nanoparticle preparation and film formation. Polymer Chemistry 2016, 7, 184. (53) Bayer, I. S.; Megaridis, C. M.; Jie, Z.; Gamota, D.; Biswas, A. Analysis and surface energy estimation of various model polymeric surfaces using contact angle hysteresis. Journal of Adhesion Science & Technology 2007, 21, 1439. (53) Bayer, I. S.; Megaridis, C. (55) Martin, R. T.; Camargo, L. P.; Miller, S. A. Marine-degradable polylactic acid. Green Chem. 2014, 16, 1768. (54) Sarapas, J. M.; Tew, G. N. Poly(ether–thioethers) by Thiol–Ene Click and Their Oxidized Analogues as Lithium Polymer Electrolytes. Macromolecules 2016, 49, 1154. (55) Martin, R. T.; Camargo, L. P.; Miller, S. A. Marine-degradable polylactic acid. Green Chem. 2014, 16, 1768. (56) Samanta, S.; Bogdanowicz, D. R.; Lu, H. H.; Koberstein, J. T. Polyacetals: Water- Soluble, pH-Degradable Polymers with Extraordinary Temperature Response. Macromolecules 2016, 49, 1858. LIST OF REFERENCES M.; Jie, Z.; Gamota, D.; Biswas, A. Analysis and surface energy estimation of various model polymeric surfaces using contact angle hysteresis. Journal of Adhesion Science & Technology 2007, 21, 1439. (54) Sarapas, J. M.; Tew, G. N. Poly(ether–thioethers) by Thiol–Ene Click and Their Oxidized Analogues as Lithium Polymer Electrolytes. Macromolecules 2016, 49, 1154. (55) Martin, R. T.; Camargo, L. P.; Miller, S. A. Marine-degradable polylactic acid. Green Chem. 2014, 16, 1768. 53 54 Table 2.1. Summation of hydrophobicity trends for small molecules and polymers. Scenario Data Trend Aliphatic acyclic model compounds Figure 2.1 Acetal > carbonate > ester > anhydride Cyclic and tricyclic monomers Figure 2.2 Carbonate > acetal > ester > anhydride HPLC retention time of model compounds Figure 2.3 Acetal > ester > anhydride Effect of alkene on lmax Figure 2.4 Acetal > ester Electronic effect of benzene Figure 2.7 Carbonate > ester > anhydride Polyethers versus polyacetal Figure 2.5a polyacetals (POM) > PTHF > polyoxetane > PEO Ring opening polymerization Figure 2.5b polyPC > polyacetal > poly(g-valerolactone) Influence of OH chain end Figure 2.6 PCL > PVL; hydrophobicity increases as DP increases Table 2.1. Summation of hydrophobicity trends for small molecules and polymers. 55 Scheme 2.1. Comparison of sulfur-based functional groups with ether and ketone analogues. me 2.1. Comparison of sulfur-based functional groups with ether and ketone Scheme 2.1. Comparison of sulfur-based functional groups with ether and ketone analogues. Figure 2.1. UV/Vis spectroscopy data for absorbance (lmax) of Nile Red in molecules with oxygen-containing functional groups. Increasing lmax values indicate a decrease in hydrophobicity. Figure 2.1. UV/Vis spectroscopy data for absorbance (lmax) of Nile Red in molecules with oxygen-containing functional groups. Increasing lmax values indicate a decrease in hydrophobicity. 56 Figure 2.2 Computational octanol-water partition coefficients (LogPoct) for cyclic and t i li ) i f t l d t f C O ti 4 2 d b) C O Figure 2.2 Computational octanol-water partition coefficients (LogPoct) for cyclic and tricyclic monomers: a) comparison of acetals and ester for C:O ratio = 4:2 and b) C:O ratio = 5:2. Increasing LogPoct/SA values indicate an increase in hydrophobicity. Figure 2.2 Computational octanol-water partition coefficients (LogPoct) for cyclic and tricyclic monomers: a) comparison of acetals and ester for C:O ratio = 4:2 and b) C:O ratio = 5:2. Increasing LogPoct/SA values indicate an increase in hydrophobicity. 57 Figure 2.3. LIST OF REFERENCES Correlation between experimental HPLC retention times and computational LogPoct values normalized by surface area (SA). Line represents linear regression. Figure 2.3. Correlation between experimental HPLC retention times and computational LogPoct values normalized by surface area (SA). Line represents linear regression. Figure 2.3. Correlation between experimental HPLC retention times and computational LogPoct values normalized by surface area (SA). Line represents linear regression. 58 UV/Vis spectroscopy data for solvatochromatic behavior of Nile Re cyclic carbonates, esters, and anhydrides with the corresponding acycl b) evaluation of functional groups with and without vinyl groups. Low Figure 2.4. UV/Vis spectroscopy data for solvatochromatic behavior of Nile Red comparing a) cyclic carbonates, esters, and anhydrides with the corresponding acyclic molecule and b) evaluation of functional groups with and without vinyl groups. Lower lmax values indicate more hydrophobicity. 59 omparison of hydrophobicity values for a) ethers and acetals and b) acetals, arbonates versus number of monomer units. Examples shown include ofuran (PTHF) (●), polyoxymethylene (POM) (■), polyoxetane (x), oxide) (PEO) (▲), poly(propylene carbonate) (♦), and poly(g- ) (●). For comparison, a hypothetic polyacetal () in Figure 2.5b is given. Figure 2.5. Comparison of hydrophobicity values for a) ethers and acetals and b) a esters, and carbonates versus number of monomer units. Examples shown i polytetrahydrofuran (PTHF) (●), polyoxymethylene (POM) (■), polyoxetan poly(ethylene oxide) (PEO) (▲), poly(propylene carbonate) (♦), and valerolactone) (●) For comparison a hypothetic polyacetal () in Figure 2 5b is Figure 2.5. Comparison of hydrophobicity values for a) ethers and acetals and b) acetals, esters, and carbonates versus number of monomer units. Examples shown include polytetrahydrofuran (PTHF) (●), polyoxymethylene (POM) (■), polyoxetane (x), poly(ethylene oxide) (PEO) (▲), poly(propylene carbonate) (♦), and poly(g- valerolactone) (●). For comparison, a hypothetic polyacetal () in Figure 2.5b is given. The dashed lines represent logarithmic regression. 60 Figure 2.6. UV/Vis spectroscopy data for solvatochromatic behavior of Nile Red comparing oligomers of PVL and PCL initiated with benzyl alcohol. Mn values measured by NMR end group analysis. Figure 2.6. UV/Vis spectroscopy data for solvatochromatic behavior of Nile Red comparing oligomers of PVL and PCL initiated with benzyl alcohol. Mn values measured by NMR end group analysis. 61 Figure 2.7. Electronic effect of benzene reduced the gap between carbonate and anhydride functional groups. For the 3 dibenzyl compounds shown above, the hydrophobicity trend (carbonate > ester > anhydride) follows the expectations given in Figures 2.1-2.3. LIST OF REFERENCES When replacing benzyl groups with phenyl, the gap between carbonates and anhydrides lessens. Figure 2.7. Electronic effect of benzene reduced the gap between carbonate and anhydride functional groups. For the 3 dibenzyl compounds shown above, the hydrophobicity trend (carbonate > ester > anhydride) follows the expectations given in Figures 2.1-2.3. When replacing benzyl groups with phenyl, the gap between carbonates and anhydrides lessens. 62 Figure 2.8. Comparison of hydrophobicity values for polyesters with different C:O ratios. The decamers incorporated OCH3 initiators and OH endgroups. PCL = polycaprolactone, PVL = polyvalerolactone, PGBL = poly(g-butyrlactone). LogPoct and Connolly surface area (SA) calculated with Materials Studio after Forcite geometry optimization. Figure 2.8. Comparison of hydrophobicity values for polyesters with different C:O ratios. The decamers incorporated OCH3 initiators and OH endgroups. PCL = polycaprolactone, PVL = polyvalerolactone, PGBL = poly(g-butyrlactone). LogPoct and Connolly surface ( ) l l d i h i l di f i i i i Figure 2.8. Comparison of hydrophobicity values for polyesters with different C:O ratios. The decamers incorporated OCH3 initiators and OH endgroups. PCL = polycaprolactone, PVL = polyvalerolactone, PGBL = poly(g-butyrlactone). LogPoct and Connolly surface area (SA) calculated with Materials Studio after Forcite geometry optimization. Figure 2.8. Comparison of hydrophobicity values for polyesters with different C:O ratios. The decamers incorporated OCH3 initiators and OH endgroups. PCL = polycaprolactone, PVL = polyvalerolactone, PGBL = poly(g-butyrlactone). LogPoct and Connolly surface area (SA) calculated with Materials Studio after Forcite geometry optimization. 63 Figure 2.9. Contact angle of static drop (~10 µL) on Teflon coated septa (orange bars) and Restek butyl rubber septa (blue bars) for liquid carbonates and anhydrides at ambient temperature. A comparison of acyclic and cyclic functional groups indicate anhydrides, such as AA and MSA, exhibit a smaller difference in contact angles (~4° ± 2°) than carbonate functional groups, like DMC (~37° ± 3°) and PC (~49° ± 3°). Since the substrates are hydrophobic, larger contact angles indicate more polarity. Figure 2 9 Contact angle of static drop (~10 µL) on Teflon coated septa (orange bars) and Figure 2.9. Contact angle of static drop (~10 µL) on Teflon coated septa (orange bars) and Figure 2.9. Contact angle of static drop (~10 µL) on Teflon coated septa (orange bars) and Restek butyl rubber septa (blue bars) for liquid carbonates and anhydrides at ambient Figure 2.9. LIST OF REFERENCES Contact angle of static drop (~10 µL) on Teflon coated septa (orange bars) and Restek butyl rubber septa (blue bars) for liquid carbonates and anhydrides at ambient temperature. A comparison of acyclic and cyclic functional groups indicate anhydrides, such as AA and MSA, exhibit a smaller difference in contact angles (~4° ± 2°) than carbonate functional groups, like DMC (~37° ± 3°) and PC (~49° ± 3°). Since the substrates are hydrophobic, larger contact angles indicate more polarity. Figure 2.9. Contact angle of static drop (~10 µL) on Teflon coated septa (orange bars) and Restek butyl rubber septa (blue bars) for liquid carbonates and anhydrides at ambient temperature. A comparison of acyclic and cyclic functional groups indicate anhydrides, such as AA and MSA, exhibit a smaller difference in contact angles (~4° ± 2°) than carbonate functional groups, like DMC (~37° ± 3°) and PC (~49° ± 3°). Since the substrates are hydrophobic, larger contact angles indicate more polarity. 64 Figure 2.10. TGA data (20 °C/min) (top) and first derivative (bottom) PCL DP 5, and c) PCL DP 10. Figure 2.10. TGA data (20 °C/min) (top) and first derivative (bottom) for a) PCL DP 2, b) PCL DP 5, and c) PCL DP 10. Figure 2.10. TGA data (20 °C/min) (top) and first derivative (bottom) for a) PCL DP 2, b) PCL DP 5, and c) PCL DP 10. 65 Figure 2.11. TGA data (20 °C/min) (top) and first derivative (bottom PVL DP 5, and c) PVL DP 10. Figure 2.11. TGA data (20 °C/min) (top) and first derivative (bottom) for a) PVL DP 2, b) PVL DP 5, and c) PVL DP 10. Figure 2.11. TGA data (20 °C/min) (top) and first derivative (bottom) for a) PVL DP 2, b) PVL DP 5, and c) PVL DP 10. Figure 2.11. TGA data (20 °C/min) (top) and first derivative (bottom) for a) PVL DP 2, b) PVL DP 5, and c) PVL DP 10. 66 Figure 2.12. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PCL DP2, b) PCL DP5, and c) PCL DP11. Data taken from second heating cycle. Figure 2.12. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PCL DP2, b) PCL DP5, and c) PCL DP11. Data taken from second heating cycle. Figure 2.12. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PCL DP2, b) PCL DP5, and c) PCL DP11. LIST OF REFERENCES Data taken from second heating cycle. 67 Figure 2.13. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PVL DP2, b) PVL DP5, and c) PVL DP10. Data taken from second heating cycle. Figure 2.13. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PVL DP2, b) PVL DP5, and c) PVL DP10. Data taken from second heating cycle. Figure 2.13. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PVL DP2, b) PVL DP5, and c) PVL DP10. Data taken from second heating cycle. Figure 2.13. Differential scanning calorimetry (DSC) data (20 °C/min) for a) PVL DP2, b) PVL DP5, and c) PVL DP10. Data taken from second heating cycle. 68 MANUSCRIPT – III Published in ACS Macromolecules Bis- and Tris-Urea H-Bond Donors for Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst Bis- and Tris-Urea H-Bond Donors for Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst Kurt V. Fastnacht, Samuel S. Spink, Nayanthara U. Dharmaratne, Jinal U. Pothupitiya, Partha P. Datta, Elizabeth T. Kiesewetter and Matthew K. Kiesewetter Chemistry, University of Rhode Island, Kingston, RI, USA Corresponding Author: Matthew Kiesewetter, Ph.D. Chemistry University of Rhode Island 140 Flagg Road Kingston, RI, 02881, USA Email address: mkiesewetter@chm.uri.edu Kurt V. Fastnacht, Samuel S. Spink, Nayanthara U. Dharmaratne, Jinal U. Pothupitiya, Partha P. Datta, Elizabeth T. Kiesewetter and Matthew K. Kiesewetter Chemistry, University of Rhode Island, Kingston, RI, USA ABSTRACT A new class of H-bond donating ureas was developed for the ring-opening polymerization (ROP) of lactone monomers, and they exhibit dramatic rate acceleration versus previous H-bond mediated polymerization catalysts. The most active of these new catalysts, a tris-urea H-bond donor, is among the most active organocatalysts known for ROP, yet it retains the high selectivity of H-bond mediated organocatalysts. The urea cocatalyst, along with an H-bond accepting base, exhibits the characteristics of a “living” ROP, is highly active, in one case, accelerating a reaction from days to minutes, and remains active at low catalyst loadings. The rate acceleration exhibited by this H-bond donor occurs for all base cocatalysts examined. A mechanism of action is proposed, and the new catalysts are shown to accelerate small molecule transesterifications versus currently known mono-thiourea catalysts. It is no longer necessary to choose between a highly active or highly selective organocatalyst for ROP. 70 70 INTRODUCTION The H-bonding catalysts for ring-opening polymerization (ROP) stand out among the highly controlled polymerization methods for their ability to tolerate functional groups while precisely controlling molecular weight and polydispersity.1–7 H-bond donating cocatalysts are believed to effect a “living” ROP via dual activation of monomer by a H- bond donor, usually a thiourea (TU), and activation of alcohol chain end by base cocatalyst.8,9 The exquisite and remarkable combination of rate and selectivity present in other fields (e.g., olefin polymerization catalysis)10,11 has yet to be paralleled in organocatalytic ROP, especially H-bond mediated transformations. The development of organocatalysts for polymerization has largely proceeded along divergent pathways toward highly selective1,9,12–15 or highly active16–19 catalysts. Indeed, the low activity of organocatalysts for ROP has been specifically identified as a shortcoming of the field, whereas highly active metal-containing catalysts for ROP are well-known.20,21 We recently disclosed a bisthiourea (bisTU) H-bond donating cocatalyst, 2-S in Figure 3.1, for the ROP of L-lactide (LA), which displayed enhanced catalytic activity (over mono-TU), but no reduction in reaction control.22 During the process of extending the utility of this system to other lactone monomers, we developed a trisurea (trisU, 3-O in Figure 3.1) H-bond donor featuring remarkable activity for the ROP of lactones. Not only does this cocatalyst demonstrate the utility of the under-explored urea motif (c.f. thiourea) of H-bond donors, but when applied with a H-bond accepting cocatalyst, it is the most active ROP organocatalyst known, and one whose enhanced rate does not come at the expense of reaction control Scheme 3 1 71 EXPERIMENTAL SECTION General Considerations. All manipulations were performed in an MBRAUN stainless steel glovebox equipped with a gas purification system or using Schlenk technique under a nitrogen atmosphere. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Tetrahydrofuran and dichloromethane were dried on an Innovative Technologies solvent purification system with alumina columns and nitrogen working gas. Benzene-d6 and chloroform-d were purchased from Cambridge Isotope Laboratories and distilled from CaH2 under a nitrogen atmosphere. δ-valerolactone (VL; 99%), ε-caprolactone (CL; 99%) and benzyl alcohol were distilled from CaH2 under reduced pressure. 1,3-diaminopropane, 3,5-bis(trifluoromethyl)phenyl isocyanate and cyclohexylamine were purchased from Acros Organics. 3,5-bis(trifluoromethyl)phenyl isothiocyanate was purchased from Oakwood Products. 7-methyl-1,5,7- triazabicyclo[4.4.0]dec-5-ene (MTBD) was purchased from TCI. Tris(2-aminoethyl)amine was purchased from Alpha Aesar. The H-bond donors 1-S, 1-O and 2-S were prepared according to published procedures.23–25 NMR experiments were performed on Bruker Avance III 300 MHz or 400 MHz spectrometers. Size exclusion chromatography (SEC) was performed at 40 °C using dichloromethane eluent on an Agilent Infinity GPC system equipped with three Agilent PLGel columns 7.5 mm × 300 mm (5 μm, pore sizes: 103, 104, 105 Å). Mn and Mw/Mn were determined versus PS standards (500 g/mol-3150 kg/mol, Polymer Laboratories). Water and acetonitrile were all Optima HPLC grade solvents from Fisher Chemical (Fair Lawn, NJ, USA). 72 Mass spectrometry was performed using a Thermo Electron (San Jose, CA, USA) LTQ Orbitrap XL mass spectrometer affixed with either an atmospheric-pressure chemical ionization (APCI) or electrospray ionization (ESI) interface, positive ions were produced and introduced into the instrument. Tune conditions for infusion experiments (10 μL/min flow, sample concentration <20 µg/mL in 50/50 v/v water/acetonitrile) were as follows: ionspray voltage, 5,000 V; capillary temperature, 275 °C; sheath gas (N2, arbitrary units), 8; auxiliary gas (N2, arbitrary units), 0; capillary voltage, 35 V; and tube lens, 110 V. Prior to analysis, the instrument was calibrated for positive ions using Pierce LTQ ESI positive ion calibration solution (lot #PC197784). Ion trap experiments used N2 as a collision gas with normalized collision energies (NCE) between 10-25 eV for multistage fragmentation. High-energy collision (HCD) experiments were performed with He as the collision gas with a NCE of 25 eV. Computational Details. The Spartan ’14 package for Windows 7 was used for all computations. Computed structures were geometry optimized at the B3LYP/6-31G level of theory. EXPERIMENTAL SECTION Reported energies were calculated in CH2Cl2 solvent and were calculated at the B3LYP/6-31G level of theory from the DFT-optimized structures. Energies, structures and coordinates are given below. Synthesis of 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane. A dried 50 mL Schlenk flask was charged with a stir bar, dichloromethane (15.0 mL) and 1,3- diaminopropane (0.45 mL, 5.40 mmol). 3,5-bis(trifluoromethyl)phenyl isothiocyanate (1.00 mL, 5.495 mmol) was added dropwise to the round bottom flask. The solution was stirred for 24 hours, and the solvent was removed under reduced pressure. The resulting solid was purified via silica gel column chromatography with 90 : 10, dichloromethane : 73 methanol mobile phase. Yield: 21%. 1H NMR (300 MHz, DMSO) spectrum below. Product was carried on without full characterization. 1H NMR (300 MHz, C2D6OS) δ 1.6 (p, J = 6, 2H) 2.65 (t, J = 6, 2H) 3.54 (br, 2H) 7.69 (s, 1H) 8.23 (s, 2H). Synthesis of 2-OS. 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane (100.8 mg, 0.292 mmol) was added to a dried 10 mL Schlenk flask containing dichloromethane (1 mL), 3,5-bis(trifluoromethyl)phenyl isocyanate (74.0 mL, 0.290 mmol). Product precipitated from solution and was isolated by decanting the solvent. Solid was recrystallized from dichloromethane and dried under high vacuum overnight. Yield: 70%. HRMS m/z calcd (C21H16F12N4OS + H+) 601.0926, found 601.0893. 1H NMR (300 MHz, DMSO-d6) δ 1.74 (p, J = 6, 2H) 3.19 (q, J = 6, 2H) 3.55 (br, 2H) 6.75 (t, J = 6, 1H) 7.53 (s, 1H) 7.73 (s, 1H) 8.08 (s, 2H) 8.24 (s, 2H) 9.33 (s, 1H) 10.15 (s, 1H). 13C NMR (75 MHz, acetone-d6) δ 29.0, 36.8, 41.4, 113.0, 115.7, 116.8, 121.1, 121.5, 123.0 (q), 124.8, 130.2 (q), 141.5, 142.2, 154.5, 180.1. Synthesis of 2-O. A dried 10 mL Schlenk flask was charged with a stir bar, dichloromethane (7 mL), 1,3-diaminopropane (35.9 μL, 0.43 mmol). 3,5- bis(trifluoromethyl)phenyl isocyanate (148.6 μL, 0.86 mmol) was added dropwise to the round bottom flask. The resulting slurry was stirred for 1 hr, filtered and washed with cold dichloromethane. Yield: 97%. HRMS m/z calcd (C21H16F12N4O2 + H+) 585.1154, found 585.1100. 1H NMR (300 MHz, DMSO-d6) δ 1.68 (p, J = 6 Hz, 2H) 3.22 (q, J = 6 Hz, 4H) 6.59 (t, J = 6, 2H) 7.58 (s, 2H) 8.14 (s, 4H) 9.39 (s, 2H). 13C NMR (75 MHz, DMSO-d6) δ 30.3, 36.6, 113.3, 117.1, 123.3 (q), 130.5 (q), 142.6, 154.9. Synthesis of 3-S. EXPERIMENTAL SECTION A dried 100 mL Schlenk flask was charged with a stir bar, tetrahydrofuran (50 mL), tris(2-aminoethyl) amine (1.05 mL, 6.84 mmol), 3,5- 74 bis(triflouromethyl)phenyl isocyanate (3.90 mL, 21.20 mmol). The solution was left to stir for 24 hrs and the solvent was subsequently removed in vacuo. The resulting solid product was purified using a silica gel column with a 90 : 10, hexanes : ethyl acetate mobile phase. Product was removed of volatiles under high vacuum overnight. Yield: 87%. HRMS m/z calcd (C33H27F18N7S3 + H+) 960.1275, found 960.1262. 1H NMR (300 MHz, acetone-d6) δ 2.82 (t, J = 6, 6H) 3.68 (m, 6H) 7.44 (s, 3H) 7.71 (br, 2H) 8.04 (s, 6H) 9.40 (br, 2H). 13C NMR (75 MHz, acetone-d6) δ 43.7, 53.7, 117.6, 123.3, 124.2 (q), 131.8 (q), 142.5, 182.1. Synthesis of 3-O. A dried 100 mL Schlenk flask was charged with a stir bar, tetrahydrofuran (50 mL), tris(2-aminoethyl) amine (1.03 mL, 6.84 mmol), 3,5- bis(triflouromethyl)phenylisocyanate (3.6 mL, 21.20 mmol). The solution was stirred for 24 hrs. The solvent was removed in vacuo. Resulting solid was purified using a silica gel column with a 96:4 dichloromethane:methanol mobile phase. Yield: 88%. HRMS m/z calcd (C33H27F18N7O3 + H+) 912.1961, found 912.1933. 1H NMR (300 MHz, acetone-d6) δ 2.58 (t, J = 3, 6H) 3.21 (m, 6H) 6.32 (m, 2 H) 7.29 (s, 3H) 7.86 (s, 6H) 8.58 (s, 2H). 13C NMR (75 MHz, acetone-d6) δ 39.3, 55.8, 114.9, 118.3, 124.4 (q), 132.3 (q), 143.3, 156.3. Example VL Polymerization Experiment. A 7 mL vial was charged with 3-O (15.2 mg, 0.0167 mmol), MTBD (2.4 μL, 0.0167 mmol), benzyl alcohol (2.08 μL, 0.01999 mmol) and C6D6 (250 μL). In a second 7 mL vial, VL (0.100 g, 0.999 mmol) was dissolved in C6D6 (249 μL). The contents of the second vial were transferred to the first via pipette and stirred until homogenous, approximately 1 min. The contents were transferred to an NMR tube via pipette, and the reaction was monitored by 1H NMR. The reaction was quenched using benzoic acid (4.06 mg, 0.0333 mmol). Polymer was precipitated with the 75 addition of hexanes. Supernatant was decanted and solid PVL was dried in vacuo. Yield: 89%, Mn = 7,500, Mw/Mn = 1.07. Chain Extension Experiment. A 7 mL vial was loaded with 3-O (13.3 mg, 0.015 mmol), MTBD (2.2 mg, 0.015 mmol), 1-pyrenebutanol (9.6 mg, 0.035 mmol), and C6D6 (219 μL). EXPERIMENTAL SECTION In a second 7mL vial, CL (100 mg, 0.876 mmol) and C6D6 (219 μL) were loaded. The contents of the second vial were added to the first and stirred. After 15 min, a 150 μL aliquot was taken from the reaction vial, quenched with benzoic acid (1.2mg, 0.010 mmol), and additional CL (197.3 mg, 1.723 mmol) was added to the reaction vial. After another 50 min, a second aliquot was quenched with benzoic acid (1.2 mg, 0.010 mmol). Samples from both the first and second aliquots were then transferred to NMR tubes and conversion was determined via 1H NMR analysis. The remainder of the aliquots was precipitated with the addition of hexane, and the supernatants were decanted. Each solid PCL sample was dried in vacuo, and GPC analysis was performed. Example Copolymerization Experiment. A 7 mL vial was charged with 3-O (15.2 mg, 0.0167 mmol), MTBD (2.4 μL, 0.0167 mmol), benzyl alcohol (1.04 μL, 0.00999 mmol) and C6D6 (250 μL). In a second 7 mL vial, VL (0.100 g, 0.999 mmol) and CL (0.144 g, 0.999 mmol) were dissolved in C6D6 (249 μL). The contents of vial 2 were transferred to the first via pipette and stirred until homogenous, approximately 5 sec. The contents were transferred to an NMR tube via pipette, and the reaction was monitored by 1H NMR. The reaction was quenched using benzoic acid (4.06 mg, 0.0333 mmol). Polymer was precipitated with the addition of hexanes. Supernatant was decanted and solid polymer was dried in vacuo, 91% yield (196 mg), Mn = 21,400; Mw/Mn = 1.21. 76 Example ROP of Lactide. L-lactide (72 mg, 0.5 mmol) and o-dichlorobenzene (0.5 mL) were added into a 7 mL vial and stirred until a homogenous solution was obtained. To a second 7 mL vial, benzyl alcohol (2.163 mg, 0.02 mmol), Me6TREN (0.008 mmol) and 3-O (0.008 mmol) were added. Contents from the first vial were transferred into vial 2 via Pasteur pipette. The contents were mixed and transferred to an NMR tube. Reaction progression was monitored by 1H NMR. After 30 min, the reaction had reached 55% conversion and was quenched with benzoic acid. The reaction was removed of volatiles and treated with hexanes/isopropanol (1:1) to dissolve monomer. The residual polymer was subjected to dialysis in DCM against methanol. Yield: 38 mg, 52%; Mn = 2,700; Mw/Mn = 1.11. 1.11. Example Transesterification Experiment. Ethyl acetate (100 mg. 1.14 mmol), 1-S (0.057 mmol) and C6D6 (0.22 mL) were added to a 7 ml glass vial. To a second 7 mL glass vial, benzyl alcohol (122.7 mg, 1.14 mmol), MTBD (0.057 mmol) and C6D6 (0.22 mL) were added. The contents of vial 2 were transferred via Pasteur pipette to vial 1, and the solution was stirred until homogeneous (1 min). The solution was transferred to an NMR tube, and reaction progression was monitored by 1H NMR. 77 77 RESULTS AND DISCUSSION The effects of bisTU on the ROP of δ-valerolactone (VL) and ε-caprolactone (CL) were evaluated, and the rate acceleration in the presence of 2-S versus 1-S is general to both lactone monomers. For the ROP of either VL or CL (2 M, 100 mg) from benzyl alcohol in C6D6, the application of 2-S/MTBD (2.5 mol % each) produces a rate acceleration over the traditional monothiourea (1-S/MTBD 5 mol % each) that is not associated with loss of reaction control, Table 3.1. The reactions retain the characteristics of “living” polymerizations, exhibiting a linear evolution of Mn versus conversion, first order consumption of monomer, Mn that is predictable by [M]o/[I]o and a living chain end that is susceptible to chain extension, see Figures 3.2-7. The imine base, DBU, and phosphazene base, BEMP, are also effective cocatalysts for the ROP of lactones (with 2-S), but the reaction is more active with MTBD cocatalyst, Table 3.1. ROP involving 2-S is suggested to proceed through an activated-TU mechanism, whereby one TU moiety activates the other, which in turn activates the monomer. The ROPs of VL and CL are first order in the consumption of monomer (Figure 3.3 and 3.10), which suggests one bisTU (2-S) molecule activating one monomer in the transition state. This is consistent with previous suggestions that H-bond-mediated ROP operates via dual activation of monomer by 1 and of alcohol chain end by base.1 Because H-bonds require no orbital overlap and are electrostatic in nature,26 we cannot rule out a dual-thiourea activated mechanism, Eq. 2.1. However, computational studies for the activation of lactones by 2-S suggest an activated-TU mechanism is preferred over a dual-thiourea 78 activation mechanism, Eq. 2.1; this assertion is also supported by the 2-S/alkylamine cocatalyzed ROP of lactide.22,27 The series of thiourea H-bond donating catalysts was extended to a trisTU H-bond donor, 3-S, but this catalyst exhibits significantly reduced activity versus 1-S or 2-S in the TU/base cocatalyzed ROP of lactones, Table 3.1. This suggests that simply adding TU moieties does not result in faster ROP. Geometry optimized DFT computations suggest that a stable conformation of 3-S is the C3 symmetric structure, see Figure 3.15 and 3.16. This calculated structure features a cyclic arrangement of the three TU moieties, each serving as a H-bond donor and a H-bond acceptor to each of the adjacent TU moieties with H-bond lengths of 2.61 ± 0.07 Å. RESULTS AND DISCUSSION We hypothesize that the added stability due to the three intramolecular H-bonds attenuates the activity of 3-S (versus 2-S). In contrast, the intramolecular H-bond activation in 2-S leaves a TU moiety available for catalysis. Additive effects from multiple TU moieties are found in nature,28 and such constructs have been observed to be beneficial to catalysis,22,29,30 although not universally so.24,31 Interested in extending the suite of H-bond-mediated catalysts, we noted that changing the C=S to the shorter C=O bond would be expected to disrupt the intramolecular H-bond network, freeing one urea moiety for catalysis. The trisurea H-bond donor (3-O) is predicted by DFT calculations to have much longer average H-bond lengths versus 3-S, 2.92 ± 0.81 Å. The application of the trisU (3-O) catalyst in combination with organic bases effects the fastest organocatalytic ROP of lactones that has been reported, yet the reaction remains highly controlled.3,17–21 The 3-O/MTBD (1.67 mol % each) catalyzed ROP of VL (2 M, 100 mg) from benzyl alcohol (2 mol %) proceeds to full conversion in 3 min, Table 3.2. The comparable reactions with 2-S/MTBD (2.5 mol % each) or 1-S/MTBD (5 mol % each) 79 achieve full conversion in 102 min or 2 h, respectively. The rate acceleration for the ROP of CL with 3-O/MTBD is even more remarkable; this reaction achieves full conversion in 26 min. This constitutes a marked rate acceleration versus 2-S or 1-S with MTBD, which achieves full conversion in 10 or 45 h, respectively, and the polydispersities for the 3- O/MTBD catalyzed ROP of VL or CL remain less than Mw/Mn = 1.07, Table 3.2. The 3- O mediated ROPs of both monomers are highly controlled, exhibiting the characteristics of “living” polymerizations, (see Figures 3.10 and 3.11). Initiation of a CL ROP from 1- pyrenebutanol produces PCL with overlapping refractive index and UV traces in the GPC, suggesting end-group fidelity; the “living” alcohol chain end is susceptible to chain extension by repeated additions of monomer, (see Figure 3.13). The 3-O/MTBD cocatalysts remain active at low concentration; full conversion for the ROP of VL (2 M, C6D6) from benzyl alcohol ([M]o/[I]o = 50) was achieved in 5 h at 0.25 mol % 3-O/MTBD loading, (see Table 3.4). The efficacy of 3-O/base cocatalysts for the ROP of other ester and carbonate monomers was evaluated. The 3-O/MTBD (1.67 mol %) cocatalysts are effective for the ROP of trimethylene carbonate (TMC). RESULTS AND DISCUSSION This reaction (100 mg TMC, 1 M in CH2Cl2) reaches 97% conversion in 1 min (Mn = 9,000; Mw/Mn = 1.05; [M]o/[I]o = 50), which is more active than the 1-S/DBU catalyzed ROP of TMC.5 For the ROP of LA, 3-O (with tris[2- (dimethylamino)ethyl]amine) exhibits a solvent incompatibility with LA and PLA, resulting in the precipitation of polymer or catalyst prior to full conversion (see Figure 3.17). The best conversion was achieved in o-dichlorobenzene, 55% in 30 min (Mn = 2,700; Mw/Mn = 1.11; [M]o/[I]o = 25; 52% yield). This is less active than our previously reported catalyst, 2-S, which reaches full conversion in minutes.22 MALDI analysis of the PLA 80 resulting from the ROP of LA shows only minor transesterification (m/z = ±72n; see Figure 3.17). A copolymerization of VL and CL was conducted with 3-O/MTBD. As determined by 1H NMR, the consumption of VL is almost complete prior to the incorporation of CL units, suggesting the formation of a gradient-copolymer (see Figure 3.12 and Experimental Section; Mn = 21,400; Mw/Mn = 1.29; 91% yield). The H-bond donor 3-O with MTBD is not active for the ROP of β-butyrolactone, which is consistent with other H-bonding ROP catalysts.8 resulting from the ROP of LA shows only minor transesterification (m/z = ±72n; see Figure 3.17). A copolymerization of VL and CL was conducted with 3-O/MTBD. As determined by 1H NMR, the consumption of VL is almost complete prior to the incorporation of CL units, suggesting the formation of a gradient-copolymer (see Figure 3.12 and Experimental Section; Mn = 21,400; Mw/Mn = 1.29; 91% yield). The H-bond donor 3-O with MTBD is not active for the ROP of β-butyrolactone, which is consistent with other H-bonding ROP catalysts.8 It is proposed here that 3-O/MTBD cocatalyzed ROP occurs via an activated-urea mechanism, whereby a single 3-O activates a lactone and MTBD activates an alcohol chain end through H-bonding, Scheme 3.2. A plot of observed rate constant (kobs) versus [3-O] for the ROP of VL from benzyl alcohol suggests that the ideal stoichiometry of the 3- O/MTBD catalyzed reaction is 1:1 (see Figure 3.14). Further, the 3-O/MTBD cocatalyzed ROP of VL is first order in monomer (see Figure 3.9), which suggests that a single 3-O molecule acting at one monomer is present in the transition state. RESULTS AND DISCUSSION This is consistent with previous reports that suggest that H-bond donors featuring multiple (thio)urea moieties activate one reagent prior to the TU-reagent complex undergoing further chemistry,22,32 and it is also consistent with a report of a urea-thiourea H-bond donating catalyst, which was proposed to be operative via an activated-(thio)urea mechanism.28 Indeed, 1H NMR spectra (in acetone) of 1-O, 2-O, and 3-O show a progressive downfield shift of the N−H protons, which can be interpreted to arise from stronger intramolecular H-bonding in 3-O and 2-O versus 1-O. A multiurea activated mechanism (e.g., Eq. 2.1), which is reminiscent of a solvophobic pocket, cannot be ruled out. However, the marked inefficacy toward ROP of 3-S which is geometrically able to adopt a conformation featuring strong intramolecular It is proposed here that 3-O/MTBD cocatalyzed ROP occurs via an activated-urea mechanism, whereby a single 3-O activates a lactone and MTBD activates an alcohol chain end through H-bonding, Scheme 3.2. A plot of observed rate constant (kobs) versus [3-O] for the ROP of VL from benzyl alcohol suggests that the ideal stoichiometry of the 3- O/MTBD catalyzed reaction is 1:1 (see Figure 3.14). Further, the 3-O/MTBD cocatalyzed ROP of VL is first order in monomer (see Figure 3.9), which suggests that a single 3-O molecule acting at one monomer is present in the transition state. This is consistent with previous reports that suggest that H-bond donors featuring multiple (thio)urea moieties activate one reagent prior to the TU-reagent complex undergoing further chemistry,22,32 81 H-bonds (see Figure 3.15 and 3.16), suggests that the activated-urea mechanism is the more robust proposal. Among catalysts for the ROP of lactones, the 3-O/base cocatalysts stand out due to the extremely rapid rate that they exhibit at room temperature. For comparison, we conducted the ROP of CL (2 M) from benzyl alcohol (1 mol %) with the bifunctional catalyst TBD, Table 3.2. The guanidine base, TBD (Figure 3.1), has been regarded as one of the most active organocatalysts available for the ROP of lactones.16 The TBD catalyzed ROP of CL from benzyl alcohol (Table 3.2, entry 12) proceeds to 93% conversion in 140 min (Mw/Mn = 1.37), whereas the same ROP with 3-O/MTBD (Table 3.2, entry 8) achieves 97% conversion in 26 min (Mw/Mn = 1.05). RESULTS AND DISCUSSION In small molecule transformations, urea H- bond donating catalysts have been observed to possess similar activity to their heavy chalcogen counterparts.33 The development of urea and thiourea H-bond donating catalysts continued apace until the turn of the millennium when several reports emerged that extolled the operational (e.g., increased solubility)34,35 and synthetic (e.g., higher yields and enantioselectivities)35–37 benefits of thioureas over ureas. In our estimation, the ubiquity of the thiourea motif in H-bond mediated transformations may be more due to the coincidental timing of these reports than any general superiority of thioureas over urea H-bonding catalysts. Indeed, ureas are more polar than thioureas and should be expected to be better H-bond activators,33 and in some catalysis applications, urea catalysts are clearly superior.38,39 The late Margaret Etter may have presaged our observation of 3-O as an effective H-bond donating catalyst in her characterization of aryl ureas featuring meta- electron withdrawing groups by noting that urea carbonyls are good H-bond acceptors.38 The urea versions of 2 and 1 were synthesized and evaluated for their efficacy in the ROP of VL (2 M, 100 mg, 1 equiv.) from benzyl alcohol (2 mol %) in C6D6. In general, all n-O The urea versions of 2 and 1 were synthesized and evaluated for their efficacy in the ROP of VL (2 M, 100 mg, 1 equiv.) from benzyl alcohol (2 mol %) in C6D6. In general, all n-O 82 (n = 1, 2, or 3) catalysts were more active than the corresponding n-S H-bond donors, Tables 3.1 and 3.2. For the 2-X (X = O, S, or OS) H-bond donors, the rate of ROP increases with the progressive substitution of O (versus S) and Mw/Mn remains low. These results suggest the increased utility of ureas versus thioureas for H-bond-mediated ROP. All reported urea catalysts are soluble under the desired reaction conditions with the exception of 2-O, which requires an extra equivalent of MTBD to become homogeneous in C6D6.40 A plot of the observed rate constant (kobs) versus [MTBD] for the ROP of CL from benzyl alcohol increases linearly under conditions [MTBD] ≤ [2-S], but becomes zero order in [MTBD] when [MTBD] > [2-S], (see Figure 3.7). This suggests that the proper stoichiometry of the 2-S/MTBD catalyzed reaction is 1:1. The catalysts (1−3 with MTBD) are all operative in CH2Cl2, CHCl3, and THF albeit with slightly reduced reaction rates or Mw/Mn (see Table 3.5). RESULTS AND DISCUSSION Preliminary studies suggest that these catalysts exhibit the same reactivity trends in small molecule transesterification and, hence, may have general applicability beyond ROP. The transesterification of ethyl acetate (1.6 M) with benzyl alcohol (1.6 M) was conducted in C6D6. Observed rate constants (kobs) at early reaction time were measured for each H-bond donor/MTBD cocatalyzed transesterification. These rate constants show the same trends in catalyst activity that were observed for the ROP reactions: 3-O is the most rapid catalyst and it is 1−2 orders of magnitude more rapid than 1-S, (see Table 3.3). This suggests a general role for the increased activation of esters by urea H-bond donors (versus thioureas), yet the slower rates for the transesterification of s- trans (versus s cis) esters accounts for the low rate of transesterification post A plot of the observed rate constant (kobs) versus [MTBD] for the ROP of CL from benzyl alcohol increases linearly under conditions [MTBD] ≤ [2-S], but becomes zero order in [MTBD] when [MTBD] > [2-S], (see Figure 3.7). This suggests that the proper stoichiometry of the 2-S/MTBD catalyzed reaction is 1:1. The catalysts (1−3 with MTBD) are all operative in CH2Cl2, CHCl3, and THF albeit with slightly reduced reaction rates or Mw/Mn (see Table 3.5). Preliminary studies suggest that these catalysts exhibit the same reactivity trends in small molecule transesterification and, hence, may have general applicability beyond ROP. The transesterification of ethyl acetate (1.6 M) with benzyl alcohol (1.6 M) was conducted in C6D6. Observed rate constants (kobs) at early reaction time were measured for each H-bond donor/MTBD cocatalyzed transesterification. These rate constants show the same trends in catalyst activity that were observed for the ROP reactions: 3-O is the most rapid catalyst and it is 1−2 orders of magnitude more rapid than 1-S, (see Table 3.3). This suggests a general role for the increased activation of esters by urea H-bond donors (versus thioureas), yet the slower rates for the transesterification of s- trans (versus s-cis) esters accounts for the low rate of transesterification post polymerization, (see Table 3.6). 83 CONCLUSION Urea H-bond donors in combination with base cocatalysts have been shown to be highly effective for the ROP of lactones. Despite being among the most rapid organocatalysts for ROP, the 3-O/MTBD cocatalyzed ROPs of VL and CL are among the most controlled polymerizations, exhibiting the characteristics of “living” polymerizations and producing polymers with narrow Mw/Mn. The source of the rate acceleration versus mono- and bisurea H-bond donors is proposed to arise from successively increased intramolecular H-bond activation with each additional urea moiety. The reintroduction of the urea motif of H-bond donors to the lexicon of organocatalytic (ROP) chemistry provides a rich diversity of catalyst scaffolds to explore in mono-, bis-, tris-, and poly-H- bond donors. Previous to the discovery of trisurea cocatalyzed ROP, one was forced to choose between a highly active or highly selective organocatalyst; this age is over. 84 LIST OF REFERENCES (1) Dove, A. P.; Pratt, R. C.; Lohmeijer, B. G. G.; Waymouth, R. M.; Hedrick, J. L. J. Am. Chem. Soc. 2005, 127 (40), 13798–13799. (2) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Chem. Rev. 2007, 107 (12). (3) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R. M. Macromolecules 2010, 43 (5), 2093–2107. (4) McKinlay, C. J.; Waymouth, R. M.; Wender, P. A. J. Am. Chem. Soc. 2016, 138 (10), 3510–3517. (5) Pratt, R. C.; Nederberg, F.; Waymouth, R. M.; Hedrick, J. L. Chem. Commun. 2008, No. 1, 114–116. (6) Sanders, D. P.; Fukushima, K.; Coady, D. J.; Nelson, A.; Fujiwara, M.; Yasumoto, M.; Hedrick, J. L. J. Am. Chem. Soc. 2010, 132 (42), 14724–14726. (7) Blake, T. R.; Waymouth, R. M. J. Am. Chem. Soc. 2014, 136 (26), 9252–9255. (8) Bas G. G. Lohmeijer Frank Leibfarth, John W. Logan, R. C. P.; David A. Long Fredrik Nederberg, Jeongsoo Choi, A. P. D.; Charles Wade and James L. Hedrick, R. M. W.; Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.; Waymouth, R. M.; Hedrick, J. L. Macromolecules 2006, 39 (25), 8574–8583. R. M. W.; Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D (9) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Lundberg, P. N. P.; Dove, A. P.; Li, H.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L. Macromolecules 2006, 39 (23), 7863–7871. 85 (10) Kiesewetter, E. T.; Randoll, S.; Radlauer, M.; Waymouth, R. M. J. Am. Chem. Soc. 2010, 132 (16), 5566–5567. (11) Cohen, A.; Kopilov, J.; Lamberti, M.; Venditto, V.; Kol, M. Macromolecules 2010, 43 (4), 1689–1691. (12) Miyake, G. M.; Chen, E. Y. X. Macromolecules 2011, 44 (11), 4116–4124. (13) Zhu, J.-B.; Chen, E. Y.-X. J. Am. Chem. Soc. 2015, 137 (39), 12506–12509. (14) Delcroix, D.; Martín-Vaca, B.; Bourissou, D.; Navarro, C. Macromolecules 2010, 43 (21), 8828–8835. (15) Makiguchi, K.; Kikuchi, S.; Satoh, T.; Kakuchi, T. J. Polym. Sci. Part A Polym. Chem. 2013, 51 (11), 2455–2463. (16) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Waymouth, R. M.; Hedrick, J. L. J. Am. Chem. Soc. 2006, 128 (14), 4556–4557. (17) Jeong, W.; Shin, E. J.; Culkin, D. A.; Hedrick, J. L.; Waymouth, R. M. J. LIST OF REFERENCES Am. Chem. Soc. 2009, 131 (13), 4884–4891. (18) Naumann, S.; Scholten, P. B. V; Wilson, J. A.; Dove, A. P. J. Am. Chem. Soc. 2015, 137 (45), 14439–14445. (19) Li, X.; Zhang, Q.; Li, Z.; Xu, S.; Zhao, C.; Chen, C.; Zhi, X.; Wang, H.; Zhu, N.; Guo, K. Polym. Chem. 2016, 7 (7), 1368–1374. (20) Guillaume, S. M.; Kirillov, E.; Sarazin, Y.; Carpentier, J.-F. Chem. – A Eur. J. 2015, 21 (22), 7988–8003. (21) Du, G.; Wei, Y.; Zhang, W.; Dong, Y.; Lin, Z.; He, H.; Zhang, S.; Li, X. Dalt. Trans. 2013, 42 (4), 1278–1286. 86 (22) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. T.; Kiesewetter, M. K. Macromolecules 2015, 48 (17), 6127–6131. (23) Bas G. G. Lohmeijer Frank Leibfarth, John W. Logan, R. C. P.; David A. Long Fredrik Nederberg, Jeongsoo Choi, A. P. D.; Charles Wade and James L. Hedrick, R. M. W.; Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.; Waymouth, R. M.; Hedrick, J. L. Macromolecules 2006, 39 (25), 8574–8583. L. Macromolecules 2006, 39 (25), 8574–8583. (24) Bertucci, M. A.; Lee, S. J.; Gagne, M. R. Chem. Commun. 2013, 49 (20), 2055– 2057. (25) Tripathi, C. B.; Mukherjee, S. J. Org. Chem. 2012, 77 (3), 1592–1598. (26) Anslyn, E. V.; Dougherty, D. A. In Modern Physical Organic Chemistry; University Science, 2006; pp 145–222. (27) Kazakov, O. I.; Kiesewetter, M. K. Macromolecules 2015, 48 (17), 6121–6126. (28) Jones, C. R.; Dan Pantoş, G.; Morrison, A. J.; Smith, M. D. Angew. Chemie Int. Ed. 2009, 48 (40), 7391–7394. (29) Shi, Y.; Lin, A.; Mao, H.; Mao, Z.; Li, W.; Hu, H.; Zhu, C.; Cheng, Y. Chem. – A Eur. J. 2013, 19 (6), 1914–1918. (30) Sohtome, Y.; Tanatani, A.; Hashimoto, Y.; Nagasawa, K. Tetrahedron Lett. 2004, 45 (29), 5589–5592. (31) Li, X.; Deng, H.; Luo, S.; Cheng, J.-P. European J. Org. Chem. 2008, 2008 (25), 4350–4356. (32) Breugst, M.; Houk, K. N. J. Org. Chem. 2014, 79 (13), 6302–6309. (33) Doyle, A. G.; Jacobsen, E. N. Chem. Rev. 2007, 107 (12), 5713–5743. 87 (34) Curran, D. P.; Kuo, L. H. Tetrahedron Lett. 1995, 36 (37), 6647–6650. (35) Wittkopp, A.; Schreiner, P. R. Chem. - A Eur. J. 2003, 9 (2), 407–414. (36) Wenzel, A. G.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124 (44), 12964–12965. (36) Wenzel, A. (40) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. K. Macromolecules 2014, 47, 7463–7468. LIST OF REFERENCES G.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124 (44), 12964–12965. (37) Vachal, P.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124 (34), 10012–10014. (37) Vachal, P.; Jacobsen, E. N. J. Am. Chem. Soc. 2002, 124 (34), 10012–10014. (38) Etter, M. C.; Panunto, T. W. J. Am. Chem. Soc. 1988, 110 (17), 5896–5897. (38) Etter, M. C.; Panunto, T. W. J. Am. Chem. Soc. 1988, 110 (17), 5896–5897. (39) Berkessel, A.; Cleemann, F.; Mukherjee, S.; Mueller, T. N.; Lex, J. Angew. Chemie - Int. Ed. 2005, 44 (5), 807–811. (40) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. K. Macromolecules 2014, 47, 7463–7468. 88 Entry Monomer TU (mol%) Base (mol%) Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 VL 1-S (5%) MTBD (5%) 110 94 8,300 1.06 2 2-S (2.5%) MTBD (2.5%) 80 90 6,800 1.07 3 2-S (2.5%) BEMP (2.5%) 84 91 8,900 1.06 4 2-S (2.5%) DBU (2.5%) 90 86 8,400 1.05 5 3-S (1.67%) MTBD (1.67%) 230 90 7 600 1.06 6 CL 1-S (5%) MTBD (5%) 45 h 90 7,200 1.09 7 2-S (2.5%) MTBD (2.5%) 10 h 89 7,200 1.11 8 3-S (1.67%) MTBD (1.67%) 42 h 55 6,100 1.07 Table 3.1. MTBD and 1-S, 2-S or 3-S catalyzed ROP of VL and CL. Reaction conditions: VL or CL (1.0 mmol, 1 equiv., 2M), benzyl alcohol (2 mol%), C6D6. a) monomer conversion was determined via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. LIST OF REFERENCES 89 Entry Monomer TU or U (mol%) [M]o/[I]o Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 VL 1-O (5%) 50 70 90 6,100 1.08 2 2-OS (2.5%) 50 88 90 8,100 1.07 3 2-O (2.5%) d 50 34 90 8,000 1.07 4 3-O (1.67%) 50 3 89 7,500 1.07 5 100 6 90 15,000 1.04 6 200 10 92 28,600 1.02 7 500 16 92 41,500 1.02 8 CL 3-O (1.67%) 50 26 97 7,900 1.05 9 100 57 94 18,500 1.02 10 200 116 94 30,700 1.03 11 500 166 93 58,600 1.03 12d TBD (1.67%) 50 140 93 10,400 1.37 Entry Monomer TU or U (mol%) [M]o/[I]o Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 VL 1-O (5%) 50 70 90 6,100 1.08 2 2-OS (2.5%) 50 88 90 8,100 1.07 3 2-O (2.5%) d 50 34 90 8,000 1.07 4 3-O (1.67%) 50 3 89 7,500 1.07 5 100 6 90 15,000 1.04 6 200 10 92 28,600 1.02 7 500 16 92 41,500 1.02 8 CL 3-O (1.67%) 50 26 97 7,900 1.05 9 100 57 94 18,500 1.02 10 200 116 94 30,700 1.03 11 500 166 93 58,600 1.03 12d TBD (1.67%) 50 140 93 10,400 1.37 Table 3.2. 1-O, 2-O or 3-O and MTBD cocatalyzed ROP of lactones. Reaction conditions: VL or CL (1.0 mmol, 1 equiv., 2M), urea or thiourea (given mol%), MTBD (mol% matched to H-bond donor). a) Monomer conversion monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. c) 2-O (2.5 mol%) and MTBD (5 mol%) cocatalysts. d) no (thio)urea or MTBD cocatalysts were used in this run. Table 3.2. 1-O, 2-O or 3-O and MTBD cocatalyzed ROP of lactones. Reaction conditions: VL or CL (1.0 mmol, 1 equiv., 2M), urea or thiourea (given mol%), MTBD (mol% matched to H-bond donor). a) Monomer conversion monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. c) 2-O (2.5 mol%) and MTBD (5 mol%) cocatalysts. d) no (thio)urea or MTBD cocatalysts were used in this run. LIST OF REFERENCES 90 Entry TU or U (mol%) kobs (1/min)a [EA]eq (M)b 1 1-S (5%) 0.000 80 1.08 2 1-O (5%) 0.003 57 0.88 3 2-S (2.5%) 0.000 55 0.99 4 2-O (2.5%) 0.004 10 0.99 5 3-S (1.67%) 0.000 61 1.19 6 3-O (1.67%) 0.002 11 0.89 Entry TU or U (mol%) kobs (1/min)a [EA]eq (M)b 1 1-S (5%) 0.000 80 1.08 2 1-O (5%) 0.003 57 0.88 3 2-S (2.5%) 0.000 55 0.99 4 2-O (2.5%) 0.004 10 0.99 5 3-S (1.67%) 0.000 61 1.19 6 3-O (1.67%) 0.002 11 0.89 Entry TU or U (mol%) kobs (1/min)a [EA]eq (M)b 1 1-S (5%) 0.000 80 1.08 2 1-O (5%) 0.003 57 0.88 3 2-S (2.5%) 0.000 55 0.99 4 2-O (2.5%) 0.004 10 0.99 5 3-S (1.67%) 0.000 61 1.19 6 3-O (1.67%) 0.002 11 0.89 Table 3.3. Transesterification of ethyl acetate. a) Observed rate constant for the first order disappearance of [EA] versus time. Rate constant was extracted from the linear portion of the data, up to ~20% conversion. b) Concentration of ethyl acetate remaining at equilibrium. 91 Entry mol% cats. (each) Time (min) Conva Mnb Mw/Mnb 1 1.67 3 89 7,500 1.07 2 1 10 91 7,100 1.07 3 0.5 40 93 7,700 1.07 4 0.25 300 93 7,200 1.07 5 0.1 24hr 0 NA NA Table 3.4. Low 3-O/MTBD Cocatalyst Loadings in the ROP of VL. Reaction conditions: VL(0.998 mmol, 1 equiv., 2M), C6D6 and benzyl alcohol (2 mol%). a) Monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 92 Table 3.5. Solvent Screen in the 3-O/MTBD Cocatalyzed ROP of VL. LIST OF REFERENCES Reaction conditions: VL (0.998 mmol, 1 equiv., 2M), 1 mol% benzyl alcohol, a) monomer conversion was Entry Solvent Time (min) Conva Mnb Mw/Mnb 1 C6D6 4 91 12,200 1.04 2 CH2Cl2 5 90 14,800 1.05 3 CHCl3 5 90 7,000 1.07 4 Cl-C6H5 4 93 10,000 1.08 5 THF 5 89 13,600 1.05 Entry Solvent Time (min) Conva Mnb Mw/Mnb 1 C6D6 4 91 12,200 1.04 2 CH2Cl2 5 90 14,800 1.05 3 CHCl3 5 90 7,000 1.07 4 Cl-C6H5 4 93 10,000 1.08 5 THF 5 89 13,600 1.05 Entry Solvent Time (min) Conva Mnb Mw/Mnb 1 C6D6 4 91 12,200 1.04 2 CH2Cl2 5 90 14,800 1.05 3 CHCl3 5 90 7,000 1.07 4 Cl-C6H5 4 93 10,000 1.08 5 THF 5 89 13,600 1.05 Table 3.5. Solvent Screen in the 3-O/MTBD Cocatalyzed ROP of VL. Reaction conditions: VL (0.998 mmol, 1 equiv., 2M), 1 mol% benzyl alcohol, a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC. Table 3.5. Solvent Screen in the 3-O/MTBD Cocatalyzed ROP of VL. Reaction conditions: VL (0.998 mmol, 1 equiv., 2M), 1 mol% benzyl alcohol, a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC. 93 e 3.6. Post-polymerization Transesterification in 3-O/MTBD Cocatalyzed ROP. tion conditions: VL (0.998 mmol, 1 equiv., 2M), 2 mol% benzyl alcohol, a) monomer ersion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC. Entry Monomer Time (min) Conva Mnb Mw/Mnb 1 VL 3 93 6,200 1.10 2 VL 6 93 6,300 1.12 3 VL 60 94 6,600 1.21 4 CL 25 91 9,000 1.04 5 CL 60 98 10,000 1.05 6 CL 120 99 10,000 1.09 Entry Monomer Time (min) Conva Mnb Mw/Mnb 1 VL 3 93 6,200 1.10 2 VL 6 93 6,300 1.12 3 VL 60 94 6,600 1.21 4 CL 25 91 9,000 1.04 5 CL 60 98 10,000 1.05 6 CL 120 99 10,000 1.09 Table 3.6. Post-polymerization Transesterification in 3-O/MTBD Cocatalyzed ROP. Reaction conditions: VL (0.998 mmol, 1 equiv., 2M), 2 mol% benzyl alcohol, a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC. Table 3.6. Post-polymerization Transesterification in 3-O/MTBD Cocatalyzed ROP. Reaction conditions: VL (0.998 mmol, 1 equiv., 2M), 2 mol% benzyl alcohol, a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC. LIST OF REFERENCES Eq. 2.1. Intramolecular conformational arrangement of 2-S done computationally. Eq. 2.1. Intramolecular conformational arrangement of 2-S done computationally. 94 Scheme 3.1. Highly active and highly selective H-bond donor 3-O. Scheme 3.1. Highly active and highly selective H-bond donor 3-O. Scheme 3.2. Proposed mechanism for 3-O/MTBD catalyzed ROP. Scheme 3.2. Proposed mechanism for 3-O/MTBD catalyzed ROP. 95 Figure 3.1. Base and (thio)urea cocatalysts evaluated for ROP. Figure 3.1. Base and (thio)urea cocatalysts evaluated for ROP. Figure 3.2. Mn versus conversion for the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (2.994 mmol, 1 equiv., 1M in C6D6), benzyl alcohol (2 mol%, 0.0598 mmol), MTBD (5 mol%, 0.1497 mmol) and 2-S (5 mol%, 0.1496 mmol). (blue is Mn, red is Mw/Mn). Figure 3.2. Mn versus conversion for the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (2.994 mmol, 1 equiv., 1M in C6D6), benzyl alcohol (2 mol%, 0.0598 mmol), MTBD (5 mol%, 0.1497 mmol) and 2-S (5 mol%, 0.1496 mmol). (blue is Mn, red is Mw/Mn). 96 Figure 3.3. First order evolution of [VL] versus time for the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2.0 mol%, 0.0199 mmol), MTBD (5.0 mol%, 0.0499 mmol) and 2-S (5.0 mol%, 0.0499 mmol). Figure 3.3. First order evolution of [VL] versus time for the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2.0 mol%, 0.0199 mmol), MTBD (5.0 mol%, 0.0499 mmol) and 2-S (5.0 mol%, 0.0499 mmol). Figure 3.4. Mn versus [VL]o/[I]o for the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 1M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (5.0 mol%, 0.0499 mmol) and 2-S (5.0 mol%, 0.0499 mmol). Figure 3.4. Mn versus [VL]o/[I]o for the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 1M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (5.0 mol%, 0.0499 mmol) and 2-S (5.0 mol%, 0.0499 mmol). 97 Figure 3.5. GPC traces of the polymer resulting from the 2-S/MTBD (5 mol% each, 0.0499 mmol) cocatalyzed ROP and subsequent chain extension of VL (0.999 mmol, then 0.999 mmol more) from 1-pyrenebutanol (0.0199 mmol) in C6D6 (999 μL). Figure 3.5. GPC traces of the polymer resulting from the 2-S/MTBD (5 mol% each, 0.0499 mmol) cocatalyzed ROP and subsequent chain extension of VL (0.999 mmol, then 0.999 mmol more) from 1-pyrenebutanol (0.0199 mmol) in C6D6 (999 μL). LIST OF REFERENCES 98 Figure 3.6. Observed rate constant (kobs, min-1) versus [MTBD] in the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv, 1M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (2.5 mol%, 0.025M). Figure 3.6. Observed rate constant (kobs, min-1) versus [MTBD] in the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv, 1M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (2.5 mol%, 0.025M). Figure 3.6. Observed rate constant (kobs, min-1) versus [MTBD] in the 2-S/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv, 1M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (2.5 mol%, 0.025M). Figure 3.7. Observed rate constant (kobs, h-1) versus [2-S] in the 2-S/MTBD catalyzed ROP of CL. Conditions: CL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), 2-S (0.05M). Figure 3.7. Observed rate constant (kobs, h-1) versus [2-S] in the 2-S/MTBD catalyzed ROP of CL. Conditions: CL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), 2-S (0.05M). 99 Figure 3.8. Mn vs conversion of VL for the 3-O/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (1mol%, 0.0199 mmol), MTBD (1.67 mol%, 0.0166 mmol) and 3-O (1.67 mol%, 0.0166 mmol). (blue is Mn, red is Mw/Mn). Figure 3.8. Mn vs conversion of VL for the 3-O/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (1mol%, 0.0199 mmol), MTBD (1.67 mol%, 0.0166 mmol) and 3-O (1.67 mol%, 0.0166 mmol). (blue is Mn, red is Mw/Mn). Figure 3.8. Mn vs conversion of VL for the 3-O/MTBD catalyzed ROP of VL. Co Figure 3.9. First order evolution of [VL] vs time for the 3-O/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (1.67 mol%, 0.0166 mmol) and 3-O (1.67 mol%, 0.0166 mmol). Figure 3.9. First order evolution of [VL] vs time for the 3-O/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (1.67 mol%, 0.0166 mmol) and 3-O (1.67 mol%, 0.0166 mmol). 100 Figure 3.10. Mn vs conversion for the 3-O/MTBD catalyzed ROP of CL. LIST OF REFERENCES Conditions: CL (1.752 mmol, 1 equiv, 2M in C6D6), benzyl alcohol (2 mol%, 0.035 mmol), MTBD (1.67 mol%, 0.029 mmol) and 3-O (1.67 mol%, 0.029 mmol). (blue is Mn, red is Mw/Mn). Figure 3.10. Mn vs conversion for the 3-O/MTBD catalyzed ROP of CL. Conditions: CL (1.752 mmol, 1 equiv, 2M in C6D6), benzyl alcohol (2 mol%, 0.035 mmol), MTBD (1.67 mol%, 0.029 mmol) and 3-O (1.67 mol%, 0.029 mmol). (blue is Mn, red is Mw/Mn). Figure 3.11. First order evolution of [CL] vs time for the 3-O/MTBD catalyzed ROP of CL. Conditions: CL (1.752 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.035 mmol), MTBD (1.67 mol%, 0.029 mmol) and 3-O (1.67 mol%, 0.029 mmol). Figure 3.11. First order evolution of [CL] vs time for the 3-O/MTBD catalyzed ROP of CL. Conditions: CL (1.752 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.035 mmol), MTBD (1.67 mol%, 0.029 mmol) and 3-O (1.67 mol%, 0.029 mmol). 101 Figure 3.12. First order evolution of [CL] and [VL] vs time for the 3-O/MTBD catalyzed copolymerization of CL. Conditions: CL (1.752 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.035 mmol), MTBD (1.67 mol%, 0.029 mmol) and 3-O (1.67 mol%, 0.029 mmol). Figure 3.12. First order evolution of [CL] and [VL] vs time for the 3-O/MTBD catalyzed copolymerization of CL. Conditions: CL (1.752 mmol, 1 equiv., 2M in C6D6), benzyl alcohol (2 mol%, 0.035 mmol), MTBD (1.67 mol%, 0.029 mmol) and 3-O (1.67 mol%, 0.029 mmol). Figure 3.13. GPC traces of the polymer resulting from the 3-O/MTBD (1.67 mol% each, 0.015 mmol) cocatalyzed ROP and subsequent chain extension of CL (0.876 mmol, then 1.1723 mmol more) from 1-pyrenebutanol (0.035 mmol) in C6D6 (219 μL). Figure 3.13. GPC traces of the polymer resulting from the 3-O/MTBD (1.67 mol% each, 0.015 mmol) cocatalyzed ROP and subsequent chain extension of CL (0.876 mmol, then 1.1723 mmol more) from 1-pyrenebutanol (0.035 mmol) in C6D6 (219 μL). 102 Figure 3.14. Observed rate constant (kobs, min-1) vs [3-O] in the 3-O/MTBD catalyzed ROP of VL. Conditions: VL (0.999 mmol, 1 equiv., 0.5M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (1.67 mol%, 0.0166 mmol, 0.008 M). Figure 3.14. Observed rate constant (kobs, min-1) vs [3-O] in the 3-O/MTBD catalyzed ROP of VL. Figure 3.20. 13C NMR (75 MHz, DMSO-d6) of 2-O. N H N H N H N H O O CF3 CF3 CF3 F3C Figure 3.22. 1H NMR (300 MHz, DMSO-d6) of 2-OS. Figure 3.23. 13C NMR (75 MHz, acetone-d6) of 2-OS. N H N H N H N H S O CF3 CF3 CF3 F3C LIST OF REFERENCES Conditions: VL (0.999 mmol, 1 equiv., 0.5M in C6D6), benzyl alcohol (2 mol%, 0.0199 mmol), MTBD (1.67 mol%, 0.0166 mmol, 0.008 M). 103 Figure 3.15. DFT B3LYP//6-31G geometry optimized structures of 3-S. Figure 3.15. DFT B3LYP//6-31G geometry optimized structures of 3-S. Figure 3.16. DFT B3LYP//6-31G geometry optimized structures of 3-O. Figure 3.16. DFT B3LYP//6-31G geometry optimized structures of 3-O. 104 Figure 3.17. MALDI-TOF of the PLA resulting from the 3-O/(tris[2- (dimethylamino)ethyl]amine) catalyzed ROP of L-LA. Figure 3.17. MALDI-TOF of the PLA resulting from the 3-O/(tris[2- (dimethylamino)ethyl]amine) catalyzed ROP of L-LA. 105 Figure 3.18. Downfield half of the 1H NMR spectra (acetone + trace benzene-d6 (lock), 400 MHz) of (upper) 1-O, (middle) 2-O, and (lower) 3-O. The progressive downfield shift of the NH protons is indicative of increased (2-O versus 3-O) intramolecular H-bonding. Figure 3.18. Downfield half of the 1H NMR spectra (acetone + trace benzene-d6 (lock), 400 MHz) of (upper) 1-O, (middle) 2-O, and (lower) 3-O. The progressive downfield shift of the NH protons is indicative of increased (2-O versus 3-O) intramolecular H-bonding. 106 gure 3.19. 1H NMR (300 MHz, DMSO-d6) of 2-O. N H N H N H N H O O CF3 CF3 CF3 F3C 107 Figure 3.19. 1H NMR (300 MHz, DMSO-d6) of 2-O. N H N H N H N H O O CF3 CF3 CF3 F3C Figure 3.19. 1H NMR (300 MHz, DMSO-d6) of 2-O. Figure 3.19. 1H NMR (300 MHz, DMSO-d6) of 2-O. 107 Figure 3.20. 13C NMR (75 MHz, DMSO-d6) of 2-O. N H N H N H N H O O CF3 CF3 CF3 F3C Figure 3.20. 13C NMR (75 MHz, DMSO-d6) of 2-O. 108 Figure 3.21. 1H NMR (300 MHz, DMSO-d6) of 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane. N H N H NH2 S CF3 F3C Figure 3.21. 1H NMR (300 MHz, DMSO-d6) of 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane. Figure 3.21. 1H NMR (300 MHz, DMSO-d6) of 1-[3,5-bis(trifluoromethyl)phenyl thiourea]-3-aminopropane. 109 igure 3.22. 1H NMR (300 MHz, DMSO-d6) of 2-OS. N H N H N H N H S O CF3 CF3 CF3 F3C Figure 3.22. 1H NMR (300 MHz, DMSO-d6) of 2-OS. 110 Figure 3.23. 13C NMR (75 MHz, acetone-d6) of 2-OS. N H N H N H N H S O CF3 CF3 CF3 F3C 111 4. 1H NMR (300 MHz, acetone-d6) of 3-O. N N H N H CF3 F3C O N H N H CF3 CF3 O NH NH F3C CF3 O Figure 3.24. Figure 3.26. 1H NMR (300 MHz, acetone-d6) of 3-S. LIST OF REFERENCES 1H NMR (300 MHz, acetone-d6) of 3-O. N N H N H CF3 F3C O N H N H CF3 CF3 O NH NH F3C CF3 O Figure 3.24. 1H NMR (300 MHz, acetone-d6) of 3-O. 112 Figure 3.25. 13C NMR (75 MHz, acetone-d6) of 3-O. N N H N H CF3 F3C O N H N H CF3 CF3 O NH NH F3C CF3 O F3C Figure 3.25. 13C NMR (75 MHz, acetone-d6) of 3-O. 113 114 Figure 3.26. 1H NMR (300 MHz, acetone-d6) of 3-S. N N H N H CF3 F3C S N H N H CF3 CF3 S NH NH F3C CF3 S F3C Figure 3.26. 1H NMR (300 MHz, acetone-d6) of 3-S. 114 114 N N H N H CF3 F3C S N H N H CF3 CF3 S NH NH F3C CF3 S Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. N N H N H CF3 F3C S N H N H CF3 CF3 S NH NH F3C CF3 S Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. 115 Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Solvation: dichloromethane [SM8] Free Energy of Solvation : -111.5381226 kJ/mol SCF total energy: -3369.3171898 hartrees SPARTAN '14 Properties Program: (Win/64b) Release 1.1.8 Use of molecular symmetry disabled Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C1 3.0236320 1.8782697 -1.5793812 2 S S1 2.4865855 3.2691914 -2.3431680 3 N N1 4.1615953 1.8195168 -0.8199347 4 H H4 4.3807088 0.9334757 -0.3754557 5 N N2 2.4147858 0.6395117 -1.6440053 6 H H3 2.8892309 -0.0770555 -1.1048418 7 C C2 1.1325107 0.2212543 -2.0380878 8 C C4 -1.4114202 -0.8709745 -2.5894579 9 C C3 0.9092347 -1.1643141 -1.9771559 10 C C6 0.0704432 1.0587976 -2.4057200 11 C C5 -1.1820850 0.5015117 -2.6687443 12 C C7 -0.3459752 -1.7004420 -2.2407846 13 H H6 1.7244618 -1.8277314 -1.7058464 14 H H7 0.2227144 2.1268062 -2.4722793 Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Job type: Single point. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Solvation: dichloromethane [SM8] Free Energy of Solvation : -77.8518861 kJ/mol SCF total energy: -3369.3245007 hartrees SPARTAN '14 Properties Program: (Win/64b) Release 1.1.8 Use of molecular symmetry disabled Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C1 0.6186602 3.4506147 -0.4942311 2 S S1 -0.1839783 4.1767400 0.8241190 3 N N1 1.9063132 3.7081440 -0.8145935 4 H H4 2.3364448 3.0985513 -1.5079477 5 N N2 0.0531547 2.5251892 -1.3346437 6 H H3 0.7188179 2.0849533 -1.9741357 7 C C2 -1.1547538 1.8052874 -1.2340656 8 C C4 -3.4580090 0.1851075 -1.1642309 9 C C3 -1.1316466 0.5028274 -1.7487208 10 C C6 -2.3542935 2.3143430 -0.7172609 11 C C5 -3.4821723 1.4924413 -0.6759791 12 C C7 -2.2707223 -0.2955969 -1.7112994 13 H H6 -0.2138590 0.1114105 -2.1743285 14 H H7 -2.4085778 3.3304717 -0.3544325 15 H H10 -4.3392201 -0.4410508 -1.1138825 Activated-TU plus VL in DCM Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Solvation: dichloromethane [SM8] Free Energy of Solvation : -77.8518861 kJ/mol SCF total energy: -3369.3245007 hartrees SPARTAN '14 Properties Program: (Win/64b) Release 1.1.8 Use of molecular symmetry disabled Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C1 0.6186602 3.4506147 -0.4942311 2 S S1 -0.1839783 4.1767400 0.8241190 3 N N1 1.9063132 3.7081440 -0.8145935 4 H H4 2 3364448 3 0985513 1 5079477 Activated-TU plus VL in DCM Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Activated-TU plus VL in DCM Activated-TU plus VL in DCM Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Job type: Single point. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Solvation: dichloromethane [SM8] Free Energy of Solvation : -111.5381226 kJ/mol SCF total energy: -3369.3171898 hartrees SPARTAN '14 Properties Program: (Win/64b) Release 1.1.8 Use of molecular symmetry disabled Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C1 3.0236320 1.8782697 -1.5793812 2 S S1 2.4865855 3.2691914 -2.3431680 3 N N1 4.1615953 1.8195168 -0.8199347 4 H H4 4.3807088 0.9334757 -0.3754557 5 N N2 2.4147858 0.6395117 -1.6440053 6 H H3 2.8892309 -0.0770555 -1.1048418 7 C C2 1.1325107 0.2212543 -2.0380878 8 C C4 -1.4114202 -0.8709745 -2.5894579 9 C C3 0.9092347 -1.1643141 -1.9771559 10 C C6 0.0704432 1.0587976 -2.4057200 11 C C5 -1.1820850 0.5015117 -2.6687443 12 C C7 -0.3459752 -1.7004420 -2.2407846 13 H H6 1.7244618 -1.8277314 -1.7058464 14 H H7 0.2227144 2.1268062 -2.4722793 116 15 H H10 -2.3930742 -1.2817497 -2.7895526 16 C C10 4.9143846 2.9692703 -0.3332953 17 H H11 4.5344106 3.8436800 -0.8652759 18 H H14 5.9699149 2.8446241 -0.6074915 19 C C12 3.4027906 3.4954232 1.7271696 20 H H15 2.9363627 4.2861218 1.1280701 21 H H18 3.4772296 3.8782160 2.7517105 22 N N3 2.5369424 2.3192087 1.7215030 23 H H20 2.9203006 1.4691364 1.3284096 24 C C13 1.2067155 2.3544009 1.9901580 25 N N4 0.6109505 1.1232290 1.7699492 26 H H22 1.2548396 0.3613405 1.5797141 27 C C14 -0.7331750 0.7225280 1.7506543 28 C C15 -3.3546714 -0.3089651 1.5817556 29 C C16 -1.8288593 1.5927093 1.6175975 30 C C17 -0.9686276 -0.6575333 1.7958613 31 C C18 -2.2638136 -1.1643464 1.7001179 32 C C19 -3.1155194 1.0670046 1.5420799 33 H H21 -1.6699138 2.6600289 1.5790835 34 H H23 -0.1285243 -1.3372930 1.8897131 35 H H26 -4.3634555 -0.6988767 1.5185011 36 S S2 0.4404338 3.7350223 2.5568843 37 C C9 4.8032256 3.1640599 1.1907223 38 H H5 5.2007974 2.2820671 1.7151565 39 H H66 5.4672919 3.9953370 1.4584577 40 C C8 -2.3348816 1.4323499 -2.9566642 41 C C11 -0.5287562 -3.1927444 -2.1923754 42 C C20 -2.4486148 -2.6540574 1.7596442 43 C C21 -4.2949114 1.9896479 1.3633098 44 F F1 -1.5855148 -3.2933792 0.9172092 45 F F2 -3.6909519 -3.0390037 1.4192127 46 F F3 -2.1937480 -3.1488662 2.9941791 47 F F4 -5.2867990 1.6844871 2.2341956 48 F F5 -4.8239995 1.8788063 0.1239271 49 F F6 -3.9711423 3.2821442 1.5505404 50 F F7 -1.9478724 2.4963754 -3.6880106 51 F F8 -2.8730393 1.9085369 -1.8120448 52 F F9 -3.3238961 0.8070089 -3.6323087 53 F F10 -1.7988703 -3.5416596 -1.9232477 54 F F11 -0.1879272 -3.7820470 -3.3600940 55 F F12 0.2611471 -3.7628518 -1.2373203 56 O O1 3.4367224 -0.7462457 0.7948371 57 C C22 3.2789749 -1.9494030 0.9874045 58 O O2 2.0981603 -2.3565783 1.4442692 59 C C23 4.3678470 -2.9558988 0.6627870 60 H H1 5.3107962 -2.5327712 1.0237134 61 C C25 1.7615096 -3.7600728 1.7091266 62 H H2 1.8528809 -3.8888231 2.7924874 63 C C24 4.1149410 -4.3742363 1.1861986 64 C C26 2.6466391 -4.7344629 0.9547784 65 H H9 4.4431817 -2.9637701 -0.4338844 66 H H12 0.7120346 -3.8367938 1.4267382 67 H H16 4.7842542 -5.0791288 0.6833204 68 H H17 4.3433540 -4.4264428 2.2584168 69 H H19 2.4193356 -5.7464532 1.3082089 70 H H24 2.4038772 -4.7036553 -0.1143083 117 Activated-TU plus VL in DCM Job type: Single point. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Method: RB3LYP Basis set: 6-31G Number of shells: 258 Number of basis functions: 818 Multiplicity: 1 Solvation: dichloromethane [SM8] Free Energy of Solvation : -77.8518861 kJ/mol SCF total energy: -3369.3245007 hartrees SPARTAN '14 Properties Program: (Win/64b) Release 1.1.8 Use of molecular symmetry disabled Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C1 0.6186602 3.4506147 -0.4942311 2 S S1 -0.1839783 4.1767400 0.8241190 3 N N1 1.9063132 3.7081440 -0.8145935 4 H H4 2.3364448 3.0985513 -1.5079477 5 N N2 0.0531547 2.5251892 -1.3346437 6 H H3 0.7188179 2.0849533 -1.9741357 7 C C2 -1.1547538 1.8052874 -1.2340656 8 C C4 -3.4580090 0.1851075 -1.1642309 9 C C3 -1.1316466 0.5028274 -1.7487208 10 C C6 -2.3542935 2.3143430 -0.7172609 11 C C5 -3.4821723 1.4924413 -0.6759791 12 C C7 -2.2707223 -0.2955969 -1.7112994 13 H H6 -0.2138590 0.1114105 -2.1743285 14 H H7 -2.4085778 3.3304717 -0.3544325 15 H H10 -4.3392201 -0.4410508 -1.1138825 Solvation: dichloromethane [SM8] 118 16 C C10 2.8027843 4.6914441 -0.2147555 17 H H11 2.2071707 5.3416025 0.4302030 18 H H14 3.1967392 5.3095899 -1.0312663 19 C C12 3.6869285 3.6006471 1.9812784 20 H H15 3.4926792 4.4774002 2.6093305 21 H H18 4.5555740 3.0769033 2.3864867 22 N N3 2.5249512 2.7335360 2.1290191 23 H H20 1.6271786 3.2234830 2.1204578 24 C C13 2.5094968 1.3974966 1.9057752 25 N N4 1.2256902 0.8893739 1.9753434 26 H H22 0.5039979 1.5998622 2.0720812 27 C C14 0.6984366 -0.4058478 1.8631176 28 C C15 -0.5781053 -2.9147839 1.6314277 29 C C16 1.4548726 -1.5782301 1.7537946 30 C C17 -0.7080836 -0.5063312 1.8791934 31 C C18 -1.3281747 -1.7434602 1.7746406 32 C C19 0.8068657 -2.8102748 1.6216030 33 H H21 2.5337814 -1.5241645 1.7735517 34 H H23 -1.3157391 0.3883894 1.9709278 35 H H26 -1.0674423 -3.8761542 1.5331906 36 S S2 3.9137885 0.5130018 1.5863986 37 C C9 3.9823780 4.0499365 0.5409167 38 H H5 4.3622406 3.1976526 -0.0342216 39 H H66 4.7977458 4.7836087 0.5904183 40 C C8 -4.7350832 2.0133800 -0.0202682 41 C C11 -2.1500066 -1.7036538 -2.2285185 42 C C20 -2.8297569 -1.8589628 1.8102446 43 C C21 1.6636565 -4.0277290 1.4065496 44 F F1 -3.4379221 -0.6675959 1.9789275 45 F F2 -3.2341444 -2.6701329 2.8116355 46 F F3 -3.3103570 -2.3975544 0.6612719 47 F F4 2.2021271 -4.0279014 0.1506788 48 F F5 0.9720086 -5.1765982 1.5310462 49 F F6 2.7028943 -4.0758489 2.2624221 50 F F7 -4.7381778 1.7499162 1.3069433 51 F F8 -5.8465881 1.4468122 -0.5385132 52 F F9 -4.8560965 3.3509353 -0.1561754 53 F F10 -3.3293734 -2.3455705 -2.2770265 54 F F11 -1.6167016 -1.7291299 -3.4757367 55 F F12 -1.3143128 -2.4466695 -1.4544742 56 C C27 3.2831624 -2.6893528 -2.7877261 57 C C26 4.2042838 -1.8989580 -1.8562757 58 C C23 1.8607816 -2.1796104 -2.6467576 59 H H1 1.4433504 -2.4372747 -1.6707839 60 O O3 1.7549042 -0.7255085 -2.7677807 61 C C24 4.1439872 -0.4106729 -2.2166165 62 H H2 4.5853662 0.2219220 -1.4415431 63 C C25 2.7476791 0.1196308 -2.4682591 64 O O4 2.4921707 1.3173809 -2.4670203 65 H H19 4.7032344 -0.2167170 -3.1438153 66 H H24 1.1890437 -2.5577159 -3.4195913 67 H H25 3.8858802 -2.0439632 -0.8171413 68 H H27 5.2374791 -2.2545185 -1.9257839 69 H H29 3.6188117 -2.5973989 -3.8292343 70 H H30 3.2857759 -3.7555044 -2.5347410 119 3-S vacuum 3-S vacuum Job type: Single point Job type: Single point. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Method: RB3LYP Basis set: 6-31G Number of shells: 328 Number of basis functions: 1062 Multiplicity: 1 SCF total energy: -4648.8994977 hartrees Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C2 2.7640909 1.5991396 -1.8183489 2 S S1 2.7485580 3.1262167 -1.0447103 3 N N1 3.9056042 0.9614619 -2.1515710 4 H H5 3.8171067 0.0156768 -2.5219318 5 N N2 1.6401143 0.8812915 -2.1324001 6 H H6 1.7976801 -0.1118807 -2.3241979 7 C C3 0.3013170 1.2854719 -2.3201684 8 C C4 -2.3985949 1.8898648 -2.8274305 9 C C5 -0.6568049 0.2633989 -2.3905650 10 C C6 -0.0979427 2.6145578 -2.5214042 11 C C7 -1.4411826 2.8999161 -2.7585829 12 C C8 -1.9912618 0.5700573 -2.6483517 13 H H7 -0.3550712 -0.7708386 -2.2578626 14 H H8 0.6245689 3.4167849 -2.4885809 15 H H10 -3.4400894 2.1260593 -3.0079652 16 C C10 5.2607492 1.4663694 -1.9949113 SCF total energy: -4648.8994977 hartrees 120 17 H H11 5.2004357 2.3893175 -1.4161725 18 H H14 5.6613762 1.7333310 -2.9831061 19 C C12 5.2614840 1.0846459 2.2224675 20 H H15 5.6552614 1.8030174 2.9553352 21 H H18 5.2238331 0.1128850 2.7169096 22 N N3 3.8972486 1.4512082 1.8763081 23 H H20 3.7898009 2.2354648 1.2335711 24 C C13 2.7688079 0.8382435 2.2910562 25 N N4 1.6323822 1.4610792 1.8454412 26 H H22 1.7761138 2.1258750 1.0805754 27 C C14 0.2952149 1.3955875 2.2898815 28 C C15 -2.4076716 1.4772893 3.0662460 29 C C16 -0.1001634 0.8644173 3.5247431 30 C C17 -0.6697046 1.9795466 1.4542447 31 C C18 -2.0048731 2.0203529 1.8476473 32 C C19 -1.4452920 0.9000425 3.8901120 33 H H21 0.6266397 0.4189908 4.1874706 34 H H23 -0.3707644 2.4090036 0.5031227 35 H H26 -3.4498786 1.4932500 3.3590253 36 S S2 2.7872540 -0.5873993 3.2373224 37 C C9 6.1953142 0.4363981 -1.3527427 38 H H66 6.1888855 -0.4701462 -1.9683179 39 C C11 -1.8674470 4.3253073 -3.0057754 40 C C20 -3.0179504 -0.5330077 -2.6738316 41 C C21 -3.0397027 2.5973303 0.9165316 42 C C22 -1.8526707 0.3468162 5.2322748 43 F F1 -3.5124705 1.6644908 0.0579004 44 F F2 -2.5334476 3.6035368 0.1693208 45 F F3 -4.1002269 3.0852263 1.5924522 46 F F4 -1.7965399 1.2987129 6.1932143 47 F F5 -1.0460150 -0.6600559 5.6273026 48 F F6 -3.1163949 -0.1241728 5.2138338 49 F F7 -3.0493461 4.5979143 -2.4148542 50 F F8 -2.0241180 4.5658245 -4.3285172 51 F F9 -0.9633317 5.2130042 -2.5444791 52 F F10 -4.0540158 -0.2298086 -3.4835079 53 F F11 -2.4869431 -1.6987577 -3.1065017 54 F F12 -3.5312931 -0.7659805 -1.4445504 55 H H2 7.2278878 0.8305944 -1.4034350 56 N N5 5.8050047 0.0710587 0.0077225 57 C C1 5.3031898 -2.3644845 -0.2073505 58 H H12 5.7187878 -3.3500893 0.0462233 59 N N6 3.9444816 -2.2719799 0.3039059 60 H H13 3.8481195 -2.1103761 1.3062523 61 C C23 2.8079859 -2.3397588 -0.4200681 62 N N7 1.6771503 -2.2902165 0.3516485 63 H H16 1.8200060 -1.9728928 1.3139804 64 C C24 0.3476586 -2.6736662 0.0707355 65 C C25 -2.3355466 -3.4584339 -0.2512243 66 C C26 -0.0128164 -3.5441791 -0.9645950 67 C C27 -0.6411242 -2.2147546 0.9557636 68 C C28 -1.9660771 -2.6107041 0.7933034 69 C C29 -1.3494205 -3.9144412 -1.1196988 121 70 H H17 0.7365180 -3.9304652 -1.6412965 71 H H19 -0.3679088 -1.5593844 1.7769814 72 H H24 -3.3673969 -3.7561716 -0.3826470 73 S S3 2.8039171 -2.4388198 -2.1286046 74 C C30 -3.0225496 -2.0715427 1.7241143 75 C C31 -1.6964523 -4.8350935 -2.2620423 76 F F13 -3.4712995 -0.8605647 1.3225534 77 F F14 -2.5472708 -1.9197359 2.9816806 78 F F15 -4.0938615 -2.8882522 1.7904528 79 F F16 -3.0010486 -5.1748762 -2.2628214 80 F F17 -0.9777024 -5.9799826 -2.2020364 81 F F18 -1.4211603 -4.2665814 -3.4560206 82 H H25 5.2463778 -2.3130804 -1.2958413 83 C C32 6.2160596 -1.2822943 0.3756853 84 H H9 7.2550846 -1.5046887 0.0680968 85 H H27 6.2005151 -1.3631277 1.4678909 86 C C33 6.1849094 1.0733869 1.0009875 87 H H1 6.1513581 2.0605765 0.5272716 88 H H33 7.2243617 0.9408821 1.3561298 122 3-O in vacuum Job type: Single point. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Method: RB3LYP Basis set: 6-31G Number of shells: 325 Number of basis functions: 1050 Multiplicity: 1 SCF total energy: -3680.0562311 hartrees Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C2 -2.8191769 -1.4459723 1.9139004 2 N N1 -3.6837952 -2.1887576 1.1546312 3 H H5 -3.3678728 -2.4709288 0.2289292 4 N N2 -1.4891232 -1.8088459 1.7201386 5 H H6 -1.3189721 -2.6347168 1.1595125 6 C C3 -0.3503171 -1.2225691 2.2807562 7 C C4 2.0493416 -0.0862620 3.2500984 8 C C5 0.8981553 -1.7653417 1.9324703 9 C C6 -0.3880865 -0.1375919 3.1660749 10 C C7 0.8077738 0.4196093 3.6248772 11 C C8 2.0751581 -1.2017548 2.4093743 12 H H7 0.9430649 -2.6421333 1.2986243 13 H H8 -1.3417631 0.2382549 3.5069152 14 H H10 2.9675834 0.3607257 3.6085817 3-O in vacuum Job type: Single point 3-O in vacuum Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 325 Number of basis functions: 1050 Multiplicity: 1 SCF total energy: -3680.0562311 hartrees Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C2 -2.8191769 -1.4459723 1.9139004 2 N N1 -3.6837952 -2.1887576 1.1546312 3 H H5 -3.3678728 -2.4709288 0.2289292 4 N N2 -1.4891232 -1.8088459 1.7201386 5 H H6 -1.3189721 -2.6347168 1.1595125 6 C C3 -0.3503171 -1.2225691 2.2807562 7 C C4 2.0493416 -0.0862620 3.2500984 8 C C5 0.8981553 -1.7653417 1.9324703 9 C C6 -0.3880865 -0.1375919 3.1660749 10 C C7 0.8077738 0.4196093 3.6248772 11 C C8 2.0751581 -1.2017548 2.4093743 12 H H7 0.9430649 -2.6421333 1.2986243 13 H H8 -1.3417631 0.2382549 3.5069152 14 H H10 2.9675834 0.3607257 3.6085817 Job type: Single point. Method: RB3LYP Basis set: 6-31G Number of shells: 325 Number of basis functions: 1050 Multiplicity: 1 Job type: Single point. Figure 3.27. 13C NMR (75 MHz, acetone-d6) of 3-S. Method: RB3LYP Basis set: 6-31G Number of shells: 325 Number of basis functions: 1050 Multiplicity: 1 SCF total energy: -3680.0562311 hartrees Cartesian Coordinates (Angstroms) Atom X Y Z --------- ------------- ------------- ------------- 1 C C2 -2.8191769 -1.4459723 1.9139004 2 N N1 -3.6837952 -2.1887576 1.1546312 3 H H5 -3.3678728 -2.4709288 0.2289292 4 N N2 -1.4891232 -1.8088459 1.7201386 5 H H6 -1.3189721 -2.6347168 1.1595125 6 C C3 -0.3503171 -1.2225691 2.2807562 7 C C4 2.0493416 -0.0862620 3.2500984 8 C C5 0.8981553 -1.7653417 1.9324703 9 C C6 -0.3880865 -0.1375919 3.1660749 10 C C7 0.8077738 0.4196093 3.6248772 11 C C8 2.0751581 -1.2017548 2.4093743 12 H H7 0.9430649 -2.6421333 1.2986243 13 H H8 -1.3417631 0.2382549 3.5069152 14 H H10 2.9675834 0.3607257 3.6085817 SCF total energy: -3680.0562311 hartrees 123 16 H H11 -5.2929735 -1.5776720 2.2812241 17 H H14 -5.5727231 -3.0456003 1.3512156 18 C C12 -5.0034839 2.3026999 1.1529605 19 H H15 -5.4141876 2.8510338 2.0107121 20 H H18 -5.1638710 2.9229996 0.2706689 21 N N3 -3.5626078 2.1839246 1.3289192 22 H H20 -3.2552643 1.5176639 2.0341170 23 C C13 -2.7075480 2.4561068 0.2947921 24 N N4 -1.3695918 2.4426817 0.6796475 25 H H22 -1.1808298 2.3856234 1.6730415 26 C C14 -0.2468647 2.6493585 -0.1300865 27 C C15 2.1192833 2.9594669 -1.6481134 28 C C16 -0.3128602 2.8382121 -1.5177245 29 C C17 1.0130801 2.6622575 0.4901460 30 C C18 2.1732307 2.8145091 -0.2614320 31 C C19 0.8652741 2.9766362 -2.2530596 32 H H21 -1.2765799 2.9083830 -1.9996946 33 H H23 1.0841378 2.5679058 1.5672095 34 H H26 3.0246836 3.0602805 -2.2319525 35 C C9 -5.8077784 -1.2796189 0.1690927 36 H H66 -5.6264522 -1.8391170 -0.7522703 37 C C11 0.7017128 1.6374689 4.4992039 38 C C20 3.3929018 -1.7795935 1.9584829 39 C C21 3.4964690 2.7661755 0.4591288 40 C C22 0.7286742 3.0647507 -3.7472031 41 F F1 3.7883104 1.5106720 0.8698746 42 F F2 3.4783570 3.5405298 1.5685110 43 F F3 4.5149046 3.1852040 -0.3163414 44 F F4 -0.1933562 3.9728184 -4.1229471 45 F F5 0.3041618 1.8626279 -4.2545865 46 F F6 1.8827161 3.3625955 -4.3652558 47 F F7 1.8768286 2.0056603 5.0334528 48 F F8 -0.1816778 1.4698727 5.5017425 49 F F9 0.2438951 2.7088936 3.7708375 50 F F10 4.3632951 -1.5832125 2.8730045 51 F F11 3.3009087 -3.1113309 1.7348205 52 F F12 3.8139235 -1.2205603 0.8016466 53 H H2 -6.9020118 -1.3084428 0.3481495 54 N N5 -5.3242305 0.0878459 -0.0254913 55 C C1 -4.9805141 -0.0027500 -2.5265910 56 H H12 -5.3647749 0.4806546 -3.4344254 57 N N6 -3.5335039 0.1592466 -2.4979075 58 H H13 -3.2003848 1.0939897 -2.2725799 59 C C23 -2.7136441 -0.8963614 -2.2004932 60 N N7 -1.3641707 -0.5966324 -2.3694008 61 H H16 -1.1452379 0.2902806 -2.8060456 62 C C24 -0.2640861 -1.4257205 -2.1210107 63 C C25 2.0520690 -2.9541406 -1.5814598 64 C C26 -0.3764353 -2.7544235 -1.6878494 65 C C27 1.0182818 -0.8888841 -2.3224627 66 C C28 2.1546345 -1.6471006 -2.0585374 67 C C29 0.7775112 -3.4885720 -1.4103902 124 68 H H17 -1.3550647 -3.2014190 -1.5908365 69 H H19 1.1224511 0.1238865 -2.6959210 70 H H24 2.9366803 -3.5290567 -1.3426359 71 C C30 3.5015073 -1.0197297 -2.3155295 72 C C31 0.5869700 -4.8638394 -0.8351782 73 F F13 3.6211695 0.1677586 -1.6766378 74 F F14 3.6770187 -0.7671852 -3.6332394 75 F F15 4.5179740 -1.8043213 -1.9139979 76 F F16 1.7373992 -5.5432264 -0.7016732 77 F F17 -0.2571635 -5.6138996 -1.5707294 78 F F18 0.0265986 -4.7772247 0.4138683 79 H H25 -5.1789666 -1.0708885 -2.6262966 80 C C32 -5.7176408 0.6174481 -1.3316820 81 H H9 -6.8064816 0.5068331 -1.5117772 82 H H27 -5.5059394 1.6890298 -1.3407401 83 C C33 -5.7494111 0.9645796 1.0671436 84 H H1 -5.5772680 0.4377785 2.0084475 85 H H33 -6.8350147 1.1876081 1.0191279 86 O O1 -3.1162551 -2.0179532 -1.8707390 87 O O2 -3.1675940 -0.5810756 2.7256196 88 O O3 -3.0718983 2.7369773 -0.8530852 125 MANUSCRIPT – IV Published in ACS Macro Letters ABSTRACT A series of conformationally flexible bis(thio)urea H-bond donors plus base cocatalyst were applied to the ring-opening polymerization of lactones. The rate of the ROP displays a strong dependence upon the length and identity of the tether, where a circa five methylene-unit long tether exhibits the fastest ROP. Any constriction to conformational freedom is deleterious to catalysis. For valerolactone and caprolactone, the bisurea H-bond donors are more effective, but for lactide the bisthioureas are the better catalysts. The ROP reactions are rapid and controlled across a wide range of reaction conditions, including solvent-free. The active mechanism is highly dependent upon the identity of the base cocatalyst, and a mechanistic rationale for the observations is discussed. Implications for the design of future generation catalysts are discussed. 127 INTRODUCTION H-bonding organocatalysts for ring-opening polymerization (ROP) are a facile means of generating precisely tailored macromolecules.1–3 Among the larger class of H-bonding organocatalysts, (thio)urea H-bond donors stand out for the remarkable level of control they can rendered in polymer synthesis.4,5 The thiourea plus base mediated ROP of lactone and carbonate monomers are thought to effect enchainment by H-bond activation of monomer by thiourea and initiating/propagating chain end by base; these catalysts are most active in non-polar solvent.5 The urea plus base class of H-bonding catalysts offer no apology in terms of rate and are among the most active catalysts for the ROP of lactones.6– 8 Several mechanistic studies by our group and others have shown that (thio)urea/base mediated ROP can proceed by one of two mechanisms: neutral H-bonding or (thio)imidate mediated ROP (Scheme 4.1).7–10 Which mechanism is operative depends largely on reaction conditions (high temperature,11 polar solvent,10,12,13 strong electron-withdrawing groups on H-bond donor,10 early reaction time and strong bases favor imidate)7,8 though generally ureas are more active than thioureas and imidate mediated ROP is far more active than neutral H-bonding.3 Remarkably, these ‘hyperactive’ catalysts for ROP remain controlled. INTRODUCTION The synthetic addition of one or more (thio)urea H-bond donating arms to the parent (thio)urea has been shown to substantially increase the activity of (thio)urea H-bond donors.6,14 Our group first disclosed bis- and tris-(thio)urea H-bond donors for ROP,6,14 and other intramolecular Lewis acid donors have been used.15 In general, the bis- (thio)ureas (2-O and 2-S, Figure 4.1) are more active than mono-(thio)urea (1-O and 1-S), and ureas are more active than thioureas for the bis-(thio)urea plus base mediated ROP of H-bonding organocatalysts for ring-opening polymerization (ROP) are a facile means of generating precisely tailored macromolecules.1–3 Among the larger class of H-bonding organocatalysts, (thio)urea H-bond donors stand out for the remarkable level of control they can rendered in polymer synthesis.4,5 The thiourea plus base mediated ROP of lactone and carbonate monomers are thought to effect enchainment by H-bond activation of monomer by thiourea and initiating/propagating chain end by base; these catalysts are most active in non-polar solvent.5 The urea plus base class of H-bonding catalysts offer no apology in terms of rate and are among the most active catalysts for the ROP of lactones.6– 8 Several mechanistic studies by our group and others have shown that (thio)urea/base mediated ROP can proceed by one of two mechanisms: neutral H-bonding or (thio)imidate mediated ROP (Scheme 4.1).7–10 Which mechanism is operative depends largely on reaction conditions (high temperature,11 polar solvent,10,12,13 strong electron-withdrawing groups on H-bond donor,10 early reaction time and strong bases favor imidate)7,8 though generally ureas are more active than thioureas and imidate mediated ROP is far more active than neutral H-bonding.3 Remarkably, these ‘hyperactive’ catalysts for ROP remain controlled. INTRODUCTION The synthetic addition of one or more (thio)urea H-bond donating arms to the parent (thio)urea has been shown to substantially increase the activity of (thio)urea H-bond donors.6,14 Our group first disclosed bis- and tris-(thio)urea H-bond donors for ROP,6,14 and other intramolecular Lewis acid donors have been used.15 In general, the bis- (thio)ureas (2-O and 2-S, Figure 4.1) are more active than mono-(thio)urea (1-O and 1-S), and ureas are more active than thioureas for the bis-(thio)urea plus base mediated ROP of 128 lactone and carbonate monomers.3,6,13 However, this rule of thumb does not apply for the (thio)urea plus base mediated ROP of lactide (LA), where the higher rates are displayed by (bis)-thioureas (versus (bis)-ureas) of like substitution.12 Again, the high rates exhibited by 2-O and 2-S plus base for ROP occur without the reduction of reaction control. Pan et al. synthesized bisurea H-bond donors featuring rigid linkers,16 which were less active for ROP than the flexible 3-carbon tethered 2-O and 2-S.6,14 In the pantheon of conformationally flexible linkers that can be envisaged, only one has been reported (n- propyl).6 In light of the recent interest in these catalysts, we disclose here several bisurea and bisthiourea H-bond donors for ROP with flexible linkers, most with higher activity and control than the parent 2-X system. We extend previously proposed mechanisms to the bis(thio)urea plus alkylamine base mediated ROP of LA to explain why thioureas have been observed to be more effective (versus ureas). 129 EXPERIMENTAL SECTION General Considerations. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Benzene-d6 and chloroform-d were purchased from Cambridge Isotope Laboratories, distilled from calcium hydride and stored under N2. Acetone-d6 was purchased from Cambridge Isotope Laboratories, distilled from calcium sulfate and stored under N2. δ-valerolactone (VL), ε-caprolactone (CL) and benzyl alcohol were distilled under high vacuum from calcium hydride prior to use. Dry CH2Cl2 was obtained from an Innovative Technology solvent purification system. All experiments were conducted in a stainless-steel glovebox under N2 unless stated otherwise. NMR experiments were performed on a Bruker Avance III 300 or 400 MHz spectrometer. Mass spectrometry experiments were performed using a Thermo Electron (San Jose, CA, USA) LTQ Orbitrap XL mass spectrometer affixed with electrospray ionization (ESI) interface in a positive ion mode. Collected mass spectra were averaged for at least 50 scans. Tune conditions for infusion experiments (10 µL/min flow, sample concentration 5 µg/mL in 50/50 v/v water/ methanol) were as follows: ion spray voltage, 5000 V; capillary temperature, 275oC; sheath gas (N2, arbitrary units), 11; auxiliary gas (N2, arbitrary units), 2; capillary voltage, 21 V; and tube lens, 90 V; multipole 00 offset, -4.25 V; lens 0 voltage, - 5.00; multipole 1 offset, - 8.50 V; Multipole RF Amplitude, 400 V; Ion trap’s AGC target settings for Full MS was 3.0e4 and FT’s 2.0e5 (with 3 and 2 averaged microscans , respectively). Prior to analysis, the instrument was calibrated for positive ions using Pierce LTQ ESI positive ion calibration solution (lot #PC197784). 130 Example ROP of VL in benzene-d6. To a 7 mL vial, 2-O5 (14.69 mg, 0.024 mmol), VL (100.00 mg, 0.998 mmol) and benzene-d6 (250 µL) were added. The contents were stirred until the solution became homogenous. To a second 7 ml vial, benzyl alcohol (2.16 mg, 0.019 mmol), MTBD (3.67 mg, 0.024 mmol) and benzene-d6 (250 µL) were added. The contents in the second vial were transferred to the first vial via Pasteur pipette, and the contents were agitated to mix. The reaction solution was then transferred to an NMR tube, and the progress of the reaction monitored by 1H NMR. The reaction was quenched by the addition of benzoic acid (3.00 mg, 0.024 mmol). Polymer isolated by precipitation with hexanes, and the volatiles were removed under high vacuum before characterization via GPC. Example solvent-free ROP of VL. Synthesis of 1,1'-(butane-1,4-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O4). Synthesis of 1,1'-(butane-1,4-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O4). Synthesis of 1,1'-(butane-1,4-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O4). A flame dried 25 ml Schlenk flask was charged with a stir bar, dichloromethane (25 mL), 3,5-bistrifluoromethylphenyl isocyanate (1 g, 3.91 mmol), and 1,4-diaminobutane (0.20 mL, 1.95 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 70%. NMR spectra given below. HRMS: calc. (C22H19F12N4O2+H)+= 599.1238; found m/z =599.1311. Synthesis of 1,1'-(pentane-1,5-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5). A flame dried 25 ml Schlenk flask was charged with a stir bar, dichloromethane (25 mL), 3,5-bistrifluoromethylphenyl isocyanate (1 g, 3.91 mmol), and 1,5-diaminopentane (0.22 mL, 1.95 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. EXPERIMENTAL SECTION A 1.5 mL vial was charged with 2-O5 (12.23 mg, 0.019 mmol), benzyl alcohol (2.15 mg, 0.019 mmol), VL (200.00 mg, 1.99 mmol), magnetic stir bar and stirred until homogeneous. A second vial was charged with MTBD (3.05 mg, 0.019 mmol) and VL (200 mg, 1.99 mmol) and agitated to mix. The contents of the second vial were transferred to the first vial using a Pasteur pipette, and the solution was stirred. Reaction progress was monitored by taking aliquots of the reaction mixture – either ~1.5 µL solution or a small amount of solid extracted via spatula – at different time intervals and quenched in a solution of benzoic acid in chloroform-d. Conversion was determined via 1H NMR. The polymer samples in the aliquots were isolated by precipitating with hexanes, and the volatiles were removed under high vacuum before characterization via GPC. 131 A flame dried 25 ml Schlenk flask was charged with a stir bar, dichloromethane (25 mL), 3,5-bistrifluoromethylphenyl isocyanate (1 g, 3.91 mmol), and ethylenediamine (0.13 mL, 1.95 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 65%. NMR spectra given below. HRMS: calc. (C20H15F12N4O2+H)+= 571.0998; found m/z = 571.0998. Synthesis of 1,1'-(pentane-1,5-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5). Synthesis of 1,1'-(pentane-1,5-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5). 132 Yield: 74%. NMR spectra given below. HRMS: calc. (C23H21F12N4O2+H)+= 613.1467; found m/z = 613.1467. Synthesis of 1,1'-(hexane-1,6-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O6). A flame dried 25 ml Schlenk flask was charged with a stir bar, dichloromethane (25 mL), 3,5-bistrifluoromethylphenyl isocyanate (1 g, 3.91 mmol), and hexamethylenediamine (0.25 mL, 1.95 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 80%. NMR spectra given below. HRMS: calc. (C24H23F12N4O2+H)+= 627.1624; found m/z = 627.1624. A flame dried 100 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bistrifluoromethylphenyl isocyanate (0.63 g, 2.5 mmol), and 1,12-diaminododecane (0.25 g, 1.25 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 92%. NMR spectra given below. HRMS: calc. (C30H34F12N4O2+H)+= 711.2382; found m/z = 711.2563. 133 Synthesis of 1,1'-((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)urea) (2-O5-N). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bistrifluoromethylphenyl isocyanate (1.08 g, 4.26 mmol), and N1-(2-aminoethyl)-N1- methylethane-1,2-diamine (0.27 mL, 2.13 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 47%. NMR spectra given below. HRMS: calc. (C23H21F12N5O2+H)+= 628.1564; found m/z = 628.1594. Synthesis of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2- O5-O). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bistrifluoromethylphenyl isocyanate (2.2 g, 8.14 mmol), and 2,2'-oxybis(ethan-1- amine) (0.44mL, 4.07 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 90%. NMR spectra given below. HRMS: calc. (C22H19F12N4O3+H)+= 615.1260; found m/z = 615.1260. 134 Synthesis of 1,1'-(2,2-dimethylpropane-1,3-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)urea) (2-O3-diMe). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bistrifluoromethylphenyl isocyanate (0.63 g, 2.5 mmol) 2,2-Dimethyl-1,3- propanediamine and (0.15 mL, 1.25 mmol). Synthesis of 1,1'-(pentane-1,5-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5). After stirring overnight, the reaction mixture was filtered and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 47%. NMR spectra given below. HRMS :calc. (C23H21F12N4O2)+ = 613.1467 found m/z =613.1467. Synthesis of 1,1’-(butane-1,4-diyl)bis(trifluoromethyl)phenyl)thiourea) (2-S4). Synthesis of 1,1’-(butane-1,4-diyl)bis(trifluoromethyl)phenyl)thiourea) (2-S4). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bis(trifluoromethyl)phenyl isothiocyanate (1.49 g, 5.5 mmol), and 1,4-diaminobutane (0.27 mL, 2.7 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 35%. NMR spectra given below. HRMS: calc. (C22H19F12N4S2+H)+= 631.0845; found m/z =631.0825. 135 Synthesis of 1,1'-(pentane-1,5-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bis(trifluoromethyl)phenyl isothiocyanate (1.49 g, 5.5 mmol), and 1,5-diaminopentane (0.32 mL, 2.7 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 57%. NMR spectra given below. HRMS: calc. (C23H21F12N4S2+H)+= 645.1011; found m/z = 645.1016. Synthesis of 1,1'-(hexane-1,6-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S6). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bis(trifluoromethyl)phenyl isothiocyanate (1.87 g, 6.88 mmol), and hexamethylenediamine (0.44 mL, 3.44 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 76%. NMR spectra given below. HRMS: calc. (C24H23F12N4S2+H)+= 659.1167; found m/z =659.1148. Synthesis of 1,1'-(dodecane-1,12-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2- S12). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bis(trifluoromethyl)phenyl isothiocyanate (0.68 g, 2.52 mmol), and 1,12- diaminododecane (0.25 g, 1.25 mmol) was added dropwise to the Schlenk flask. After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 136 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 91%. NMR spectra given below. HRMS: calc. (C30H35F12N4S2+H)+= 743.2033; found m/z = 743.2086. Synthesis of 1,1'-((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea) (2-S5-N). Synthesis of 1,1'-((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea) (2-S5-N). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (20 mL), 3,5-bistrifluoromethylphenyl isothiocyanate (1.48 g, 4.26 mmol), and N1-(2-aminoethyl)- N1-methylethane-1,2-diamine (0.27 mL, 2.13 mmol). Synthesis of 1,1’-(butane-1,4-diyl)bis(trifluoromethyl)phenyl)thiourea) (2-S4). After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 67%. NMR spectra given below. HRMS: calc. (C23H21F12N5S2+H)+= 660.1120; found m/z = 660.1120. Synthesis of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-O). A flame dried 50 ml Schlenk flask was charged with a stir bar, dichloromethane (25 mL), 3,5-bistrifluoromethylphenyl isothiocyanate (2.21 g, 8.14 mmol), and 2,2'-oxybis(ethan-1- amine) (0.44 mL, 4.07 mmol). After stirring overnight, the reaction mixture was filtered via suction filtration and rinsed with 3 portions of cold CH2Cl2 to provide a pure white powder that was freed of volatiles under high vacuum. Yield: 86%. NMR spectra given below. HRMS: calc. (C22H19F12N4OS2+H)+= 647.0803; found m/z = 647.0803. 137 RESULTS AND DISCUSSION In the bis(thio)urea plus MTBD cocatalyzed (0.024 mmol) ROP of VL (1.0 mmol, 2 M) from benzyl alcohol (0.02 mmol) in C6D6, the bis(thio)ureas featuring a 5-carbon (methylene) tether were most active. Using established procedures,6 electron deficient bis(thio)ureas were synthesized featuring linear aliphatic tethers ranging from two to twelve methylene units, bisthioureas (2-Sn) and bisureas (2-On) in Figure 4.1 (n = 2, 3, 4, 5, 6, 12), see Supplemental Information, SI. The 2-O2 H-bond donor was insoluble in solvents relevant for ROP. Especially versus the rigid (thio)urea tethers,16 our results here suggest that the most effective catalysis arises when the (thio)urea moieties are allowed to interact with one another, lending credence to the originally proposed mechanism whereby bis(thio)urea moieties bring about ROP through an activated-(thio)urea mechanism characterized by one (thio)urea stabilizing through internal H-bond activation the (thio)urea which activates the lactone for enchainment, Figure 4.1.6,14 We propose that the increased efficacy of 2-S5 and 2-O5 plus MTBD (versus other linker lengths) arises from the stability of the pseudo-7-membered cycle formed by intramolecular H-bonding – the (thio)urea moiety being largely rigid. The enhanced rates displayed by 2-O5 and 2-S5 are enhanced by a factor of two versus their respective ‘parent’ 2-X3 H-bond donor, and this enhanced rate does not result in increased Mw/Mn. The ROP are living in behavior, both 2-S5 and 2-O5 plus MTBD produce linear evolution of Mn versus conversion (Figure 4.2) and Mn that is predictable by [M]o/[I]o Table 4.1. In C6D6 (and other non-polar solvents), a H-bond mediated 138 mechanism of enchainment has been proposed;3,6,9 urea plus base mediated ROP have been shown to display faster rates than the analogous thiourea.6,9 Bisurea catalysts plus MTBD remain highly active for the ROP of VL in polar solvent and solvent-free conditions. In polar solvent (including solvent-free), the imidate mechanism of enchainment has been shown to be favored.7,10,13 This mechanism is characterized by proton transfer from urea to MTBD, forming a highly active urea anion (imidate).9 In the bisurea system, we propose that the incipient anion is stabilized via intramolecular hydrogen bonding by the ‘extra’ urea moiety; hence, an activated-(thio)urea anion mechanism analogous to the neutral activated (thio)urea H-bonding mechanism can be envisaged, Figure 4.3.14 This mechanism is corroborated by the observation that the five- methylene tether in the 2-O5/MTBD cocatalyzed ROP of VL produces the most active ROP, just as in the H-bonded system. RESULTS AND DISCUSSION Reproducing an established experiment,13 when 2- O5 is treated with 1 equivalent of MTBD in acetone-d6, an upfield chemical shift is observed, consistent with anion formation and an imidate mechanism (Figure 4.4). The individual urea moieties are indistinguishable, which suggests that the anion/neutral urea exchange is rapid on the NMR timescale (400 MHz). Treatment with an additional equivalent of MTBD (0.096 M MTBD, 0.048 M 2-O5) does not further shift the bisurea resonances, which suggest that bisimidate is not formed and corroborates previous observations of bis-(thio)ureas operating as a single H-bond donating species.6 Further, the rate of the 2-O5/MTBD (0.048 M 2-O5; 0.096 M MTBD) cocatalyzed ROP of VL under the respective conditions are identical (Figure 4.5), suggesting that the ideal stoichiometry for bisurea/base mediated ROP in an imidate mechanism is 1:1. Similar Bisurea catalysts plus MTBD remain highly active for the ROP of VL in polar solvent and solvent-free conditions. In polar solvent (including solvent-free), the imidate mechanism of enchainment has been shown to be favored.7,10,13 This mechanism is characterized by proton transfer from urea to MTBD, forming a highly active urea anion (imidate).9 In the bisurea system, we propose that the incipient anion is stabilized via intramolecular hydrogen bonding by the ‘extra’ urea moiety; hence, an activated-(thio)urea anion mechanism analogous to the neutral activated (thio)urea H-bonding mechanism can be envisaged, Figure 4.3.14 This mechanism is corroborated by the observation that the five- methylene tether in the 2-O5/MTBD cocatalyzed ROP of VL produces the most active ROP, just as in the H-bonded system. Reproducing an established experiment,13 when 2- O5 is treated with 1 equivalent of MTBD in acetone-d6, an upfield chemical shift is observed, consistent with anion formation and an imidate mechanism (Figure 4.4). The individual urea moieties are indistinguishable, which suggests that the anion/neutral urea exchange is rapid on the NMR timescale (400 MHz). Treatment with an additional equivalent of MTBD (0.096 M MTBD, 0.048 M 2-O5) does not further shift the bisurea resonances, which suggest that bisimidate is not formed and corroborates previous observations of bis-(thio)ureas operating as a single H-bond donating species.6 Further, the rate of the 2-O5/MTBD (0.048 M 2-O5; 0.096 M MTBD) cocatalyzed ROP of VL under the respective conditions are identical (Figure 4.5), suggesting that the ideal stoichiometry for bisurea/base mediated ROP in an imidate mechanism is 1:1. RESULTS AND DISCUSSION Similar 139 relative rates (for the bisureas) are observed under solvent free conditions (Table 4.3) as in acetone-d6. Two bisurea H-bond donors featuring heteroatom-containing tethers were synthesized and indicate a sensitive relationship between cocatalyst geometry and reaction rate. Both of the 2-O5-N or 2-O5-O (Figure 4.1) plus MTBD cocatalyzed ROP of VL showed slightly reduced reaction times versus the ‘parent’ 2-O5 under all conditions: benzene-d6, acetone- d6 and solvent-free (Table 4.4). The reactions remained well-controlled, especially solvent-free where Mw/Mn < 1.03. The relative reaction times in each of C6D6, acetone-d6 and solvent-free fall in the order 2-O5-O (fastest) < 2-O5-N < 2-O5 (slowest). We attribute the subtle changes in reaction time to minute changes in the tether length, where the normal bond lengths are C-O < C-N < C-C. This suggests that the most active bis-(thio)urea tether length is somewhere between four and five methylene units long, which is a parameter that our group has found useful in our ongoing development of advanced H-bond donating catalyst systems. However, these relative rates may be coincidence and could be easily attributed to increased conformational flexibility due to the heteroatom, but these results suggest that there is no stark change in mechanism due to the presence of the heteroatom. As opposed to ROP in solution, bisurea plus MTBD cocatalyzed ROP under solvent-free conditions provide the best weight control (by [M]o/[I]o ≤ 500), narrowest distributions (Mw/Mn ≤ 1.05) and access to the highest molecular weights (Table 4.1), consistent with previous observations;12 the 2-O5-O H-bond donor is especially active and well controlled. The ROP of CL with these cocatalysts (plus MTBD) exhibit the same relative reaction times and display good control, Table 4.4. Additionally, we synthesized a symmetric analogue of bisurea 2-O3 featuring a 3,3-dimethyl substitution, 2-O3-diMe (see SI), which Two bisurea H-bond donors featuring heteroatom-containing tethers were synthesized and indicate a sensitive relationship between cocatalyst geometry and reaction rate. Both of the 2-O5-N or 2-O5-O (Figure 4.1) plus MTBD cocatalyzed ROP of VL showed slightly reduced reaction times versus the ‘parent’ 2-O5 under all conditions: benzene-d6, acetone- d6 and solvent-free (Table 4.4). The reactions remained well-controlled, especially solvent-free where Mw/Mn < 1.03. The relative reaction times in each of C6D6, acetone-d6 and solvent-free fall in the order 2-O5-O (fastest) < 2-O5-N < 2-O5 (slowest). RESULTS AND DISCUSSION We attribute the subtle changes in reaction time to minute changes in the tether length, where the normal bond lengths are C-O < C-N < C-C. This suggests that the most active bis-(thio)urea tether length is somewhere between four and five methylene units long, which is a parameter that our group has found useful in our ongoing development of advanced H-bond donating catalyst systems. However, these relative rates may be coincidence and could be easily attributed to increased conformational flexibility due to the heteroatom, but these results suggest that there is no stark change in mechanism due to the presence of the heteroatom. As opposed to ROP in solution, bisurea plus MTBD cocatalyzed ROP under solvent-free conditions provide the best weight control (by [M]o/[I]o ≤ 500), narrowest distributions (Mw/Mn ≤ 1.05) and access to the highest molecular weights (Table 4.1), consistent with previous observations;12 the 2-O5-O H-bond donor is especially active and well controlled. The ROP of CL with these cocatalysts (plus MTBD) exhibit the same relative reaction times and display good control, Table 4.4. Additionally, we synthesized a symmetric analogue of bisurea 2-O3 featuring a 3,3-dimethyl substitution, 2-O3-diMe (see SI), which 140 is less active as a cocatalyst (with MTBD) for the ROP of VL (C6D6, 90 % conv in 1 hour). This suggests that any steric compression or hindered bond rotation arising from the geminal dimethyl substitution (i.e. Thorpe-Ingold effect)17 is deleterious to ROP. The bisthiourea analogues of these bisurea H-bond donors were also synthesized, but they displayed reduced rates and control versus the bisureas, as expected (see Table 4.5). These modified bis(thio)urea H-bond donors emphasize the sensitive interplay of catalyst structure towards ROP activity. is less active as a cocatalyst (with MTBD) for the ROP of VL (C6D6, 90 % conv in 1 hour). This suggests that any steric compression or hindered bond rotation arising from the geminal dimethyl substitution (i.e. Thorpe-Ingold effect)17 is deleterious to ROP. The bisthiourea analogues of these bisurea H-bond donors were also synthesized, but they displayed reduced rates and control versus the bisureas, as expected (see Table 4.5). These modified bis(thio)urea H-bond donors emphasize the sensitive interplay of catalyst structure towards ROP activity. RESULTS AND DISCUSSION 141 ROP of Lactide The most active bis(thio)urea H-bond donors from the VL studies were applied for the ROP of lactide in CH2Cl2 and solvent-free with Me6TREN cocatalyst.18 Low solubility of bisureas under reaction conditions limited some direct comparisons, but this and previous studies12 show that the bisthioureas are more effective than the corresponding bisureas for the ROP of LA (Table 4.6). In the case of lactide, weak alkylamine base cocatalysts are used because stronger imine bases (e.g. MTBD) will polymerize lactide in the absence of H-bond donor in a less-controlled ROP.5,18,19 We speculated that the increased rate observed for bisthiourea (versus bisurea) plus Me6TREN mediated ROP of lactide was due to a change in mechanism between the two species. Indeed, the 1H NMR of 2-S5-O (acetone-d6) shows an upfield shift for the aromatic resonances in 2-S5-O in the presence (versus absence) of Me6TREN, suggesting the formation of an imidate species, whereas the chemical shifts for 2-O5-O with and without Me6TREN are negligibly different, suggesting H-bonding (see Figure 4.9). The same experiment when conducted with 1-S or TCC shows downfield chemical shifts consistent with H-bonding.20 Similar to the acetone-d6 experiment, the relative reaction times for the ROP results in solvent-free conditions (Table 4.6) suggest that the 2-S5-O plus Me6TREN mediated ROP of lactide proceeds via an imidate mechanism while 2-O5-O is an H-bond mediated enchainment. For identically substituted ureas and thioureas in the ROP of LA, the thiourea is the more active catalyst, and this is attributed to the pKa of the H-bond donor. The difference in mechanism for the two H-bond donating catalysts presumably arises because any thiourea will be more acidic than its identically substituted (e.g. 3,5- bistrifluoromethyl phenyl) urea.21,22 When a pair of mono-H-bond donors (urea or 142 thiourea) of the same pKa are used as cocatalysts with Me6TREN for the ROP of lactide, the urea is the more active catalyst, Table 4.7. Having identical pKa, such a pair of urea and thiourea will effect enchainment by the same mechanism, and hence, the more polar urea (or imidate) is the more active H-bond donor. When a highly acidic H-bond donor is employed (Table 4.7, last entry), the incipient (thio)imidate displays reduced activity arising from its low basicity, as previously observed.10,21 These observations are seemingly contrary to the (thio)urea plus strong base mediated ROP of other lactones (e.g. RESULTS AND DISCUSSION valerolactone or caprolactone).7,9,11,21 However, in this latter scenario, the stronger base cocatalyst (versus Me6TREN) can deprotonate either the urea or thiourea.13 In that event, the urea (or resulting imidate) will be more active than the thiourea (or thioimidate).7 thiourea) of the same pKa are used as cocatalysts with Me6TREN for the ROP of lactide, the urea is the more active catalyst, Table 4.7. Having identical pKa, such a pair of urea and thiourea will effect enchainment by the same mechanism, and hence, the more polar urea (or imidate) is the more active H-bond donor. When a highly acidic H-bond donor is employed (Table 4.7, last entry), the incipient (thio)imidate displays reduced activity arising from its low basicity, as previously observed.10,21 These observations are seemingly contrary to the (thio)urea plus strong base mediated ROP of other lactones (e.g. valerolactone or caprolactone).7,9,11,21 However, in this latter scenario, the stronger base cocatalyst (versus Me6TREN) can deprotonate either the urea or thiourea.13 In that event, the urea (or resulting imidate) will be more active than the thiourea (or thioimidate).7 143 CONCLUSION A series of conformationally flexible bis(thio)urea H-bond donors were applied with the appropriate base cocatalysts for the ROP of lactones. Conformational flexibility is essential for catalyst activity, and the (thio)urea moieties separated by circa five methylene units displays the most rapid ROP. Synthetic polymer chemists should hew towards 2-S5-O for the ROP of lactide; it is readily soluble, easily accessible and is among the most active organocatalysts for the synthesis of polylactide. That bisthioureas (versus bisureas) are more active for the ROP of LA is contrary to what is observed for VL and CL, and the higher activity of the thioureas is rendered by the alkylamine cocatalyst, which is unable to deprotonate the (bis)urea catalyst and enter the highly active imidate mediated ROP. For VL and CL, the bisurea 2-O5-O plus MTBD cocatalyst system is the most active bis(thio)urea examined, and the reaction is well controlled, especially under the easily- employable solvent-free conditions. We trust that the results of this study will be informative for the synthesis of advanced H-bond donating catalysts for ROP, and such work is ongoing in our lab. 144 (6) Fastnacht, K. V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; LIST OF REFERENCES (1) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Organocatalytic Ring-Opening Polymerization. Chem. Rev. 2007, 107 (12), 5813–5840. (2) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R. M. Organocatalysis: Opportunities and Challenges for Polymer Synthesis. Macromolecules 2010, 43 (5), 2093–2107. (3) Fastnacht, K. V.; Datta, P. P.; Kiesewetter, M. K. Bifunctional and Supramolecular Organocatalysts for Polymerization. In Organic Catalysis for Polymerization; Dove, A. P., Sardon, H., Naumann, S., Eds.; Royal Society of Chemistry: London, 2019; pp 87–120. (4) Bas G. G. Lohmeijer Frank Leibfarth, John W. Logan, R. C. P.; David A. Long Fredrik Nederberg, Jeongsoo Choi, A. P. D.; Charles Wade and James L. Hedrick, R. M. W.; Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; et al. Guanidine and Amidine Organocatalysts for Ring-Opening Polymerization of Cyclic Esters. Macromolecules 2006, 39 (25), 8574–8583. (5) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. a.; Lundberg, P. N. P.; Dove, A. P.; Li, H.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L. Exploration, Optimization, and Application of Supramolecular Thiourea−Amine Catalysts for the Synthesis of Lactide (Co)Polymers. Macromolecules 2006, 39 (23), 7863–7871. (6) Fastnacht, K. V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; 145 Kiesewetter, E. T.; Kiesewetter, M. K. Bis- and Tris-Urea H-Bond Donors for Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst. ACS Macro Lett. 2016, 5 (8), 982–986. Kiesewetter, E. T.; Kiesewetter, M. K. Bis- and Tris-Urea H-Bond Donors for (7) Lin, B.; Waymouth, R. M. Urea Anions: Simple, Fast, and Selective Catalysts for Ring-Opening Polymerizations. J. Am. Chem. Soc. 2017, 139 (4), 1645–1652. (8) Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Fast and Selective Ring- Opening Polymerizations by Alkoxides and Thioureas. Nat. Chem. 2016, 8, 1047– 1053. (9) Dharmaratne, N. U.; Pothupitiya, J. U.; Kiesewetter, M. K. The Mechanistic Duality of (Thio)Urea Organocatalysts for Ring-Opening Polymerization. Org. Biomol. Chem. 2019, 17 (13), 3305–3313. (10) Pothupitiya, J. U.; Hewawasam, R. S.; Kiesewetter, M. K. Urea and Thiourea H- Bond Donating Catalysts for Ring-Opening Polymerization: Mechanistic Insights via (Non)Linear Free Energy Relationships. Macromolecules 2018, No. 51, 3203– 3211. (11) Coderre, D. N.; Fastnacht, K. V.; Wright, T. J.; Dharmaratne, N. U.; Kiesewetter, M. K. H-Bonding Organocatalysts for Ring-Opening Polymerization at Elevated Temperatures. LIST OF REFERENCES Macromolecules 2018, 51, 10121–10126. (12) Pothupitiya, J. U. J. U.; Dharmaratne, N. U. N. U.; Jouaneh, T. M. M. T. M. M.; Fastnacht, K. V. K. V.; Coderre, D. N.; Kiesewetter, M. K. M. K. H-Bonding Organocatalysts for the Living, Solvent-Free Ring-Opening Polymerization of Lactones: Toward an All-Lactones, All-Conditions Approach. Macromolecules 2017, 50 (22), 8948–8954. 146 (13) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, M. K. Triclocarban: Commercial Antibacterial and Highly Effective H-Bond Donating Catalyst for Ring-Opening Polymerization. ACS Macro Lett. 2017, 6 (4), 421–425. (14) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. T.; Kiesewetter, M. K. Rate Accelerated Organocatalytic Ring-Opening Polymerization of l-Lactide via the Application of a Bis(Thiourea) H-Bond Donating Cocatalyst. Macromolecules 2015, 48 (17), 6127–6131. (15) Xu, S.; Sun, H.; Liu, J.; Xu, J.; Pan, X.; Dong, H.; Liu, Y.; Li, Z.; Guo, K. Internal Lewis Pair Enhanced H-Bond Donor: Boronate-Urea and Tertiary Amine Co- Catalysis in Ring-Opening Polymerization. Polym. Chem. 2016, 7 (44), 6843– 6853. (16) Du, F.; Zheng, Y.; Yuan, W.; Shan, G.; Bao, Y.; Jie, S.; Pan, P. Solvent-Free Ring- Opening Polymerization of Lactones with Hydrogen-Bonding Bisurea Catalyst. J. Polym. Sci. Part A Polym. Chem. 2019, 57 (2), 90–100. (17) Ansyln, E. V.; Dougherty, D. A. Catalysis. In Modern Physical Organic Chemistry; 2006; pp 490–530. (18) Kazakov, O. I. I.; Kiesewetter, M. K. K. Cocatalyst Binding Effects in Organocatalytic Ring-Opening Polymerization of l -Lactide. Macromolecules 2015, 48 (17), 6121–6126. (19) Coady, D. J.; Engler, A. C.; Horn, H. W.; Bajjuri, K. M.; Fukushima, K.; Jones, G. O.; Nelson, A.; Rice, J. E.; Hedrick, J. L. Catalyst Chelation Effects in Organocatalyzed Ring-Opening Polymerization of Lactide. ACS Macro Lett. 2012, 147 1 (1), 19–22. (20) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. K. Cooperative Hydrogen-Bond Pairing in Organocatalytic Ring- Opening Polymerization. 2014, No. Figure 4.1. (21) Lin, B.; Waymouth, R. M. Organic Ring-Opening Polymerization Catalysts: Reactivity Control by Balancing Acidity. Macromolecules 2018, 51 (8), 2932– 2938. (22) Jakab, G.; Tancon, C.; Zhang, Z.; Lippert, K. M.; Schreiner, P. R. (Thio)Urea Organocatalyst Equilibrium Acidities in DMSO. Org. Lett. 2012, 14 (7), 1724– 1727. 148 Entry Solvent [M]o/[I]o Mna (g/mol) Mw/Mna 2-O5 1 benzene 50 9500 1.05 2 100 19600 1.05 3 200 30900 1.07 4 acetone 50 9100 1.05 5 100 12000 1.05 6 200 16900 1.15 7 solvent-freeb 50 10500 1.10 8 100 21500 1.07 9 200 42300 1.03 10 500 96200 1.16 2-O5-O 11 benzene 50 8000 1.06 12 100 20000 1.07 13 200 38700 1.10 14 acetone 50 7700 1.10 15 100 19700 1.04 16 200 35700 1.03 17 solvent-freeb 50 10500 1.10 18 100 23600 1.10 19 200 43500 1.02 20 500 110500 1.03 Entry Solvent [M]o/[I]o Mna (g/mol) Mw/Mna Table 4.1. Bis(thio)urea plus MTBD cocatalyzed ROP of VL. Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), cocatalyst (0.024 mmol each) (a) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. (b) Reaction conditions: VL (3.99 mmol) cocatalyst (0.019 mmol each). 149 Table 4.2. Bis(thio)urea and MTBD cocatalyzed ROP of VL. 1 (1), 19–22. Reaction conditio VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), bis(thio)urea/MT H-bond donor (2-S/On) time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 2-S2 92 88 8 300 1.06 2-S3 80 89 9 000 1.06 2-S4 53 90 8 200 1.06 2-S5 50 91 9 500 1.05 2-S6 69 89 9 200 1.04 2-S12 250 87 8 200 1.04 2-O3 20 89 8 900 1.07 2-O4 20 92 9 600 1.06 2-O5 12 89 9 500 1.05 2-O6 15 90 9 900 1.06 2-O12 35 90 9 900 1.07 H-bond donor (2-S/On) time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 2-S2 92 88 8 300 1.06 2-S3 80 89 9 000 1.06 2-S4 53 90 8 200 1.06 2-S5 50 91 9 500 1.05 2-S6 69 89 9 200 1.04 2-S12 250 87 8 200 1.04 2-O3 20 89 8 900 1.07 2-O4 20 92 9 600 1.06 2-O5 12 89 9 500 1.05 2-O6 15 90 9 900 1.06 2-O12 35 90 9 900 1.07 H-bond donor (2-S/On) time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 2-S2 92 88 8 300 1.06 2-S3 80 89 9 000 1.06 2-S4 53 90 8 200 1.06 2-S5 50 91 9 500 1.05 2-S6 69 89 9 200 1.04 2-S12 250 87 8 200 1.04 2-O3 20 89 8 900 1.07 2-O4 20 92 9 600 1.06 2-O5 12 89 9 500 1.05 2-O6 15 90 9 900 1.06 2-O12 35 90 9 900 1.07 Table 4.2. Bis(thio)urea and MTBD cocatalyzed ROP of VL. Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), bis(thio)urea/MTBD (0.024 mmol each) in C6D6. a. Monomer conversion was monitored via 1H NMR. b. M d M /M d t i d b GPC (CH Cl ) l t t d d Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 150 Table 4.3. 1 (1), 19–22. Bis(thio)urea Plus MTBD cocatalyzed ROP of VL in Acetone-d6 and solvent H-bond donor (2-S/On) time (min) conv.c (%) Mnd (g/mol) Mw/Mnd acetone-d6a 2-S5 180 90 8 900 1.08 2-O5 12 84 8 200 1.05 2-O12 40 84 8 200 1.10 solvent-freeb 2-O3 20 99 42 300 1.05 2-O4 22 99 49 600 1.04 2-O5 12 99 42 300 1.03 2-O6 19 99 39 400 1.02 2-O12 29 99 39 200 1.02 solvent H-bond donor (2-S/On) time (min) conv.c (%) Mnd (g/mol) Mw/Mnd acetone-d6a 2-S5 180 90 8 900 1.08 2-O5 12 84 8 200 1.05 2-O12 40 84 8 200 1.10 solvent-freeb 2-O3 20 99 42 300 1.05 2-O4 22 99 49 600 1.04 2-O5 12 99 42 300 1.03 2-O6 19 99 39 400 1.02 2-O12 29 99 39 200 1.02 Table 4.3. Bis(thio)urea Plus MTBD cocatalyzed ROP of VL in Acetone-d6 and Solvent-free conditions. a) Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), bis(thio)urea/MTBD (0.024 mmol each) in acetone-d6. b) Solvent free-reaction conditions: VL (3.99 mmol, 1 equiv), benzyl alcohol (0.019 mmol), cocatalyst (0.019 mmol each). c) Monomer conversion was monitored via 1H NMR. d) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 151 VL or CL Solvent H-bond donor (2-O5-N/O) Time (min) conv. d (%) Mne (g/mol) Mw/Mne VL benzene-d6 a 2-O5-N 10 88 8 000 1.05 2-O5-O 5 90 8 000 1.06 acetone-d6 b 2-O5-N 8 86 7 800 1.10 2-O5-O 5 86 7 700 1.11 solvent-free 2-O5-Nd 5 91 37 500 1.02 2-O5-Oe 4 99 110 500 1.03 CL solvent-freec 2-O5 30 99 50 500 1.14 2-O5-N 18 99 51 100 1.20 2-O5-O 8 99 47 000 1.13 VL or CL Solvent H-bond donor (2-O5-N/O) Time (min) conv. d (%) Mne (g/mol) Mw/Mne VL benzene-d6 a 2-O5-N 10 88 8 000 1.05 2-O5-O 5 90 8 000 1.06 acetone-d6 b 2-O5-N 8 86 7 800 1.10 2-O5-O 5 86 7 700 1.11 solvent-free 2-O5-Nd 5 91 37 500 1.02 2-O5-Oe 4 99 110 500 1.03 CL solvent-freec 2-O5 30 99 50 500 1.14 2-O5-N 18 99 51 100 1.20 2-O5-O 8 99 47 000 1.13 Table 4.4. ROP of VL or CL cocatalyzed by MTBD plus Bisureas with Heteroatom- containing Tethers. a) Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), cocatalyst (0.024 mmol each) in C6D6 or acetone-d6. 1 (1), 19–22. b) Solvent free- reaction conditions: VL or CL (3.99 mmol, 1 equiv), benzyl alcohol (0.019 mmol), cocatalyst (0.019 mmol each) c) Solvent free-reaction conditions: VL (3.99 mmol, 1 equiv), benzyl alcohol (0.008 mmol), cocatalyst (0.019 mmol each). d) Monomer conversion was monitored via 1H NMR. e) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 152 Table 4.5. ROP of VL cocatalyzed by MTBD plus Bisthioureas with Heteroat containing Tethers. a) Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alc (0.019 mmol), cocatalyst (0.024 mmol each) in C6D6 or acetone-d6, b) Reaction conditi VL (3.99 mmol, 1 equiv), benzyl alcohol (0.019 mmol), cocatalyst (0.019 mmol each Monomer conversion was monitored via 1H NMR. d) Mn and Mw/Mn were determine Solvent H-bond donor (2-S5-N/O) Time (min) conv. c (%) Mnd (g/mol) Mw/Mnd benzene-d6 a 2-S5-N 100 89 8 600 1.05 2-S5-O 50 90 10 600 1.03 acetone-d6 a 2-S5-N 240 84 7 700 1.07 2-S5-O 210 83 10 500 1.04 solvent-freeb 2-S5-N 90 97 42 600 1.06 2-S5-O 60 99 43 600 1.06 Table 4.5. ROP of VL cocatalyzed by MTBD plus Bisthioureas with Heteroatom containing Tethers. a) Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcoh (0.019 mmol), cocatalyst (0.024 mmol each) in C6D6 or acetone-d6, b) Reaction condition VL (3.99 mmol, 1 equiv), benzyl alcohol (0.019 mmol), cocatalyst (0.019 mmol each). c Monomer conversion was monitored via 1H NMR. d) Mn and Mw/Mn were determined b GPC (CH2Cl2) versus polystyrene standards. Solvent H-bond donor (2-S5-N/O) Time (min) conv. c (%) Mnd (g/mol) Mw/Mnd benzene-d6 a 2-S5-N 100 89 8 600 1.05 2-S5-O 50 90 10 600 1.03 acetone-d6 a 2-S5-N 240 84 7 700 1.07 2-S5-O 210 83 10 500 1.04 solvent-freeb 2-S5-N 90 97 42 600 1.06 2-S5-O 60 99 43 600 1.06 Table 4 5 ROP of VL cocatalyzed by MTBD plus Bisthioureas with Heteroa Solvent H-bond donor (2-S5-N/O) Time (min) conv. c (%) Mnd (g/mol) Mw/Mnd benzene-d6 a 2-S5-N 100 89 8 600 1.05 2-S5-O 50 90 10 600 1.03 acetone-d6 a 2-S5-N 240 84 7 700 1.07 2-S5-O 210 83 10 500 1.04 solvent-freeb 2-S5-N 90 97 42 600 1.06 2-S5-O 60 99 43 600 1.06 Table 4.5. ROP of VL cocatalyzed by MTBD plus Bisthioureas with Heteroatom- containing Tethers. 1 (1), 19–22. a) Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), cocatalyst (0.024 mmol each) in C6D6 or acetone-d6, b) Reaction conditions: VL (3.99 mmol, 1 equiv), benzyl alcohol (0.019 mmol), cocatalyst (0.019 mmol each). c) Monomer conversion was monitored via 1H NMR. d) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 153 solvent H-bond donor (2-S/O5-N/O) Time (min) conv.c (%) Mnd (g/mol) Mw/Mnd CH2Cl2 a 2-S5 12 90 17 400 1.05 2-S5-N 20 90 17 600 1.04 2-S5-O 5 90 18 800 1.04 solvent-free b 2-O5-O 105 90 15 800 1.13 2-S5-O 15 90 18 500 1.06 solvent H-bond donor (2-S/O5-N/O) Time (min) conv.c (%) Mnd (g/mol) Mw/Mnd CH2Cl2 a 2-S5 12 90 17 400 1.05 2-S5-N 20 90 17 600 1.04 2-S5-O 5 90 18 800 1.04 solvent-free b 2-O5-O 105 90 15 800 1.13 2-S5-O 15 90 18 500 1.06 solvent H-bond donor (2-S/O5-N/O) Time (min) conv.c (%) Mnd (g/mol) Mw/Mnd CH2Cl2 a 2-S5 12 90 17 400 1.05 2-S5-N 20 90 17 600 1.04 2-S5-O 5 90 18 800 1.04 solvent-free b 2-O5-O 105 90 15 800 1.13 2-S5-O 15 90 18 500 1.06 Table 4.6. Bis(thio)urea and Me6TREN cocatalyzed ROP of L-LA. a) Reaction conditions: L- LA (0.693 mmol, 1 equiv, 1 M), benzyl alcohol (0.0069 mmol,), cocatalyst (0.017 mmol each) in CH2Cl2. b) Solvent-free reaction conditions: L-LA (1.38 mmol, 1 equiv), benzyl alcohol (0.0138 mmol), cocatalyst (0.007 mmol each) at 100 °C. c) Monomer conversion was monitored via 1H NMR. d) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 154 Table 4.7. Mono(thio)urea and Me6TREN cocatalyzed ROP of L-LA. Reaction conditions L-LA (0 50 mmol 1 equiv 1 M) benzyl alcohol (0 005 mmol) cocatalyst (0 025 mmo H-bond donor pKaa time (min) conv.b (%) Mnc (g/mol) Mw/Mnc 1-S-H 16.8 1440 - - - 1-O-CF3 16.1 35 90 17 900 1.04 1-S 13.2 48 90 18 900 1.07 Schreiner’s urea 13.8 1 92 19 300 1.07 Schreiner’s thiourea 8.5 600 90 18 100 1.04 Table 4.7. Mono(thio)urea and Me6TREN cocatalyzed ROP of L-LA. Reaction conditions: L-LA (0.50 mmol, 1 equiv, 1 M), benzyl alcohol (0.005 mmol), cocatalyst (0.025 mmol each) in CH2Cl2. a) pKa values in DMSO21 b) Monomer conversion was monitored via 1H NMR. c) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. 1 (1), 19–22. L-LA (0.50 mmol, 1 equiv, 1 M), benzyl alcohol (0.005 mmol), cocatalyst (0.025 mmol each) in CH2Cl2. a) pKa values in DMSO21 b) Monomer conversion was monitored via 1H NMR. c) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. L-LA (0.50 mmol, 1 equiv, 1 M), benzyl alcohol (0.005 mmol), cocatalyst (0.025 mmol each) in CH2Cl2. a) pKa values in DMSO21 b) Monomer conversion was monitored via 1H NMR. c) Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. Scheme 4.1. Neutral H-bond versus imidate mediated ROP of VL. Scheme 4.1. Neutral H-bond versus imidate mediated ROP of VL. 155 Figure 4.1. Mono(thio)urea, bis(thio)urea donors evaluated for the 1-X/2-Xn plus MTBD and Me6TREN mediated ROP of VL, CL, L-LA and proposed activated-thiourea mode activation for bis-(thio)urea H-bond donors. Figure 4.1. Mono(thio)urea, bis(thio)urea donors evaluated for the 1-X/2-Xn plu Figure 4.1. Mono(thio)urea, bis(thio)urea donors evaluated for the 1-X/2-Xn plus MTBD and Me6TREN mediated ROP of VL, CL, L-LA and proposed activated-thiourea mode activation for bis-(thio)urea H-bond donors. Figure 4.1. Mono(thio)urea, bis(thio)urea donors evaluated for the 1-X/2-Xn plus MTBD and Me6TREN mediated ROP of VL, CL, L-LA and proposed activated-thiourea mode activation for bis-(thio)urea H-bond donors. Figure 4.2. Mn and Mw/Mn versus conversion for the H-bond donor plus MTBD cocatalyzed ROP of VL using (left) 2-S5 and (right) 2-O5. Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), 2-X5/MTBD (0.024 mmol each) in C6D6. 1 1.2 1.4 1.6 1.8 2 0 2000 4000 6000 8000 10000 0 20 40 60 80 100 Mw/Mn Mn % conversion 1 1.2 1.4 1.6 1.8 2 0 2000 4000 6000 8000 10000 0 20 40 60 80 100 Mw/Mn Mn % conversion 1 1.2 1.4 1.6 1.8 2 0 2000 4000 6000 8000 10000 0 20 40 60 80 100 Mw/Mn Mn % conversion 1 1.2 1.4 1.6 1.8 2 0 2000 4000 6000 8000 10000 0 20 40 60 80 100 Mw/Mn Mn % conversion Figure 4.2. Mn and Mw/Mn versus conversion for the H-bond donor plus MTBD cocatalyzed ROP of VL using (left) 2-S5 and (right) 2-O5. Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), 2-X5/MTBD (0.024 mmol each) in C6D6. C6D6. 156 Figure 4.3. Proposed activated (thio)urea anion mechanism for the bisurea plus MTBD mediated ROP of VL. Figure 4.3. 1 (1), 19–22. Proposed activated (thio)urea anion mechanism for the bisurea plus MTBD mediated ROP of VL. Figure 4.4. Downfield portion of 1H NMR spectra (400 MHz, ppm) of 2-O5 plus MTBD in acetone-d6. Figure 4.4. Downfield portion of 1H NMR spectra (400 MHz, ppm) of 2-O5 plus MTBD in acetone-d6. 157 Figure 4.5. First order evolution of VL versus time for the 2-O5/MTBD catalyzed ring- opening polymerization of VL. Conditions: VL (2 M, 1 mmol), benzyl alcohol (2 mol%, 0.019 mmol), 2-O5 (0.024 mmol), MTBD orange - 0.024 mmol, blue- 0.048 mmol) in acetone-d6. y = 0.2973x + 0.0046 R² = 0.9895 y = 0.2979x - 0.0058 R² = 0.9765 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0 0.2 0.4 0.6 0.8 1 1.2 ln ([VL]0/[VL]) Time (min) 2-O5:MTBD (1:1) 2-O5+MTBD (1:2) y = 0.2973x + 0.0046 R² = 0.9895 y = 0.2979x - 0.0058 R² = 0.9765 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0 0.2 0.4 0.6 0.8 1 1.2 ln ([VL]0/[VL]) Time (min) 2-O5:MTBD (1:1) 2-O5+MTBD (1:2) Figure 4.5. First order evolution of VL versus time for the 2-O5/MTBD catalyzed ring- opening polymerization of VL. Conditions: VL (2 M, 1 mmol), benzyl alcohol (2 mol%, 0.019 mmol), 2-O5 (0.024 mmol), MTBD orange - 0.024 mmol, blue- 0.048 mmol) in acetone-d6. Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), cocatalyst (0.024 mmol each) in acetone-d6. y = 104.51x + 126.67 R² = 0.9831 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 20 40 60 80 100 Mw/Mn Mn % conversion Reaction conditions: VL (1.0 mmol, 1 equiv, 2 M), benzyl alcohol (0.019 mmol), cocatalyst (0.024 mmol each) in acetone-d6. 158 Figure 4.7. Mn and Mw/Mn versus conversion for 2-O5-O catalyst. Reaction conditions: VL (3.99 mmol, 1 equiv), benzyl alcohol (0.008 mmol), cocatalyst (0.024 mmol each) under solvent free conditions. Figure 4.7. Mn and Mw/Mn versus conversion for 2-O5-O catalyst. Reaction conditions: VL (3.99 mmol, 1 equiv), benzyl alcohol (0.008 mmol), cocatalyst (0.024 mmol each) under solvent free conditions. Figure 4.8. (left) Mn and Mw/Mn versus conversion for 2-S5, (right) Mn versus conversion for 2-O5 catalyst. Reaction conditions: VL (3.99 mmol, 1 equiv.), benzyl alcohol (0.019 mmol,), cocatalyst (0.019 mmol each) under solvent free conditions. 1 (1), 19–22. y = 718.11x - 2880.5 R² = 0.9722 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 10000 20000 30000 40000 50000 60000 70000 80000 0 20 40 60 80 100 Mw/Mn Mn % Conversion y = 246.67x + 710.84 R² = 0.9966 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 50 100 Mw/Mn Mn % Conversion y = 718.11x - 2880.5 R² = 0.9722 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 10000 20000 30000 40000 50000 60000 70000 80000 0 20 40 60 80 100 Mw/Mn Mn % Conversion y = 246.67x + 710.84 R² = 0.9966 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 50 100 Mw/Mn Mn % Conversion Figure 4.8. (left) Mn and Mw/Mn versus conversion for 2-S5, (right) Mn versus conversion for 2-O5 catalyst. Reaction conditions: VL (3.99 mmol, 1 equiv.), benzyl alcohol (0.019 mmol,), cocatalyst (0.019 mmol each) under solvent free conditions. 159 Figure 4.9. Downfield portion of 1 H NMR spectra (400 MHz, ppm) of 2-O5-O and 2-S5- O with and without Me6TREN in acetone- d6. + Me6TREN + Me6TREN a a a b b b c c c d d d c c c d a b c c c c c c d d d d a d a a b b b Figure 4.9. Downfield portion of 1 H NMR spectra (400 MHz, ppm) of 2-O5-O and 2-S5- O with and without Me6TREN in acetone- d6. 160 Figure 4.10. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(ethane-1,2- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O2), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(ethane-1,2-diyl)bis(3-(3,5- Figure 4.10. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(ethane-1,2- Figure 4.10. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(ethane-1,2- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O2), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(ethane-1,2-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)urea). Figure 4.10. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(ethane-1,2- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O2), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(ethane-1,2-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)urea). 161 Figure 4.11. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O4), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O4), Figure 4.11. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O4), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O4), Figure 4.11. 1 (1), 19–22. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O4), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O4), Figure 4.11. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O4), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O4), 162 Figure 4.12. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O5), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O5). Figure 4.12. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O5), (Lower) 13 Figure 4.12. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O5), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O5). Figure 4.12. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O5), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O5). 163 Figure 4.13. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O6), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O6). Figure 4.13. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O6), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O6). 164 Figure 4.14. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O12), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O12). Figure 4.14. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O12), (Lower) Figure 4.14. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of (2-O12), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of (2-O12). 165 Figure 4.15. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2- O5-N), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea). Figure 4.15. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2- O5-N), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea). Figure 4.15. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2- O5-N), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea). 166 Figure 4.16. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)urea). Figure 4.16. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)urea). Figure 4.16. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)urea). 167 Figure 4.17. 1 (1), 19–22. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(2,2- dimethylpropane-1,3-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O3-diMe) , (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(2,2-dimethylpropane- 1,3-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea). Figure 4.17. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(2,2- dimethylpropane-1,3-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O3-diMe) , (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(2,2-dimethylpropane- 1,3-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea). Figure 4.17. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(2,2- dimethylpropane-1,3-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea) (2-O3-diMe) , (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(2,2-dimethylpropane- 1,3-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)urea). 168 Figure 4 18 (Upper) 1H NMR (acetone d6 400 MHz ppm) spectrum of 1 1' (butane 1 4 Figure 4.18. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(butane-1,4 iyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S4), (Lower) 13C NMR (acetone 6, 100 MHz, ppm) spectrum of 1,1'-(butane-1,4-diyl)bis(3-(3,5 i (t ifl th l) h l)thi ) Figure 4.18. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(butane-1,4- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S4), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(butane-1,4-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). 169 Figure 4.19. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(pentane-1,5- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(pentane-1,5-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea) Figure 4.19. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(pentane-1,5- Figure 4.19. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(pentane-1,5- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(pentane-1,5-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). 170 Figure 4.20. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(hexane-1,6- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S6), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(hexane-1,6-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). Figure 4.20. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(hexane-1,6- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S6), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(hexane-1,6-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). Figure 4.20. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(hexane-1,6- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S6), (Lower) 13C NMR (acetone- Figure 4.20. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(hexane-1,6- diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S6), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(hexane-1,6-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). 171 Figure 4.21. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(dodecane- 1,12-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S12), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(dodecane-1,12-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). Figure 4.21. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(dodecane- 1,12-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S12), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(dodecane-1,12-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). Figure 4.21. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum of 1,1'-(dodecane- 1,12-diyl)bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S12), (Lower) 13C NMR (acetone-d6, 100 MHz, ppm) spectrum of 1,1'-(dodecane-1,12-diyl)bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). 172 Figure 4.22. 1 (1), 19–22. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-N), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) Figure 4.22. (Upper) 1H NMR (acetone-d6, 400 MHz, ppm) spectrum 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-N), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'- ((methylazanediyl)bis(ethane-2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) 173 Figure 4.23. (Upper) 1H NMR (acetone- d6, 400 MHz, ppm) spectrum 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). CF3 F3C N H N H O N H N H S CF3 CF3 S CF3 F3C N H N H O N H N H S CF3 CF3 S Figure 4.23. (Upper) 1H NMR (acetone- d6, 400 MHz, ppm) spectrum 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). CF3 F3C N H N H O N H N H S CF3 CF3 S Figure 4.23. (Upper) 1H NMR (acetone- d6, 400 MHz, ppm) spectrum 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). Figure 4.23. (Upper) 1H NMR (acetone- d6, 400 MHz, ppm) spectrum 1,1'-(oxybis(ethane- 2,1-diyl))bis(3-(3,5-bis(trifluoromethyl)phenyl)thiourea) (2-S5-O), (Lower) 13C NMR (acetone- d6, 100 MHz, ppm) spectrum of 1,1'-(oxybis(ethane-2,1-diyl))bis(3-(3,5- bis(trifluoromethyl)phenyl)thiourea). 174 MANUSCRIPT – V Published in ACS Macro Letters Triclocarban: commercial antibacterial and highly effective H-bond donating Email address: mkiesewetter@chm.uri.edu INTRODUCTION H-bond mediated ring-opening polymerization (ROP) has attracted interest due to the highly controlled nature of these transformations.1−4 These mild, highly functional group tolerant catalysts, especially the bimolecular systems consisting of (thio)urea H-bond donor plus H-bond accepting base, have facilitated the construction of precise polymer architectures, multiblocks, and well-defined systems.3,5−8 Targeted efforts by several groups toward rate-accelerated, H-bond mediated ROP seek to address a critical shortcoming of the field: low activity.9−12 For example, our group has recently disclosed the utility of urea H-bond donors for rate accelerated ROP;13 thiourea H-bond donors have been used in organocatalytic ROP for more than a decade, but are less active.3 Another barrier to the wide implementation of this chemistry is the paucity of commercially available H-bond donors. Most (thio)urea catalysts are synthesized via a “click” reaction of an appropriate amine and iso(thio)cyanate.3,14,15 While simple, this stands in contrast to the wide array of readily available H-bond accepting base cocatalysts and adds a synthetic step prior to conducting polymerization chemistry. Certainly, the ready availability of chemical reagents and catalysts facilitates the wide implementation of chemical transformations. In this context, the antibacterial compound, triclocarban (TCC, Figure 1), recently banned as a hand soap additive by the FDA, captured our attention.16 It is an electron-deficient biaryl urea, similar to the slate of urea and multiurea H-bond donating catalysts that we recently showed to be highly active for ROP.13 While TCC has attracted considerable scientific interest as an antibacterial compound, possible bioaccumulate, and H-bond mediated ring-opening polymerization (ROP) has attracted interest due to the highly controlled nature of these transformations.1−4 These mild, highly functional group tolerant catalysts, especially the bimolecular systems consisting of (thio)urea H-bond donor plus H-bond accepting base, have facilitated the construction of precise polymer architectures, multiblocks, and well-defined systems.3,5−8 Targeted efforts by several groups toward rate-accelerated, H-bond mediated ROP seek to address a critical shortcoming of the field: low activity.9−12 For example, our group has recently disclosed the utility of urea H-bond donors for rate accelerated ROP;13 thiourea H-bond donors have been used in organocatalytic ROP for more than a decade, but are less active.3 Another barrier to the wide implementation of this chemistry is the paucity of commercially available H-bond donors. ABSTRACT The antibacterial compound, triclocarban (TCC), is shown to be a highly effective H-bond donating catalyst for ring-opening polymerization (ROP) when applied with an H- bond accepting base cocatalyst. These ROPs exhibit the characteristics of ‘living’ polymerizations. TCC is shown to possess the high activity characteristic of urea (vs thiourea) H-bond donors. The urea class of H-bond donors is shown to remain highly active in H-bonding solvents, a trait that is not displayed by the corresponding thiourea H- bond donors. Two H-bond donating ureas that are electronically similar to TCC are evaluated for their efficacy in ROP, and a mechanism of action is proposed. This ‘off-the- shelf’ H-bond donor is among the most active and most controlled organocatalysts for the ROP of lactones. 176 INTRODUCTION Most (thio)urea catalysts are synthesized via a “click” reaction of an appropriate amine and iso(thio)cyanate.3,14,15 While simple, this stands in contrast to the wide array of readily available H-bond accepting base cocatalysts and adds a synthetic step prior to conducting polymerization chemistry. Certainly, the ready availability of chemical reagents and catalysts facilitates the wide implementation of chemical transformations. In this context, the antibacterial compound, triclocarban (TCC, Figure 1), recently banned as a hand soap additive by the FDA, captured our attention.16 It is an electron-deficient biaryl urea, similar to the slate of urea and multiurea H-bond donating catalysts that we recently showed to be highly active for ROP.13 While TCC has attracted considerable scientific interest as an antibacterial compound, possible bioaccumulate, and 177 possible environmental toxin, we believe that this readily available compound has not previously been employed as a catalyst.17−19 le environmental toxin, we believe that this readily available compound has no 178 EXPERIMENTAL SECTION General Considerations. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Triclocarban (TCC), 7-methyl-1,5,7- triazabicyclo[4.4.0]dec-5-ene (MTBD) and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) were purchased from TCI. Tris[2-(dimethylamino)ethyl]amine (Me6TREN) and 2-tert- butylimino-2-diethylamino-1,3-dimethylperhydro-1,3,2-diazaphosphorine (BEMP) were purchased from Alfa Aesar. Benzyl alcohol and 1,2-dichlorobenzene were distilled under high vacuum from calcium hydride. THF was dried on an Innovative Technology solvent purification system. DMF was dried over 4 Å molecular sieves for 48 h prior to use. 1- pyrenebutanol was purchased from Sigma Aldrich. δ-valerolactone (VL), ε-caprolactone (CL) and β-butyrolactone (BL) were distilled from calcium hydride under high vacuum. L-Lactide (L-LA) was purchased from Acros Organics and recrystallized from dry toluene. Benzene-d6 and chloroform-d were purchased from Cambridge Isotope Laboratories and distilled from calcium hydride. Acetone-d6 was purchased from Cambridge Isotope Laboratories and stored over 4 Å molecular sieves for 48 h prior to use. Experiments were conducted using pre-dried glassware in an MBRAUN or INERT stainless steel glovebox or using a Schlenk line under nitrogen atmosphere. NMR experiments were conducted on a Bruker Avance III 300 MHz or 400 MHz spectrometer or a Varian 500 MHz spectrometer. Gel Permeation Chromatography (GPC) was performed at 40 °C using HPLC grade dichloromethane eluent on an Agilent Infinity GPC system equipped with three Agilent PLGel columns 7.5 mm × 300 mm (5 μm, pore sizes: 103, 104, 50 Å). Mn and Mw/Mn were determined versus polystyrene standards (500 g/mol-3150 kg/mol, General Considerations. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Triclocarban (TCC), 7-methyl-1,5,7- triazabicyclo[4.4.0]dec-5-ene (MTBD) and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) were purchased from TCI. Tris[2-(dimethylamino)ethyl]amine (Me6TREN) and 2-tert- butylimino-2-diethylamino-1,3-dimethylperhydro-1,3,2-diazaphosphorine (BEMP) were purchased from Alfa Aesar. Benzyl alcohol and 1,2-dichlorobenzene were distilled under high vacuum from calcium hydride. THF was dried on an Innovative Technology solvent purification system. DMF was dried over 4 Å molecular sieves for 48 h prior to use. 1- pyrenebutanol was purchased from Sigma Aldrich. δ-valerolactone (VL), ε-caprolactone (CL) and β-butyrolactone (BL) were distilled from calcium hydride under high vacuum. L-Lactide (L-LA) was purchased from Acros Organics and recrystallized from dry toluene. Benzene-d6 and chloroform-d were purchased from Cambridge Isotope Laboratories and distilled from calcium hydride. Acetone-d6 was purchased from Cambridge Isotope Laboratories and stored over 4 Å molecular sieves for 48 h prior to use. EXPERIMENTAL SECTION Experiments were conducted using pre-dried glassware in an MBRAUN or INERT stainless steel glovebox or using a Schlenk line under nitrogen atmosphere. NMR experiments were conducted on a Bruker Avance III 300 MHz or 400 MHz spectrometer or a Varian 500 MHz spectrometer. Gel Permeation Chromatography (GPC) was performed at 40 °C using HPLC grade dichloromethane eluent on an Agilent Infinity GPC system equipped with three Agilent PLGel columns 7.5 mm × 300 mm (5 μm, pore sizes: 103, 104, 50 Å). Mn and Mw/Mn were determined versus polystyrene standards (500 g/mol-3150 kg/mol, 179 Polymer Laboratories). Mass spectrometry was performed using a Thermo Scientific (San Jose, CA, USA) LTQ Orbitrap XL mass spectrometer affixed with either an atmospheric- pressure chemical ionization (APCI) or electrospray ionization (ESI) interface, positive ions were produced and introduced into the S2 instrument. Tune conditions for infusion experiments (5 μL/min flow, sample concentration 5 μg/mL in 50/50 v/v water/methanol) were as follows: ionspray voltage, 5 kV; capillary temperature, 275 °C; sheath gas (N2, arbitrary units), 8; auxiliary gas (N2, arbitrary units), 2; capillary voltage, 35 V; and tube lens, 90 V. Prior to analysis, the instrument was calibrated for positive ions using Pierce LTQ ESI positive ion calibration solution (lot #PC197784). Ion trap experiments used N2 as a collision gas with normalized collision energies (NCE) between 10-25 eV for multistage fragmentation. High-energy collision (HCD) experiments were performed with He as the collision gas with a NCE of 25 eV. Polymer Laboratories). Mass spectrometry was performed using a Thermo Scientific (San Jose, CA, USA) LTQ Orbitrap XL mass spectrometer affixed with either an atmospheric- pressure chemical ionization (APCI) or electrospray ionization (ESI) interface, positive ions were produced and introduced into the S2 instrument. Tune conditions for infusion experiments (5 μL/min flow, sample concentration 5 μg/mL in 50/50 v/v water/methanol) were as follows: ionspray voltage, 5 kV; capillary temperature, 275 °C; sheath gas (N2, arbitrary units), 8; auxiliary gas (N2, arbitrary units), 2; capillary voltage, 35 V; and tube lens, 90 V. Prior to analysis, the instrument was calibrated for positive ions using Pierce LTQ ESI positive ion calibration solution (lot #PC197784). Ion trap experiments used N2 as a collision gas with normalized collision energies (NCE) between 10-25 eV for multistage fragmentation. High-energy collision (HCD) experiments were performed with He as the collision gas with a NCE of 25 eV. Example ring-opening polymerization of VL. EXPERIMENTAL SECTION To a 7 mL vial, TCC (15.7 mg, 0.05 mmol), VL (100 mg, 1.0 mmol) and benzene-d6 (250 μL) were added. The contents were stirred until the solution became homogenous. To a second 7 ml vial, benzyl alcohol (4.3 mg, 0.04 mmol), MTBD (7.6 mg, 0.05 mmol) and benzene-d6 (250 μL) were added. The contents in the second vial were transferred to the first vial via Pasteur pipette, and the contents were agitated to mix. The reaction solution was then transferred to an NMR tube, and the progress of the reaction monitored by 1H NMR. The reaction was quenched by the addition of benzoic acid (6.1 mg, 0.05 mmol). Polymer isolated by precipitation with hexanes contains residual TCC that can be removed by repeated precipitation or washing 180 with methanol. PVL was removed of volatiles under high vacuum prior to characterization. Yield 89%, Mn (GPC)= 5,400, Mw/Mn = 1.09, Mn (NMR) = 2,700. with methanol. PVL was removed of volatiles under high vacuum prior to characterization. Yield 89%, Mn (GPC)= 5,400, Mw/Mn = 1.09, Mn (NMR) = 2,700. Example post-polymerization transesterification. To a 7 mL vial, TCC (15.7 mg, 0.05 mmol), VL (100 mg, 1.0 mmol) and benzene-d6 (250 μL) were added. The contents were stirred until the solution became homogenous. To a second 7 ml vial, benzyl alcohol (1.1 mg, 0.01 mmol), MTBD (7.6 mg, 0.05 mmol) and benzene-d6 (250 μL) were added. The contents in the second vial were transferred to the first vial via Pasteur pipette, and the contents were agitated to mix. Three 50 μL aliquots from the reaction were quenched at 20 min, 45 min and 60 min using benzoic acid (6.2 mg, 0.05 mmol). Polymer in each aliquot was then isolated by precipitation with hexanes. PVL was removed of volatiles under high vacuum prior to characterization by GPC: Mn = 22,300, 23,900, 23,900, Mw/Mn = 1.02, 1.03, 1.03 respectively. Example chain extension experiment. To a 7 mL vial, TCC (15.7 mg, 0.05 mmol), VL (100 mg, 1.0 mmol) and benzene-d6 (250 μL) were added. The contents were agitated until the solution became homogenous. To a second 7 ml vial, 1-pyrenebutanol (5.5 mg, 0.02 mmol), MTBD (7.6 mg, 0.05 mmol) and benzene-d6 (250 μL) were added. The contents of the second vial were transferred to the first vial via Pasteur pipette, and the contents were agitated to mixed. EXPERIMENTAL SECTION After 13 min, a 100 μL aliquot from the reaction was quenched using benzoic acid (6.2 mg, 0.05 mmol), and VL (100 mg, 1.0 mmol) was added to the reaction vial. A second 100 μL aliquot from the reaction vial was quenched in 27 min Example chain extension experiment. To a 7 mL vial, TCC (15.7 mg, 0.05 mmol), VL 181 using benzoic acid (6.2 mg, 0.05 mmol). Conversion of VL in the two aliquots were then determined by 1H NMR, followed by the isolation of PVL and characterization by GPC. Example ring-opening polymerization of L-Lactide. A first 7 mL vial was charged with TCC (15.8 mg, 0.05 mmol), Me6TREN (13.4 μL, 0.05 mmol) and benzyl alcohol (1.0 μL, 0.01 mmol). A second 7 mL vial was charged with L-LA (144.1 mg, 1 mmol) and acetone- d6 (1000 μL). The contents of the second vial were added to the first vial, and the resulting mixture was vigorously shaken until homogenous. The reaction mixture was transferred to an NMR tube via pipette, and the reaction progress was monitored by 1H NMR. The reaction was quenched with benzoic acid (0.1 mmol). The reaction mixture was removed of volatiles under reduced pressure, dissolved in minimal dichloromethane, and the polylactide (PLA) was precipitated with the addition of hexanes. The supernatant was decanted, and the precipitate was subjected to high vacuum to remove volatiles. Example binding experiment. For the titration of 1-O with CL, stock solutions of 1-O and CL were prepared in benzene-d6. Into several NMR tubes, varying amounts of each solution were added to each tube along with neat benzene-d6 such that the final volume of each sample was 0.4 mL. The final concentrations were [1-O] = 0.005M and 0.25M < [CL] < 2.25 M. 1H-NMR spectra (referenced to residual benzene-H) were acquired for each tube at 300 K and the chemical shift of the ortho-protons of 1-O was noted. Binding constants were determined by the curve fitting method,33-35 and these values match those determined from the Lineweaver-Burke method.36,37 Binding curves are shown below (Figures 5.15). 182 Example synthesis of 1-(4-chlorophenyl)-3-phenylurea (mono-CC). A dried Schlenk flask was charged with 4-chlorophenylisocyanate (598.2 mg, 3.90 mmol) and ~10 mL dried DCM. Next, aniline (0.36 mL, 3.95 mmol) was added via syringe. Immediately upon addition of aniline, a white precipitate formed. The reaction mixture was filtered and rinsed 3 times with cold DCM to provide a pure white powder (846.1 mg, 3.43 mmol, 88.1 % yield). Characterization matches literature;38 NMR spectra below; 13C NMR (CD3OD, 100 MHz): δ = 155.2, 140.3, 139.5, 129.5, 129.4, 128.5, 124.0, 121.6, 120.5. Example synthesis of 1-(3,4-dichlorophenyl)-3-phenylurea (di-CC). A dried Schlenk flask was charged with 3,4-dichlorophenylisocyanate (731.8 mg, 3.89 mmol) and ~10 mL dried DCM. Next, aniline (0.36 mL, 3.95 mmol) was added via syringe. Immediately upon addition of aniline a white precipitate formed. The reaction mixture was filtered and rinsed 3 times with cold DCM to provide a pure white powder (1.01 g, 3.59 mmol, 92.7 % yield). Characterization matches literature;39 NMR spectra below; 13C NMR (CD3OD, 100 MHz): δ = 154.87, 140.81, 140.16, 133.32, 131.48, 129.89, 126.14, 124.12, 121.39, 120.53, 119.60. 183 RESULTS AND DISCUSSION H-bond mediated ring-opening polymerization (ROP) has attracted interest due to the highly controlled nature of these transformations.1−4 These mild, highly functional group tolerant catalysts, especially the bimolecular systems consisting of a (thio)urea H-bond donor plus H-bond accepting base, have facilitated the construction of precise polymer architectures, multiblocks, and well-defined systems.3,5−8 Targeted efforts by several groups toward rate-accelerated, H-bond mediated ROP seek to address a critical shortcoming of the field: low activity.9−12 For example, our group has recently disclosed the utility of urea H-bond donors for rate accelerated ROP;13 thiourea H-bond donors have been used in organocatalytic ROP for more than a decade, but are less active.3 Another barrier to the wide implementation of this chemistry is the paucity of commercially available H-bond donors. Most (thio)urea catalysts are synthesized via a “click” reaction of an appropriate amine and iso(thio)cyanate.3,14,15 While simple, this stands in contrast to the wide array of readily available H-bond accepting base cocatalysts and adds a synthetic step prior to conducting polymerization chemistry. Certainly, the ready availability of chemical reagents and catalysts facilitates the wide implementation of chemical transformations. In this context, the antibacterial compound, triclocarban (TCC, Figure 5.1), recently banned as a hand soap additive by the FDA, captured our attention.16 It is an electron-deficient biaryl urea, similar to the slate of urea and multiurea H-bond donating catalysts that we recently showed to be highly active for ROP.13 While TCC has attracted considerable scientific interest as an antibacterial compound, possible bioaccumulate, and possible environmental toxin, we believe that this readily available compound has not H-bond mediated ring-opening polymerization (ROP) has attracted interest due to the highly controlled nature of these transformations.1−4 These mild, highly functional group tolerant catalysts, especially the bimolecular systems consisting of a (thio)urea H-bond donor plus H-bond accepting base, have facilitated the construction of precise polymer architectures, multiblocks, and well-defined systems.3,5−8 Targeted efforts by several groups toward rate-accelerated, H-bond mediated ROP seek to address a critical shortcoming of the field: low activity.9−12 For example, our group has recently disclosed the utility of urea H-bond donors for rate accelerated ROP;13 thiourea H-bond donors have been used in organocatalytic ROP for more than a decade, but are less active.3 Another barrier to the wide implementation of this chemistry is the paucity of commercially available H-bond donors. RESULTS AND DISCUSSION Most (thio)urea catalysts are synthesized via a “click” reaction of an appropriate amine and iso(thio)cyanate.3,14,15 While simple, this stands in contrast to the wide array of readily available H-bond accepting base cocatalysts and adds a synthetic step prior to conducting polymerization chemistry. Certainly, the ready availability of chemical reagents and catalysts facilitates the wide implementation of chemical transformations. In this context, the antibacterial compound, triclocarban (TCC, Figure 5.1), recently banned as a hand soap additive by the FDA, captured our attention.16 It is an electron-deficient biaryl urea, similar to the slate of urea and multiurea H-bond donating catalysts that we recently showed to be highly active for ROP.13 While TCC has attracted considerable scientific interest as an antibacterial compound, possible bioaccumulate, and possible environmental toxin, we believe that this readily available compound has not 184 previously been employed as a catalyst.17−19 The efficacy of TCC/amidine base combinations for the ROP of lactone monomers was evaluated, Table 5.1. All reactions were conducted in C6D6 and conversion monitored by 1H NMR. The guanidine base, MTBD, exhibited faster rates than the imine base, DBU, and it was used for further experimentation. The ROP of δ-valerolactone (VL) from benzyl alcohol is highly controlled, exhibiting the characteristics of a living polymerization: linear evolution of Mn vs conversion, first order consumption of monomer and Mn predictable from [M]0 /[I]0, (Figures 5.2 and 4.3). This behavior is typical among organocatalysts for ROP. Initiation of a VL (1.0 mmol) ROP catalyzed by TCC/MTBD (0.05 mmol each) from 1- pyrenebutanol (0.02 mmol) and subsequent addition of a second monomer portion (1.0 mmol) exhibits overlapping UV and refractive index traces in the gel permeation chromatogram (GPC) of the resulting polymer (Figure 5.4), suggesting end group fidelity and a chain end that is susceptible to chain-extension. The TCC/MTBD (5 mol %) cocatalysts are also effective for the ROP of ε-caprolactone (CL), producing a similarly well-behaved ROP. The ROP rates exhibited by TCC/MTBD represent a significant advance over those exhibited by 1-S/MTBD, yet the reactions remain highly controlled. By comparison, for [M]0/[I]0 = 50 from benzyl alcohol in C6D6, the 1-S/MTBD-catalyzed ROPs of VL and CL achieve full conversion in 110 min and 45 h, respectively (c.f. Table 5.1).13 Entry 2 (Table 5.1) was attempted on a 200 mg scale, producing nearly identical polyvalerolactone (24 min, 90% conv, Mn = 18100, Mw/Mn = 1.04), which suggests that scale-up is feasible. RESULTS AND DISCUSSION We have embarked on a research program aimed at mitigating the low activity of H-bond mediated transformations without sacrificing the precise control typical of these catalysts. 185 In this vein, electron deficient aryl ureas have proved to be particularly efficacious; our lab previously disclosed the rapid rates exhibited by mono-, bis-, and tris-urea H-bond donors for the ROP of lactones.13 In general, urea H-bond donors are more active for ROP than their corresponding thioureas. This trend extends to the urea anions which, besides being remarkably active and controlled catalysts for ROP, are much more active than the corresponding thiourea anions.10,12 The uncharged H-bond donor 3-O, in combination with MTBD (0.017 mmol each), effects the ROP of VL (1.0 mmol) from benzyl alcohol (0.02 mmol) in C6D6 in 3 min.13 While the analogous reaction with TCC/MTBD (0.05 mmol each) achieves full conversion in a slower 14 min, the commercial availability of the TCC catalyst is expected to be a boon to the wider application of this and similar systems. Additionally, the TCC/MTBD cocatalysts exhibit high selectivity for monomer (vs polymer). When a fully converted PVL reaction solution remains unquenched, the Mn and Mw/Mn are minimally altered over an hour: 20 min, Mn = 22300, Mw/Mn = 1.02; 60 min, Mn = 23900, Mw/Mn = 1.03 (c.f. Table 5.1, entry 2), which may constitute an advantage versus other highly active systems for ROP.10,12-13 In this vein, electron deficient aryl ureas have proved to be particularly efficacious; our lab previously disclosed the rapid rates exhibited by mono-, bis-, and tris-urea H-bond donors for the ROP of lactones.13 In general, urea H-bond donors are more active for ROP than their corresponding thioureas. This trend extends to the urea anions which, besides being remarkably active and controlled catalysts for ROP, are much more active than the corresponding thiourea anions.10,12 The uncharged H-bond donor 3-O, in combination with MTBD (0.017 mmol each), effects the ROP of VL (1.0 mmol) from benzyl alcohol (0.02 mmol) in C6D6 in 3 min.13 While the analogous reaction with TCC/MTBD (0.05 mmol each) achieves full conversion in a slower 14 min, the commercial availability of the TCC catalyst is expected to be a boon to the wider application of this and similar systems. Additionally, the TCC/MTBD cocatalysts exhibit high selectivity for monomer (vs polymer). RESULTS AND DISCUSSION When a fully converted PVL reaction solution remains unquenched, the Mn and Mw/Mn are minimally altered over an hour: 20 min, Mn = 22300, Mw/Mn = 1.02; 60 min, Mn = 23900, Mw/Mn = 1.03 (c.f. Table 5.1, entry 2), which may constitute an advantage versus other highly active systems for ROP.10,12-13 Urea H-bond donors remain active in polar, H-bond accepting solvent. A long-standing limitation of H-bond mediated catalysis is the often narrow window of nonpolar solvents in which these catalysts are operable.20,21 We had previously observed that the urea H- bond donor 3-O remains active in THF and hypothesized that TCC would exhibit similar behavior, and a solvent screen was conducted for the TCC/MTBD catalyzed ROP of VL (Table 5.4). In DMF, the reaction time is extremely attenuated, and the reaction does not achieve >83% conversion. In THF, the ROP remains highly active (90% conv in 30 min), but Mw/Mn (=1.23) broadens. The result in acetone is surprising in that the reaction rate 186 does not slow versus C6D6, and the Mw/Mn remains narrow, Table 5.2. The ROP rates for all thiourea H-bond donors drop considerably versus their rates in C6D6,13 Table 5.2. The TCC/MTBD catalyzed ROP of VL in acetone-d6 remains controlled and exhibits the characteristics of a “living” polymerization (Figures 5.7 and 4.8). The polymer samples resulting from the initiation of a VL (1.0 mmol, 2 M) ROP from 1-pyrenebutanol (0.02 mmol) catalyzed by TCC/MTBD (0.05 mmol each) and subsequent chain extension show overlapping UV and RI traces in the GPC (Figure 5.9), which suggests end-group fidelity and that there is no initiation from the enol form of acetone-d6. When TCC and Me6TREN cocatalysts (5 mol % each) are applied for the ROP of L-lactide (1.0 mmol) from benzyl alcohol (0.01 mmol) in acetone-d6, the ROP reaction exhibits “living” behavior (Figures 5.12 and 5.13). In contrast to the ROP of VL, CL, or carbonate monomers, mild base cocatalysts are required for the ROP of lactide.21−24 The poly(lactide) was isolated and analyzed by selectively decoupled 1H NMR, revealing the polylactide (PLA) to be ∼ 90% isotactic (Figure 5.17), which suggests minor epimerization. The MALDITOF analysis of the same PLA sample shows the presence of ± 72 m/z repeat units, indicating that postpolymerization transesterification is occurring to a minor extent. RESULTS AND DISCUSSION This latter observation is in contrast to 2-S H-bond donating catalyst, which effects the ROP of LA in the virtual absence of postpolymerization transesterifiation.23 H-bond donating biaryl ureas were synthesized and applied in catalytic ROP to determine the origin of the enhanced rates of TCC (vs 1-O). These catalysts, here dubbed monoclocarban (mono-CC)25,26 and diclocarban (di-CC)26 in Figure 5.1, were applied to the ROP of VL in C6D6; we believe these molecules have not previously been used as catalysts. The TCC/MTBD (5 mol % each) cocatalyzed ROP of VL (1.0 mmol, 2 M) from 187 benzyl alcohol (0.01 mmol) in C6D6 reaches 91% conversion in 22 min (Table 5.1). The H-bond donors di-CC or mono-CC plus MTBD (5 mol % each) exhibit similar activity to TCC, but di-CC is the most active of the three H-bond donors (88% conversion in 15 min for di-CC and 37 min for mono-CC). The ROP of VL catalyzed by di-CC/MTBD exhibits the characteristics of a “living” polymerization (Figure 5.18). The similar rates exhibited by TCC and di-CC toward ROP may suggest that the additional chlorine atom in TCC (vs di-CC) is not essential for catalysis or that the additional electron withdrawing effects from the “extra” chlorine atom in TCC versus di-CC are inhibitory to catalysis. The latter possibility recalls similar effects that have been observed for extremely electron deficient thioureas,27,28 and these observations suggest that the augmented activity of the biaryl TCC (vs 1-O) can be approximated by functionalization at a single aryl ring. Certainly, the increased efficacy of TCC (vs 1-O) for ROP calls into question the primacy of the bis(trifluoromethyl)aryl group, at least for urea H-bond donors.27 While the commercial availability of TCC may be a boon to the application of H-bond mediated transformations in polymer synthesis laboratories, we expect that the development of advanced catalysts architectures will benefit from the more synthetically modular catalyst scaffold of di-CC. The enhanced efficacy of TCC and all urea H-bond donors in C6D6 could be attributed to the stronger binding of ureas vs thioureas to monomer.20 The limited solubility of TCC and n-O in nonpolar solvent in the absence of base cocatalyst limits the extent to which we can quantitatively probe this hypothesis by measuring binding constants to monomer. RESULTS AND DISCUSSION For The enhanced efficacy of TCC and all urea H-bond donors in C6D6 could be attributed to the stronger binding of ureas vs thioureas to monomer.20 The limited solubility of TCC and n-O in nonpolar solvent in the absence of base cocatalyst limits the extent to which we can quantitatively probe this hypothesis by measuring binding constants to monomer. For example, TCC is insoluble in benzene in the absence of H-bond acceptor, and binding constants for this compound could not be measured. However, the binding constants of 1- O and 1-S to CL were independently measured in C6D6 and are consistent with the long- 188 held hypothesis: for 1-O, Keq = 41 ± 1 (300 K) and for 1-S, Keq = 28 ± 1 (300 K).24 However, a binding constant rationale cannot be used to explain the ROP activity observed in acetone. As expected, when the 1-O/monomer binding study is repeated in acetone-d6, there is no observed change in chemical shift of 1-O up to ∼ 1000 equiv of monomer, which suggests very weak (Keq ∼ 1) or no binding in acetone-d6. While we have previously observed 1-S to exhibit a marked effect on a ROP reaction in the near absence of binding to monomer,29,30 these questions collectively reinforce a recently proposed mechanism.12 While this study was ongoing, “hyperactive” urea anions for ROP, generated by the action of alkoxides upon aryl and alkyl ureas, were disclosed; these systems are incredibly active yet controlled, exhibiting rates that rival traditional metal-based systems.12 The proposed mechanism of action whereby an active urea anion catalyst is generated by the deprotonation of a urea by alkoxide is distinct from traditional H-bond mediated ROP by neutral catalysts, and we sought to investigate the feasibility of this mechanism for TCC/imine bases. As opposed to the quantitative deprotonation of TCC by potassium methoxide, one could envisage an equilibrium established between urea plus base and the corresponding salt, eq 1. 1H NMR spectra in acetone-d6 of TCC and TCC plus MTBD or DBU (5 mM each species) show an upfield shift of the TCC resonances upon treatment with base that would be associated with the formation of an anionic character at the urea (Figure 5.21). RESULTS AND DISCUSSION Repeating this experiment with highly basic BEMP (Figure 5.1, BEMP-H+ pKaMeCN = 27.6)31 establishes a pattern of increased upfield shift with increasing pKa (MTBD-H+ pKaMeCN = 25.4; DBU-H+ pKaMeCN = 24.3).32 Repeating the TCC/BEMP 1H NMR experiment with a deficient amount of BEMP (2.5 mM) shows only one set of held hypothesis: for 1-O, Keq = 41 ± 1 (300 K) and for 1-S, Keq = 28 ± 1 (300 K).24 However, a binding constant rationale cannot be used to explain the ROP activity observed in acetone. As expected, when the 1-O/monomer binding study is repeated in acetone-d6, there is no observed change in chemical shift of 1-O up to ∼ 1000 equiv of monomer, which suggests very weak (Keq ∼ 1) or no binding in acetone-d6. While we have previously observed 1-S to exhibit a marked effect on a ROP reaction in the near absence of binding to monomer,29,30 these questions collectively reinforce a recently proposed mechanism.12 189 resonances for TCC, suggesting that the equilibrium in eq 1 is dynamic on the 1H NMR time scale. The 1H NMR experiments suggest that TCC/BEMP would be the most imidate-like species (i.e., eq 1 further to the right) and presumably the most active TCC/organic base catalyst pair yet examined herein. Indeed, the BEMP/TCC (0.05 mmol) catalyzed ROP of VL (1 M, 1 mmol) from benzyl alcohol (0.01 mmol) in benzene achieves full conversion in 3 min (Table 5.3). Higher reaction concentrations can be employed, but the reaction becomes difficult to monitor, fully converting within seconds at 2 M VL. The same ROP of VL fails to reach full conversion in THF or acetone-d6 within 30 min. In C6D6, the ROP is highly controlled and exhibits the characteristics of a “living” polymerization (Figure 5.10), and the [M]0/[I]0 series (Table 5.3) is notable for the high predictability of Mn even when considered against other organocatalytic systems. Further, Mw/Mn broadens slowly postpolymerization (Table 5.3, entry 2: 3 min, Mn = 22400, Mw/Mn = 1.04; 6 min, Mn = 24100, Mw/Mn = 1.07; 15 min, Mn = 24700, Mw/Mn = 1.15; 90% conv. for all aliquots). TCC/BEMP is ineffective for the ROP of β-butyrolactone, consistent with other urea and thiourea H-bond donors.13,20 We propose that the TCC/base cocatalyzed ROP of ester monomers proceeds through a mixed mechanism where the identity of the dominate catalyst largely depends on the pKa of the cocatalysts. RESULTS AND DISCUSSION The 1H NMR spectrum of TCC plus Me6TREN shows very slight downfield shift of the TCC resonances and broadening of the N−H resonances which could be attributed to H-bonding; there is no evidence to suggest the formation of imidate character at the urea for this cocatalyst pair (c.f. TCC/BEMP, Figure 5.21). Accordingly, we propose that TCC/base cocatalyzed ROP is capable of effecting ROP through a classic 190 dual H-bond mechanism mediated by neutral catalysts or an imidate mediated mechanism, the primary determination of which mode is dominate rests with the pKa of the base. In the case of TCC plus Me6TREN, we proposed a primarily neutral catalyst mechanism versus BEMP, which may proceed primarily through an imidate mechanism, Scheme 5.1. Certainly, the rate of the TCC/BEMP ROP recalls that of the alkoxide-generated urea anions.12 This mechanistic proposal is an extension of the recent work with “hyperactive” urea anion catalysts for ROP, taking into account weakly basic cocatalysts.12 For the present system, it is unclear if the conjugate acid of the base serves as a H-bond donor or primarily serves to deprotonate the urea. The complicated and sensitive interplay of cocatalyst/reagent interactions requires more study to be thoroughly understood. 191 CONCLUSION The antibacterial TCC has been shown to be a highly-effective cocatalyst for ring- opening polymerization. The commercially available H-bond donor, when applied with an H-bond accepting base cocatalyst, is among the most active organic catalysts for the ROP of esters, yet it exhibits the characteristics of a ‘living’ polymerization, producing well- defined polymers. The activity of this catalyst can be approximated by other mono- and di-chloro biaryl urea H-bond donor(s), which adds synthetic flexibility for the generation of future H-bond donating ureas. We suspect that the ROP of lactone monomers is just one application that can offer new roles to old reagents, in this case the antibacterial compound now banned in hand soap, TCC. 192 LIST OF REFERENCES (1) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R. M. Organocatalysis: Opportunities and Challenges for Polymer Synthesis. Macromolecules 2010, 43, 2093−2107. (2) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Organocatalytic Ring-Opening Polymerization. Chem. Rev. 2007, 107, 5813−5840. (3) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Lundberg, P. N. P.; Dove, A. P.; Li, H.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L. Exploration, Optimization, and Application of Supramolecular Thiourea−Amine Catalysts for the Synthesis of Lactide (Co)polymers. Macromolecules 2006, 39, 7863−7871. (4) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. K. Cooperative Hydrogen-Bond Pairing in Organocatalytic Ring-Opening Polymerization. Macromolecules 2014, 47, 7463−7468. (5) Csihony, S.; Culkin, D. A.; Sentman, A. C.; Dove, A. P.; Waymouth, R. M.; Hedrick, J. L. Single-Component Catalyst/initiators for the Organocatalytic Ring- Opening Polymerization of Lactide. J. Am. Chem. Soc. 2005, 127, 9079−9084. (6) McKinlay, C. J.; Waymouth, R. M.; Wender, P. A. Cell-Penetrating, Guanidinium- Rich Oligophosphoesters: Effective and Versatile Molecular Transporters for Drug and Probe Delivery. J. Am. Chem. Soc. 2016, 138, 3510−3517. (6) 193 (7) Blake, T. R.; Waymouth, R. M. Organocatalytic Ring-Opening Polymerization of Morpholinones: New Strategies to Functionalized Polyesters. J. Am. Chem. Soc. 2014, 136, 9252−9255. (7) (8) Miyake, G. M.; Chen, E. Y. X. Cinchona Alkaloids as Stereoselective Organocatalysts for the Partial Kinetic Resolution Polymerization of Rac-Lactide. Macromolecules 2011, 44, 4116−4124. (9) (a) Naumann, S.; Scholten, P. B. V; Wilson, J. A.; Dove, A. P. Dual Catalysis for Selective Ring-Opening Polymerization of Lactones: Evolution toward Simplicity. J. Am. Chem. Soc. 2015, 137, 14439−14445. (b) Xu, S.; Sun, H.; Liu, J.; Xu, J.; Pan, X.; Dong, H.; Liu, Y.; Li, Z.; Guo, K. Internal Lewis pair enhanced H-bond donor: boronate-urea and tertiary amine co-catalysis in ring-opening polymerization. Polym. Chem. 2016, 7, 6843. (10) Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Fast and Selective Ring- Opening Polymerizations by Alkoxides and Thioureas. Nat. Chem. 2016, 8, 1047−1053. (11) Guillaume, S. M.; Kirillov, E.; Sarazin, Y.; Carpentier, J. F. Beyond Stereoselectivity, Switchable Catalysis: Some of the Last Frontier Challenges in Ring-Opening Polymerization of Cyclic Esters. Chem. - Eur. J. 2015, 21, 7988−8003. (12) Lin, B.; Waymouth, R. M. Urea Anions: Simple, Fast, and Selective Catalysts for Ring-Opening Polymerizations. J. Am. Chem. Soc. 2017, 139, 1645−1652. (13) Fastnacht, K. LIST OF REFERENCES V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; Kiesewetter, E. T.; Kiesewetter, M. K. Bis- and Tris-Urea H-Bond Donors for 194 Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst. ACS Macro Lett. 2016, 5, 982−986. Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst. ACS Macro Lett. 2016, 5, 982−986. (14) Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Click Chemistry: Diverse Chemical Function from a Few Good Reactions. Angew. Chem., Int. Ed. 2001, 40, 2004−2021. (15) Dove, A. P.; Pratt, R. C.; Lohmeijer, B. G. G.; Waymouth, R. M.; Hedrick, J. L. Thiourea-Based Bifunctional Organocatalysis: Supramolecular Recognition for Living Polymerization. J. Am. Chem. Soc. 2005, 127, 13798−13799. (16) Erickson, B. E. FDA Bans Antibacterials in Consumers Soaps. Chem. Eng. News 2016, 94, 16. (17) DeLeo, P. C.; Sedlak, R. I. Comment on “On the Need and Speed of Regulating Triclosan and Triclocarban in the United States. Environ. Sci. Technol. 2014, 48, 11021−11022. (18) Pycke, B. F. G.; Roll, I. B.; Brownawell, B. J.; Kinney, C. A.; Furlong, E. T.; Kolpin, D. W.; Halden, R. U. Transformation Products and Human Metabolites of Triclocarban and Triclosan in Sewage Sludge across the United States. Environ. Sci. Technol. 2014, 48, 7881−7890. (19) Pycke, B. F. G.; Geer, L. A.; Dalloul, M.; Abulafia, O.; Jenck, A. M.; Halden, R. U. Human Fetal Exposure to Triclosan and Triclocarban in an Urban Population from Brooklyn, New York. Environ. Sci. Technol. 2014, 48, 8831−8838. (20) Doyle, A. G.; Jacobsen, E. N. Small-Molecule H-Bond Donors in Asymmetric Catalysis. Chem. Rev. 2007, 107, 5713−5743. 195 (21) Kazakov, O. I.; Kiesewetter, M. K. Cocatalyst Binding Effects in Organocatalytic Ring-Opening Polymerization of L-Lactide. Macromolecules 2015, 48, 6121−6126. (22) Coady, D. J.; Engler, A. C.; Horn, H. W.; Bajjuri, K. M.; Fukushima, K.; Jones, G. O.; Nelson, A.; Rice, J. E.; Hedrick, J. L. Catalyst Chelation Effects in Organocatalyzed Ring-Opening Polymerization of Lactide. ACS Macro Lett. 2012, 1, 19−22. (23) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. T.; Kiesewetter, M. K. Rate Accelerated Organocatalytic Ring-Opening Polymerization of L-Lactide via the Application of a Bis(thiourea) H-Bond Donating Cocatalyst. Macromolecules 2015, 48, 6127−6131. (24) Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.; Waymouth, R. M.; Hedrick, J. L. Guanidine and Amidine Organocatalysts for Ring-Opening Polymerization of Cyclic Esters. LIST OF REFERENCES Macromolecules 2006, 39, 8574−8583. (25) Kulkarni, A. R.; Garai, S.; Thakur, G. A. Scalable, One-Pot, Microwave- Accelerated Tandem Synthesis of Unsymmetrical Urea Derivatives. J. Org. Chem. 2017, 82, 992. (26) Beaver, D. J.; Roman, D. P.; Stoffel, P. J. The Preparation and Bacteriostatic Activity of Substituted Ureas. J. Am. Chem. Soc. 1957, 79, 1236−1245. (27) Lippert, K. M.; Hof, K.; Gerbig, D.; Ley, D.; Hausmann, H.; Guenther, S.; Schreiner, P. R. Hydrogen-Bonding Thiourea Organocatalysts: The Privileged 3,5- Bis(trifluoromethyl)phenyl Group. Eur. J. Org. Chem. 2012, 30, 5919−5927. 196 (28) Zhang, Z.; Schreiner, P. R. (Thio)urea Organocatalysis–What Can Be Learnt from Anion Recognition? Chem. Soc. Rev. 2009, 38, 1187−1198. (29) Datta, P. P.; Kiesewetter, M. K. Controlled Organocatalytic Ring-Opening Polymerization of ε-Thionocaprolactone. Macromolecules 2016, 49, 774−780. (30) Bannin, T. J.; Kiesewetter, M. K. Poly(thioester) by Organocatalytic Ring-Opening Polymerization. Macromolecules 2015, 48 (16), 5481−5486. (31) Schwesinger, R.; Schlempep, H.; Hasenfratz, C.; Willaredt, J.; Dambacher, T.; Breuer, T.; Ottaway, C.; Fletschinger, M.; Boele, J.; Fritz, H.; Putzas, D.; Rotter, H. W.; Bordwell, F. G.; Satish, A. V.; Ji, G.; Peters, E.; Peters, K.; Schnering, H. G. V.; Walz, L. Extremely Strong, Uncharged Auxiliary Bases; Monomeric and Polymer-Supported Polyaminophosphazenes (P2-P5). Liebigs Ann. 1996, 1996, 1055−1081. . W.; Bordwell, F. G.; Satish, A. V.; Ji, G.; Peters, E.; Peters, K.; Schnering, H. (32) Kaljurand, I.; Kütt, A.; Sooväli, L.; Rodima, T.; Mäemets, V.; Leito, I.; Koppel, I. A. Extension of the Self-Consistent Spectrophotometric Basicity Scale in Acetonitrile to a Full Span of 28 pKa Units: Unification of Different Basicity Scales. J. Org. Chem. 2005, 70, 1019−1028. (33) Webb, J. E. A.; Crossley, M. J.; Turner, P.; Thordarson, P. Pyromellitamide Aggregates and Their Response to Anion Stimuli. J. Am. Chem. Soc. 2007, 129, 7155. (34) Thordarson, P. Determining association constants from titration experiments in supramolecular chemistry. Chem. Soc. Rev. 2011, 40, 1305. (35) Deranleau, D. A. Theory of the measurement of weak molecular complexes. II. Consequences of multiple equilibria. J. Am. Chem. Soc. 1969, 91, 4044. 197 (36) Horman, I.; Dreux, B. Estimation of association constants of bimolecular complexes. Reply to comments. Anal. Chem. 1983, 55, 1219. (37) Peters, S. J.; Stevenson, C. D. The Complexation of Na+ by 18-crown -6 studied via nuclear magnetic resonance. J. Chem. Educ. 2004, 81, 715. (38) Lee, H-G.; Kim, M-J.; Park, S-E.; Kin, J-J.; Kim, B. R.; Lee, S-G.; Yoon, Y-J. LIST OF REFERENCES Phenyl 4,5-dichloro-6-oxopyridazine-1(6H)-carboxylate as carbonyl source: facile and selective synthesis of carbamates and ureas under mild conditions. Synlett 2009, 17, 2809. (39) Khan, K. M.; Saeed, S.; Ali, M.; Gohar, M.; Zahid, J.; Khan, A.; Perveen, S.; Choudhary, M. I. Unsymmetrically disubstituted urea derivatives: A potent class of antiglycating agents. Bioorg. Med. Chem. 2009, 17, 2447. (39) 198 entry mon. [M]o/[I]o time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 1d VL 100 81 90 18 900 1.06 2 100 22 91 19 900 1.05 3 50 14 90 8 500 1.08 4 200 46 90 35 900 1.09 5 500 125 90 72 900 1.02 6 CL 100 132 90 21 200 1.06 Table 5.1. MTBD and TCC Catalyzed ROP of VL and CL. Reaction conditions: VL or CL (1.0 mmol, 1 equiv, 2M), benzyl alcohol, C6D6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. d) DBU (5 mol%, 0.05 mmol) cocatalyst was employed (no MTBD). Table 5.1. MTBD and TCC Catalyzed ROP of VL and CL. Reaction conditions: VL or CL (1.0 mmol, 1 equiv, 2M), benzyl alcohol, C6D6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. d) DBU (5 mol%, 0.05 mmol) cocatalyst was employed (no MTBD). 199 Table 5.2. Urea or Thiourea Plus MTBD Cocatalyzed ROP of VL in Acetone. Reaction conditions: VL (1.0 mmol, 1 equiv, 2M), benzyl alcohol (2 mol%), and MTBD (same mol% as U/TU), acetone-d6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. Table 5.2. Urea or Thiourea Plus MTBD Cocatalyzed ROP of VL in Acetone. Reaction conditions: VL (1.0 mmol, 1 equiv, 2M), benzyl alcohol (2 mol%), and MTBD (same mol% as U/TU), acetone-d6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. Table 5.3. Triclocarban Plus BEMP Cocatalyzed ROP of VL and CL. Reaction conditions: VL or CL (1.0 mmol, 1 equiv, 1M), benzyl alcohol, C6D6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. d) CL (1.0 mmol, 1 equiv, 2M). LIST OF REFERENCES entry TU or U (mol%) time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 1 TCC (5%) 13 89 10 000 1.09 2 1-S (5%) 1200 89 9 500 1.21 3 1-O (5%) 60 91 11 900 1.08 4 2-S (2.5%) 1020 90 11 400 1.28 5 2-O (2.5%) 20 90 10 800 1.15 6 3-S (1.7%) 7440 89 12 100 1.16 7 3-O (1.7%) 20 89 10 300 1.13 entry TU or U (mol%) time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 1 TCC (5%) 13 89 10 000 1.09 2 1-S (5%) 1200 89 9 500 1.21 3 1-O (5%) 60 91 11 900 1.08 4 2-S (2.5%) 1020 90 11 400 1.28 5 2-O (2.5%) 20 90 10 800 1.15 6 3-S (1.7%) 7440 89 12 100 1.16 7 3-O (1.7%) 20 89 10 300 1.13 Table 5.2. Urea or Thiourea Plus MTBD Cocatalyzed ROP of VL in Acetone. Reaction conditions: VL (1.0 mmol, 1 equiv, 2M), benzyl alcohol (2 mol%), and MTBD (same mol% as U/TU), acetone-d6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. Table 5.2. Urea or Thiourea Plus MTBD Cocatalyzed ROP of VL in Acetone. Reaction conditions: VL (1.0 mmol, 1 equiv, 2M), benzyl alcohol (2 mol%), and MTBD (same mol% as U/TU), acetone-d6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. 200 entry mon. [M]o/[I]o time (min) conv.a (%) Mnb (g/mol) Mw/Mnb 1 VL 50 1 87 11 900 1.04 2 100 3 90 22 400 1.04 3 200 6 90 47 900 1.06 4 500 10 90 108 800 1.05 5d CL 100 6 90 16 500 1.04 Table 5.3. Triclocarban Plus BEMP Cocatalyzed ROP of VL and CL. Reaction conditions: VL or CL (1.0 mmol, 1 equiv, 1M), benzyl alcohol, C6D6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. d) CL (1.0 mmol, 1 equiv, 2M). Table 5.3. Triclocarban Plus BEMP Cocatalyzed ROP of VL and CL. Reaction conditions: VL or CL (1.0 mmol, 1 equiv, 1M), benzyl alcohol, C6D6. a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. d) CL (1.0 mmol, 1 equiv, 2M). 201 Entry urea [M]o/[I]o time (min) conv. LIST OF REFERENCES (%)a Mnb Mw/Mnb 1 TCC 50 14 90 8500 1.08 2 100 22 91 19900 1.05 3 200 46 90 35900 1.0 4 500 125 90 72900 1.02 5 di-CC 50 15 88 6000 1.04 6 100 20 89 12000 1.04 7 200 78 94 25000 1.03 8 500 180 89 64000 1.06 Entry urea [M]o/[I]o time (min) conv. (%)a Mnb Mw/Mnb 1 TCC 50 14 90 8500 1.08 2 100 22 91 19900 1.05 3 200 46 90 35900 1.0 4 500 125 90 72900 1.02 5 di-CC 50 15 88 6000 1.04 6 100 20 89 12000 1.04 7 200 78 94 25000 1.03 8 500 180 89 64000 1.06 le 5.4. Chain Length Variation for the TCC or di-CC plus MTBD cocatalyzed ROP a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. Entry Solvent time (min) conv. (%)a Mnb Mw/Mnb 1 benzene-d6 22 91 19,900 1.06 2 acetone-d6 22 89 19,400 1.11 3 chloroform-d 273 89 19,100 1.08 4 THF 30 89 14,700 1.23 5 DMF 600 83 9,000 1.41 Table 5.5. Solvent Screen of TCC/MTBD cocatalyzed ROP of VL. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. Entry Solvent time (min) conv. (%)a Mnb Mw/Mnb 1 benzene-d6 22 91 19,900 1.06 2 acetone-d6 22 89 19,400 1.11 3 chloroform-d 273 89 19,100 1.08 4 THF 30 89 14,700 1.23 5 DMF 600 83 9,000 1.41 Table 5.5. Solvent Screen of TCC/MTBD cocatalyzed ROP of VL. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. Entry Solvent time (min) conv. (%)a Mnb Mw/Mnb 1 benzene-d6 22 91 19,900 1.06 2 acetone-d6 22 89 19,400 1.11 3 chloroform-d 273 89 19,100 1.08 4 THF 30 89 14,700 1.23 5 DMF 600 83 9,000 1.41 Entry Solvent time (min) conv. (%)a Mnb Mw/Mnb 1 benzene-d6 22 91 19,900 1.06 2 acetone-d6 22 89 19,400 1.11 3 chloroform-d 273 89 19,100 1.08 4 THF 30 89 14,700 1.23 5 DMF 600 83 9,000 1.41 Table 5.5. Solvent Screen of TCC/MTBD cocatalyzed ROP of VL. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. Table 5.5. Solvent Screen of TCC/MTBD cocatalyzed ROP of VL. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. 202 Entry [M]o/[I]o time (min) conv. LIST OF REFERENCES (%)a Mnb Mw/Mnb 1 50 13 88 7400 1.11 2 100 20 88 14100 1.10 3 200 32 89 22600 1.09 4 500 45 89 44700 1.08 Table 5.6. Chain Length Variation for the TCC/MTBD cocatalyzed ROP of VL in acetone- d6. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. Table 5.6. Chain Length Variation for the TCC/MTBD cocatalyzed ROP of VL in acetone- d6. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. d6. a) Conversion determined by 1H NMR. b) Mn and Mw were obtained by GPC. Scheme 5.1. Proposed mechanism for TCC/base cocatalyzed ROP. Scheme 5.1. Proposed mechanism for TCC/base cocatalyzed ROP. 203 Figure 5.1. Base and (thio)urea cocatalysts evaluated for ROP. Figure 5.1. Base and (thio)urea cocatalysts evaluated for ROP. Figure 5.2. First order evolution of VL vs time for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene-d6. y = 0.1151x + 0.0498 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 2 4 6 8 10 12 14 16 18 ln[VL]o/[VL] time (min) y = 0.1151x + 0.0498 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 0 2 4 6 8 10 12 14 16 18 ln[VL]o/[VL] time (min) Figure 5.2. First order evolution of VL vs time for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene-d6. 204 Figure 5.3. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene- d6. 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion % 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion % Figure 5.3. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL. LIST OF REFERENCES Conditions: VL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene- d6. Figure 5.3. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene- d6. Figure 5.3. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/MTBD c d6. Figure 5.4. GPC traces of the polymers resulting from the chain extension experiment of VL. Conditions: VL (2 M, 1mmol), 1-pyrenebutanol (2mol%, 0.02mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene-d6, and subsequent chain extension by the addition of VL (1mmol). 18 19 20 21 22 23 Retention time (min) RI UV Chain extended RI Chain extended UV 18 19 20 21 22 23 Retention time (min) RI UV Chain extended RI Chain extended UV Retention time (min) Figure 5.4. GPC traces of the polymers resulting from the chain extension experiment of VL. Conditions: VL (2 M, 1mmol), 1-pyrenebutanol (2mol%, 0.02mmol), TCC (5mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene-d6, and subsequent chain extension by the addition of VL (1mmol). 205 Figure 5.5. First order evolution of CL vs time for the TCC/MTBD catalyzed ring-opening polymerization of CL. Conditions: CL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01mmol), TCC (5mol%, 0.04 mmol), MTBD (5 mol%, 0.04 mmol) in benzene-d6. y = 0.0089x + 0.0129 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0 50 100 150 200 250 300 350 400 ln[CL]o/[CL] time (min) y = 0.0089x + 0.0129 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0 50 100 150 200 250 300 350 400 ln[CL]o/[CL] time (min) Figure 5.5. First order evolution of CL vs time for the TCC/MTBD catalyzed ring-opening polymerization of CL. Conditions: CL (2 M, 1 mmol), benzyl alcohol (1mol%, 0.01mmol), TCC (5mol%, 0.04 mmol), MTBD (5 mol%, 0.04 mmol) in benzene-d6. Figure 5.6. Mn (blue) and Mw/Mn (orange) catalyzed ring-opening polymerization of CL. Conditions: CL (2 M, 1mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.04 mmol), MTBD (5 mol%, 0.04 mmol) in benzene-d6. LIST OF REFERENCES 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 0 5000 10000 15000 20000 25000 0 20 40 60 80 100 MW/Mn Mn, g/mol Conversion % 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 0 5000 10000 15000 20000 25000 0 20 40 60 80 100 MW/Mn Mn, g/mol Conversion % Figure 5.6. Mn (blue) and Mw/Mn (orange) catalyzed ring-opening polymerization of CL. Conditions: CL (2 M, 1mmol), benzyl alcohol (1mol%, 0.01 mmol), TCC (5mol%, 0.04 mmol), MTBD (5 mol%, 0.04 mmol) in benzene-d6. 206 Figure 5.7. Approach to equilibrium evolution of [VL] vs time for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2.1 M, 2mmol, 1 equiv.), benzyl alcohol (1mol%, 0.02 mmol), TCC (5mol%, 0.1 mmol), MTBD (5 mol%, 0.1 mmol) in acetone-d6. [VL]eq = 0.22 M 0 0.5 1 1.5 2 2.5 3 3.5 4 0 5 10 15 ln([VL]o-[VL]eq)/([VL]-[VL]eq) time (min) 0 0.5 1 1.5 2 2.5 3 3.5 4 0 5 10 15 ln([VL]o-[VL]eq)/([VL]-[VL]eq) time (min) Figure 5.7. Approach to equilibrium evolution of [VL] vs time for the TCC catalyzed ring-opening polymerization of VL. Conditions: VL (2.1 M, 2mmol, 1 equiv.), benzyl alcohol (1mol%, 0.02 mmol), TCC (5mol%, 0.1 mmol), MTBD (5 mol%, 0.1 mmol) in acetone-d6. [VL]eq = 0.22 M Figure 5.8. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2.1 M, 2 mmol, 1 equiv.), benzyl alcohol (1 mol. %, 0.02 mmol), TCC (5 mol. %, 0.1 mmol), MTBD (5 mol. %, 0.1 mmol) in acetone-d6. 1 1.2 1.4 1.6 1.8 2 0 3000 6000 9000 12000 15000 18000 0 20 40 60 80 100 Mw/Mn Mn, g/mol conversion % 1 1.2 1.4 1.6 1.8 2 0 3000 6000 9000 12000 15000 18000 0 20 40 60 80 100 Mw/Mn Mn, g/mol conversion % Figure 5.8. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/MTBD catalyzed ring-opening polymerization of VL. Conditions: VL (2.1 M, 2 mmol, 1 equiv.), benzyl alcohol (1 mol. %, 0.02 mmol), TCC (5 mol. %, 0.1 mmol), MTBD (5 mol. %, 0.1 mmol) in acetone-d6. 207 Figure 5.9. GPC traces of the polymers resulting from the chain extension of PVL in acetone. LIST OF REFERENCES Conditions: VL (2 M, 1 mmol), 1-pyrenebutanol (2 mol%, 0.02mmol), TCC (5 mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in acetone-d6, and subsequent chain extension by the addition of VL (1 mmol). 18 19 20 21 22 23 retention time (min) RI UV Chain extended RI Chain extended UV 18 19 20 21 22 23 retention time (min) RI UV Chain extended RI Chain extended UV Figure 5.9. GPC traces of the polymers resulting from the chain extension of PVL in acetone. Conditions: VL (2 M, 1 mmol), 1-pyrenebutanol (2 mol%, 0.02mmol), TCC (5 mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in acetone-d6, and subsequent chain extension by the addition of VL (1 mmol). 208 Figure 5.10. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/BEMP catalyzed ring-opening polymerization of VL. Conditions: VL (1 M, 1 mmol), benzyl alcohol (1 mol%, 0.01 mmol), TCC (5 mol%, 0.05 mmol), BEMP (5 mol%, 0.05 mmol) in benzene- d6. 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 MW/Mn Mn, g/mol Conversion % 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 MW/Mn Mn, g/mol Conversion % Figure 5.10. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/BEMP catalyzed ring-opening polymerization of VL. Conditions: VL (1 M, 1 mmol), benzyl alcohol (1 mol%, 0.01 mmol), TCC (5 mol%, 0.05 mmol), BEMP (5 mol%, 0.05 mmol) in benzene- d6 d6. Figure 5.11. GPC traces of the polymers resulting from the chain extension experiment of VL. Conditions: VL (1 M, 1 mmol), 1-pyrenebutanol (2 mol%, 0.02mmol), TCC (5 mol%, 0.05 mmol), BEMP (5 mol%, 0.05 mmol) in benzene-d6, and subsequent chain extension by the addition of VL (1 mmol). 17.5 18 18.5 19 19.5 20 20.5 21 21.5 Retention time (min) RI UV Chain extended RI Chain extended UV 17.5 18 18.5 19 19.5 20 20.5 21 21.5 Retention time (min) RI UV Chain extended RI Chain extended UV Figure 5.11. GPC traces of the polymers resulting from the chain extension experiment of VL. Conditions: VL (1 M, 1 mmol), 1-pyrenebutanol (2 mol%, 0.02mmol), TCC (5 mol%, 0.05 mmol), BEMP (5 mol%, 0.05 mmol) in benzene-d6, and subsequent chain extension by the addition of VL (1 mmol). Figure 5.11. LIST OF REFERENCES GPC traces of the polymers resulting from the chain extension experiment of VL. Conditions: VL (1 M, 1 mmol), 1-pyrenebutanol (2 mol%, 0.02mmol), TCC (5 mol%, 0.05 mmol), BEMP (5 mol%, 0.05 mmol) in benzene-d6, and subsequent chain extension by the addition of VL (1 mmol). 209 Figure 5.12. First order evolution of [L-LA] vs time for the TCC/Me6TREN catalyzed ring- opening polymerization. Conditions: L-LA (1 M, 1 mmol), benzyl alcohol (1 mol %, 0.01 mmol), TCC (5 mol %, 0.05 mmol), Me6TREN (5 mol %, 0.05 mmol) in acetone-d6. 0.0 0.5 1.0 1.5 2.0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 ln[L-Lac]o/[L-Lac] time (hr) 0.0 0.5 1.0 1.5 2.0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 ln[L-Lac]o/[L-Lac] time (hr) Figure 5.12. First order evolution of [L-LA] vs time for the TCC/Me6TREN catalyzed ring- opening polymerization. Conditions: L-LA (1 M, 1 mmol), benzyl alcohol (1 mol %, 0.01 mmol), TCC (5 mol %, 0.05 mmol), Me6TREN (5 mol %, 0.05 mmol) in acetone-d6. Figure 5.12. First order evolution of [L-LA] vs time for the TCC/Me6TREN catalyzed ring- opening polymerization. Conditions: L-LA (1 M, 1 mmol), benzyl alcohol (1 mol %, 0.01 mmol), TCC (5 mol %, 0.05 mmol), Me6TREN (5 mol %, 0.05 mmol) in acetone-d6. 210 Figure 5.13. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/Me6TREN catalyzed ring-opening polymerization of L-LA. Conditions: L-LA (1 M, 1 mmol), benzyl alcohol (1 mol. %, 0.01 mmol), TCC (5 mol. %, 0.05 mmol), Me6TREN (5 mol. %, 0.05 mmol) in acetone-d6. 1 1.2 1.4 1.6 1.8 2 0 5000 10000 15000 20000 25000 30000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion, % 1 1.2 1.4 1.6 1.8 2 0 5000 10000 15000 20000 25000 30000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion, % Figure 5.13. Mn (blue) and Mw/Mn (orange) vs conversion for the TCC/Me6TREN catalyzed ring-opening polymerization of L-LA. Conditions: L-LA (1 M, 1 mmol), benzyl alcohol (1 mol. %, 0.01 mmol), TCC (5 mol. %, 0.05 mmol), Me6TREN (5 mol. %, 0.05 mmol) in acetone-d6. 211 Figure 5.14. MALDI-TOF of the PLLA resulting from TCC/Me6TREN cocatalyzed ROP of L-lactide. LIST OF REFERENCES The major pattern (blue line) is due to whole repeat units m/z = (Na+ + benzyl alcohol + nLA) while the minor pattern (red line) is due to half repeat units generated by post-polymerization transesterification m/z = (Na+ + benzyl alcohol + (n+1/2)LA). All m/z bear a benzyl alcohol initiator. 4884.452 5028.501 4740.410 5316.590 4596.361 5460.631 5604.667 4452.311 5748.697 4308.266 5892.733 4164.214 6036.765 6180.796 4020.161 6324.815 6468.813 3876.084 6756.914 7044.875 0 1000 2000 3000 Intens. [a.u.] 4000 4500 5000 5500 6000 6500 7000 m/z Figure 5.14. MALDI-TOF of the PLLA resulting from TCC/Me6TREN cocataly Figure 5.14. MALDI-TOF of the PLLA resulting from TCC/Me6TREN cocatalyzed ROP of L-lactide. The major pattern (blue line) is due to whole repeat units m/z = (Na+ + benzyl alcohol + nLA) while the minor pattern (red line) is due to half repeat units generated by post-polymerization transesterification m/z = (Na+ + benzyl alcohol + (n+1/2)LA). All m/z bear a benzyl alcohol initiator. 212 Figure 5.15. Titration binding curve for the CL/1-O binding in benzene-d6. Chemical shift 8.160 8.180 8.200 8.220 8.240 8.260 8.280 8.300 8.320 8.340 8.360 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 𝛅, ppm [CL], M 8.160 8.180 8.200 8.220 8.240 8.260 8.280 8.300 8.320 8.340 8.360 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 𝛅, ppm [CL], M Figure 5.15. Titration binding curve for the CL/1-O binding in benzene-d6. Chemical shift Figure 5.15. Titration binding curve for the CL/1-O binding in benzene-d6. Chemical shift of the o-phenyl protons vs [CL]; solid line is the fit from the binding equation. Figure 5.16. Titration binding curve for the CL/1-S binding in benzene-d6. Chemical shift of the o-phenyl protons vs [CL]; solid line is the fit from the binding equation. 7 7.2 7.4 7.6 7.8 8 8.2 8.4 8.6 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 𝛅, ppm [CL], M 7 7.2 7.4 7.6 7.8 8 8.2 8.4 8.6 0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350 0.400 0.450 𝛅, ppm [CL], M Figure 5.16. Titration binding curve for the CL/1-S binding in benzene-d6. Chemical shift of the o-phenyl protons vs [CL]; solid line is the fit from the binding equation. Figure 5.16. Titration binding curve for the CL/1-S binding in benzene-d6. Chemical shift of the o-phenyl protons vs [CL]; solid line is the fit from the binding equation. Figure 5.16. Titration binding curve for the CL/1-S binding in benzene-d6. LIST OF REFERENCES Chemical shift of the o-phenyl protons vs [CL]; solid line is the fit from the binding equation. 213 Figure 5.17. Methine region of the methyl-decoupled 1H NMR spectrum of PLLA obtained via TCC/Me6TREN cocatalyzed ROP of L-LA (500 MHz, 25 °C). 17. Methine region of the methyl-decoupled 1H NMR spectrum of PLLA obtained Figure 5.17. Methine region of the methyl-decoupled 1H NMR spectrum of PLLA obtained via TCC/Me6TREN cocatalyzed ROP of L-LA (500 MHz, 25 °C). Figure 5.17. Methine region of the methyl-decoupled 1H NMR spectrum of PLLA obtained via TCC/Me6TREN cocatalyzed ROP of L-LA (500 MHz, 25 °C). g g y p p via TCC/Me6TREN cocatalyzed ROP of L-LA (500 MHz, 25 °C). via TCC/Me6TREN cocatalyzed ROP of L-LA (500 MHz, 25 °C). via TCC/Me6TREN cocatalyzed ROP of L-LA (500 MHz, 25 °C). 214 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion % Figure 5.18. (upper) First order evolution of [VL] vs time for the di-CC/MTBD catalyzed ROP of VL. (lower) The ROP displays a linear evolution of Mn (blue) vs conversion and narrow Mw/Mn (orange). Conditions: VL (2 M, 1.0 mmol), benzyl alcohol (2.0 mol%, 0 02mmol) di CC (5 0 mol% 0 05 mmol) MTBD (5 mol% 0 05 mmol) in benzene d 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion % 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 0 20 40 60 80 100 Mw/Mn Mn, g/mol Conversion % Figure 5.18. (upper) First order evolution of [VL] vs time for the di-CC/MTBD catalyzed ROP of VL. (lower) The ROP displays a linear evolution of Mn (blue) vs conversion and narrow Mw/Mn (orange). Conditions: VL (2 M, 1.0 mmol), benzyl alcohol (2.0 mol%, 0.02mmol), di-CC (5.0 mol%, 0.05 mmol), MTBD (5 mol%, 0.05 mmol) in benzene-d6. 215 Figure 5.19. (upper) 1H NMR (CD3OD, 400 MHz) spectrum of mono-CC. (lower) 13C NMR (CD3OD, 100 MHz) spectrum of mono-CC. Figure 5.19. (upper) 1H NMR (CD3OD, 400 MHz) spectrum of mono-CC. (lower) 13C NMR (CD3OD, 100 MHz) spectrum of mono-CC. Figure 5.19. LIST OF REFERENCES (upper) 1H NMR (CD3OD, 400 MHz) spectrum of mono-CC. (lower) 13C NMR (CD3OD, 100 MHz) spectrum of mono-CC. 216 Figure 5.20. (upper) 1H NMR (CD3OD, 400 MHz) spectrum of di-CC. (lower) 13C NM (CD3OD, 100 MHz) spectrum of di-CC. Figure 5.20. (upper) 1H NMR (CD3OD, 400 MHz) spectrum of di-CC. (lower) 13C NMR (CD3OD, 100 MHz) spectrum of di-CC. 217 Figure 5.21. Downfield portion of the 1H NMR spectra of TCC plus base ([TCC] = [base] = 5 mM) in acetone-d6. Figure 5.21. Downfield portion of the 1H NMR spectra of TCC plus base ([TCC] = [base] Figure 5.21. Downfield portion of the 1H NMR spectra of TCC plus base ([TCC] = [base] = 5 mM) in acetone-d6. = 5 mM) in acetone-d6. 218 MANUSCRIPT – VI Published in Macromolecules Published in Macromolecules H-bonding Organocatalysts for the Living, Solvent-free Ring-Opening H-bonding Organocatalysts for the Living, Solvent-free Ring-Opening Polymerization of Lactones: Towards an All-Lactones, All-Conditions Approach erization of Lactones: Towards an All-Lactones, All-Conditions Approach Jinal U. Pothupitiya, Nayanthara U. Dharmaratne, Terra Marie M. Jouaneh, Kurt V. Fastnacht, Danielle N. Coderre and Matthew K. Kiesewetter Chemistry, University of Rhode Island, Kingston, RI, USA ABSTRACT The developing urea class of H-bond donors facilitates the solvent-free ROP of lactones at ambient and elevated temperatures, displaying enhanced rates and control versus other known organocatalysts for ROP under solvent-free conditions. The ROPs retain the characteristics of living polymerizations despite solidifying prior to full conversion, and copolymers can be accessed in a variety of architectures. One-pot block copolymerizations of lactide and valerolactone, which had previously been inaccessible in solution phase organocatalytic ROP, can be achieved under these reaction conditions, and one-pot triblock copolymers are also synthesized. For the ROP of lactide, however, thioureas remain the more effective H-bond donating class. For all (thio)urea catalysts under solvent-free conditions and in solution, the more active catalysts are generally more controlled. A rationale for these observations is proposed. The triclocarban (TCC) plus base systems are particularly attractive in the context of solvent-free ROP due to their commercial availability which could facilitate the adoption of these catalysts. 220 EXPERIMENTAL SECTION General Considerations. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Ethylene Brassylate was purchased from Sigma Aldrich. Benzyl alcohol was distilled from calcium hydride under high vacuum. 1-pyrenebutanol was purchased from Sigma Aldrich. δ-valerolactone (VL) and ε-caprolactone (CL) were distilled from calcium hydride under high vacuum. L-Lactide (L-LA) was purchased from Acros Organics and recrystallized from dry toluene. 7-Methyl-1,5,7- triazabicyclo[4.4.0]dec-5-ene (MTBD), 1,8-Diazabicyclo(5.4.0)undec-7-ene (DBU) was purchased from Tokyo Chemical Industry Co. LTD. and 2-tert-butylimino-2-diethylamino- 1,3-dimethylperhydro-1,3,2-diazaphosphorine (BEMP) from Acros Organics. Benzene-d6 and chloroform-d were purchased from Cambridge Isotope Laboratories and distilled from calcium hydride. Experiments were conducted using pre-dried glassware in an MBRAUN or INERT stainless steel glovebox under N2 atmosphere. NMR experiments were conducted on a Bruker Avance III 300 MHz or 400 MHz spectrometer, and 1H decoupled spectra were acquired on a Varian 500 MHz spectrometer. Gel Permeation Chromatography (GPC) was performed at 40 °C using HPLC grade dichloromethane eluent on an Agilent Infinity GPC system equipped with three Agilent PLGel columns 7.5 mm × 300 mm (5 μm, pore sizes: 103, 104, 50 Å). Mn and Mw/Mn were determined versus polystyrene standards (500 g/mol-3150 kg/mol, Polymer Laboratories). DSC experiments were conducted using a Shimadzu DSC-60A instrument, calibrated with an indium standard using aluminum pans under inert conditions. General Considerations. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Ethylene Brassylate was purchased from Sigma Aldrich. Benzyl alcohol was distilled from calcium hydride under high vacuum. 1-pyrenebutanol was purchased from Sigma Aldrich. δ-valerolactone (VL) and ε-caprolactone (CL) were distilled from calcium hydride under high vacuum. L-Lactide (L-LA) was purchased from Acros Organics and recrystallized from dry toluene. 7-Methyl-1,5,7- triazabicyclo[4.4.0]dec-5-ene (MTBD), 1,8-Diazabicyclo(5.4.0)undec-7-ene (DBU) was purchased from Tokyo Chemical Industry Co. LTD. and 2-tert-butylimino-2-diethylamino- 1,3-dimethylperhydro-1,3,2-diazaphosphorine (BEMP) from Acros Organics. Benzene-d6 and chloroform-d were purchased from Cambridge Isotope Laboratories and distilled from calcium hydride. Experiments were conducted using pre-dried glassware in an MBRAUN or INERT stainless steel glovebox under N2 atmosphere. NMR experiments were conducted on a Bruker Avance III 300 MHz or 400 MHz spectrometer, and 1H decoupled spectra were acquired on a Varian 500 MHz spectrometer. Gel Permeation Chromatography (GPC) was performed at 40 °C using HPLC grade dichloromethane eluent on an Agilent Infinity GPC system equipped with three Agilent PLGel columns 7.5 mm × 300 mm (5 μm, pore sizes: 103, 104, 50 Å). INTRODUCTION Conducting ring-opening polymerization under solvent-free conditions is an appealing strategy from several perspectives. Such situations include industrial polymerizations,1 ‘green’ processes2 and other applications where use and disposal of solvent is of concern as well as the ROP of macrolactones and other monomers with high equilibrium monomer concentration [M]eq where neat conditions are suggested by reaction thermodynamics.3–5 The H-bonding class of organocatalysts – consisting of urea or thiourea plus base – stand out among the controlled methods for ROP in their precise control for polymerization over transesterification,6,7 but they have not been widely applied to solvent-free ROP. These catalysts have facilitated the construction of highly tailored polymers including highly functionalized monomer feeds.8–13 Thiourea/base systems are widely viewed as operating through an H-bond mediated pathway whereby thiourea H- bond activates monomer and base cocatalyst activates the initiating/propagating chain end, Scheme 6.1.12–14 Nascent urea/base systems are believed to effect their highly active ROP via an imidate mediated mechanism that is more analogous to that of the guanidine organocatalyst 1,5,7-triazabicyclodec-5-ene (TBD) than their heavy chalcogen counterparts, Scheme 6.1.15–17 Among the larger pantheon of organic catalysts, TBD, which can operate via an H-bond mediated mechanism,18–20 has become popular for solvent-free ROP of strained and macrolactones,20–22 and certainly, metal-containing23 and enzymatic catalysts24–26 are often used under solvent free conditions. The amidine organocatalyst, 1,8-diazabicycloundec-7-ene (DBU), has also been applied for the ROP of lactide in eutectic monomer blends.27 Regardless, (thio)urea systems have not been applied 221 in solvent free conditions because, ironically, highly polar lactone monomers are poor solvents for the H-bond ROP mediated by thioureas.13,28 However, several urea/base cocatalysts have been recently shown to be effective in polar solvent,16,17 which led us to speculate that these systems may remain active under solvent-free conditions. The triclocarban (TCC)17,29 plus base systems seemed particularly attractive in this context due to their commercial availability which could facilitate wider adoption by the polymer community. in solvent free conditions because, ironically, highly polar lactone monomers are poor solvents for the H-bond ROP mediated by thioureas.13,28 However, several urea/base cocatalysts have been recently shown to be effective in polar solvent,16,17 which led us to speculate that these systems may remain active under solvent-free conditions. The triclocarban (TCC)17,29 plus base systems seemed particularly attractive in this context due to their commercial availability which could facilitate wider adoption by the polymer community. 222 EXPERIMENTAL SECTION Mn and Mw/Mn were determined versus polystyrene standards (500 g/mol-3150 kg/mol, Polymer Laboratories). DSC experiments were conducted using a Shimadzu DSC-60A instrument, calibrated with an indium standard using aluminum pans under inert conditions. 223 Example solvent-free ROP. A 1 mL vial was charged with TCC (6.3 mg, 0.019 mmol), benzyl alcohol (0.86 mg, 0.008 mmol), VL (200 mg, 1.995 mmol), magnetic stir bar and stirred until homogeneous. A second vial was charged with MTBD (3.06 mg, 0.019 mmol) and VL (200 mg, 1.99 mmol) and agitated to mix. The contents of the second vial were transferred to the first vial using a Pasteur pipette, and the solution was stirred. Reaction progress was monitored by taking aliquots of the reaction mixture – either ~1.5 µL solution or a small amount of solid extracted via spatula – at different time intervals and quenched in a solution of benzoic acid in chloroform-d. Conversion was determined via 1H NMR. The polymer was isolated by precipitating with hexanes, and the volatiles were removed under high vacuum before characterization via GPC. Example solvent-free ROP with TBD. A 1 mL vial was charged with TBD (2.8 mg, 0.019 mmol), benzyl alcohol (0.86 mg, 0.008 mmol), VL (400 mg, 3.99 mmol), magnetic stir bar and stirred vigorously to mix. The TBD does not completely dissolve under these reaction conditions. Reaction progress was monitored by taking aliquots of the reaction mixture – either ~1.5 µL solution or a small amount of solid extracted via spatula – at different time intervals and quenched in a solution of benzoic acid in chloroform-d. Conversion was determined via 1H NMR. The polymer was isolated by precipitating with hexanes, and the volatiles were removed under high vacuum before characterization via GPC. Example solution ROP of LA. A 7 mL vial was charged with 2-S (10.7 mg, 0.0174 mmol), L-LA (100 mg, 0.694 mmol), benzyl alcohol (0.72 μL, 0.00694 mmol) and toluene (0.25M, toluene 2.77mL). PMDTA (3.63 μL, 0.0174 mmol) was loaded into a 2 mL vial equipped with septa cap. The vials were transferred from the glovebox to an oil bath adjusted to the specific temp. The contents of vial 1 were transferred into vial 2. Aliquots were taken via 224 syringe and quenched using a CH2Cl2 solution of benzoic acid and conversion was monitored by 1H NMR. Example one-pot copolymerization. EXPERIMENTAL SECTION A 7 mL vial was charged with TCC (6.3 mg, 0.019 mmol), benzyl alcohol (1.72 mg, 0.016 mmol), VL (400 mg, 3.99 mmol), magnetic stir bar and stirred until homogeneous. A second vial was charged with BEMP (5.48 mg, 0.019 mmol) and CL (455.8 mg, 3.99 mmol) and mixed well. The contents of the second vial was transferred to the first vial using a Pasteur pipette, and the mixture was left to stir. The contents of the second vial were transferred to the first vial using a Pasteur pipette, and the solution was stirred. Reaction progress was monitored by taking aliquots of the reaction mixture – either ~1.5 µL solution or a small amount of solid extracted via spatula – at different time intervals and quenched in a solution of benzoic acid in chloroform-d. Conversion was determined via 1H NMR. The polymer was isolated by precipitating with hexanes, and the volatiles were removed under high vacuum before characterization via GPC. GPC. Example ROP of L-LA. A 7 mL vial was charged with 2-S (17.1 mg, 0.028 mmol), benzyl alcohol (3 mg, 0.028 mmol), L-LA (400 mg, 2.77 mmol), stir bar, and the contents were heated to 100oC to melt the sample. A second vial was charged with PMDTA (4.80 mg, 0.028 mmol), and PMDTA was transferred via a 10 µL syringe to the first vial and the mixture was left to stir. The reaction was monitored by taking aliquots of the reaction mixture via spatula and quenching in a solution of benzoic acid in chloroform-d. Reaction progress was monitored via 1H NMR. The polymer was then isolated by precipitating with methanol, and the volatiles were removed under high vacuum before characterization via GPC. 225 Determination of percent isotacticity. The 13C and 1H decoupled NMR spectra of the isolated polymer were acquired on a Varian 500 MHz at 50oC. The samples for 13C and 1H NMR were prepared as 10 % w/v and 1% w/v solutions, respectively, in CDCl3. The 1H NMR spectrum of the polymer was obtained by selective decoupling by irradiating the methyl region, and tacticity was determined from the methine region according to published procedures, see manuscript for references. Determination of percent isotacticity. The 13C and 1H decoupled NMR spectra of the isolated polymer were acquired on a Varian 500 MHz at 50oC. The samples for 13C and 1H NMR were prepared as 10 % w/v and 1% w/v solutions, respectively, in CDCl3. The 1H NMR spectrum of the polymer was obtained by selective decoupling by irradiating the methyl region, and tacticity was determined from the methine region according to published procedures, see manuscript for references. Direct-from-monomer negative mold. A 7 ml polypropylene vial was charged with TCC (13.7 mg, 0.049 mmol), benzyl alcohol (15.6 mg, 0.049 mmol), VL (1000 mg, 10 mmol) and agitated to mixed. The top inside wall of a clean vial cap was charged with BEMP (13.7 mg, 0.049 mmol), which clings to the surface of the cap. The vial was capped with the BEMP containing cap and shaken vigorously to coat the inner wall of the vial during the course of the ROP (~30 sec). The reaction vessel was removed from the glove box. The PVL vial negative is easily freed from the reaction vial. 226 RESULTS AND DISCUSSION ROP of Strained Lactones. The solvent-free TCC/base cocatalyzed ROP of δ- valerolactone (VL) from benzyl alcohol at room temperature exhibits the characteristics of a living polymerization. Initial studies were performed on VL (solvent-free, 3.99 mmol) using benzyl alcohol initiator (0.2 mol %), TCC H-bond donor and one of three base cocatalysts (DBU, MTBD or BEMP in Table 6.1). The TCC/base (0.02 mmol each) catalyst systems displayed the same rate trends in neat VL as in solution17: BEMP > MTBD > DBU, Table 6.1. Polymerizations were conducted in a glovebox, and aliquots were withdrawn, quenched in a CDCl3 solution of benzoic acid and conversion monitored by 1H NMR (see Experimental Section). Despite solidification of the reaction mixture during the ROP (~60% conversion), the TCC/MTBD cocatalyzed ROP of VL from benzyl alcohol displays the characteristics of a living polymerization: linear evolution of Mn vs conversion, narrow Mw/Mn, first order evolution of [monomer] (Figure 6.1) and Mn predictable from [M]o/[I]o (Table 6.1). 1H NMR analysis of the polyvalerolactone (PVL) reaction mixture of each system (at [M]o/[I]o =100) confirms the consumption of benzyl alcohol initiator suggesting good initiator efficiency. When the same polymerization is initiated from pyrenebutanol, the resulting PVL displays overlapping UV/vis and RI traces in the GPC chromatogram (Figure 6.2). The TCC/BEMP system displays the same living behavior, but it was too active to effectively monitor at 0.02 mmol catalysts (4 mmol VL), although reduced catalyst loadings remain active (>0.004 mmol, the lowest evaluated) and controlled, see Table 6.5 and Figures 6.3 and 6.2. The TCC/BEMP and TCC/MTBD systems were also applied for the ROP of ε-caprolactone (CL). The reaction rates and ROP of Strained Lactones. The solvent-free TCC/base cocatalyzed ROP of δ- valerolactone (VL) from benzyl alcohol at room temperature exhibits the characteristics of a living polymerization. Initial studies were performed on VL (solvent-free, 3.99 mmol) using benzyl alcohol initiator (0.2 mol %), TCC H-bond donor and one of three base cocatalysts (DBU, MTBD or BEMP in Table 6.1). The TCC/base (0.02 mmol each) catalyst systems displayed the same rate trends in neat VL as in solution17: BEMP > MTBD > DBU, Table 6.1. Polymerizations were conducted in a glovebox, and aliquots were withdrawn, quenched in a CDCl3 solution of benzoic acid and conversion monitored by 1H NMR (see Experimental Section). RESULTS AND DISCUSSION Despite solidification of the reaction mixture during the ROP (~60% conversion), the TCC/MTBD cocatalyzed ROP of VL from benzyl alcohol displays the characteristics of a living polymerization: linear evolution of Mn vs conversion, narrow Mw/Mn, first order evolution of [monomer] (Figure 6.1) and Mn predictable from [M]o/[I]o (Table 6.1). 1H NMR analysis of the polyvalerolactone (PVL) reaction mixture of each system (at [M]o/[I]o =100) confirms the consumption of benzyl alcohol initiator suggesting good initiator efficiency. When the same polymerization is initiated from pyrenebutanol, the resulting PVL displays overlapping UV/vis and RI traces in the GPC chromatogram (Figure 6.2). The TCC/BEMP system displays the same living behavior, but it was too active to effectively monitor at 0.02 mmol catalysts (4 mmol VL), although reduced catalyst loadings remain active (>0.004 mmol, the lowest evaluated) and controlled, see Table 6.5 and Figures 6.3 and 6.2. The TCC/BEMP and TCC/MTBD systems were also applied for the ROP of ε-caprolactone (CL). The reaction rates and 227 molecular weight dispersities are attenuated versus the ROP of VL, but both reactions remain controlled and display living behavior (Figure 6.4 and Table 6.6). PVL and PCL samples can be freed of catalysts impurities by washing with methanol, but PVL samples containing residual catalysts showed no alteration of their materials properties up to 0.5 mol% catalysts loading (the highest loading tested). Both cocatalysts are required for efficient ROP. Solutions of TCC plus benzyl alcohol (0.1 mol % each) in VL (1 equiv) were stored at room temperature and periodically monitored for 60 days and showed no conversion. Solutions of BEMP or MTBD (0.5 mol%) in VL (1 equiv) were less stable towards conversion at room temperature, reaching 7% and 17% conversion, respectively, after 21 days. Both VL solutions of base were more stable when stored at -10°C, the BEMP solution showing zero conversion to polymer after 20 days, see Figure 6.5. We presume that the observed conversions are due to initiation from base.30 Despite being inert separately, the combined solutions can yield an ROP so rapid, that the combined solvent-free solutions (10 mmol VL, 0.049 mmol TCC/BEMP each, 0.020 mmol benzyl alcohol) can be used to make a negative mold of the reaction vessel in 30 sec (Figure 6.6) directly from monomer. Potential application can be envisaged. Among H-bond mediated catalysts for ROP, urea/base cocatalysts stand out for the activity and control they exhibit in solvent-free ROP conditions. RESULTS AND DISCUSSION Among other organic catalysts that have been applied for neat ROP,31–33 TBD is of particular interest in the context of the present studies.21,22 Some reports of TBD-mediated solvent-free ROP are conducted in the melt or describe ROP to amorphous polymers,21,22 which would be expected to produce the narrow molecular weight distributions typical of solution 228 processes. In our hands, TBD (Table 6.1, entry 7) exhibits similar rates as TCC/MTBD for solvent-free ROP of VL but with broadened Mw/Mn versus the ROP with the urea/base cocatalysts. However, TCC/BEMP produces the narrowest Mw/Mn and is the most active examined under these conditions. When used alone, strong bases have also been shown to effect the ROP of lactones. For example, BEMP has been applied to the room temperature, solvent-free ROP of VL without an H-bond donating cocatalyst.31 The ROP appeared to be living in nature but sluggish, reaching full conversion to poly(valerolactone) (PVL) in days and displaying a broadened Mw/Mn.31 We believe the ability to conduct rapid and highly controlled ROP of lactones like VL and CL under solvent-free and non-melt conditions constitutes an advantage of the TCC/base cocatalysts over other (organo)catalyst systems. The observation of highly active TCC/base cocatalyzed ROP under solvent-free conditions corroborates an imidate mechanism of action, Scheme 6.1. Several H-bond donors were evaluated for the solvent-free ROP of VL (Table 6.2), and these results suggest that urea H-bond donors are more effective than thiourea H-bond donors for the ROP of strained lactones. This observation is consistent with a cocatalyst binding argument for thiourea/base cocatalyzed ROP,28,34 as the cocatalyst interactions would be greatly attenuated in polar VL solvent whereas thiourea-VL binding should remain active in neat monomer. For the urea/base cocatalyzed ROP under an imidate mechanism, the efficacy of the catalysts would not be disturbed by the polar reaction environment. This is confirmed by NOESY NMR experiments of acetone-d6 solutions of TCC/MTBD or TCC/BEMP which show intermolecular contact, indicating the formation of the imidate (Figure 6.7). NOESY experiments on 1-S/MTBD and 1-S/BEMP show no intermolecular 229 communication in acetone-d6 or C6D6, which suggests that thioimidate formation is not a prominent mechanism of action for these catalyst pairs. RESULTS AND DISCUSSION These experiments are corroborated by previous studies performed by our group that show strong H-bonding (no proton transfer) between cocatalyst pairs of thiourea and base,28 but evidence of proton transfer is observed between urea and base.17 Previous studies have shown that 1:1 mole ratios of H-bond donor and base are optimal for ROP (in solution) no matter how many H- bond donating moieties are present in the donor molecule.28,35,36 ROP of Macrolactones. The TCC/BEMP cocatalyzed ROP of macrolactones, ethylene brassylate and pentadecalactone, under solvent-free conditions proceeds at 80°C. The TCC/BEMP (0.06 mmol) cocatalyzed ROP of EB (2.95 mmol) at room temperature and solvent-free is exceedingly slow (16 h, 60% conversion), but the same ROP at 80°C proceeds in hours to full conversion, Table 6.3. The ROP reactions in Table 6.3 display moderate control of Mn by [M]o/[I]o and broad Mw/Mn ~1.5; however, the Mn evolves linearly with conversion and Mw/Mn remains narrow early in the ROP (Figure 6.8). These observations are consistent with previous reports of the entropically-controlled ROP of macrolactones, which often require heating to favor the formation of polymer,22,24,25,37 although enzymatic catalysts do not require excessive heating.3,24,26 The results with TCC/BEMP stand in stark contrast to the ROP of EB (0.4 g, 1.47 mmol) mediated by 1- S/BEMP (0.03 mmol each) from benzyl alcohol (0.03 mmol) which achieves only 25% conversion to polymer in 10 hours. When combined with the 1-S vs TCC result for the ROP of VL (see ROP of Strained Lactones), these results suggest that ureas are generally more effective than thiourea H-bond donors for ROP. 230 H-bonding catalysts are effective and thermally stable at 80°C. The ROP of macrolactones are usually conducted at elevated temperatures so that the entropically controlled enchainment favors the formation of polymer.22,37 Hence, the H-bond mediated, solvent-free ROP of macrolactones presents a distinct challenge over that of VL or CL because the H-bonding interactions of thiourea and urea mediated ROP weaken at high temperature.38 Further, organic catalysts are susceptible to charring/decomposition at high temperature.39 However, neither deactivation nor decomposition appear to be a concern for TCC/base cocatalyzed ROP at 80°C (Table 6.3). TCC/BEMP were also applied for the solvent-free ROP of PDL from benzyl alcohol (Table 6.3). Versus EB, the polymerization times (6-8 h) and Mw/Mn (> 2) are attenuated, which may be due to the elevated viscosity of the PDL and PPDL vs EB and PEB. ROP of Lactide. RESULTS AND DISCUSSION Contrary to other lactones, thiourea catalysts are more active than their urea analogues in the solvent-free ROP of L-LA conducted at 100°C. L-LA and isotactic PLA are crystalline which requires elevated temperatures to melt the monomer (Tm = 97°C) ROP of Lactide. Contrary to other lactones, thiourea catalysts are more active than their urea analogues in the solvent-free ROP of L-LA conducted at 100°C. L-LA and isotactic PLA are crystalline which requires elevated temperatures to melt the monomer (Tm = 97°C) and, ostensibly, polymer (Tm = 180°C, PLLA). Several H-bond donors with PMDTA cocatalyst (0.0277 mmol each, see Table 6.4) were evaluated for the solvent-free ROP of L-LA (0.40 g, 2.77 mmol) from benzyl alcohol (0.055 mmol) at 100°C. As opposed to the (thio)urea mediated ROP of VL or CL, weak base cocatalysts are optimal for the ROP of LA.13,40 Every thiourea H-bond donor is more active for the ROP of LA than its corresponding urea H-bond donor. Indeed, the only urea that stands out in the series is TCC, which is more active in the ROP than the other urea H-bond donors. A similar trend has been observed in solution where TCC is more active than 1-O,17 and ureas with fewer electron withdrawing substituents have been observed to be more active.16 231 A screen of base cocatalysts with 2-S showed PMDTA cocatalyst to exhibit a good combination of high rate and control (Table 6.7). The reaction solution solidifies during the course of the polymerization, ~80% conversion, Figure 6.9, and the first order evolution of concentration of monomer exhibits deviation from linearity that may be associated with limited molecular mobility in the crystalline polymer. However, the other characteristics of a living polymerization persist: Mn predictable from [M]o/[I]o, linear evolution of Mn vs conversion, and narrow Mw/Mn (Figure 6.9). When initiated from pyrenebutanol, the 2- S/PMDTA cocatalyzed ROP in solvent-free conditions produces PLA that exhibits overlapping UV and RI traces in the GPC (Figure 6.2), which suggests that the polymer chains are initiated from the fluorescent alcohol. Certainly, the results in Table 6.4 suggest that conducting the H-bond mediated ROP of LA in the polymer melt (i.e. >180°C) is not necessary to retain the high level of control associated with a living organocatalytic ROP. The solvent-free ROP of LA mediated by 2-S/PMDTA remains active at elevated temperatures. RESULTS AND DISCUSSION Solution ROPs of LA (0.694 mmol, 0.25 M) from benzyl alcohol (0.0025 M) using 2-S/PMDTA (0.0063 M each) were conducted in toluene. At 50°C, the ROP proceeds to 24% conversion in 125 min (kobs = 0.0021, Figure 6.10), but at 80°C and above, the reaction proceeds to polymer very slowly (2% conversion at 60 min). Analogous solvent-free ROP remains active up to 180°C (the highest temperature examined), and the ROPs of macrolactones do not experience deactivation at elevated temperature (see ROP of Macrolactones). The low concentrations required to fully dissolve LA in toluene is not the source of the catalyst deactivation at high temperature. A solvent free ROP of LA (6.244 mmol) catalyzed by 2-S/PMDTA at reduced catalyst loadings (0.006 mmol each) from benzyl alcohol (0.125 mmol) at 100°C achieves 90% conversion in 24 h. Further, the 232 solution ROP in CDCl3 (1 M LA; 0.025 M cats; 0.010 M benzyl alcohol) allows for higher reagent concentrations, but a ROP at 40°C in this solvent is not appreciably faster than the low concentration toluene run discussed above (kCDCl3/ktoluene = 2). Even these reaction conditions experience reduced activity at elevated temperatures. These observations suggest greater synthetic flexibility in the solvent-free (versus solution) ROP of LA. For the ROP of lactide, the effects of reaction conditions on polymer tacticity must also be considered. For each polymerization in Table 6.4, the percent isotacticity was determined from the isolated polymer by 1H decoupled 13C NMR using previously established tacticity-dependent chemical shifts (Experimental Section).13,41,42 A small temperature screen was conducted, and running the 2-S/PMDTA (0.028 mmol each) cocatalyzed ROP of L-LA at or below ~97°C (the melting point of LA) results in drastically reduced polymerization rates, and reaction temperatures at or above 140°C erode stereochemistry. In the ROP of L-LA, the retention of stereochemistry is important due to the highly tacticity-dependent material properties of PLA.39 The 2-S/PMDTA cocatalyst system (Table 6.4, entry 1) is not only the most active catalyst of the systems examined, but it exhibits the highest isotacticity (0.94). This observation suggests that 2-S is highly selective for chain extension vs non-productive reactions and begs for the further optimization of this platform, which will be the focus of future work. In a recent touchstone on the challenges of solvent-free ROP of LA,39 it was noted that commercial samples of PLA are ideally >0.97 isotacticity. RESULTS AND DISCUSSION This suggests that 2-S/PMDTA at 0.94 isotacticity is not a ready-made solution to the problem that is the solvent-free, organocatalytic ROP of LA; however, our results suggest that these H-bond mediated catalysts may be able to provide the answer upon further optimization. Indeed, in a comparable solution experiment 233 (c.f. Table 6.4, entry 1), the 2-S/PMDTA (0.05 mmol) cocatalyzed ROP of L-LA (1 mmol, 1 M) from benzyl alcohol (0.01 mmol) in CDCl3 at room temperature yields highly isotactic polymer (%iso = 0.97; Tm = 169°C). In addition to tacticity, the introduction of color to PLA samples from catalyst (or decomposition) impurities can be a concern.39 The 2- S/PMDTA cocatalyzed ROP (Table 6.4, entry 1) produces an off-white, yellow color at high conversion, but the discoloration is very minimal if freshly-distilled PMDTA is used. Future catalysts with enhanced thermal stability or augmented activity (i.e. lower catalyst loadings) may prevent discoloration, but the color is easily removed by washing the polymer with methanol. Last, the comparable TBD-catalyzed (0.014 mmol) ROP of L-LA (0.40 g, 2.78 mmol) from benzyl alcohol (0.028 mmol) was conducted under solvent-free conditions (at 100oC) resulting in 90% conversion to polymer in 4 h and yielding PLLA that exhibits lower isotacticity (%iso = 0.78) than the 2-S/PMDTfA cocatalyzed reaction (TBD polymer: Mn = 19,900; Mw/Mn = 1.30). These results suggest that 2-S, and indeed most thioureas, plus amine base cocatalysts are more effective than TBD for the solvent- free ROP of LA. Unlike other monomers examined, thioureas (vs ureas) are superior catalysts – in terms of both activity and control – for the H-bond mediated ROP of lactide. We conducted a rate comparison for the 1-S vs 1-O (with PMDTA) mediated ROP of L-LA in acetone- d6, and the thiourea catalyst is the more active of the two: k1-S/k1-O (acetone-d6) = 4.4. Contrary to LA, ureas are always more active and controlled than the corresponding thioureas for the ROP of VL and CL regardless of the reaction solvent.17,36 This suggests that the relative activity of urea vs thiourea is not dictated by solvent, and the various monomers seem to exhibit a preference for urea vs thiourea. Our group previously 234 described the activity of thiourea/amine base cocatalysts in the ROP of LA as being related to the nature of cocatalyst binding (i.e. RESULTS AND DISCUSSION enthalpic vs solvophobic binding),34 and understanding the preference exhibited by LA for thioureas vs ureas may require a full study of the solution interactions at play during an ROP catalyzed by the various catalysts. We are unable to measure a urea/LA binding constant due to poor solubility in non- hydrogen bonding solvents. Copolymerizations. The generation of copolymers is possible through a one-pot, solvent- free approach. In the one-pot ROP of VL (3.99 mmol) and CL (3.99 mmol) from benzyl alcohol (0.016 mmol) (solvent-free, room temperature), the TCC/MTBD (0.02 mmol) cocatalyst system fully converts VL to polymer in 10 min, but the homopolymer precipitates from CL solution prior to conversion of the slower opening monomer. The TCC/BEMP system, however, allows for the full conversion to PVL-co-PCL (Mn = 94,000; Mw/Mn = 1.41) in 5 h. The first order evolution of [monomer]s versus time suggests the formation of a gradient-block copolymer (Figure 6.11). The successful formation of block- copolymer with TCC/BEMP versus TCC/MTBD which does not produce copolymer suggests that the former cocatalysts are able to conduct ROP of the slower monomer on a time scale that is competitive with precipitation of the homo-PVL from CL solvent. Copolymerizations were also performed with VL/EB (3.99 mmol/3.99 mmol) and CL/EB (3.5 mmol/3.5 mmol) employing the TCC/BEMP (2 mol%) cocatalysts. In these ROPs, conducted entirely at room temperature and solvent free (see Experimental Section) VL and CL quickly achieve full conversion in 5 min and 6 min, respectively, and the EB blocks grow slowly over the next ~11 h to give PVL-co-PEB (Mn = 27,000; Mw/Mn = 1.60) and PCL-co-PEB (Mn = 28,700; Mw/Mn = 1.48). The relative kinetics (Figure 6.12) suggest 235 the formation of block copolymers. This same approach was used to generate a one-pot triblock copolymer: VL (3.99 mmol), CL (3.99 mmol) and EB (3.99 mmol) were grown from benzyl alcohol (0.1197 mmol) using TCC/BEMP (0.1197 mmol each). The conversion versus time of this ROP suggests a gradient-block polymer (see SI, the previous VL/CL copolymerization produced a gradient-copolymer), and 1H and 13C NMR of the isolated polymer (see Figure 6.13) indicate that all blocks are present (Mn = 40,000, Mw/Mn = 1.53). A solvent-free approach to the copolymerization of LA and VL allows for the one- pot synthesis of block copolymers. RESULTS AND DISCUSSION In the one-pot synthesis of diblock copolymers of LA and VL in solution, the ROP has been observed to proceed to full conversion of LA when reaction progress halts, resulting in the incorporation of no comonomer.11 To confirm this report, a copolymerization of LA (0.33 mmol, 0.66 M) and VL (0.99 mmol, 2 M) from benzyl alcohol using trimethyl-1,4,7-triazacyclononane (t-TACN, see Table 6.7)/TCC cocatalysts (0.0265 mmol each) was attempted in C6D6.34,40 The LA achieved full conversion to polymer in 18 h, but the VL does not convert over the next 24 hours; the expected result. Under solvent-free conditions, however, the copolymerization under otherwise same conditions results in full conversion of both monomer portions in 3.5 h yielding a single peak in the GPC trace (Mn = 27,600; Mw/Mn = 1.57) and two phase transitions (DSC: Tm =52°C and 148°C). First order evolution of [monomer] vs time suggests the formation of a block copolymer (Figure 6.14). When the one-pot copolymerization of LA and CL is attempted, the LA achieves full conversion in 2 min, but the CL does not undergo any enchainment over the next 24 h. Typically, alkylamine base cocatalysts are not effective for the ROP of VL or CL, but these results suggest that 236 solvent-free, reaction conditions may provide new opportunities in catalyst development as well as materials synthesis. 237 CONCLUSION Thiourea and urea catalysts have been shown to be effective for the solvent-free ROP of lactones at ambient and elevated temperatures. The urea class of H-bond donors facilitates solvent-free ROP for most monomers (VL, CL and EB), the thioureas being exceedingly slow in the ROP of these highly polar lactones. Solvent polarity is not the primary determining factor, however, as thioureas (not ureas) are more effective for the ROP of LA in solvent and the monomer melt. The ROPs retain the characteristics of a living polymerization despite solidifying prior to full conversion, and copolymers can be accessed in a variety of combinations. For those seeking to employ organocatalysts of this class in polymer synthesis, we offer a succinct summation: 1) urea (vs thiourea) H-bond donors plus base are the most active and most controlled organocatalysts for ROP under any reaction conditions (lactide excluded); 2) TCC/BEMP is the most active, most controlled organocatalytic system but TCC/MTBD is almost as active and probably more readily available; 3) for the ROP of lactide, the bisthiourea 2-S plus PMDTA is the most active and controlled organic cocatalyst system that we are aware of. In catalyst development, the community has come a long way in terms of catalytic activity from 1-S (PLA: 0.7/min; PVL: 0.2/min) to more active H-bond donors (PLA: 18.2/min (2-S); PVL: 63/min (TCC)), but catalyst productivity has so far been limited by catalyst deactivation at reduced loadings. Further, there is no ostensible cost – in terms of reaction control – for employing more active (thio)urea H-bond donors. For (thio)urea ROP catalysts under solvent-free conditions and in solution,35,36 more active catalysts are generally more ll d 238 LIST OF REFERENCES (1) Lim, L. T.; Auras, R.; Rubino, M. Prog. Polym. Sci. 2008, 33, 820–852. (2) Mathers, R. T.; Meier, M. A. R. In Green Polymerization Methods: Renewable Starting Materials, Catalysis and Waste Reduction; Mathers, R. T., Meier, M. a R., Eds.; Wiley-VCH: Federal Republic of Germany, 2011; pp 1–7. (3) Albertsson, A.-C.; Varma, I. K.; Srivastava, R. K. In Handbook of Ring-Opening Polymerization; Dubois, P., Coulembier, O., Raquez, J.-M., Eds.; Wiley-VCH: Federal Republic of Germany, 2009; pp 287–302. (4) Olsén, P.; Odelius, K.; Albertsson, A. C. Biomacromolecules 2016, 17 (3), 699–709. (5) James, S. L.; Adams, C. J.; Bolm, C.; Braga, D.; Collier, P.; Friščić, T.; Grepioni, F.; Harris, K. D. M.; Hyett, G.; Jones, W.; Krebs, A.; Mack, J.; Maini, L.; Orpen, a G.; Parkin, I. P.; Shearouse, W. C.; Steed, J. W.; Waddell, D. C. Chem. Soc. Rev. 2012, 41 (1), 413–447. (6) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R. M. Macromolecules 2010, 43 (5), 2093–2107. (7) Kamber, N. E. N. E.; Jeong, W.; Waymouth, R. M. R. M.; Pratt, R. C. R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Chem. Rev. 2007, 107 (12), 5813–5840. (8) Cooley, C. B.; Trantow, B. M.; Nederberg, F.; Kiesewetter, M. K.; Hedrick, J. L.; Waymouth, R. M.; Wender, P. A. J. Am. Chem. Soc. 2009, 131, 16401–16403. (9) Geihe, E. I.; Cooley, C. B.; Simon, J. R.; Kiesewetter, M. K.; Edward, J. a; Hickerson, R. P.; Kaspar, R. L.; Hedrick, J. L.; Waymouth, R. M.; Wender, P. a. Proc. Natl. Acad. Sci. U. S. A. 2012, 109 (33), 13171–13176. 239 (10) Pratt, R. C.; Nederberg, F.; Waymouth, R. M.; Hedrick, J. L. Chem. Commun. 2008, 114–116. (11) Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.; Waymouth, R. M.; Hedrick, J. L. Macromolecules 2006, 39 (25), 8574–8583. (12) Dove, A. P.; Pratt, R. C.; Lohmeijer, B. G. G.; Waymouth, R. M.; Hedrick, J. L. J. Am. Chem. Soc. 2005, 127 (40), 13798–13799. (13) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. a.; Lundberg, P. N. P.; Dove, A. P.; Li, H.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L. Macromolecules 2006, 39 (23), 7863–7871. (14) Thomas, C.; Bibal, B. Green Chem. 2014, 16 (4), 1687–1699. (15) Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Nat. Chem. (21) Martello, M. T.; Burns, A.; Hillmyer, M. ACS Macro Lett. 2012, 1 (1), 131–135. LIST OF REFERENCES 2016, 8 (11), 1047–1053. (16) Lin, B.; Waymouth, R. M. J. Am. Chem. Soc. 2017, 139 (4), 1645–1652. (17) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, M. K. ACS Macro Lett. 2017, 6 (4), 421–425. (17) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, (17) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, M. K. ACS Macro Lett. 2017, 6 (4), 421–425. (18) Chuma, A.; Horn, H. W.; Swope, W. C.; Pratt, R. C.; Zhang, L.; Lohmeijer, B. G. G.; Wade, C. G.; Waymouth, R. M.; Hedrick, J. L.; Rice, J. E. J. Am. Chem. Soc. 2008, 130 (21), 6749–6754. (18) Chuma, A.; Horn, H. W.; Swope, W. C.; Pratt, R. C.; Zhang, L.; Lohmeijer, B. G. (19) Simón, L.; Goodman, J. M. J. Org. Chem. 2007, 72 (25), 9656–9662. (20) Pratt, R. C.; Lohmeijer, B. G. G.; Long, D. A.; Waymouth, R. M.; Hedrick, J. L. J. Am. Chem. Soc. 2006, 128 (14), 4556–4557. (21) Martello, M. T.; Burns, A.; Hillmyer, M. ACS Macro Lett. 2012, 1 (1), 131–135. 240 (22) Pascual, A.; Sardon, H.; Veloso, A.; Ruiperez, F.; Mecerreyes, D. ACS Macro Lett. 2014, 3 (9), 849–853. (23) Ding, L.; Jin, W.; Chu, Z.; Chen, L.; Lü, X.; Yuan, G.; Song, J.; Fan, D.; Bao, F. Inorg. Chem. Commun. 2011, 14 (8), 1274–1278. (24) Van Der Mee, L.; Helmich, F.; De Bruijn, R.; Vekemans, J. A. J. M.; Palmans, A. R. A.; Meijer, E. W. Macromolecules 2006, 39 (15), 5021–5027. (25) Duda, A.; Kowalski, A.; Penczek, S.; Uyama, H.; Kobayashi, S. Macromolecules 2002, 35 (11), 4266–4270. (26) Heise, A.; Duxbury, C. J.; Palmans, A. R. A. In Handbook of Ring-Opening Polymerization; Dubois, P., Coulembier, O., Raquez, J.-M., Eds.; Wiley-VCH: Federal Republic of Germany, 2009; pp 379–394. (27) Coulembier, O.; Lemaur, V.; Josse, T.; Minoia, A.; Cornil, J.; Dubois, P. Chem. Sci. 2012, 3 (3), 723–726. (28) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. K. Macromolecules 2014, 47 (21), 7463–7468. (29) Erickson, B. E. Chem. Eng. News 2016, 94 (36), 16. (30) Stukenbroeker, T. S.; Bandar, J. S.; Zhang, X.; Lambert, T. H.; Waymouth, R. M. ACS Macro Lett. 2015, 4 (8), 853–856. (31) Zhang, L.; Nederberg, F.; Pratt, R. C.; Waymouth, R. M.; Hedrick, J. L.; Wade, C. G.; V, S. U.; February, R. LIST OF REFERENCES V; Re, V.; Recei, M.; April, V. Macromolecules 2007, 40, 4154–4158. (32) Ladelta, V.; Bilalis, P.; Gnanou, Y.; Hadjichristidis, N. Polym. Chem. 2017, 8 (3), 511–515. 241 (33) Kadota, J.; Pavlović, D.; Hirano, H.; Okada, A.; Agari, Y.; Bibal, B.; Deffieux, A.; Peruch, F. RSC Adv. 2014, 4 (28), 14725. (34) Kazakov, O. I.; Kiesewetter, M. K. Macromolecules 2015, 48 (17), 6121–6126. (35) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. T.; Kiesewetter, M. K. Macromolecules 2015, 48 (17), 6127–6131. (36) Fastnacht, K. V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; Kiesewetter, E. T.; Kiesewetter, M. K. ACS Macro Lett. 2016, 5 (8), 982–986. (37) Duda, A.; Kowalski, A. In Handbook of Ring-Opening Polymerization; Dubois, P., Coulembier, O., Raquez, J.-M., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA, 2009; pp 1–52. (38) Ansyln, E. V.; Dougherty, D. A. In Modern Physical Organic Chemistry; 2006; pp 162–194. (39) Mezzasalma, L.; Dove, A. P.; Coulembier, O. Eur. Polym. J. 2017, No. May, 0–1. (40) Coady, D. J.; Engler, A. C.; Horn, H. W.; Bajjuri, K. M.; Fukushima, K.; Jones, G. O.; Nelson, A.; Rice, J. E.; Hedrick, J. L. ACS Macro Lett. 2012, 1 (1), 19–22. (41) Ovitt, T. M.; Coates, G. W. J. Am. Chem. Soc. 2002, 124 (7), 1316–1326. (42) Thakur, K. M.; Kean, R. T.; Hall, E. S.; Kolstad, J. J.; Lindgren, T. ; Doscotch, M. (42) Thakur, K. M.; Kean, R. T.; Hall, E. S.; Kolstad, J. J.; Lindgren, T. ; Doscotch, M. ; Siepmann, J. I.; Munson, E. J. Macromolecules 1997, 30 (8), 2422–2428. ; Siepmann, J. I.; Munson, E. J. Macromolecules 1997, 30 (8), 2422–2428. 242 O O OH + TCC and base (0.02 mmol each) O O H O n N H N H O Cl Cl Cl N N N N N N H N N N P N N N t-Bu MTBD TBD DBU BEMP TCC solvent-free, r.t. Table 6.1. TCC plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 mmol, 1 equiv, neat), TCC and base (0.02 mmol, each). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. c) no TCC, only TBD (0.02 mmol). LIST OF REFERENCES VL (neat) Entry Base [M]o/[I]o Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 DBU 500 65 97 99,500 1.12 2 BEMP 500 3 95 108,000 1.04 3 MTBD 500 31 99 100,500 1.08 4 200 15 96 43,900 1.07 5 100 15 98 22,000 1.16 6 50 10 98 10,300 1.10 7c TBD 500 27 99 115,500 1.21 O O OH + TCC and base (0.02 mmol each) O O H O n N H N H O Cl Cl Cl N N N N N N H N N N P N N N t-Bu MTBD TBD DBU BEMP TCC solvent-free, r.t. VL (neat) Entry Base [M]o/[I]o Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 DBU 500 65 97 99,500 1.12 2 BEMP 500 3 95 108,000 1.04 3 MTBD 500 31 99 100,500 1.08 4 200 15 96 43,900 1.07 5 100 15 98 22,000 1.16 6 50 10 98 10,300 1.10 7c TBD 500 27 99 115,500 1.21 Table 6.1. TCC plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 mmol, 1 equiv, neat), TCC and base (0.02 mmol, each). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. c) no TCC, only TBD (0.02 mmol). Table 6.1. TCC plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 mmol, 1 equiv, neat), TCC and base (0.02 mmol, each). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. c) no TCC, only TBD (0.02 mmol). 243 O O OH + Donor and MTBD (0.02 mmol each) O O H O n solvent-free, r.t. LIST OF REFERENCES VL (3.9 mmol, neat) (0.008 mmol) N N N N H N H X CF3 F3C N H N H X CF3 F3C N H N H X CF3 CF3 HN HN X CF3 F3C NH NH X F3C F3C N H N H X F3C F3C N 2-S: X = S 2-O: X = O 1-S: X = S 1-O: X = O 3-S: X = S 2-O: X = O MTBD (0.008 mmol) Entry Donor Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 1-S 1,200 94 71,500 1.29 2 1-O 440 96 97,800 1.17 3 2-S 1,420 96 82300 1.17 4 2-O 30 97 101,000 1.13 5 3-S 1,900 99 85,700 1.19 6 3-O 6 98 94,500 1.07 Table 6.2. H-bond donor plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 mmol, 1 equiv, solvent-free), benzyl alcohol (0.008 mmol) and (thio)urea/MTBD (0.02 mmol each). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. Table 6.2. H-bond donor plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 mmol, 1 equiv, solvent-free), benzyl alcohol (0.008 mmol) and (thio)urea/MTBD (0.02 mmol each). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. . H-bond donor plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 Table 6.2. H-bond donor plus base cocatalyzed ROP of VL. Reaction conditions: VL (3.99 mmol, 1 equiv, solvent-free), benzyl alcohol (0.008 mmol) and (thio)urea/MTBD (0.02 mmol each). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. 244 O O O O O O OH TCC and BEMP (2.0 mol% each) solvent-free, 80°C poly(ethylene breassylate) or poly(petnadecalactone) EB (neat) or PDL (neat) + Entry Mon. [M]o/[I]o Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 EB 50 90 99 28,200 1.64 2 100 130 97 42,800 1.62 3 200 330 96 51,000 1.60 4 PDL 50 360 96 24,800 2.23 5 100 900 97 33,000 2.46 O O O O O O OH TCC and BEMP (2.0 mol% each) solvent-free, 80°C poly(ethylene breassylate) or poly(petnadecalactone) EB (neat) or PDL (neat) + OH Entry Mon. LIST OF REFERENCES [M]o/[I]o Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb 1 EB 50 90 99 28,200 1.64 2 100 130 97 42,800 1.62 3 200 330 96 51,000 1.60 4 PDL 50 360 96 24,800 2.23 5 100 900 97 33,000 2.46 Table 6.3. TCC plus base cocatalyzed ROP of macrolactones. Reaction conditions: EB and PDL (2.95 and 1.66 mmol respectively, 1 equiv, solvent free), benzyl alcohol, TCC/BEMP (0.06 (for EB) and 0.033 (for PDL) mmol). a) monomer conversion was monitored via 1H NMR. b) Mn and Mw/Mn were determined by GPC (CH2Cl2) vs polystyrene standards. 245 245 Entry Donor Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb %Isoc 1 2-S 5 90 10,700 1.06 0.94 2 3-S 20 90 14,600 1.07 0.80 3 TCC 33 90 15,800 1.09 0.82 4 3-O 102 91 11,300 1.16 0.85 5 1-S 130 90 11,300 1.11 0.83 6 2-O 230 90 10,400 1.18 0.83 7 1-O 540 92 12,100 1.11 0.82 O O OH donor and PMDTA (0.028 mmol each) solvent-free, 100°C O O O O O H n O O L-LA 2.8 mmol 0.055 mmol + N N N PMDTA O O OH donor and PMDTA (0.028 mmol each) solvent-free, 100°C O O O O O H n O O L-LA 2.8 mmol 0.055 mmol + N N N PMDTA OH donor and PMDTA (0.028 mmol each) solvent-free, 100°C O O O O O H n N N N PMDTA Entry Donor Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb %Isoc 1 2-S 5 90 10,700 1.06 0.94 2 3-S 20 90 14,600 1.07 0.80 3 TCC 33 90 15,800 1.09 0.82 4 3-O 102 91 11,300 1.16 0.85 5 1-S 130 90 11,300 1.11 0.83 6 2-O 230 90 10,400 1.18 0.83 7 1-O 540 92 12,100 1.11 0.82 Entry Donor Time (min) Conv.a (%) Mnb (g/mol) Mw/Mnb %Isoc 1 2-S 5 90 10,700 1.06 0.94 2 3-S 20 90 14,600 1.07 0.80 3 TCC 33 90 15,800 1.09 0.82 4 3-O 102 91 11,300 1.16 0.85 5 1-S 130 90 11,300 1.11 0.83 6 2-O 230 90 10,400 1.18 0.83 7 1-O 540 92 12,100 1.11 0.82 Table 6.4. H-bond Mediated Solvent-free ROP of L-LA. Reaction conditions: L-LA (400 mg, 2.77 mmol), benzyl alcohol (2 mol%, 0.055 mmol), donor (1 mol%, 0.028 mmol), PMDTA (1 mol%, 0.028 mmol) in the monomer melt at 100oC. a. Conversion determined by 1H NMR. b. Table 6.6. TCC plus MTBD or BEMP cocatalyzed ROP of CL. a. Conversion determined by 1H NMR b Mn and Mw/Mn were obtained by GPC by 1H NMR. b. Mn and Mw/Mn were obtained by GPC. Table 6.7. Base Screen in the 2-S Mediated ROP of L-LA. a. Conversion determined by 1H NMR. b. Isotacticity determined by selectively decoupled 1 H NMR at 50 oC. Mn and Mw/Mn were obtained by GPC. LIST OF REFERENCES Mn and Mw/Mn determined by GPC versus PS standards. c. %iso = factional percent isotactic, see Experimental section. 246 O O OH + n 0.400 g 4 mmol 1 equiv. 0.2 mol% O O OH + n 0.400 g 4 mmol 1 equiv. 0.2 mol% TCC / BEMP (0.1 - 0.5 mol% solvent-free, 20°C O O H O n Table 6.5. Solvent free ROP of VL with TCC/BEMP. a. Conversion determined by 1 H NMR b M and M were obtained by GPC O O OH + n 0.400 g 4 mmol 1 equiv. 0.2 mol% TCC / BEMP (0.1 - 0.5 mol% solvent-free, 20°C O O H O n Entry Base loading (mol%) Time (min) Conv. (%)a Mn (g/mol)b Mw/Mnb 1 0.1 60 95 100,500 1.21 2 0.2 10 99 103,200 1.12 3 0.4 8 99 114,500 1.06 4 0.5 3 96 108,400 1.04 5 0.6 2 98 113,200 1.23 TCC / BEMP (0.1 - 0.5 mol% solvent-free, 20°C O O H O n Entry Base loading (mol%) Time (min) Conv. (%)a Mn (g/mol)b Mw/Mnb 1 0.1 60 95 100,500 1.21 2 0.2 10 99 103,200 1.12 3 0.4 8 99 114,500 1.06 4 0.5 3 96 108,400 1.04 5 0.6 2 98 113,200 1.23 Table 6.5. Solvent free ROP of VL with TCC/BEMP. a. Conversion determined by 1 H NMR. b. Mn and Mw were obtained by GPC. NMR. b. Mn and Mw were obtained by GPC. O O OH + n 0.400 g 1 equiv. 0.2 mol% O O n 0.400 g 1 equiv. O O OH + n 0.400 g 1 equiv. 0.2 mol% TCC / BEMP or MTBD (0.1 - 0.5 mol% solvent-free, 20°C O O H O n TCC / BEMP or MTBD (0.1 - 0.5 mol% solvent-free, 20°C O O H O n Entry Base Base loading (mol%) Time (min) Conv. (%)a Mn (g/mol)b Mw/Mnb 1 BEMP 0.1 47 160 n/a n/a 2 0.2 95 142 83,000 1.17 3 0.3 98 60 85,000 1.22 4 0.4 98 40 89,100 1.20 5 0.5 97 30 82,500 1.20 6 MTBD 0.1 9 450 n/a n/a 7 0.3 40 240 n/a n/a 8 0.5 98 300 92,000 1.28 247 Entry Base Time (min) Conv. LIST OF REFERENCES (%)a Isotacticityb Mnc Mw/Mnc 1 t-TACN 1 95 0.82 12,500 1.06 2 Me6TREN 2 87 0.88 13,900 1.04 3 PMDTA 5 90 0.94 10,600 1.07 4 (+)-sparteine 10 87 0.89 12,600 1.16 5 DMAP 40 93 0.72 16,500 1.30 6 TMEDA 90 90 0.76 13,800 1.10 7 pyridine 24 h 3 - - - O O OH 2-S / base (1 mol%) solvent-free, 100°C O O O O O H n O O 0.400 1 equiv. 2 mol% + N N N N N N N N N N N H H N N N TMEDA Me6TREN t-TACN spartiene PMDTA Entry Base Time (min) Conv. (%)a Isotacticityb Mnc Mw/Mnc 1 t-TACN 1 95 0.82 12,500 1.06 2 Me6TREN 2 87 0.88 13,900 1.04 3 PMDTA 5 90 0.94 10,600 1.07 4 (+)-sparteine 10 87 0.89 12,600 1.16 5 DMAP 40 93 0.72 16,500 1.30 6 TMEDA 90 90 0.76 13,800 1.10 7 pyridine 24 h 3 - - - O O OH 2-S / base (1 mol%) solvent-free, 100°C O O O O O H n O O 0.400 1 equiv. 2 mol% + N N N N N N N N N N N H H N N N TMEDA Me6TREN t-TACN spartiene PMDTA O O OH 2-S / base (1 mol%) solvent-free, 100°C O O O O O H n O O 0.400 1 equiv. 2 mol% + N N N N N N N N N N N H H N N N TMEDA Me6TREN t-TACN spartiene PMDTA Entry Base Time (min) Conv. (%)a Isotacticityb Mnc Mw/Mnc 1 t-TACN 1 95 0.82 12,500 1.06 2 Me6TREN 2 87 0.88 13,900 1.04 3 PMDTA 5 90 0.94 10,600 1.07 4 (+)-sparteine 10 87 0.89 12,600 1.16 5 DMAP 40 93 0.72 16,500 1.30 6 TMEDA 90 90 0.76 13,800 1.10 7 pyridine 24 h 3 - - - 248 248 Scheme 6.1. Mechanism for the urea or thiourea plus base cocatalyzed ROP. LIST OF REFERENCES R N H N H R' S + N N N + N N N H O O Ph O H R N H N R' O MTBD-H O O Ph O H O O N H N O N N N H CF3 CF3 1-O imidate OH O O Evidence of Proton Transfer Observed R N H N R' O H O Ph N H N H S CF3 CF3 N N N + OH O O cocastalys H-bonding Observed Ureas: IMIDATE MEDIATED MECHANISM Thioureas: NEUTRAL H-BONDING MECHANISM Ph O O H O n 1-S MTBD Thioureas: Thioureas: NEUTRAL H-BONDING MECHANISM Thioureas: NEUTRAL H-BONDING MECHANISM OH cocastalys H-bonding Observed CF O Ureas: IMIDATE MEDIATED MECHANISM CF3 OH O imidate Evidence of Proton Transfer Observed Evidence of Proton Transfer Observed Scheme 6.1. Mechanism for the urea or thiourea plus base cocatalyzed ROP. 249 Figure 6.1. (upper) First order evolution of [monomer] versus time and (lower) Mn and Mw/Mn versus conversion for the TCC/MTBD cocatalyzed ROP of VL. Conditions: VL (3.99 mmol), TCC (0.02 mmol), MTBD (0.02 mmol) and benzyl alcohol (0.008 mmol) at room temperature. Figure 6.1. (upper) First order evolution of [monomer] versus time and (lower) Mn and Mw/Mn versus conversion for the TCC/MTBD cocatalyzed ROP of VL. Conditions: VL (3.99 mmol), TCC (0.02 mmol), MTBD (0.02 mmol) and benzyl alcohol (0.008 mmol) at room temperature. 250 Figure 6.2. RI and UV GPC traces of the ROP initiated from pyrenebutanol for (top) PVL (TCC/BEMP) and (bottom) PLA (2-S/PMDTA). Figure 6.2. RI and UV GPC traces of the ROP initiated from pyrenebutanol for (top) PVL (TCC/BEMP) and (bottom) PLA (2-S/PMDTA). 251 igure 6.3. (upper) First order evolution of [monomer] versus time; and (lower) Mn an Mw/Mn versus conversion for the TCC/BEMP cocatalyzed ROP of VL. Conditions: VL 3.99 mmol), TCC (0.019 mmol), BEMP (0.019 mmol) and benzyl alcohol (0.008 mmol t room temperature Figure 6.3. (upper) First order evolution of [monomer] versus time; and (lower) Mn and Mw/Mn versus conversion for the TCC/BEMP cocatalyzed ROP of VL. Conditions: VL (3.99 mmol), TCC (0.019 mmol), BEMP (0.019 mmol) and benzyl alcohol (0.008 mmol) at room temperature. Figure 6.3. (upper) First order evolution of [monomer] versus time; and (lower) Mn and Mw/Mn versus conversion for the TCC/BEMP cocatalyzed ROP of VL. Conditions: VL (3.99 mmol), TCC (0.019 mmol), BEMP (0.019 mmol) and benzyl alcohol (0.008 mmol) at room temperature. 252 e 6.4. LIST OF REFERENCES (upper) First order evolution of [monomer] versus time and (lower) Mn and Mn versus conversion for the TCC/BEMP cocatalyzed ROP of CL. Conditions: CL mmol), TCC (0.018 mmol), BEMP (0.018 mmol) and benzyl alcohol (0.007 mmol) t t y = 0.1216x - 0.0856 R² = 0.9948 0 0.5 1 1.5 2 2.5 3 3.5 4 0 5 10 15 20 25 30 35 Ln([M]o/[M]) Time, min Stirring Stops 150 rpm Hard Solid (no deformation) y = 0.1216x - 0.0856 R² = 0.9948 0 0.5 1 1.5 2 2.5 3 3.5 4 0 5 10 15 20 25 30 3 Ln([M]o/[M]) Time, min Stirring Stops 150 rpm Hard Solid (no deformation) Hard Solid (no deformation) Figure 6.4. (upper) First order evolution of [monomer] versus time and (lower) Mn and Mw/Mn versus conversion for the TCC/BEMP cocatalyzed ROP of CL. Conditions: CL (3.50 mmol), TCC (0.018 mmol), BEMP (0.018 mmol) and benzyl alcohol (0.007 mmol) at room temperature. Figure 6.4. (upper) First order evolution of [monomer] versus time and (lower) Mn and Mw/Mn versus conversion for the TCC/BEMP cocatalyzed ROP of CL. Conditions: CL (3.50 mmol), TCC (0.018 mmol), BEMP (0.018 mmol) and benzyl alcohol (0.007 mmol) at room temperature. 253 Figure 6.5. Percent conversion to polymer of VL solutions of MTBD, BEMP and TC plus benzyl alcohol (0.5 mol%, 0.02 mmol for all catalysts) at (upper) -10°C and (lowe room temperature. Conversions were determined via aliquot by 1H NMR (CDCl3). Figure 6.5. Percent conversion to polymer of VL solutions of MTBD, BEMP and TCC plus benzyl alcohol (0.5 mol%, 0.02 mmol for all catalysts) at (upper) -10°C and (lower) room temperature. Conversions were determined via aliquot by 1H NMR (CDCl3). 254 Figure 6.6. Solvent-free ROP allows for the direct-from-monomer creation of a negativ mold in seconds. Above, the hollow, PVL negative mold is of a polypropylene reaction vial. Conditions: 10 mmol VL, 0.049 mmol TCC/BEMP (each), 0.020 mmol benzy alcohol (see Experimental Section). Figure 6.6. Solvent-free ROP allows for the direct-from-monomer creation of a negative mold in seconds. Above, the hollow, PVL negative mold is of a polypropylene reaction vial. Conditions: 10 mmol VL, 0.049 mmol TCC/BEMP (each), 0.020 mmol benzyl alcohol (see Experimental Section). 255 Figure 6.7. 400 MHz 1H NOESY in acetone-d6 of (upper) TCC/MTBD (0.05 mmol each and (lower) TCC/BEMP (0.05 mmol each). 1-S/MTBD show no cross peaks in aceton d6 or C6D6 at 0.05 mmol. Figure 6.7. LIST OF REFERENCES 400 MHz 1H NOESY in acetone-d6 of (upper) TCC/MTBD (0.05 mmol each), and (lower) TCC/BEMP (0.05 mmol each). 1-S/MTBD show no cross peaks in acetone- d6 or C6D6 at 0.05 mmol. Figure 6.7. 400 MHz 1H NOESY in acetone-d6 of (upper) TCC/MTBD (0.05 mmol each), and (lower) TCC/BEMP (0.05 mmol each). 1-S/MTBD show no cross peaks in acetone- d6 or C6D6 at 0.05 mmol. 256 Figure 6.8. Mn and Mw/Mn vs conversion for the TCC/BEMP cocatalyzed ROP of EB. Conditions: EB (2.95 mmol), TCC (0.059 mmol), BEMP (0.059 mmol) and benzyl alcohol (0.015 mmol) at 80oC. Figure 6.8. Mn and Mw/Mn vs conversion for the TCC/BEMP cocatalyzed ROP of EB. Conditions: EB (2.95 mmol), TCC (0.059 mmol), BEMP (0.059 mmol) and benzyl alcohol (0.015 mmol) at 80oC. 6.8. Mn and Mw/Mn vs conversion for the TCC/BEMP cocatalyzed ROP of EB Figure 6.8. Mn and Mw/Mn vs conversion for the TCC/BEMP cocatalyzed ROP of EB. Conditions: EB (2.95 mmol), TCC (0.059 mmol), BEMP (0.059 mmol) and benzyl alcohol (0.015 mmol) at 80oC. 257 0 20 40 60 80 100 120 0 10 20 30 40 50 60 70 ln[L-LA ]o/[L-LA ] time, min stirring stops 150 rpm Hard Solid gure 6.9. (upper) First order evolution of [LA] vs time, and (lower) Mn and Mw/Mn v onversion. Reaction conditions: L-LA (400 mg, 2.77 mmol), benzyl alcohol (1mol% 028 mmol), 2-S (1 mol%, 0.028 mmol), PMDTA (1 mol%, 0.028 mmol) at 100oC. 0 20 40 60 80 100 120 0 10 20 30 40 50 60 70 ln[L-LA ]o/[L-LA ] time, min stirring stops 150 rpm Hard Solid 0 20 40 60 80 100 120 0 10 20 30 40 50 60 70 ln[L-LA ]o/[L-LA ] time, min stirring stops 150 rpm Hard Solid Figure 6.9. (upper) First order evolution of [LA] vs time, and (lower) Mn and Mw/Mn vs conversion. Reaction conditions: L-LA (400 mg, 2.77 mmol), benzyl alcohol (1mol%, 0.028 mmol), 2-S (1 mol%, 0.028 mmol), PMDTA (1 mol%, 0.028 mmol) at 100oC. 258 e 6.10. First order evolution of [LA] vs time for the 2-S/PMDTA (2.5 mol% each) alyzed ROP from benzyl alcohol (1 mol%) in (upper) toluene at 50°C (0.25 M, 0.694 l), and (lower) CDCl3 at 40°C (1 M, 0.694 mmol). Figure 6.10. LIST OF REFERENCES First order evolution of [LA] vs time for the 2-S/PMDTA (2.5 mol% each) cocatalyzed ROP from benzyl alcohol (1 mol%) in (upper) toluene at 50°C (0.25 M, 0.694 mmol), and (lower) CDCl3 at 40°C (1 M, 0.694 mmol). 259 Figure 6.11. First order evolution of [VL] and [CL] vs time for the one-pot copolymerization catalyzed by TCC/BEMP. Conditions: VL (3.99 mmol), CL (3.99 mmol), TCC (0.02 mmol), BEMP (0.02 mmol) and benzyl alcohol (0.016 mmol) Figure 6.11. First order evolution of [VL] and [CL] vs time for the one-pot copolymerization catalyzed by TCC/BEMP. Conditions: VL (3.99 mmol), CL (3.99 mmol), TCC (0.02 mmol), BEMP (0.02 mmol) and benzyl alcohol (0.016 mmol) 260 Figure 6.12. First order evolution of [monomer] vs time for the copolymerization of: (upper) VL and EB. Conditions: VL (3.99 mmol), EB (3.99 mmol), TCC (0.079 mmol), BEMP (0.079 mmol) and benzyl alcohol (0.16 mmol). VL had reached full conversion by first interrogation. (middle) CL and EB. Conditions: CL (3.99 mmol), EB (3.99 mmol), TCC (0.070 mmol), BEMP (0.070 mmol) and benzyl alcohol (0.140 mmol). CL had reached full conversion by first interrogation.(lower) VL and LA. Conditions: VL (3.99 mmol), L-LA (1.33 mmol), TCC (0.133 mmol), t-TACN (0.133 mmol) and benzyl alcohol (0.0532mmol). Figure 6.12. First order evolution of [monomer] vs time for the copolymerization of: (upper) VL and EB. Conditions: VL (3.99 mmol), EB (3.99 mmol), TCC (0.079 mmol), BEMP (0.079 mmol) and benzyl alcohol (0.16 mmol). VL had reached full conversion by first interrogation. (middle) CL and EB. Conditions: CL (3.99 mmol), EB (3.99 mmol), TCC (0.070 mmol), BEMP (0.070 mmol) and benzyl alcohol (0.140 mmol). CL had reached full conversion by first interrogation.(lower) VL and LA. Conditions: VL (3.99 mmol), L-LA (1.33 mmol), TCC (0.133 mmol), t-TACN (0.133 mmol) and benzyl alcohol (0.0532mmol). Figure 6.12. First order evolution of [monomer] vs time for the copolymerization of: (upper) VL and EB. Conditions: VL (3.99 mmol), EB (3.99 mmol), TCC (0.079 mmol), BEMP (0.079 mmol) and benzyl alcohol (0.16 mmol). VL had reached full conversion by first interrogation. (middle) CL and EB. Conditions: CL (3.99 mmol), EB (3.99 mmol), TCC (0.070 mmol), BEMP (0.070 mmol) and benzyl alcohol (0.140 mmol). CL had reached full conversion by first interrogation.(lower) VL and LA. Conditions: VL (3.99 mmol), L-LA (1.33 mmol), TCC (0.133 mmol), t-TACN (0.133 mmol) and benzyl alcohol (0.0532mmol). 261 Figure 6.13. LIST OF REFERENCES 1H and 13C NMR (400 MHz 1H, CDCl3) of the poly(VL-co-CL-co-E NMR suggests a mole ratio of the monomers in the polymer to be 1:1 (VL+CL:EB O O O O O O O O O H n m l O O O O O O O O O H n m l O O O O O O O O O H n m l Figure 6.13. 1H and 13C NMR (400 MHz 1H, CDCl3) of the poly(VL-co-CL-co-EB). 1H NMR suggests a mole ratio of the monomers in the polymer to be 1:1 (VL+CL:EB). Figure 6.13. 1H and 13C NMR (400 MHz 1H, CDCl3) of the poly(VL-co-CL-co-EB). 1H NMR suggests a mole ratio of the monomers in the polymer to be 1:1 (VL+CL:EB). Figure 6.13. 1H and 13C NMR (400 MHz 1H, CDCl3) of the poly(VL-co-CL-co-EB). 1H NMR suggests a mole ratio of the monomers in the polymer to be 1:1 (VL+CL:EB). NMR suggests a mole ratio of the monomers in the polymer to be 1:1 (VL+CL:EB) 262 Figure 6.14. First order evolution of [monomer] vs time for the copolymerization of VL and L-LA. Conditions: VL (3.99 mmol), L-LA (1.33 mmol), TCC (0.133 mmol), t-TACN (0.133 mmol) and benzyl alcohol (0.053mmol). Figure 6.14. First order evolution of [monomer] vs time for the copolymerization of VL and L-LA. Conditions: VL (3.99 mmol), L-LA (1.33 mmol), TCC (0.133 mmol), t-TACN (0.133 mmol) and benzyl alcohol (0.053mmol). Figure 6.15. Percentage conversion of EB vs time for the copolymerization of: VL, CL and EB. Conditions: VL (3.99 mmol), CL (3.99mmol), EB (3.99 mmol), TCC (1 mol%, 0.119 mmol), BEMP (1 mol%, 0.119 mmol) and benzyl alcohol (1 mol%, 0.119 mmol). VL had reached full conversion by first interrogation (at 2 min) and CL by third interrogation (at 5 min). Figure 6.15. Percentage conversion of EB vs time for the copolymerization of: VL, CL and EB. Conditions: VL (3.99 mmol), CL (3.99mmol), EB (3.99 mmol), TCC (1 mol%, 0.119 mmol), BEMP (1 mol%, 0.119 mmol) and benzyl alcohol (1 mol%, 0.119 mmol). VL had reached full conversion by first interrogation (at 2 min) and CL by third interrogation (at 5 min). 263 MANUSCRIPT – VII Published in ACS Macromolecules H-bonding Organocatalysts for Ring-Opening Polymerization at Elevated Temperatures – Some Like it Hot Email address: mkiesewetter@chm.uri.edu ABSTRACT The ring-opening polymerization (ROP) kinetics of ε-caprolactone and lactide with various H-bonding organocatalysts, (thio)ureas paired with an amine cocatalyst, were evaluated at temperatures up to 110°C. In non-polar solvent, most cocatalyst systems exhibit decomposition at high temperatures while only two, a mono-urea and bis-urea H-bond donor plus base cocatalyst, are stable up to 110°C. The onset temperature of cocatalyst decomposition must be measured under reaction conditions. In polar solvent, when the more active imidate form of the (thio)urea is favored, most cocatalyst systems become thermally stable up to 110°C, exhibiting linear Eyring behavior, including some that were unstable in toluene. The very progress of an ROP is shown to influence the nature of the catalysts as the solution polarity changes from highly polar (at 0% conversion) to less polar at full conversion. Activation parameters are discussed, and a mechanistic explanation of the observations is proposed. 265 INTRODUCTION H-bonding organocatalysts for ring-opening polymerization (ROP) are highly controlled systems for the synthesis of macromolecules.1–3 This class of catalyst is constituted by one of a host of H-bond donating moieties (most commonly a thiourea or urea) and a base cocatalyst, which effect ROP of lactones and carbonates by simultaneous activation of monomer by (thio)urea and of initiating/propagating alcohol by base.1,4 A nascent enchainment mechanism whereby base cocatalyst abstracts a proton from (thio)urea, reversibly generating an imidate catalyst, is thought to be favored under certain conditions: acidic H-bond donors (e.g. thioureas), strong base cocatalysts and polar solvent.5–9 Urea H-bond donors have been shown to be more active than the corresponding thioureas.5,7,10,11 These trends also hold for the multi-(thio)urea H-bond donors developed for ROP of esters by our group.10,12 For these systems, the bis- or tris-donors are thought to operate by an activated (thio)urea mechanism where one bis- or tris-donor activates one monomer, and the internal H-bond stabilization provided by the extra (thio)urea moieties is thought to be the source of the augmented activity (versus mono-(thio)urea donors).10,12 The active catalytic forms – H-bonding and imidate – are in rapid equilibrium, Figure 7.1, unless a strong inorganic base (e.g. alkoxide or hydride) is applied, in which case the imidate is the catalytic species.7,9,10 The active catalytic forms – H-bonding and imidate – are in rapid equilibrium, Figure 7.1, unless a strong inorganic base (e.g. alkoxide or hydride) is applied, in which case the imidate is the catalytic species.7,9,10 One advantage of the H-bonding class of catalysts is their efficacy for room temperature ROP,1 but some applications mandate utilizing catalysts at elevated temperatures. For example, the solvent-free ROP of lactide (LA) had been identified as a challenge for organocatalysts.13 The high temperature required to melt the polymer (180°C) typically 266 results in charring when organic catalysts are applied.13 We recently obviated this problem by conducting the solvent-free ROP of LA in the monomer melt (ROP at 100°C) using a bisthiourea H-bond donor (2-S, Figure 7.1) and a commercially available alkylamine base.11 Despite the reaction ‘solution’ solidifying prior to full conversion, the ROP exhibited characteristics of a ‘living’ polymerization – linear evolution of molecular weight versus conversion and Mn predictable by [M]o/[I]o.11 Similar behavior was observed for the ROPs of valerolactone (VL, m.p = -13°C) and caprolactone (CL, m.p. = -1°C), whose solvent free-ROP can be conducted at room temperature. INTRODUCTION However, one-pot chain extension under solvent-free conditions proved difficult unless the A block polymer was melted to achieve homogeneity with the B block monomer. Surprisingly, in neither of these high temperature applications did we observe catalyst deactivation, which we had expected to happen.11 H-bonds are known to weaken at elevated temperatures, charring has been observed in the solvent-free ROP of LA, and, certainly, other organocatalytic reactions have been observed to suffer deactivation at elevated temperatures.14 Contrary to catalyst deactivation, the touchstone study of H-bond mediated ROP showed linear Arrhenius behavior for the 1-S/MTBD cocatalyzed ROP of VL up to 50°C;15 however, we sought to examine the thermal stability of a host of H-bonding catalysts far above room temperature. 267 EXPERIMENTAL SECTION General Considerations. All manipulations were performed in an INERT stainless-steel glovebox equipped with a N2 gas purification system. All chemicals were purchased from Fisher Scientific and used as received unless stated otherwise. Tetrahydrofuran and dichloromethane were dried on an Innovative Technologies solvent purification system with alumina columns and N2 working gas. Benzene-d6, chloroform-d and toluene-d8 were purchased from Cambridge Isotope Laboratories and distilled from CaH2. δ-valerolactone (VL; 99%), ε-caprolactone (CL; 99%) and benzyl alcohol were distilled from CaH2 under reduced pressure. L-lactide was purchased from Acros Organics and recrystallized from dry toluene. The H-bond donors 1-S, 1-O, 2-S, 2-O, 3-S and 3-O were prepared according to published procedures.10,12 TCC was purchased from Tokyo Chemical Company and used as received. NMR experiments were performed on Bruker Avance III 300 MHz or 400 MHz spectrometer. Size exclusion chromatography (SEC) was performed at 40°C using dichloromethane eluent on an Agilent Infinity GPC system equipped with three Agilent PLGel columns 7.5 mm × 300 mm (5 μm, pore sizes: 50, 103, 104 Å). Mn and Mw/Mn were determined versus PS standards (162 g/mol-526 kg/mol, Polymer Laboratories). Example CL Polymerization Experiment. A 7 mL vial was charged with 1-S (16.2 mg, 0.0438 mmol,), MTBD (6.7mg, 0.0438 mmol), benzyl alcohol (0.95 mg, 0.00876 mmol) and toluene (200 μL). In a second 7 mL vial, CL (100 mg, 0.876 mmol) was dissolved in toluene (238 μL). The contents of the second vial were transferred to the first via pipette and stirred until homogenous, approx. 1 min. The vial was capped with a septa cap and placed in an oil bath heated to 80°C. 50μL aliquots were taken with a fine-needle syringe. 268 The aliquots were quenched using benzoic acid (10.7 mg, 0.0874 mmol) and removed of volatiles under vacuum. The contents were dissolved in CDCl3, transferred to an NMR tube via pipette, and monomer conversion was determined by 1H NMR. Eyring plots were constructed using the observed first order rate constants (kobs) from the equations below. Yield: 89%, Mn = 7,500, Mw/Mn = 1.07. The aliquots were quenched using benzoic acid (10.7 mg, 0.0874 mmol) and removed of volatiles under vacuum. The contents were dissolved in CDCl3, transferred to an NMR tube via pipette, and monomer conversion was determined by 1H NMR. Eyring plots were constructed using the observed first order rate constants (kobs) from the equations below. Yield: 89%, Mn = 7,500, Mw/Mn = 1.07. EXPERIMENTAL SECTION (1) [ ] ( ) kobs = kp [benzyl alcohol]o [cocatalysts]o (2) Example ROP of lactide. L-lactide (126.26 mg, 0.876 mmol) and methyl isobutyl ketone (438 μL) were added into a 7 mL vial and stirred until a homogenous solution was obtained. To a second 7 mL vial, benzyl alcohol (0.95 mg, 0.00876 mmol), PMDETA (3.8 mg, 0.0219 mmol) and 2-S (13.5 mg, 0.0219 mmol) were added. Contents from the first vial were transferred into vial 2 via Pasteur pipette. The vial was capped with a septa cap and placed in an oil bath heated to 80°C. 50μL aliquots were taken with a fine-needle syringe periodically. The aliquots were quenched using benzoic acid (5.3 mg, 0.0438 mmol) and placed in the vacuum oven to remove solvent. The contents were transferred to an NMR tube via pipette, and monomer conversion was determined by 1H NMR. kobs = kp [benzyl alcohol]o [cocatalysts]o (2) (2) Example ROP of lactide under solvent free conditions. L-lactide (200 mg, 1.38 mmol), 2- S (4.3 mg, 0.007 mmol) and benzyl alcohol (1.3 mg, 0.014 mmol) were added into a 7 mL vial and the contents were stirred at 110oC. PMDETA (5.0 µL, 0.007 mmol) from a 1.4 M stock solution of PMDETA in toluene was added to the completely melted content in vial 1. Aliquots were taken periodically, quenched using benzoic acid (1.6 mg, 0.013 mmol) and analyzed by 1H NMR to determine monomer conversion. Yield 90 %, Mn = 18,000 g/mol, Mw/Mn = 1.10. 269 RESULTS AND DISCUSSION The thermal stabilities of various H-bond donors plus MTBD were determined under ROP conditions in solution. The observed rate constant (kobs) for the first order evolution of [CL] ([CL]o = 2M, 0.876 mmol) were measured for the H-bond donor/MTBD cocatalyzed ROP from benzyl alcohol (0.00876 mmol) in toluene at several temperatures from 22°C to 110°C, Table 7.1, and an Eyring plot was constructed for each cocatalyst system. These ROP have previously been shown to be first order in [cocatalysts]o and [initiator]o, resulting in the rate equation: Rate = kobs[CL], kobs = kp[cocatalysts]o[initiator]o.10,15–17 The activation parameters of enchainment are superimposed with those of catalyst dynamics/reagent binding,9,17 and these observed activation parameters (DH≠obs and DS≠obs) are also given in Table 7.1.18 The concentration of H-bond donating moiety was held constant between runs. The ROP of CL was chosen because the slower reaction kinetics (versus VL or lactide) facilitate monitoring by aliquot or 1H NMR, and the ROP of CL features a high ceiling temperature (Tceil = 261 °C)19, which suggests that any temperature dependent observations are not due to substantially diminished enchainment equilibrium constants. For 1-O and 2-O, the Eyring plots are linear over the entire temperature window whereas all thiourea H-bond donors, in addition to TCC and 3-O, exhibit curved Eyring plots with a maximum rate achieved at 80°C, Figure 7.2. In the case of 2-S, the reaction progress stops completely after 30 minutes (42 % conversion) at 110°C. For the H-bond donors that exhibit a non-linear Eyring plot, the reduction in rate at For the H-bond donors that exhibit a non-linear Eyring plot, the reduction in rate at high temperature is due to catalyst decomposition. A temperature-jump experiment was conducted for the 1-S/MTBD (0.0438 mmol) cocatalyzed ROP of CL (2 M, 0.876 mmol) 270 from benzyl alcohol (0.00876 mmol) in toluene, where after the first part of the reaction was conducted at 110°C, the temperature was changed to 80°C resulting in a reduced rate for the latter portion of the ROP, Figure 7.6. The drop in rate suggests that the catalyst deactivation observed above 80°C (Eyring plot) is not reversible and is due to decomposition. Indeed, 1H-NMR analysis after a 1-S/MTBD cocatalyzed ROP in C6D6 (2 h at 110°C) reveals additional aromatic resonances presumably originating from the decomposition of 1-S. RESULTS AND DISCUSSION Decomposition is also observed when 1-S and MTBD (0.1M each, C6D6) are heated in the absence of monomer and initiator, but 1-S is stable to 110°C when heated alone. A non-linear Eyring plot can be associated with an abrupt change in mechanism.18 However, for (thio)urea/MTBD mediated ROP, the H-bonding and imidate mechanisms are in equilibrium,7,9 which would be expected to yield a linear Eyring plot.18 Thermogravimetric analysis (TGA, under nitrogen) of H-bond donors in the presence and absence of MTBD cocatalyst show that catalyst stability must be determined under reaction conditions. Decomposition onsets for the H-bond donors in the absence of MTBD ranged from 166 - 256°C (see Table 7.2). In the presence of MTBD, a decrease in decomposition onset temperature was only seen for TCC, 1-O and 2-O (decreases of 85, 61 and 94°C, respectively). These temperatures do not correlate to the decomposition evidenced in the Eyring plots. Further, diminished thermal stability in the presence of MTBD does not indicate which catalysts are most thermally stable under reaction conditions. For example, the thioureas exhibit curved Eyring plots yet show minimal change in thermal stability with and without MTBD, and 1-O and 2-O exhibit decreased onsets in the TGA with MTBD but display linear Eyring behavior. 271 Eyring plots for both TCC and 1-S become linear when the ROP is conducted in polar solvent, suggesting that the catalysts are more thermally stable under reaction conditions in polar solvent up to 110°C. The H-bonding catalysts for ROP exhibit a mechanistic duality that is strongly dictated by solvent, where polar solvent favors imidate mediated ROP and non-polar solvent favors H-bond mediated enchainment, suggesting the polar solvent stabilizes the imidate species.7,9 In this study, the more acidic H-bond donors exhibit curved Eyring plots in toluene.5,20,21 This suggests that catalyst decomposition may be preceded by proton transfer to form, presumably, (thio)imidate, which decomposes in toluene. The extent of imidate formation is also dictated by the pKa of the base cocatalyst. When the ROP of CL from benzyl alcohol is catalyzed by TCC/DBU, an attenuation of catalyst decomposition is observed (Figure 7.15) versus the TCC/MTBD ROP (Figure 7.10). This suggests that the augmented pKa of MTBD (MTBD-H+ pKaMeCN = 25.4)22 versus that of DBU (DBU-H+ pKaMeCN = 24.3)22 favors the formation of imidate, as previously observed,7 which becomes unstable at elevated temperatures. RESULTS AND DISCUSSION Conducting the ROPs in polar solvent does not yield linear Eyring data for all catalysts systems. We employed methyl isobutyl ketone (MIBK) as the solvent for our temperature dependent studies; acetone and THF have previously been used for (thio)imidate-mediated ROP but are not high boiling.5,7 In MIBK, TCC, 1-S, and 1-O plus MTBD cocatalyzed ROPs exhibit linear Eyring plots up to 110°C, Figure 7.3 , suggesting enhanced catalyst stability in polar solvent. TCC or 1-S plus MTBD displayed non-linear Eyring behavior in toluene. However, the first order consumption of monomer plots for the 1-S, 2-S or 3-O plus MTBD cocatalyzed ROPs of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) at 110°C deviate from linearity, which suggests some catalyst decomposition. Except for 272 the 1-S, 2-S or 3-O plus MTBD cocatalyzed ROPs of CL at 110°C, the ROP display linear first order consumption of monomer to >75% conversion. The 2-O, 2-S and 3-O plus MTBD cocatalyzed ROP of CL in MIBK do not exhibit linear Eyring, suggesting decomposition is still occurring on the timescale of the ROP despite the stabilization provided by polar solvent to the imidate species. Reaction Mechanism. Temperature dependent 1H NMR studies corroborate an H- bonding mechanism for H-bond donors in non-polar solvent and an imidate mechanism for H-bond donors in polar solvent. 1H NMR spectra of a 1-S/MTBD (0.1 M) solution in C6D6 shows that the 1-S resonances are shifted downfield in the presence of MTBD, which suggests H-bonding.17 With increasing temperature (<110°C), the 1-S 1H NMR resonances (when mixed with MTBD) shift upfield, indicating an exothermic binding (DHo = - 10.7±2.0 kcal/mol)17 and weaker H-bonding at higher temperatures.17 Binding to monomer is also exothermic.15,17 These observations suggest that Arrhenius reaction acceleration outpaces the weakening of H-bonding resulting in faster ROP until catalyst decomposition >80°C. The upfield shift of 1-S resonances upon heating does not indicate the assumption of imidate character; at 110°C (the highest temperature recorded), the 1-S resonances (with MTBD) remain downfield of the 1-S resonances in the absence of MTBD. The same trend is observed for 1-O plus MTBD in C6D6 upon heating. This suggests that H-bonding remains the dominate mechanistic form for (thio)urea H-bond donors with MTBD in C6D6 at all temperatures. RESULTS AND DISCUSSION In contrast, 1-O plus MTBD in acetone-d6 show 1-O resonances with an upfield chemical shift versus 1-O in the absence of MTBD, and this chemical shift is not a function of temperature (25-60°C, Δppm < 0.1). The relatively static chemical shift 273 suggests that the extent of imidate formation is not highly sensitive to temperature and that higher rates of ROP are thermal in nature (i.e. high rates are not due to higher [imidate]). Lower activation barriers in polar solvent do not translate to faster ROP. The thiourea H-bond donors experience deactivation in polar solvent versus non-polar solvent whereas urea H-bond donors experience minimal change in rate between the two solvents or, in the case of 1-O/MTBD and particularly 3-O/MTBD, exhibit faster rates in polar solvent. This behavior has been described elsewhere.7 Oddly, the DH≠obs in toluene are generally larger than in MIBK. This suggests that despite the generally lower rates in MIBK, the activation barrier for enchainment in MIBK is lower than in toluene. One explanation for lower DH≠obs in polar (versus non-polar) solvent, associated with lower DS≠obs, is the involvement of solvent in the rate determining step.18 An alternate explanation that must be considered is that the small magnitude of the drop in activation entropy (from toluene to MIBK) is simply due to the tighter association of the cocatalyst ion-pair in an imidate mechanism. Despite the high activity of imidate mediated ROP, the reactions remain highly controlled at room temperature, and the extent of imidate formation appears to be highly sensitive to solvent polarity including the polarity of the evolving reaction solution. RESULTS AND DISCUSSION We have previously shown that the highly active imidate forms of H-bond donors are in equilibrium with the less active neutral species,7 and a larger upfield shift (more imidate character) can be observed with stronger organic bases upon a single H-bond donor or with a more acidic H-bond donor with a single base.7,9 A similar effect has been reported where polar solvent favors an imidate mechanism and non-polar solvent favors an H-bonding mechanism.9 The polarity of the bulk solution is expected to drop during the course of the 274 polymerization as closed, s-cis lactone is converted to open, s-trans lactone; indeed, the higher polarity of lactones versus s-trans esters has been used to justify the selectivity exhibited by thioureas (and presumably ureas) for the activation of monomer versus polymer.15 The 1H NMR spectra of TCC plus MTBD (0.0438 mmol each) in CDCl3 were recorded at room temperature in the presence of varying amounts of CL (2 M – 0.25 M) in the absence of initiator, conditions under which no ROP is observed, Figure 7.7. The resonances of TCC move upfield in the presence of MTBD but move back downfield with reduced monomer concentration, indicating less imidate character with reduced monomer concentrations. These observations may explain why urea imidate mediated ROP can be highly active – more active than the structurally similar TBD – but be more controlled. We proposed that the catalyst system deactivates during the course of the ROP by forming less imidate character. This hypothesis will have to be borne out by future studies. ROP of Lactide. Temperature dependent kinetics suggest that 2-S/PMDETA cocatalyzed ROP of L-lactide (LA) experiences enhanced thermal stability when solvent- free (deactivation > 110°C) versus in polar solvent (deactivation > 80°C). For the (thio)urea/base mediated ROP of LA, weak base cocatalysts are optimal for ROP. Under these weakly basic conditions, only an H-bond mediated enchainment pathway is thought to occur.11,16 Our group recently disclosed that 2-S/PMDETA was optimal – in terms of rate and isotacticity – for the solvent-free ROP of LA,11 and this was the only catalyst system whose temperature dependent kinetics were examined herein. RESULTS AND DISCUSSION We have previously shown that the (thio)urea/base cocatalyzed solvent-free ROP of LA can be conducted in the monomer melt and exhibits the characteristics of a ‘living’ ROP despite solidifying prior to full conversion.11 Consistent with previous solution studies, the Eyring analysis of 275 the 2-S/PMDETA (0.0219 mmol each) cocatalyzed ROP of LA (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) showed non-linear behavior above 80°C in MIBK, Figure 7.19, suggesting catalyst deactivation arising from decomposition. Previous studies showed deactivation at high temperature of the same polymerization in CHCl3, toluene and acetone.11 Under solvent-free conditions, the Eyring plot for the 2-S/PMDETA cocatalyzed ROP of LA from benzyl alcohol indicates deactivation at elevated temperature, with a maximum observed rate at 110°C. This coincidental elevation in decomposition onset facilitates the solvent-free ROP of LA just above the monomer melting point (Tm = 97°C)13. The solvent-free ROP of LA at 110°C remains highly controlled, [M]o/[I]o = 100 yields Mn = 18,000 g/mol, Mw/Mn = 1.10. Polymer Molecular Weight and Dispersity. The extent of reaction control is reduced at elevated temperatures, but 3-O/MTBD remains among the most active and controlled organocatalysts for ROP even at high temperature. The ‘living’ character of the H-bond donor/MTBD cocatalyzed ROP at room temperature in non-polar and polar solvent as well as solvent-free has been described previously.5–7,10 For the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol in toluene, the Mw/Mn increases with temperature (at 80°C, [M]o/[I]o = 100, Mn = 20,000, Mw/Mn = 1.10; at 110°C, [M]o/[I]o = 50 Mn = 9,900, Mw/Mn = 1.17), but both of these ROP are more controlled than the comparable ROP in MIBK (MIBK: at 80°C, [M]o/[I]o = 100, Mn = 13,500, Mw/Mn = 1.24; at 110°C: [M]o/[I]o = 50, Mn = 6,200, Mw/Mn = 1.28). This trend is also observed for 2-O/MTBD mediated ROP of CL, which produces a linear Eyring plot in toluene. This suggests that the extent of reaction control is not related to the thermal stability of a cocatalyst system. For the catalyst systems examined, the ROP retain the 276 characteristics of a ‘living’ polymerization, except for 2-O/MTBD at 110°C, whose Mn versus conversion becomes non-linear late in the reaction (Figure 7.49). Some reduction in control undoubtedly arises from non-H-bond donor mediated ROP. RESULTS AND DISCUSSION At >80°C, MTBD (0.044 mmol) effects the slow ROP of CL (2 M, 0.876 mmol) in the absence of H-bond donor to 10% conversion in 6 h, after which point conversion stalls. The MTBD mediated ROP of lactones (no H-bond donor) has been shown to exhibit less control than those in the presence of an H-bond donor.23 These observations suggest that, to some extent, reaction control (narrow Mw/Mn and predictable Mn) can be maximized if the ROP reaches full conversion before catalyst decomposition or other side reactions become prominent. The Mn and Mw/Mn produced by the several cocatalyst systems at 80°C are shown in Table 7.3. In contrast, the 3-O/MTBD cocatalyzed ROP of CL – which is one of the most active organocatalytic systems for ROP – exhibits a curved Eyring plot in toluene (and MIBK) but shows excellent reaction control even at 110°C, Figure 7.5 (toluene, 95% conversion, 60 min, Mn = 18,900 g/mol, Mw/Mn = 1.05). At a lower temperature, the reaction is more active and possibly more controlled (80°C, 99% conversion, 60 min, Mn = 21,900 g/mol, Mw/Mn = 1.03). 1H NMR analysis of 3-O plus MTBD (33 mM each) in acetone-d6 at 60oC, a temperature at which non-Arrhenius behavior is observed, provides no evidence of decomposition over 5 h. Rather, we attribute the non-linearity in the 3- O/MTBD cocatalyzed Eyring analysis to H-bond donor dynamics where the number of urea moieties engaged in catalysis, and the nature of their interactions with substrate, may be a function of temperature. 277 CONCLUSION For those seeking to employ H-bond donating catalysts for ROP, 3-O/MTBD may be the most reliable system; it exhibits fast rates, excellent control and good thermal stability with insensitivity to reaction conditions. The thermal stability of H-bond donating organocatalyst systems for ROP must be determined under reaction conditions. Except 2- S or 3-O plus MTBD in MIBK, all cocatalysts examined are thermally stable ≤80°C in polar and non-polar solvent, but conducting ROP in polar solvent can extend thermal stability to 110°C for some systems. However, reaction control – as measured by Mw/Mn – deteriorates at elevated temperatures, especially in polar solvent. The one exception appears to be 3-O/MTBD, which produces a highly controlled and rapid ROP at all temperatures in non-polar solvents and solvent-free. The 3-O/MTBD system is particularly intriguing because it produces the fastest rates and best control despite exhibiting a non-linear Eyring plot. The diversity of conditions under which the H-bonding catalysts for ROP are operable is attributable to their mechanistic duality where non-polar solvent and weakly acidic H-bond donors favor an H-bonding mechanism, and polar solvent and more acidic H-bond donors favor an imidate mechanism. In toluene (H-bonding favored), high temperatures (≥80°C) may facilitate proton transfer to form the (thio)imidate species, which is unstable and results in catalyst decomposition. In polar solvent, [imidate] is insensitive to reaction temperature. The H-bonding class of organocatalysts for ROP constitutes a platform for macromolecule synthesis that is operable under diverse experimental conditions. Indeed, the very progress of the ROP changes the effective [imidate] by modulating the polarity of 278 the reaction solution. The robustness of these catalysts belies the complicated nature of their discharge, our understanding of which and our concomitant synthetic abilities continue to evolve through targeted study. 279 (8) Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Fast and Selective Ring- Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Fast and Selective Ring LIST OF REFERENCES (1) Kamber, N. E.; Jeong, W.; Waymouth, R. M.; Pratt, R. C.; Lohmeijer, B. G. G.; Hedrick, J. L. Organocatalytic Ring-Opening Polymerization. Chem. Rev. 2007, 107 (12), 5813–5840. (2) Kiesewetter, M. K.; Shin, E. J.; Hedrick, J. L.; Waymouth, R. M. Organocatalysis: Opportunities and Challenges for Polymer Synthesis. Macromolecules 2010, 43 (5), 2093–2107. (3) Dove, A. P. Metal-Free Catalysis in Ring-Opening Polymerization. In Handbook of Ring-Opening Polymerization; Dubois, P., Coulembier, O., Raquez, J.-M., Eds.; Wiley-VCH: Federal Republic of Germany, 2009; pp 357–378. (4) Thomas, C.; Bibal, B. Hydrogen-Bonding Organocatalysts for Ring-Opening Polymerization. Green Chem. 2014, 16 (4), 1687. (5) Lin, B.; Waymouth, R. M. Organic Ring-Opening Polymerization Catalysts: Reactivity Control by Balancing Acidity. Macromolecules 2018, 51 (8), 2932– 2938. (6) Lin, B.; Waymouth, R. M. Urea Anions: Simple, Fast, and Selective Catalysts for Ring-Opening Polymerizations. J. Am. Chem. Soc. 2017, 139, 1645–1652. (7) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, M. K. Triclocarban: Commercial Antibacterial and Highly Effective H-Bond Donating Catalyst for Ring-Opening Polymerization. ACS Macro Lett. 2017, 6 (4), 421–425. (8) Zhang, X.; Jones, G. O.; Hedrick, J. L.; Waymouth, R. M. Fast and Selective Ring- (8) 280 Opening Polymerizations by Alkoxides and Thioureas. Nat. Chem. 2016, 8, 1047– 1053. (9) Pothupitiya, J. U.; Hewawasam, R. S.; Kiesewetter, M. K. Urea and Thiourea H- Bond Donating Catalysts for Ring-Opening Polymerization: Mechanistic Insights via (Non)Linear Free Energy Relationships. Macromolecules 2018, No. 51, 3203– 3211. (9) (10) Fastnacht, K. V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; Kiesewetter, E. T.; Kiesewetter, M. K. Bis- and Tris-Urea H-Bond Donors for Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst. ACS Macro Lett. 2016, 5 (8), 982–986. (11) Pothupitiya, J. U.; Dharmaratne, N. U.; Jouaneh, T. M. M.; Fastnacht, K. V.; Coderre, D. N.; Kiesewetter, M. K. H-Bonding Organocatalysts for the Living, Solvent-Free Ring-Opening Polymerization of Lactones: Toward an All-Lactones, All-Conditions Approach. Macromolecules 2017, 50, 8948–8954. All-Conditions Approach. Macromolecules 2017, 50, 8948–8954. (12) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. T.; Kiesewetter, M. K. Rate Accelerated Organocatalytic Ring-Opening Polymerization of l-Lactide via the Application of a Bis(Thiourea) H-Bond Donating Cocatalyst. Macromolecules 2015, 48 (17), 6127–6131. (13) Mezzasalma, L.; Dove, A. P.; Coulembier, O. Organocatalytic Ring-Opening Polymerization of L-Lactide in Bulk: A Long Standing Challenge. Eur. Polym. J. 2017, 95 (May), 628–634. (14) Dudziński, K.; Pakulska, A. M.; Kwiatkowski, P. LIST OF REFERENCES An Efficient Organocatalytic Method for Highly Enantioselective Michael Addition of Malonates to Enones 281 Catalyzed by Readily Accessible Primary Amine-Thiourea. Org. Lett. 2012, 14 (16), 4222–4225. (15) Lohmeijer, B. G. G.; Pratt, R. C.; Leibfarth, F.; Logan, J. W.; Long, D. A.; Dove, A. P.; Nederberg, F.; Choi, J.; Wade, C.; Waymouth, R. M.; et al. Guanidine and Amidine Organocatalysts for Ring-Opening Polymerization of Cyclic Esters. Macromolecules 2006, 39 (25), 8574–8583. (16) Kazakov, O. I.; Kiesewetter, M. K. Cocatalyst Binding Effects in Organocatalytic Ring-Opening Polymerization of l -Lactide. Macromolecules 2015, 48 (17), 6121– 6126. (17) Kazakov, O. I.; Datta, P. P.; Isajani, M.; Kiesewetter, E. T.; Kiesewetter, M. K. Cooperative Hydrogen-Bond Pairing in Organocatalytic Ring-Opening Polymerization. Macromolecules 2014, 47 (21), 7463–7468. (18) Espenson, J. H. Chemical Kinetics and Reaction Mechanism, 2nd ed.; McGraw- Hill Book Co: New York, 2002. (19) Duda, A.; Kowalski, A. Thermodynamics and Kinetics of Ring-Opening Polymerization. In Handbook of Ring-Opening Polymerization; Dubois, P., Coulembier, O., Raquez, J.-M., Eds.; Wiley-VCH Verlag GmbH & Co. KGaA, 2009; pp 1–52. (20) Bordwell, F. G.; Ji, G. Z. Effects of Structural Changes on Acidities and Homolytic Bond Dissociation Energies of the H-N Bonds in Amidines, Carboxamides, and Thiocarboxamides. J. Am. Chem. Soc. 1991, 113 (22), 8398– 8401. (21) The pKa of TCC and 3-O are not known, but we presume higher acidity of TCC (21) The pKa of TCC and 3-O are not known, but we presume higher acidity of TCC 282 versus 1-O due to Hammett effects. The internal H-bond stabilization is expected to render 3-O more acidic than the parent mono-urea. versus 1-O due to Hammett effects. The internal H-bond stabilization is expected to render 3-O more acidic than the parent mono-urea. (22) Kaljurand, I.; Kütt, A.; Sooväli, L.; Rodima, T.; Mäemets, V.; Leito, I.; Koppel, I. A. Extension of the Self-Consistent Spectrophotometric Basicity Scale in Acetonitrile to a Full Span of 28 PKa Units: Unification of Different Basicity Scales. J. Org. Chem. 2005, 70 (3), 1019–1028. (23) Datta, P. P.; Kiesewetter, M. K. Controlled Organocatalytic Ring-Opening Polymerization of ε-Thionocaprolactone. Macromolecules 2016, 49 (3), 774–780. LIST OF REFERENCES 283 toluene methyl isobutyl ketone Donor DH≠obs (kcal/mol) DS≠obs (cal/mol K) kobs (1/min)a DH≠obs (kcal/mol) DS≠obs (cal/mol K) kobs (1/min)a 1-Sb 6.13 ± 0.40 -52 ± 46 0.0018 5.22 ± 0.77 -56 ± 45 0.0008 2-Sb,c 6.21 ± 0.66 -46 ± 45 0.0065 4.24 ± 0.74 -57 ± 45 0.0027 3-Sb 8.59 ± 0.53 -46 ± 46 0.0006 -- -- -- 1-O 5.66 ± 0.54 -42 ± 46 0.0015 3.53 ± 0.26 -59 ± 46 0.0023 2-Oc 4.00 ± 0.59 -55 ± 45 0.0074 3.3 ± 0.26 -58 ± 46 0.0076 3-Ob,d 2.17 ± 0.18 -58 ± 47 0.0323 4.59 ± 1.27 -48 ± 43 0.1134 TCCb 5.32 ± 0.95 -50 ± 44 0.0121 5.37 ± 0.49 -50 ± 46 0.0115 Table 7.1. Activation Parameters for H-bond Donor/MTBD Cocatalyzed ROP of CL. Reaction conditions: CL (2M, 0.876 mmol), benzyl alcohol (0.00876 mmol), 1-X, 2-X, 3- X and TCC (0.0438 mmol, 0.0219 mmol, 0.0146 mmol and 0.0438 mmol) MTBD toluene methyl isobutyl ketone Donor DH≠obs (kcal/mol) DS≠obs (cal/mol K) kobs (1/min)a DH≠obs (kcal/mol) DS≠obs (cal/mol K) kobs (1/min)a 1-Sb 6.13 ± 0.40 -52 ± 46 0.0018 5.22 ± 0.77 -56 ± 45 0.0008 2-Sb,c 6.21 ± 0.66 -46 ± 45 0.0065 4.24 ± 0.74 -57 ± 45 0.0027 3-Sb 8.59 ± 0.53 -46 ± 46 0.0006 -- -- -- 1-O 5.66 ± 0.54 -42 ± 46 0.0015 3.53 ± 0.26 -59 ± 46 0.0023 2-Oc 4.00 ± 0.59 -55 ± 45 0.0074 3.3 ± 0.26 -58 ± 46 0.0076 3-Ob,d 2.17 ± 0.18 -58 ± 47 0.0323 4.59 ± 1.27 -48 ± 43 0.1134 TCCb 5.32 ± 0.95 -50 ± 44 0.0121 5.37 ± 0.49 -50 ± 46 0.0115 methyl isobutyl ketone Table 7.1. Activation Parameters for H-bond Donor/MTBD Cocatalyzed ROP of CL. Reaction conditions: CL (2M, 0.876 mmol), benzyl alcohol (0.00876 mmol), 1-X, 2-X, 3- X and TCC (0.0438 mmol, 0.0219 mmol, 0.0146 mmol and 0.0438 mmol) MTBD (matched to H-bond donor mmol). a) 40°C. b) Non-linear Eyring plot in toluene; activation parameters extracted from the linear portion. c) Non-linear Eyring plot in MIBK, activation parameters extracted from the linear portion. d) The rate of ROP of CL catalyzed by 3-O/MTBD in MIBK is significantly faster than in toluene even though the Eyring plot is non-linear. LIST OF REFERENCES 284 catalyst decomposition temperature onset (°C) decomposition temperature onset w/ MTBD (°C) 1-S 166 168 1-O 212 151 TCC 247 162 2-S 172 NA 2-O 256 162 3-S 187 196 catalyst decomposition temperature onset (°C) decomposition temperature onset w/ MTBD (°C) 1-S 166 168 1-O 212 151 TCC 247 162 2-S 172 NA 2-O 256 162 3-S 187 196 Table 7.2. Thermal Decomposition of H-bond Donors With and Without MTBD. TGA conditions: Sample size ~7 mg were placed in aluminum pans. Start at 25°C, ramp to 500°C at 10°C/min, held for 5 minutes. N2 gas flow rate of 10 mL/min. Catalyst samples were weighed from bulk samples. H-bond donor/base 1:1 mixtures were first dissolved in dichloromethane and solvent removed under vacuum. Remaining thick oil was used for samples. 285 285 Toluene MIBK Time (min) Conv.a Mn b Mw/Mn b Time (min) Conv.a Mn b Mw/Mn b 1-S 360 90 19,200 1.07 950 91 13,500 1.24 1-O 335 89 18,000 1.14 880 98 10,700 1.40 2-Sc 170 91 14,000 1.17 1600 83 9,600 1.17 2-O 120 89 11,500 1.19 260 91 8,800 1.24 3-S 960 89 12,600 1.08 4,900 41 4,200 1.13 3-O 40 90 19,000 1.04 15 91 14,300 1.10 TCC 61 90 16,800 1.17 65 94 11,100 1.28 Table 7.3. Mn and Mw/Mn for polymerizations conducted at 80°C. Reaction conditions: CL (2M, 0.876 mmol), benzyl alcohol (0.00876 mmol), 1-X, 2-X, 3X and TCC (0.0438 mmol 0 0219 mmol 0 0146 mmol and 0 0438 mmol) MTBD (matched to H bond donor Toluene MIBK Time (min) Conv.a Mn b Mw/Mn b Time (min) Conv.a Mn b Mw/Mn b 1-S 360 90 19,200 1.07 950 91 13,500 1.24 1-O 335 89 18,000 1.14 880 98 10,700 1.40 2-Sc 170 91 14,000 1.17 1600 83 9,600 1.17 2-O 120 89 11,500 1.19 260 91 8,800 1.24 3-S 960 89 12,600 1.08 4,900 41 4,200 1.13 3-O 40 90 19,000 1.04 15 91 14,300 1.10 TCC 61 90 16,800 1.17 65 94 11,100 1.28 Table 7.3. Mn and Mw/Mn for polymerizations conducted at 80°C. Reaction conditions: CL (2M, 0.876 mmol), benzyl alcohol (0.00876 mmol), 1-X, 2-X, 3X and TCC (0.0438 mmol, 0.0219 mmol, 0.0146 mmol and 0.0438 mmol) MTBD (matched to H-bond donor mmol). a. Conversion determined by 1H NMR. b. Mn and Mw/Mn determined by GPC. c. 2-S in MIBK at 60°C. 286 Figure 7.1. H-bonding and imidate mediated ROP of CL. LIST OF REFERENCES N H N H CF3 F3C X N H N H CF3 F3C X N H N H CF3 CF3 X N N H N H X N H N H X HN HN X CF3 CF3 CF3 F3C CF3 CF3 1-S: X = S 1-O: X = O 2-S: X = S 2-O: X = O N H N H O Cl Cl Cl N N N N N N MTBD PMDETA 3-S: X = S 3-O: X = O TCC N N DBU Figure 7.1. H-bonding and imidate mediated ROP of CL. 287 287 Figure 7.2. Eyring plots for the ROP of CL from benzyl alcohol in toluene catalyzed by (upper) TCC/MTBD and (lower) 2 O/MTBD Reaction Conditions: CL (2M 0 876 -10.8 -10.4 -10 -9.6 -9.2 -8.8 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 ln(k/T) 1/T (1/K) y = -2011.6x - 4.1662 -11.2 -10.8 -10.4 -10 -9.6 -9.2 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 ln(k/T) 1/T (1/K) 8 8 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 y = -2011.6x - 4.1662 -11.2 -10.8 -10.4 -10 -9.6 -9.2 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 ln(k/T) 1/T (1/K) Figure 7.2. Eyring plots for the ROP of CL from benzyl alcohol in toluene catalyzed by (upper) TCC/MTBD, and (lower) 2-O/MTBD. Reaction Conditions: CL (2M, 0.876 mmol), H-bond donor/MTBD (TCC 0.0438 mmol, 2-O 0.0219 mmol, [MTBD]o matched to H-bond donor) from benzyl alcohol (0.00876 mmol) in toluene. 288 -11 -10.5 -10 -9.5 -9 -8.5 -8 0.0025 0.003 0.0035 ln(k/T) 1/T (1/K) Toluene MIBK -11 -10.5 -10 -9.5 -9 -8.5 -8 0.0025 0.003 0.0035 ln(k/T) 1/T (1/K) Toluene MIBK Figure 7.3. Eyring plot for the TCC/MTBD (0.0438 mmol) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene and MIBK solvent. Figure 7.4. Eyring plot for the solvent-free ROP of LA (1.380 mmol) from benzyl alcohol (0.014 mmol) catalyzed by 2-S/PMDETA (0.007 mmol each). -8.7 -8.6 -8.5 -8.4 -8.3 -8.2 -8.1 0.0024 0.0025 0.0026 0.0027 ln(k/T) 1/T (1/K) Figure 7.4. Eyring plot for the solvent-free ROP of LA (1.380 mmol) from benzyl alcohol (0.014 mmol) catalyzed by 2-S/PMDETA (0.007 mmol each). 289 Figure 7.5. Mn and Mw/Mn versus conversion plot for the 3-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C. LIST OF REFERENCES 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 0 0.5 1 Mw/Mn Mn (g/mol) fractional conversion 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 18,000 20,000 0 0.5 1 Mw/Mn Mn (g/mol) fractional conversion Figure 7.5. Mn and Mw/Mn versus conversion plot for the 3-O/MTBD cocatalyzed ROP of CL from benzyl alcohol in toluene at 110°C. 290 Figure 7.6. (upper) Temperature drop ROP of CL (2M, 0.876 mmol) in toluene, 1-S/MTBD y = 0.0038x + 0.2028 y = 0.0013x + 0.7814 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 100 200 300 400 500 ln([M]o/[M]) Time (min) Figure 7 6 (upper) Temperature drop ROP of CL (2M 0 876 mmol) in toluene 1 y = 0.0038x + 0.2028 y = 0.0013x + 0.7814 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 100 200 300 400 500 ln([M]o/[M]) Time (min) y = 0.0038x + 0.2028 y = 0.0013x + 0.7814 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 100 200 300 400 500 ln([M]o/[M]) Time (min) y = 0.0013x + 0.7814 Time (min) Figure 7.6. (upper) Temperature drop ROP of CL (2M, 0.876 mmol) in toluene, 1-S/MTBD Figure 7.6. (upper) Temperature drop ROP of CL (2M, 0.876 mmol) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). A decrease in kobs from 0.0038 min-1 to 0.0013 min-1 is observed after the temperature change from 110°C (red) to 80°C (blue). Reaction progression tracked by aliquot, and conversion was determined by 1H NMR. (lower) 1H NMR (400 MHz, benzene-d6) of the ROP of CL (0.876 mmol, 2M), benzyl alcohol (0.00876 mmol), 1-S/MTBD (0.0438 mmol each). Top image shows the aromatic region of the 1H NMR of 1-S. Bottom image shows a second set of aromatic resonances after 2 h of heating at 110°C. Figure 7.6. (upper) Temperature drop ROP of CL (2M, 0.876 mmol) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). A decrease in kobs from 0.0038 min-1 to 0.0013 min-1 is observed after the temperature change from 110°C (red) to 80°C (blue). Reaction progression tracked by aliquot, and conversion was determined by 1H NMR. LIST OF REFERENCES (lower) 1H NMR (400 MHz, benzene-d6) of the ROP of CL (0.876 mmol, 2M), benzyl alcohol (0.00876 mmol), 1-S/MTBD (0.0438 mmol each). Top image shows the aromatic region of the 1H NMR of 1-S. Bottom image shows a second set of aromatic resonances after 2 h of heating at 110°C. (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). A decrease in kobs from 0.0038 min-1 to 0.0013 min-1 is observed after the temperature change from 110°C (red) to 80°C (blue). Reaction progression tracked by aliquot, and conversion was determined by 1H NMR. (lower) 1H NMR (400 MHz, benzene-d6) of the ROP of CL (0.876 mmol, 2M), benzyl alcohol (0.00876 mmol), 1-S/MTBD (0.0438 mmol each). Top image shows the aromatic region of the 1H NMR of 1-S. Bottom image shows a second set of aromatic resonances after 2 h of heating at 110°C. 291 Figure 7.7. 1H NMR spectra of TCC/MTBD (0.00438 mmol each) cocatalyst in the presence of varying [CL] (0.25-2M) in CDCl3. Figure 7.7. 1H NMR spectra of TCC/MTBD (0.00438 mmol each) cocatalyst in the presence of varying [CL] (0.25-2M) in CDCl3. 292 EYRING DATA IN TOLUENE y = -3124.3x - 0.7957 -11.6 -11.4 -11.2 -11 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 -9.4 0.0028 0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 ln(k/T) 1/T(1/K) 9 4 0.0028 0.0029 0.003 0.0031 0.0032 0.0033 0.0034 0.0035 Figure 7.8. Eyring plot constructed from the linear portion of the observed first order rate constants for the 2-S/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.9. Eyring plot constructed from the observed first order rate constants for the 1- O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. y = -2280.8x - 4.8548 -12.7 -12.5 -12.3 -12.1 -11.9 -11.7 -11.5 -11.3 -11.1 -10.9 -10.7 -10.5 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T(1/K) y = -2280.8x - 4.8548 -12.7 -12.5 -12.3 -12.1 -11.9 -11.7 -11.5 -11.3 -11.1 -10.9 -10.7 -10.5 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T(1/K) Figure 7.9. Eyring plot constructed from the observed first order rate constants for the 1- O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.9. Eyring plot constructed from the observed first order rate constants for the 1- O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. 293 8 8 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 -9.4 -9.2 -9 -8.8 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 ln(k/T) 1/T (1/K) Figure 7.10. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.10. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. -13 -12.8 -12.6 -12.4 -12.2 -12 -11.8 -11.6 -11.4 -11.2 -11 -10.8 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.11. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.11. EYRING DATA IN TOLUENE Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. 294 10 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 -14 -13.6 -13.2 -12.8 -12.4 -12 -11.6 -11.2 -10.8 -10.4 -10 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.00 ln(k/T) 1/T (1/K) Figure 7.12. Eyring plot constructed from the observed first order rate constants for the 3- S/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.12. Eyring plot constructed from the observed first order rate constants for the 3- S/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.13. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. y = -2011.6x - 4.1662 -11.2 -11 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 -9.4 -9.2 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) y = -2011.6x - 4.1662 -11.2 -11 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 -9.4 -9.2 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.13. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.13. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. 295 -9.4 -9.3 -9.2 -9.1 -9 -8.9 -8.8 -8.7 -8.6 -8.5 0.0024 0.0029 0.0034 0.0039 ln(k/T) 1/T (1/K) -9.4 -9.3 -9.2 -9.1 -9 -8.9 -8.8 -8.7 -8.6 -8.5 0.0024 0.0029 0.0034 0.0039 ln(k/T) 1/T (1/K) Figure 7.14. Eyring plot constructed from the observed first order rate constants for the 3- O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.14. Eyring plot constructed from the observed first order rate constants for the 3- O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. EYRING DATA IN TOLUENE -12.2 -12 -11.8 -11.6 -11.4 -11.2 -11 -10.8 -10.6 -10.4 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.15. Eyring plot constructed from the observed first order rate constants for the TCC/DBU (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.15. Eyring plot constructed from the observed first order rate constants for the TCC/DBU (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 2M) from benzyl alcohol (0.00876 mmol) in toluene. 296 10 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 -11.4 -11.2 -11 -10.8 -10.6 -10.4 -10.2 -10 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.16. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol, 1M) from benzyl alcohol (0.00876 mmol) in toluene. Figure 7.17. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol, 1M) from benzyl alcohol (0.00876 mmol) in toluene. -11.6 -11.4 -11.2 -11 -10.8 -10.6 -10.4 -10.2 -10 -9.8 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) -11.6 -11.4 -11.2 -11 -10.8 -10.6 -10.4 -10.2 -10 -9.8 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.17. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol, 1M) from benzyl alcohol (0.00876 mmol) in toluene. 297 -13.5 -13 -12.5 -12 -11.5 -11 -10.5 -10 -9.5 -9 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) 1M CL 2M CL 3M CL -13.5 -13 -12.5 -12 -11.5 -11 -10.5 -10 -9.5 -9 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) 1M CL 2M CL 3M CL Figure 7.18. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) at 1M, 2M and 3M concentrations in toluene. Figure 7.18. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) at 1M, 2M and 3M concentrations in toluene. Figure 7.18. EYRING DATA IN TOLUENE Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) at 1M, 2M and 3M concentrations in toluene. 298 EYRING DATA IN METHYL ISOBUTYL KETONE - LACTIDE -11.8 -11.7 -11.6 -11.5 -11.4 -11.3 -11.2 -11.1 -11 0.0024 0.0025 0.0026 0.0027 0.0028 0.0029 0.003 0.0031 0.0032 ln(k/T) 1/T (1/K) EYRING DATA IN METHYL ISOBUTYL KETONE - LACTIDE Figure 7.19. Eyring plot constructed from the observed first order rate constants for the 2- S/PMDETA (0.0219 mmol each) cocatalyzed ROP of L-LA (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. 299 EYRING DATA IN METHYL ISOBUTYL KETONE - CAPROLACTONE y = -2703x - 1.3888 -10.7 -10.4 -10.1 -9.8 -9.5 -9.2 -8.9 -8.6 -8.3 -8 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) DATA IN METHYL ISOBUTYL KETONE - CAPROLACTONE 8 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 EYRING DATA IN METHYL ISOBUTYL KETONE - CAPROLACTONE 0.0024 0.0026 0.0028 0.003 0.0032 0.00 Figure 7.20. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. y = -2628.8x - 4.2906 -13.6 -13.2 -12.8 -12.4 -12 -11.6 -11.2 -10.8 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.21. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. Figure 7.21. Eyring plot constructed from the observed first order rate constants for the 1- S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. 300 y = -2134.2x - 4.7594 -12.1 -12 -11.9 -11.8 -11.7 -11.6 -11.5 -11.4 -11.3 -11.2 -11.1 -11 0.002960.003040.003120.00320.003280.003360.00344 ln(k/T) 1/T (1/K) 11 0.002960.003040.003120.00320.003280.003360.00344 Figure 7.22. Eyring plot constructed from the observed first order rate constants for the 2- S/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. Figure 7.23. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. EYRING DATA IN TOLUENE y = -1270.5x - 6.4451 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 0.0025 0.0027 0.0029 0.0031 0.0033 ln(k/T) 1/T (1/K) y = -1270.5x - 6.4451 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 0.0025 0.0027 0.0029 0.0031 0.0033 ln(k/T) 1/T (1/K) y = -1270.5x - 6.4451 -10.8 -10.6 -10.4 -10.2 -10 -9.8 -9.6 0.0025 0.0027 0.0029 0.0031 0.0033 ln(k/T) 1/T (1/K) Figure 7.23. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. Figure 7.23. Eyring plot constructed from the observed first order rate constants for the 2- O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. 301 -10.5 -10 -9.5 -9 -8.5 -8 -7.5 0.0024 0.0026 0.0028 0.003 0.0032 0.0034 0.0036 ln(k/T) 1/T (1/K) Figure 7.24. Eyring plot constructed from the observed first order rate constants for the 3- O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. y = -1780.8x - 6.1151 -12.4 -12.2 -12 -11.8 -11.6 -11.4 -11.2 -11 -10.8 -10.6 0.0024 0.0029 0.0034 0.0039 ln(k/T) 1/T (1/K) Figure 7.25. Eyring plot constructed from the observed first order rate constants for the 1- O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. Figure 7.25. Eyring plot constructed from the observed first order rate constants for the 1- O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone. 302 EYRING DATA IN BENZENE -10.6 -10.4 -10.2 -10 -9.8 -9.6 -9.4 -9.2 0.0025 0.0027 0.0029 0.0031 0.0033 0.003 ln(k/T) 1/T (1/K) 9 2 0.0025 0.0027 0.0029 0.0031 0.0033 0.0035 Figure 7.26. Eyring plot constructed from the observed first order rate constants for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (0.876 mmol) from benzyl alcohol (0.00876 mmol) in benzene. FIRST ORDER PLOTS - TOLUENE Figure 7.27. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0302x + 0.025 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 20 40 60 ln([M]o/[M]) Time (min) FIRST ORDER PLOTS - TOLUENE Figure 7.27. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 303 Figure 7.28. First order evolution of [CL] versus time for the 1-O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0081x - 0.1354 0 0.5 1 1.5 2 2.5 0 50 100 150 200 250 ln([M]o/[M]) Time (min) y = 0.0081x - 0.1354 0 0.5 1 1.5 2 2.5 0 50 100 150 200 250 ln([M]o/[M]) Time (min) Figure 7.28. First order evolution of [CL] versus time for the 1-O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.28. First order evolution of [CL] versus time for the 1-O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.29. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0027x + 0.0758 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 100 200 300 400 500 ln([M]o/[M]) Time (min) y = 0.0027x + 0.0758 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 100 200 300 400 500 ln([M]o/[M]) Time (min) Figure 7.29. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.29. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 304 Figure 7.30. First order evolution of [CL] versus time for the 2-S/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 80°C. y = 0.0205x + 0.0892 0 0.5 1 1.5 2 2.5 0 20 40 60 80 100 ln([M]o/[M]) Time (min) y = 0.0205x + 0.0892 0 0.5 1 1.5 2 2.5 0 20 40 60 80 100 ln([M]o/[M]) Time (min) 7.30. First order evolution of [CL] versus time for the 2-S/MTBD (0.0219 mmol Figure 7.30. FIRST ORDER PLOTS - TOLUENE First order evolution of [CL] versus time for the 2-S/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 80°C. each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 80°C. Figure 7.31. First order evolution of [CL] versus time for the 3-S/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0036x + 0.0532 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 100 200 300 400 ln([M]o/[M]) Time (min) y = 0.0036x + 0.0532 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 100 200 300 400 ln([M]o/[M]) Time (min) Figure 7.31. First order evolution of [CL] versus time for the 3-S/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 305 Figure 7.32. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0283x - 0.0642 0 0.5 1 1.5 2 2.5 0 20 40 60 80 ln([M]o/[M]) Time (min) y = 0.0283x - 0.0642 0 0.5 1 1.5 2 2.5 0 20 40 60 80 ln([M]o/[M]) Time (min) Figure 7.32. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.32. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.33. First order evolution of [CL] versus time for the 3-O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0486x + 0.0473 0 0.5 1 1.5 2 2.5 3 3.5 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) y = 0.0486x + 0.0473 0 0.5 1 1.5 2 2.5 3 3.5 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) Figure 7.33. First order evolution of [CL] versus time for the 3-O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 306 Figure 7.34. FIRST ORDER PLOTS - TOLUENE First order evolution of [CL] versus time for the TCC/DBU (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0075x + 0.0253 0 0.1 0.2 0.3 0.4 0.5 0.6 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) y = 0.0075x + 0.0253 0 0.1 0.2 0.3 0.4 0.5 0.6 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) Figure 7.34. First order evolution of [CL] versus time for the TCC/DBU (0.0438 each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.35. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0053x - 0.0053 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) y = 0.0053x - 0.0053 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) Figure 7.35. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 307 Figure 7.36. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0148x - 0.0386 0 0.2 0.4 0.6 0.8 1 1.2 0 20 40 60 80 ln([M]o/[M]) Time (min) y = 0.0148x - 0.0386 0 0.2 0.4 0.6 0.8 1 1.2 0 20 40 60 80 ln([M]o/[M]) Time (min) Figure 7.36. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.36. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. Figure 7.37. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. FIRST ORDER PLOTS - TOLUENE y = 0.0011x + 0.0808 0 0.1 0.2 0.3 0.4 0.5 0.6 0 100 200 300 400 500 ln([M]o/[M]) Time (min) y = 0.0011x + 0.0808 0 0.1 0.2 0.3 0.4 0.5 0.6 0 100 200 300 400 500 ln([M]o/[M]) Time (min) Figure 7.37. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (1M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 308 Figure 7.38. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (3M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. y = 0.0105x - 0.0206 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) y = 0.0105x - 0.0206 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) Figure 7.38. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (3M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in toluene at 110°C. 309 FIRST ORDER PLOTS – METHYL ISOBUTYL KETONE FIRST ORDER PLOTS METHYL ISOBUTYL KETONE Figure 7.39. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. y = 0.0027x + 0.1837 0 0.5 1 1.5 2 2.5 0 200 400 600 800 ln([M]o/[M]) Time (min) y = 0.0027x + 0.1837 0 0.5 1 1.5 2 2.5 0 200 400 600 800 ln([M]o/[M]) Time (min) Figure 7.39. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. Figure 7.39. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in Figure 7.39. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol Figure 7.39. First order evolution of [CL] versus time for the 1-S/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. Figure 7.40. FIRST ORDER PLOTS - TOLUENE First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. y = 0.0868x + 0.2151 0 0.5 1 1.5 2 2.5 3 0 5 10 15 20 25 30 35 ln([M]o/[M]) Time (min) y = 0.0868x + 0.2151 0 0.5 1 1.5 2 2.5 3 0 5 10 15 20 25 30 35 ln([M]o/[M]) Time (min) Figure 7.40. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. 310 Figure 7.41. First order evolution of [CL] versus time for the 2-S/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 60°C. y = 0.0034x + 0.0087 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0 50 100 150 ln([M]o/[M] Time (min) y = 0.0034x + 0.0087 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0 50 100 150 ln([M]o/[M] Time (min) Figure 7.41. First order evolution of [CL] versus time for the 2-S/MTBD (0.0219 mmol Figure 7.41. First order evolution of [CL] versus time for the 2-S/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 60°C. each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 60°C. y = 0.0175x + 0.1956 0 0.5 1 1.5 2 2.5 3 0 50 100 150 ln([M]o/[M]) Time (min) Figure 7.42. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. y = 0.0175x + 0.1956 0 0.5 1 1.5 2 2.5 3 0 50 100 150 ln([M]o/[M]) Time (min) y = 0.0175x + 0.1956 Figure 7.42. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. Figure 7.42. First order evolution of [CL] versus time for the 2-O/MTBD (0.0219 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. 311 Figure 7.43. FIRST ORDER PLOTS - TOLUENE First order evolution of [CL] versus time for the 1-O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. y = 0.0078x + 0.1385 0 0.5 1 1.5 2 2.5 3 0 100 200 300 400 ln([M]o/[M]) Time (min) y = 0.0078x + 0.1385 0 0.5 1 1.5 2 2.5 3 0 100 200 300 400 ln([M]o/[M]) Time (min) Figure 7.43. First order evolution of [CL] versus time for the 1-O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. Figure 7.43. First order evolution of [CL] versus time for the 1-O/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. Figure 7.44. First order evolution of [CL] versus time for the 3-O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. y = 0.0218x + 0.0669 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 5 10 15 20 ln([M]o/[M]) Time (min) y = 0.0218x + 0.0669 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 5 10 15 20 ln([M]o/[M]) Time (min) Figure 7.44. First order evolution of [CL] versus time for the 3-O/MTBD (0.0146 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in methyl isobutyl ketone at 110°C. 312 FIRST ORDER PLOT –BENZENE FIRST ORDER PLOT –BENZENE Figure 7.45. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in benzene at 110°C. y = 0.0251x + 0.0415 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) y = 0.0251x + 0.0415 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 0 10 20 30 40 50 60 70 ln([M]o/[M]) Time (min) Figure 7.45. First order evolution of [CL] versus time for the TCC/MTBD (0.0438 mmol each) cocatalyzed ROP of CL (2M, 0.876 mmol) from benzyl alcohol (0.00876 mmol) in benzene at 110°C. 313 MOLECULAR WEIGHT VS CONVERSION Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction MOLECULAR WEIGHT VS CONVERSION MOLECULAR WEIGHT VS CONVERSION 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80°C. Reaction conditions: Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80°C. Reaction conditions: Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl Figure 7.46. Mn and Mw/Mn versus conversion plot for 1-S at 80 C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). Figure 7.47. Mn and Mw/Mn versus conversion plot for 1-S at 110°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.0175 mmol). FIRST ORDER PLOT –BENZENE 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction Figure 7.47. Mn and Mw/Mn versus conversion plot for 1-S at 110°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.0175 mmol). 7.47. Mn and Mw/Mn versus conversion plot for 1-S at 110°C. Reaction conditions Figure 7.47. Mn and Mw/Mn versus conversion plot for 1-S at 110°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.0175 mmol). 314 Figure 7.48. Mn and Mw/Mn versus conversion plot for 1-S at 110°C. Reaction conditions: 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 1000 2000 3000 4000 5000 6000 7000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 1000 2000 3000 4000 5000 6000 7000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction Figure 7.48. Mn and Mw/Mn versus conversion plot for 1-S at 110°C. Reaction co CL (0.876 mmol, 2M) in methyl isobutyl ketone, 1-S/MTBD (0.0438 mmol each) cocatalyzed from benzyl alcohol (0.0175 mmol). Figure 7.49. Mn and Mw/Mn Versus Conversion plot for 2-O at 110°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 2-O/MTBD (0.0219 mmol each), benzyl alcohol (0.00876 mmol). 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 14000 16000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 14000 16000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction Figure 7.49. Mn and Mw/Mn Versus Conversion plot for 2-O at 110°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 2-O/MTBD (0.0219 mmol each), benzyl alcohol (0.00876 mmol). 315 Figure 7.50. Mn and Mw/Mn Versus Conversion plot for 2-O at 90°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 2-O/MTBD (0.0219 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). FIRST ORDER PLOT –BENZENE 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 14000 16000 18000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction Figure 7.50. Mn and Mw/Mn Versus Conversion plot for 2-O at 90°C. Reaction conditions: Figure 7.50. Mn and Mw/Mn Versus Conversion plot for 2-O at 90°C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 2-O/MTBD (0.0219 mmol each) cocatalyzed from Figure 7.50. Mn and Mw/Mn Versus Conversion plot for 2-O at 90 C. Reaction conditions: CL (0.876 mmol, 2M) in toluene, 2-O/MTBD (0.0219 mmol each) cocatalyzed from benzyl alcohol (0.00876 mmol). Figure 7.51. Mn and Mw/Mn versus conversion plot for 2-O at 90°C. Reaction conditions: CL (0.876 mmol, 2M) in methyl isobutyl ketone, 2-O/MTBD (0.0219 mmol each) cocatalyzed from benzyl alcohol (0.0175 mmol). 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 2000 4000 6000 8000 10000 12000 0 0.2 0.4 0.6 0.8 1 Mw/Mn Mn Conversion Fraction CL (0.876 mmol, 2M) in methyl isobutyl ketone, 2-O/MTBD (0.0219 mmol each) 316 Stereoselective Ring Opening Polymerization of rac-lactide using Chiral Bis- thiourea Catalysts Nayanthara U. Dharmaratne, Oleg I. Kazakov, U.L.D. Inush Kalana, Nayanthara U. Dharmaratne, Thomas J. Wright and Matthew K. Kiesewetter. Email address: mkiesewetter@chm.uri.edu ABSTRACT Polylactide (PLA) is derived from biorenewable sources and is an attractive polymer due to its biocompatibility, biodegradability and the ability for it to show drastically variable physical and mechanical properties depending on its tacticity. The properties of PLA can be significantly enhanced (Tm, 230-240 oC) by forming sterecomplexes of PLA. One method of achieving this is by stereoselective polymerization of rac-lactide. In this study we developed an enantio-selective chiral bis-thiourea catalysts that not only shows high stereoselectivity at mild conditions, but also show faster rates of polymer formation with precise control in molecular weight. We believe this system follows a mixed mechanism of stereoselectivity with a higher inclination towards ESC mechanisms and catalyzes the reaction via an imidate mediated H-bonding mechanism. 318 INTRODUCTION Polylactide (PLA) which is derived from bio-renewable sources has surfaced as an attractive polymer due to its biocompatibility and biodegradability.1,2 PLA can be synthesized by either polycondensation of lactic acid or ring opening polymerization (ROP) of lactide (LA).1 ROP has been reported to produces polymers with precise control over molecular weight, which leads to polymer samples with narrow molecular weight distribution compared to polycondensation reactions.3–5 Lactide due to its two stereo- centers can be found in three forms of diastereomers, D-LA (RR), L-LA (SS) and (RS) meso- lactide.1,6 D-LA, L-LA isomers are commercially available more commonly as racemic mixture (rac-LA). ROP of rac-LA can result in a range of polymer microstructures depending on the arrangement of chiral centers along the polymer chain (Figure 8.1).1,2,6 Different microstructures of PLA give rise to polymers with different degrees of tacticity. Polymers exhibiting different degrees of tacticity show different physical and mechanical properties.6 Atactic PLA with random distribution of chiral centers is an amorphous polymer that show no melting temperature, isotactic PLLA possesses a Tm =180 oC, heterotactic PLA (Tm 130 oC) and an equimolar mixture of PLLA and PDLA show a significantly high melting temperature (Tm 230 oC) due to stereo-complexation.1,2,6,7 High melting PLA can also be obtained by the formation of stereo-block polymers by the ROP of rac-LA as a result of the two portions of PLLA and PDLA of a single polymer sample packing together forming a crystalline polymer.2,7 A method of achieving stereo-block PLA is by stereoselective ring opening polymerization of rac-lactide (sROP).2,6 Stereo- controlled ROP can follow two main mechanisms depending on the catalysts used. In the Polylactide (PLA) which is derived from bio-renewable sources has surfaced as an attractive polymer due to its biocompatibility and biodegradability.1,2 PLA can be synthesized by either polycondensation of lactic acid or ring opening polymerization (ROP) of lactide (LA).1 ROP has been reported to produces polymers with precise control over molecular weight, which leads to polymer samples with narrow molecular weight distribution compared to polycondensation reactions.3–5 Lactide due to its two stereo- centers can be found in three forms of diastereomers, D-LA (RR), L-LA (SS) and (RS) meso- lactide.1,6 D-LA, L-LA isomers are commercially available more commonly as racemic mixture (rac-LA). INTRODUCTION ROP of rac-LA can result in a range of polymer microstructures depending on the arrangement of chiral centers along the polymer chain (Figure 8.1).1,2,6 Different microstructures of PLA give rise to polymers with different degrees of tacticity. Polymers exhibiting different degrees of tacticity show different physical and mechanical properties.6 Atactic PLA with random distribution of chiral centers is an amorphous polymer that show no melting temperature, isotactic PLLA possesses a Tm =180 oC, heterotactic PLA (Tm 130 oC) and an equimolar mixture of PLLA and PDLA show a significantly high melting temperature (Tm 230 oC) due to stereo-complexation.1,2,6,7 High melting PLA can also be obtained by the formation of stereo-block polymers by the ROP of rac-LA as a result of the two portions of PLLA and PDLA of a single polymer sample packing together forming a crystalline polymer.2,7 A method of achieving stereo-block PLA is by stereoselective ring opening polymerization of rac-lactide (sROP).2,6 Stereo- controlled ROP can follow two main mechanisms depending on the catalysts used. EXPERIMENTAL SECTION All chemicals were purchased from Acros organics and used as received unless stated otherwise. chloroform-d and acetone-d6 were purchased from Cambridge Isotope Laboratories, distilled from calcium hydride and calcium sulfate respectively and stored under N2. Benzyl alcohol was distilled under high vacuum from calcium hydride prior to use. Dry CH2Cl2 was obtained from an Innovative Technology solvent purification system. 4-nitrophenyl isothiocyanate, 3,5-bis(trifluoromethyl)phenyl isothiocyanate, 3,5-dimethyl isothiocyanate was purchased from Sigma Aldrich and used as it is. L-lactide, D-lactide was obtained from Carbion and rac-lactide was obtained from Acros Organics and recrystallized from dry toluene prior to use. All experiments were conducted in a stainless- steel glovebox under N2 unless stated otherwise. NMR experiments were performed on a Bruker Avance III 300 or 400 MHz spectrometer. Decoupled experiments were performed on a Varian 500 MHz NMR spectrometer in the college of pharmacy at the University of Rhode Island. Mass spectrometry experiments were performed using a Thermo Electron (San Jose, CA, USA) LTQ Orbitrap XL mass spectrometer affixed with electrospray ionization (ESI) interface in a positive ion mode. Collected mass spectra were averaged for at least 50 scans. Tune conditions for infusion experiments (10 μL/min flow, sample concentration 5 μg/mL in 50/50 v/v water/ methanol) were as follows: ion spray voltage, 5000 V; capillary temperature, 275oC; sheath gas (N2, arbitrary units), 11; auxiliary gas (N2, arbitrary units), 2; capillary voltage, 21 V; and tube lens, 90 V; multipole 00 offset, - All chemicals were purchased from Acros organics and used as received unless stated otherwise. chloroform-d and acetone-d6 were purchased from Cambridge Isotope Laboratories, distilled from calcium hydride and calcium sulfate respectively and stored under N2. Benzyl alcohol was distilled under high vacuum from calcium hydride prior to use. Dry CH2Cl2 was obtained from an Innovative Technology solvent purification system. 4-nitrophenyl isothiocyanate, 3,5-bis(trifluoromethyl)phenyl isothiocyanate, 3,5-dimethyl isothiocyanate was purchased from Sigma Aldrich and used as it is. L-lactide, D-lactide was obtained from Carbion and rac-lactide was obtained from Acros Organics and recrystallized from dry toluene prior to use. All experiments were conducted in a stainless- steel glovebox under N2 unless stated otherwise. NMR experiments were performed on a Bruker Avance III 300 or 400 MHz spectrometer. Decoupled experiments were performed on a Varian 500 MHz NMR spectrometer in the college of pharmacy at the University of Rhode Island. INTRODUCTION In the 319 presence of bulky catalysts sROP occurs via chain end-controlled mechanism (CEC) where the initial monomer insertion happens in a random manner and thereafter monomer preference is dependent on the chirality of the chain end (figure 8.2a).1,2,7–9 These types of reactions are fast but are limited to cold reaction temperatures in order to minimize stereo error.7–9 In contrast to CEC mechanism, enantio-site control (ESC) mechanism follows a more controlled pathway where a chiral catalyst selectively activates one diastereomer over the other for polymerization and doesn’t require harsh temperatures to maintain stereo selectivity (figure 8.2b).1,2,10,11 Although numerous studies have been reported for sROP of rac-lactide that deliver high degree of stereoselectivity, these systems still suﬀer from low reaction rates (days to achieve high conversions) or require harsh conditions (high to low temperatures).7–10 Thiourea organocatalysts have had great success in producing PLA with highly controlled properties due to their high monomer selectivity.3–5,10,12 However, only a handful of studies report the use of thiourea for sROP of rac-lactide and the reported studies suffer similar drawbacks in terms of rate and/or rigorous conditions of polymerization.7,10,13 Our group has successfully designed bis, tris-thiourea catalysts that have revolutionized the field of ROP of lactones by converting cyclic monomers to polymers with high control in merely minutes.4,12,14,15 Our motivation is to translate our knowledge of bis-thioureas to design novel chiral catalysts by introducing chirality via knowledge of bis-thioureas to design novel chiral catalysts by introducing chirality via insertion of chiral amino acid derivatives that will enable the production of crystalline PLAs at faster rates under mild conditions. 320 EXPERIMENTAL SECTION Mass spectrometry experiments were performed using a Thermo Electron (San Jose, CA, USA) LTQ Orbitrap XL mass spectrometer affixed with electrospray ionization (ESI) interface in a positive ion mode. Collected mass spectra were averaged for at least 50 scans. Tune conditions for infusion experiments (10 μL/min flow, sample concentration 5 μg/mL in 50/50 v/v water/ methanol) were as follows: ion spray voltage, 5000 V; capillary temperature, 275oC; sheath gas (N2, arbitrary units), 11; auxiliary gas (N2, arbitrary units), 2; capillary voltage, 21 V; and tube lens, 90 V; multipole 00 offset, - 321 4.25 V; lens 0 voltage, - 5.00; multipole 1 offset, - 8.50 V; Multipole RF Amplitude, 400 V; Ion trap’s AGC target settings for Full MS was 3.0e4 and FT’s 2.0e5 (with 3 and 2 averaged microscans , respectively). Prior to analysis, the instrument was calibrated for positive ions using Pierce LTQ ESI positive ion calibration solution (lot #PC197784). Determination of thermal properties. Melting temperatures (Tm) of PLA synthesized in this study were determined by differential scanning calorimetry (DSC) using a shimadzu differential scanning calorimeter 60 plus that has been calibrated using high purity indium at a heating rate of 5 °C/min. Polymer sample (5 mg) was first heated to 180 oC at 5 oC/min, held at this temperature for 15 h to anneal the sample. The sample was cooled to 25 oC at 5 oC/min, held for 10 min, and reheated to 250 oC at 5 oC/min. All thermal data was obtained from the second cycle. Determination of thermal properties. Melting temperatures (Tm) of PLA synth Determination of Pm. An NMR sample of purified polymer (1mg/ml) was prepared in chloroform-d and analyzed by 1H- homo decoupled NMR at 50 °C. The procedure for determining Pm was followed as reported.16,17 HPLC measurement. The unreacted monomer at 50 % monomer conversion from the quenched reaction was isolated by washing the reaction with isopropanol : hexane 1: 1 and the soluble portion was extracted and concentrated under vacuum. Further purification by column chromatography was done to purify the unreacted monomer and to remove traces of catalysts. Isolated monomer was then analyzed by chiral HPLC using 322 hexane/isopropanol (7/3). The selectivity factor (kD/kL) was determined by kD/kL = {ln[(1- conv.)(1-ee)]}/{ln[(1-conv.)(1+ee)]}.11 Example ROP of rac-LA at room temperature. A 7 mL vial was charged with 1 (7.0 mg, 0.005 mmol), rac-LA (70 mg, 0.475 mmol) and CH2Cl2 (480 µL) and agitated to make a homogeneous solution. A stock solution of benzyl alcohol (0.5 M) was prepared using benzyl alcohol (5.4 mg, 0.05 mmol) in CH2Cl2 (100 µL) and 9.5 µL from the stock solution was transferred to the vial and the content shaken to mix well. A stock solution of Tris[2-(dimethylamino)ethyl]amine (Me6Tren) (0.5 M) was prepared using Me6TREN (11.5 mg, 0.05 mmol) in CH2Cl2 (100 µL) and 19.0 µL from the stock solution was transferred to the vial to start the reaction. The content was transferred to an NMR tube and the reaction monitored using 1H NMR and locked using a DMSO capillary. When the desired conversion was achieved, the reaction was quenched with at least 2 equivalents of benzoic acid to the amount of Me6Tren. The reaction was then concentrated under vacuum. The dried polymer was purified using a methanol wash and subjected to high vacuum. The purified polymer was characterized using NMR, DSC and GPC. Example ROP of rac-LA at -15o C. A 7 mL vial was charged with 1 (7.0 mg, 0.005 mmol), rac-LA (70 mg, 0.475 mmol) and CH2Cl2 (480 µL) and agitated to make a homogeneous solution. A stock solution of benzyl alcohol (0.5 M) was prepared using benzyl alcohol (5.4 mg, 0.05 mmol) in CH2Cl2 (100 µL) and 9.5 µL from the stock solution was transferred to the vial and the content shaken to mix well. The content in the vial was 323 allowed to equilibrate at -15o C for 30 min. Determination of thermal properties. Melting temperatures (Tm) of PLA synth A stock solution of Tris[2- (dimethylamino)ethyl]amine (Me6Tren) (0.5 M) was prepared using Me6TREN (11.5 mg, 0.05 mmol) in CH2Cl2 (100 µL) and the stock solution was also allowed to equilibrate at - 15o C for 30 min and 19.0 µL from the stock solution was transferred to the vial to start the reaction post equilibration. The reaction was monitored by obtaining aliquots. When the desired conversion was achieved, the reaction was quenched with at least 2 equivalents of benzoic acid to the amount of Me6Tren dissolved in cold DCM. The reaction was then concentrated under vacuum. The dried polymer was purified using a methanol wash and subjected to high vacuum. The purified polymer was characterized using NMR, DSC and GPC. Synthesis of bis thiourea H-bond donor TU 1-3. NH2 BocHN NHBoc O HO tBu + HBTU, DIPEA DCM. r.t Boc-DAP Boc-L-Leu I 1.1 BocHN N H O tBu NHBoc BocHN BocHN Boc-DAP Intermediate product I 1.1 was synthesized according to an adapted literature procedure.18 Intermediate product I 1.1 was synthesized according to an adapted literature procedure.18 A 25 mL flame-dried Schlenk flask was charged with dry DCM (10 mL). N-Boc-L-tert- leucine (Boc-L-Leu) (100 mg, 0.43 mmol), N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1- yl)uronium hexafluorophosphate (HBTU) (163.1 mg, 0.43 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.23 mL, 1.29 mmol) was added to the flask and the reaction was stirred under nitrogen. N-Boc-1,3-propanediamine (Boc-DAP) was then added (0.082 mL, 0.47 mmol) to the flask. The reaction was allowed to stir under nitrogen 324 overnight. Afterwards, the reaction mixture was concentrated under vacuum. The residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). I 1.1 BocHN N H O tBu NHBoc 4 N HCl in Dioxane 2h, r.t I 1.2 H2N N H O tBu NH2 . 2HCl 4 N HCl in Dioxane 2h, r.t I 1.2 H2N N H O tBu NH2 . 2HCl 4 N HCl in Dioxane 2h, r.t I 1.2 H2N N H O tBu NH2 . 2HCl I 1.1 BocHN N H O tBu NHBoc 4 N HCl in Dioxane BocHN I 1.1 I 1.2 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (1.40 mL, 5.59 mmol of HCl), followed by the addition of I 1.1 (166 mg, 0.43 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. + F3C CF3 NCS TEA, THF 4 h , r.t ITC TU1 I 1.2 H2N N H O tBu NH2 . 2HCl N H S N H N H O tBu H N CF 3 CF3 H N CF3 F3C S ITC I 1.2 A flame-dried Schlenk flask was charged with dry THF, then I 1.2 (112 mg, 0.43 mmol) was added followed by Triethylamine (TEA) (0.35 mL, 2.6 mmol). The mixture was stirred under nitrogen for about 10 minutes and Isothiocyanate (ITC) (0.16 mL, 0.86 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. Synthesis of bis thiourea H-bond donor TU 1-3. The solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1) for TU 1 and TU 3 and 100 % ethyl acetate for TU 2. TU 1, white powder. yield: 60 %. TU 2, a yellow powder. Yield: 55%. TU 3 a white powder. Yield 50%. NMR spectra given below. HRMS: TU 1 calc. (C27H27F12N5OS2+H)+= 730.15; found m/z = 730.15. TU 2 calc. 325 (C23H29N7O5S2+H)+ = 547.65; found m/z = 547.16. TU 3 calc. (C27H39N5OS2 + H)+ = 513.76; found m/z = 513.25. Synthesis of bis thiourea H-bond donor TU 4. NH2 H2N H3C CH3 NCS ITC + DCM. r.t H3C CH3 NH2 N H N H S 1,3-Diaminopropane I 4.1 H3C CH3 NCS ITC + DCM. r.t H3C CH3 NH2 N H N H S I 4.1 H3C CH3 NH2 N H N H S I 4.1 ITC I 4.1 A flame-dried Schlenk flask was charged with dry DCM, then 1,3-Diaminopropane (100 mg, 1.34 mmol), was added to the flask. The mixture was stirred under nitrogen and ITC (0.15 mL, 1.34 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using DCM/Methanol (9:1). A 25 mL flame-dried Schlenk flask was charged with dry THF (20 mL). N-Boc-L-tert- leucine (Boc-L-Leu) (309 mg, 1.34 mmol), Benzotriazole-1-yl-oxy-tris-pyrrolidino- phosphonium hexafluorophosphate (PyBOP) (572.3 mg, 1.10 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.70 mL, 4.02 mmol) was added to the flask and the reaction was stirred under nitrogen. I 4.1 was then added (317.7 mg, 1.34 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was concentrated under vacuum. The residue was used as it is for the next step. H3C CH3 NH2 N H N H S I 4.1 NHBoc O HO tBu Boc-L-Leu + PyBOP, DIPEA THF. r.t H3C CH3 N H N H N H S NHBoc O tBu I 4.2 H3C CH3 NH2 N H N H S I 4.1 NHBoc O HO tBu Boc-L-Leu + PyBOP, DIPEA THF. r.t H3C CH3 N H N H N H S NHBoc O tBu I 4.2 H3C CH3 NH2 N H N H S I 4.1 + NHBoc O HO tBu Boc-L-Leu Boc-L-Leu A 25 mL flame-dried Schlenk flask was charged with dry THF (20 mL). N-Boc-L-tert- leucine (Boc-L-Leu) (309 mg, 1.34 mmol), Benzotriazole-1-yl-oxy-tris-pyrrolidino- phosphonium hexafluorophosphate (PyBOP) (572.3 mg, 1.10 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.70 mL, 4.02 mmol) was added to the flask and the reaction was stirred under nitrogen. I 4.1 was then added (317.7 mg, 1.34 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was concentrated under vacuum. The residue was used as it is for the next step. Synthesis of bis thiourea H-bond donor TU 4. 326 H3C CH3 N H N H N H S NHBoc O tBu I 4.2 H3C CH3 N H N H N H S NHBoc O tBu I 4.2 4 N HCl in Dioxane 2h, r.t H3C CH3 N H N H N H S NH2 . HCl O tBu I 4.3 H3C CH3 N H N H N H S NH2 . HCl O tBu I 4.3 4 N HCl in Dioxane 2h, r.t I 4.2 I 4.3 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (2.10 mL, 8.71 mmol of HCl), followed by the addition of I 4.2 (603.4 mg, 1.34 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. H3C CH3 N H N H N H S NH2 . HCl O tBu I 4.3 F3C CF3 NCS TEA, THF 4 h , r.t ITC + N H S N H N H O tBu H N CF3 CF3 H N CH3 H3C S TU4 H3C CH3 N H N H N H S NH2 . HCl O tBu I 4.3 F3C CF3 NCS TEA, THF 4 h , r.t ITC + N H S N H N H O tBu H N CF3 CF3 H N CH3 H3C S TU4 TU4 ITC A flame-dried Schlenk flask was charged with dry THF, then I 4.3 (518.5 mg, 1.34 mmol) was added followed by TEA (0.6 mL, 4.02 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.5 mL, 2.71 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). TU 4, yellow white powder. yield: 60 %. NMR spectra given below. HRMS: calc. (C27H33F6N5OS2+H)+= 622.20; found m/z = 622.20. 327 Synthesis of bis thiourea H-bond donor TU 5. Synthesis of bis thiourea H-bond donor TU 5. F3C CF3 NH2 N H N H S I 5.1 NH2 H2N F3C CF3 NCS ITC + DCM. r.t F3C CF3 NH2 N H N H S 1,3-Diaminopropane I 5.1 NH2 H2N F3C CF3 NCS ITC + 1,3-Diaminopropane ITC I 5.1 A flame-dried Schlenk flask was charged with dry DCM, then 1,3-Diaminopropane (100 mg, 1.34 mmol), was added to the flask. The mixture was stirred under nitrogen and ITC (0.25 mL, 1.34 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. I 5.1 was precipitated at 0°C then filtered and washed with cold DCM and dried under vacuum. F3C CF3 NH2 N H N H S I 5.1 NHBoc O HO tBu Boc-L-Leu + HBTU, DIPEA THF. r.t F3C CF3 N H N H N H S NHBoc O tBu I 5.2 F3C CF3 NH2 N H N H S I 5.1 + NHBoc O HO tBu Boc-L-Leu HBTU, DIPEA I 5.2 A 25 mL flame-dried Schlenk flask was charged with dry THF (20 mL). N-Boc-L-tert- leucine (Boc-l-Leu) (309 mg, 1.34 mmol), N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1- yl)uronium hexafluorophosphate (HBTU) (417.2 mg, 1.10 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.70 mL, 4.02 mmol) was added to the flask and the reaction was stirred under nitrogen. I 5.1 was then added (462.7 mg, 1.34 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was concentrated under vacuum. The residue purified with column chromatography on silica gel using ethyl acetate/hexane (3:7). 328 F3C CF3 N H N H N H S NHBoc O tBu I 5.2 4 N HCl in Dioxane 2h, r.t CF3 N H N H N H S NH2 . HCl O tBu I 5.3 F3C F3C CF3 N H N H N H S NHBoc O tBu I 5.2 4 N HCl in Dioxane 2h, r.t CF3 N H N H N H S NH2 . HCl O tBu I 5.3 F3C 4 N HCl in Dioxane I 5.3 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (2.10 mL, 8.71 mmol of HCl), followed by the addition of I 5.2 (748.5 mg, 1.34 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. Synthesis of bis thiourea H-bond donor TU 5. A flame-dried Schlenk flask was charged with dry THF, then I 5.3 (663.2 mg, 1.34 mmol) was added followed by TEA (0.6 mL, 4.02 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.5 mL, 2.71 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The TEA salt was filtered and the solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). TU 5, white powder. yield: 60 %. NMR spectra given below. HRMS: calc. (C27H33F6N5OS2+H)+= 622.20; found m/z = F3C CF3 N H N H N H S NH2 . HCl O tBu I 5.3 H3C CH3 NCS TEA, THF 4 h , r.t ITC + N H S N H N H O tBu H N CH3 CH3 H N CF3 F3C S TU5 F3C CF3 N H N H N H S NH2 . HCl O tBu I 5.3 H3C CH3 NCS TEA, THF 4 h , r.t ITC + N H S N H N H O tBu H N CH3 CH3 H N CF3 F3C S TU5 I 5.3 ITC TU5 A flame-dried Schlenk flask was charged with dry THF, then I 5.3 (663.2 mg, 1.34 mmol) was added followed by TEA (0.6 mL, 4.02 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.5 mL, 2.71 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The TEA salt was filtered and the solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). TU 5, white powder. yield: 60 %. NMR spectra given below. HRMS: calc. (C27H33F6N5OS2+H)+= 622.20; found m/z = 329 Synthesis of bis thiourea H-bond donor TU 6. Synthesis of bis thiourea H-bond donor TU 6. NHBoc tBu O H N BocHN NH2 BocHN NHBoc O HO tBu + HBTU, DIPEA DCM. r.t Boc-DAE Boc-L-Leu I 6.1 NHBoc tBu O H N BocHN NH2 BocHN NHBoc O HO tBu + HBTU, DIPEA DCM. r.t Boc-DAE Boc-L-Leu I 6.1 I 6.1 I 6.1 A 25 mL flame-dried Schlenk flask was charged with dry DCM (10 mL). N-Boc-L-tert- leucine (Boc-L-Leu) (100 mg, 0.43 mmol), N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1- yl)uronium hexafluorophosphate (HBTU) (163.1 mg, 0.43 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.23 mL, 1.29 mmol) was added to the flask and the reaction was stirred under nitrogen. N-Boc-1,2-diaminoethane (Boc-DAE) was then added (75 mg, 0.47 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was concentrated under vacuum. The product was deprotected as it is. NHBoc tBu O H N BocHN I 6.1 4 N HCl in Dioxane 2h, r.t NH2 tBu O H N H2N I 6.2 . 2HCl 4 N HCl in Dioxane 2h, r.t NHBoc tBu O H N BocHN I 6.1 NH2 tBu O H N H2N I 6.2 . 2HCl I 6.2 I 6.1 I 6.1 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (1.39 mL, 5.59 mmol of HCl), followed by the addition of I 6.1 (160.5 mg, 0.43 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. NH2 tBu O H N H2N I 6.2 . 2HCl S F3C CF3 N H CF3 CF3 N H tBu O H N N H S N H + F3C CF3 NCS TEA, THF 4 h , r.t ITC TU6 I 6.2 I 6.2 ITC 330 A flame-dried Schlenk flask was charged with dry THF, then I 6.2 (75 mg, 0.43 mmol) was added followed by TEA (0.4 mL, 2.60 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.16 mL, 8.86mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The solvent was removed using a rotavap and the residue was purified using column chromatography using ethyl acetate / hexanes (1:1) as an eluant. TU 6, white powder. yield: 68 %. NMR spectra given below. HRMS: calc. (C26H25F12N5OS2+H)+= 716.13; found m/z = 716.12. Synthesis of mono thiourea H-bond donor TU 7. H N O tBu NHBoc H2N NHBoc O HO tBu Boc-L-Leu + HBTU, DIPEA DCM. r.t n-Butylamine I 7.1 n-Butylamine I 7.1 A 25 mL flame-dried Schlenk flask was charged with dry DCM (10 mL). N-Boc-L-tert- leucine (Boc-L-Leu) (100 mg, 0.43 mmol), N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1- yl)uronium hexafluorophosphate (HBTU) (163.1 mg, 0.43 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.23 mL, 1.29 mmol) was added to the flask and the reaction was stirred under nitrogen. n-Butylamine was then added (0.05 mL, 0.47 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was concentrated under vacuum. The residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). 331 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (0.8 mL, 3.05 mmol of HCl), followed by the addition of I 7.1 (123.0 mg, 0.43 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. H N O tBu BocHN I 7.1 4 N HCl in Dioxane 2h, r.t NH2 tBu O H N I 7.2 . HCl H N O tBu BocHN I 7.1 4 N HCl in Dioxane 2h, r.t NH2 tBu O H N I 7.2 . HCl I 7.1 I 7.2 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (0.8 mL, 3.05 mmol of HCl), followed by the addition of I 7.1 (123.0 mg, 0.43 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. A flame-dried Schlenk flask was charged with dry THF, then I 7.2 (95.5 mg, 0.43 mmol) NH2 tBu O H N I 7.2 F3C CF3 NCS TEA, THF 4 h , r.t ITC + H N O tBu N H F3C CF3 N H S TU7 . HCl A flame-dried Schlenk flask was charged with dry THF, then I 7.2 (95.5 mg, 0.43 mmol) was added followed by TEA (0.18 mL, 1.29 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.08 mL, 0.43 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The solvent was removed using a rotavap and the residue was purified using column chromatography using ethyl acetate / hexanes (1:2) as an eluant. Synthesis of mono thiourea H-bond donor TU 7. TU 7, yellow white powder. yield: 57 %. NMR spectra given below. HRMS: calc. (C19H25F6N3OS+H)+= 458.16; found m/z = 458.16. NH2 tBu O H N I 7.2 F3C CF3 NCS TEA, THF 4 h , r.t ITC + H N O tBu N H F3C CF3 N H S TU7 . HCl NH2 tBu O H N I 7.2 F3C CF3 NCS TEA, THF 4 h , r.t ITC + H N O tBu N H F3C CF3 N H S TU7 . HCl F3C CF3 NCS TEA, THF 4 h , r.t ITC + H N O tBu N H F3C CF3 N H S TU7 NH2 tBu O H N I 7.2 . HCl ITC I 7.2 A flame-dried Schlenk flask was charged with dry THF, then I 7.2 (95.5 mg, 0.43 mmol) was added followed by TEA (0.18 mL, 1.29 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.08 mL, 0.43 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The solvent was removed using a rotavap and the residue was purified using column chromatography using ethyl acetate / hexanes (1:2) as an eluant. TU 7, yellow white powder. yield: 57 %. NMR spectra given below. HRMS: calc. (C19H25F6N3OS+H)+= 458.16; found m/z = 458.16. 332 Synthesis of bis thiourea H-bond donor TU 9. NH2 BocHN NHBoc O HO + HBTU, DIPEA DCM. r.t Boc-DAP I 9.1 BocHN N H O NHBoc Boc-L-Phenylglycine NH2 BocHN NHBoc O HO + HBTU, DIPEA DCM. r.t Boc-DAP I 9.1 BocHN N H O NHBoc Boc-L-Phenylglycine NHBoc BocHN BocHN HO Boc-DAP I 9.1 A 25 mL flame-dried Schlenk flask was charged with dry DCM (10 mL). Boc-L-α- phenylglycine (108 mg, 0.43 mmol), N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1- yl)uronium hexafluorophosphate (HBTU) (163.1 mg, 0.43 mmol), and N,N- Diisopropylethylamine (DIPEA) (0.23 mL, 1.29 mmol) was added to the flask and the reaction was stirred under nitrogen. N-Boc-1,3-propanediamine (Boc-DAP) was then added (0.082 mL, 0.47 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was concentrated under vacuum. The residue was purified with column chromatography on silica gel using ethyl acetate. 4 N HCl in Dioxane 2h, r.t I 9.2 H2N N H O NH2 . Synthesis of mono thiourea H-bond donor TU 7. 2HCl I 9.1 BocHN N H O NHBoc 4 N HCl in Dioxane 2h, r.t I 9.1 BocHN N H O NHBoc I 9.1 BocHN N H O NHBoc I 9.2 H2N N H O NH2 . 2HCl 4 N HCl in Dioxane I 9.1 I 9.2 A round bottom flask was charged with 4 N HCl in 1,4-dioxane (1.40 mL, 5.59 mmol of HCl), followed by the addition of I 9.2 (175 mg, 0.43 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. + F3C CF3 NCS TEA, THF 4 h , r.t ITC TU9 N H S N H N H O H N CF3 CF3 H N CF3 F3C S I 9.2 H2N N H O NH2 . 2HCl ITC TU9 333 A flame-dried Schlenk flask was charged with dry THF, then I 9.2 (89 mg, 0.43 mmol) was added followed by TEA (0.35 mL, 2.6 mmol). The mixture was stirred under nitrogen for about 10 minutes and Isothiocyanate (ITC) (0.16 mL, 0.86 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). TU 9, white powder. yield: 41 %. NMR spectra given below. HRMS: calc. (C29H23F12N5OS2+H)+= 750.12; found m/z = 750.12. A flame-dried Schlenk flask was charged with dry THF, then I 9.2 (89 mg, 0.43 mmol) was added followed by TEA (0.35 mL, 2.6 mmol). The mixture was stirred under nitrogen for about 10 minutes and Isothiocyanate (ITC) (0.16 mL, 0.86 mmol) was added dropwise. C28H22F12N5OS2+H)+= 737.11; found m/z = 737.11. RESULTS AND DISCUSSIONS A leucine derivatized chiral bis thiourea, TU1, (figure 8.3) was studied for the stereo- controlled ring-opening polymerization (sROP) of rac-LA. To investigate the polymerization characteristics of TU1 as stereoselective catalysts for ROP, Me6TREN (0.0095 mmol) was dissolved in dichloromethane (DCM) and added to a solution of DCM containing TU1 (0.0095 mmol), rac-LA (0.475 mmol, 0.95 M) and benzyl alcohol (BnOH) (0.0048 mmol) at room temperature in a glovebox. Under these conditions, 90% of the monomer was converted in less than 5 h producing crystalline poly(lactide) (PLA), with predictable number average molar mass Mn = 17.5 kg/mol consistent with the monomer to initiator ratio [M] o/[I] o and narrow molecular weight distribution, Mw/Mn = 1.05; (Table 8.1, entry 1). This system produces polymers with high chain-end fidelity confirmed by Gel permeation chromatography (GPC) (figure 8.6). The polymerization exhibits “living” character showing linear increment of molecular weight with conversion (figure 8.7) and possess the ability to produce polymers with predictable molecular weights and narrow Mw/Mn £ 1.05 upto [M] o/[I] o= 200 (Table 8.1). Analysis of a sample prepared under similar conditions described above by MALDI-TOF mass spectrometry showed a single series of ions separated by m/z 144, revealing no to minimal transesterification of the polymer backbone (figure 8.8) suggesting high selectivity of the catalyst towards the monomer vs the polymer. A leucine derivatized chiral bis thiourea, TU1, (figure 8.3) was studied for the stereo- controlled ring-opening polymerization (sROP) of rac-LA. To investigate the polymerization characteristics of TU1 as stereoselective catalysts for ROP, Me6TREN (0.0095 mmol) was dissolved in dichloromethane (DCM) and added to a solution of DCM containing TU1 (0.0095 mmol), rac-LA (0.475 mmol, 0.95 M) and benzyl alcohol (BnOH) (0.0048 mmol) at room temperature in a glovebox. Under these conditions, 90% of the monomer was converted in less than 5 h producing crystalline poly(lactide) (PLA), with predictable number average molar mass Mn = 17.5 kg/mol consistent with the monomer to initiator ratio [M] o/[I] o and narrow molecular weight distribution, Mw/Mn = 1.05; (Table 8.1, entry 1). This system produces polymers with high chain-end fidelity confirmed by Gel permeation chromatography (GPC) (figure 8.6). The polymerization exhibits “living” character showing linear increment of molecular weight with conversion (figure 8.7) and possess the ability to produce polymers with predictable molecular weights and narrow Mw/Mn £ 1.05 upto [M] o/[I] o= 200 (Table 8.1). Synthesis of bis thiourea H-bond donor TU 10. NH2 BocHN + HBTU, DIPEA DCM. r.t Boc-DAP I 10.1 Boc-L-Proline N O Boc HO N O Boc H N H N Boc I 10.1 N O Boc H N H N Boc BocHN Boc-L-Proline Boc-L-Proline I 10.1 Boc-DAP Intermediate product I 10.1 was synthesized according to an adapted literature procedure. Intermediate product I 10.1 was synthesized according to an adapted literature procedure. Intermediate product I 10.1 was synthesized according to an adapted literature procedure. A 25 mL flame-dried Schlenk flask was charged with dry DCM (10 mL). Boc-L-proline (108 mg, 0.43 mmol), N,N,N′,N′-Tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate (HBTU) (163.1 mg, 0.43 mmol), and N,N-Diisopropylethylamine (DIPEA) (0.23 mL, 1.29 mmol) was added to the flask and the reaction was stirred under nitrogen. N-Boc-1,3-propanediamine (Boc-DAP) was then added (0.082 mL, 0.47 mmol) to the flask. The reaction was allowed to stir under nitrogen overnight. Afterwards, the reaction mixture was washed with water (3x 20mL), dried then concentrated under vacuum. 334 4 N HCl in Dioxane 2h, r.t I 10.2 . 2HCl I 10.1 NH O H N H2N N O Boc H N H N Boc I 10.2 . 2HCl NH O H N H2N 4 N HCl in Dioxane 2h, r.t A round bottom flask was charged with 4 N HCl in 1,4-dioxane (1.40 mL, 5.59 mmol of HCl), followed by the addition of I 10.2 (175 mg, 0.43 mmol). The content was stirred for 2 hours. Next, the volatiles were removed via vacuum transfer. The residue was subjected to high vacuum. + F3C CF3 NCS TEA, THF 4 h , r.t ITC TU10 I 10.2 . 2HCl N O H N H N H N S N H S F3C CF3 CF3 CF3 NH O H N H2N TU10 A flame-dried Schlenk flask was charged with dry THF, then I 9.2 (89 mg, 0.43 mmol) was added followed by TEA (0.35 mL, 2.6 mmol). The mixture was stirred under nitrogen for about 10 minutes and ITC (0.16 mL, 0.86 mmol) was added dropwise. The reaction mixture was allowed to react at ambient temperature under nitrogen overnight. The TEA salt was filtered and the solvent was removed using a rotavap and the residue was purified with column chromatography on silica gel using ethyl acetate / hexanes (1:1). TU 10, white powder. yield: 45.8 %. NMR spectra given below. HRMS: calc. (C28H22F12N5OS2+H)+= 737.11; found m/z = 737.11. 335 RESULTS AND DISCUSSIONS Analysis of a sample prepared under similar conditions described above by MALDI-TOF mass spectrometry showed a single series of ions separated by m/z 144, revealing no to minimal transesterification of the polymer backbone (figure 8.8) suggesting high selectivity of the catalyst towards the monomer vs the polymer. Since the main goal of this work was to establish a stereo selective catalyst for ROP of rac Since the main goal of this work was to establish a stereo-selective catalyst for ROP of rac- LA, the microstructure of the PLA produced using rac-LA with our catalyst system was investigated. This was determined by 1H homonuclear decoupled NMR spectroscopy 336 conducted at 50o C by analyzing the methine region of PLA and calculating the probability of forming meso dyads (Pm); Pm(ESC) = 0.87 using ESC statistical model and Pm(CEC) = 0.80 using CEC statistical model using non-Bernoullian statistics and Bernoullian statistics as reported elsewhere (Figures 8.9).1,16,17 Analysis of the polymer produced using differential scanning calorimetry (DSC) (Tm = 168 °C) (Figures 8.10) indicate characteristics of an isotactic crystalline polymer. Associating the information obtained from DSC and the calculated Pm values the mechanism of stereoselectivity with our catalyst system demonstrates a higher inclination towards an ESC mechanism rather than a CEC mechanism. This deduction was reached due to the high Tm of our polymer (Table 8.1, entry 1) usually shown by PLA with Pm ≥ 0.81.2,7,19 The stereoselectivity factor (s = kD/kL. = 1.8) of the polymerization in the presence of TU1/Me6TREN, was determined by analyzing the unreacted monomer at 47% monomer conversion using chiral HPLC (figure 8.12). The analysis revealed that in the presence of TU1, D-LA was favored, as displayed by the reduction of the signal associated with D-LA. A kinetic experiment was then conducted to confirm this observation by polymerizing D- LA, L-LA and rac-LA separately with TU1/ Me6TREN. The first order plot for the polymerization of D-LA, L-LA and rac-LA (figure 8.4a) show preferential rate acceleration of the polymerization of D-LA vs L-LA and rac-LA, consistent with the chiral HPLC analysis. However, the selectivity factor obtained using corresponding kobs from the first order plots (s = kobs D-LA/ kobs L-LA = 3.4) is higher than what was obtained using chiral HPLC. RESULTS AND DISCUSSIONS Conducting reactions at cold temperatures (-75oC) is another technique used to enhance selectivity in sROP. We decided to make use of this approach but at much milder conditions compared to temperatures reported thus far. Since our catalyst system already showed a moderately good stereoselectivity at room temperature, we analyzed our polymerization system at -15oC. In the presence of Me6TREN and t-TACN (relatively weak bases) the polymerization rates diminished, but still could be considered faster than other reported organocatalysts that follow ESC mechanism at room temperature,10,13 but the stereoselectivity enhanced significantly (Table 2, entry 1,2) compared to that at room temperature. When these reactions were performed with TU1 in the presence of stronger bases like BEMP and MTBD, the polymerization reached 90 % conversion in 90 min and 60 min respectively, producing highly isotactic polymers with good control over Mn and Mw/Mn (Table 2, entry 3,4). Previous studies report BEMP by itself as a successful catalyst for sROP of rac-LA to produce highly isotactic PLA within few minutes. However, the reports indicate limitations with respect to requirement of harsh conditions of reaction temperatures (-75oC) to achieve high stereoselectivity.8 This was also seen in our work, when the polymerization was conducted with BEMP in the absence of TU1 at -15oC 90 % monomer conversion was achieved in 5 min, but the resulting polymer showed no melting peak, indicating the formation of atatctic amorpous PLA. This observation indicated that the presence of TU1 is crucial in producing isotactic polymer at less harsh conditions of - 15oC or room temperature. In contrast to weak bases in the presence of strong bases, we believe the polymerization proceeds preferentially via CEC mechanism. This was depicted Conducting reactions at cold temperatures (-75oC) is another technique used to enhance selectivity in sROP. We decided to make use of this approach but at much milder conditions compared to temperatures reported thus far. Since our catalyst system already showed a moderately good stereoselectivity at room temperature, we analyzed our polymerization system at -15oC. In the presence of Me6TREN and t-TACN (relatively weak bases) the polymerization rates diminished, but still could be considered faster than other reported organocatalysts that follow ESC mechanism at room temperature,10,13 but the stereoselectivity enhanced significantly (Table 2, entry 1,2) compared to that at room temperature. RESULTS AND DISCUSSIONS Interestingly, the first order plot for the ROP of rac-LA exhibits two discrete slopes (figure 8.4a) (kobs1 = 0.009, kobs2 = 0.006), selectivity factor calculated using these kobs values (s = kobs 1/ kobs 2 = 1.5) is more similar to that obtained from chiral HPLC analysis, 337 indicating that in the case of rac-LA that comprises of 1:1 ratio of D-LA and L-LA, when TU1/Me6TREN cocatalyst system is used initially the more selective monomer (D-LA) is consumed, but when the D-LA concentration is reduced in the course of the polymerization there is a strong competition between D-LA and L-LA to bind to the catalyst and hence we observe a decline in selectivity. To determine the effect of H-bond donor : H-bond acceptor ratio in stereoselective ROP, rac-LA was polymerized with TU1/Me6TREN, where the ratio between thiourea and base was changed from 1:1 to 2:1 and 3:1 respectively, no significant change was seen in the former instance (Table 8.1, entry 2). However, a slight enhancement in Tm and Pm and a decline in rate of polymerization was observed in the latter case when the TU1/Me6TREN ratio was changed to 3:1 (Table 8.1, entry 3) indicating 1:1 ratio of thiourea and base to be ideal with respect to maintaining stereoselectivity and enhancing rates. Despite these observations it is worth to note that chiral bis-thiourea TU1 reported here in, show significantly faster rates compared to other organocatalysts reported thus far that show similar stereoselectivity and follow ESC controlled sROP at room temperature.10,13 With the motivation of further optimizing the stereoselectivity and enhancing the rate of reaction, a variety of bases from weak to strong in combination with TU1 were screened at room temperature (Table 8.1) and at cold temperatures (-15oC) (Table 8.2). When the polymerizations were conducted at room temperature with TU1 and t-TACN or PMDTA, an enhanced rate was seen in the presence of t-TACN (Table 8.11 , entry 6), and with PMDTA the reaction slowed down than in the presence of Me6TREN and reached equilibrium at 75% monomer conversion (Table 8.1 , entry 7). However, in both cases there 338 was no observed enhancement in stereoselectivity, as indicated by the Pm values and the Tm of the produced PLA. was no observed enhancement in stereoselectivity, as indicated by the Pm values and the Tm of the produced PLA. RESULTS AND DISCUSSIONS When these reactions were performed with TU1 in the presence of stronger bases like BEMP and MTBD, the polymerization reached 90 % conversion in 90 min and 60 min respectively, producing highly isotactic polymers with good control over Mn and Mw/Mn (Table 2, entry 3,4). Previous studies report BEMP by itself as a successful catalyst for sROP of rac-LA to produce highly isotactic PLA within few minutes. However, the reports indicate limitations with respect to requirement of harsh conditions of reaction temperatures (-75oC) to achieve high stereoselectivity.8 This was also seen in our work, when the polymerization was conducted with BEMP in the absence of TU1 at -15oC 90 % monomer conversion was achieved in 5 min, but the resulting polymer showed no melting peak, indicating the formation of atatctic amorpous PLA. This observation indicated that the presence of TU1 is crucial in producing isotactic polymer at less harsh conditions of - 15oC or room temperature. In contrast to weak bases in the presence of strong bases, we believe the polymerization proceeds preferentially via CEC mechanism. This was depicted 339 by the first order plots for the polymerization D-LA and L-LA with TU1/BEMP (Figure 8.5) where no significant difference between the kobs values were observed like in the case of weak bases. Taking the observation thus far in to consideration Me6TREN was used as the base for the rest of the study. To comprehend the relationship between the electronics of chiral bis thioureas and its activity in sROP, a variety of thioureas (Figure 8.3) were studied using Me6TREN as a base at room temperature. When the polymerization was conducted with achiral thiourea TU8/Me6TREN, atactic PLA was produced (suggested by DSC analysis) where no melting peak was observed, and the catalyst showed no preference towards L-LA or D-LA (Figure 8.4b), indicating that the chiral moiety is an essential component for its activity as a stereoselective catalyst. Catalyst TU9 and TU10 was synthesized bearing slightly different chiral moieties. In the presence of TU9 rac-LA was rapidly converted to polymer, but the transformation lack stereoselectivity. TU10 carrying a proline group produced highly isotactic PLA, However, the transformation was much slower than with TU1 under similar reaction conditions. When TU2 with a strong electron withdrawing (EW) group on para position was studied (Table 8.1, entry 10), the reaction reached 90% monomer conversion at a slightly slower rate than TU1 and exhibited less stereo control. RESULTS AND DISCUSSIONS Polymerization conducted with TU3 where the EW CF3 groups were replaced by electron donating (ED) CH3 groups, there was no observed conversion in 24 h, this phenomenon is likely a consequence of the diminished polarity of thiourea making it a poor H-bond donor.15,20 Making use of these observations we then designed catalyst TU4 with EW groups closer to the chiral center and ED groups away from the chiral center with the intention of forcing the monomer to selectively bind to the thiourea moiety close to the Taking the observation thus far in to consideration Me6TREN was used as the base for the rest of the study. To comprehend the relationship between the electronics of chiral bis thioureas and its activity in sROP, a variety of thioureas (Figure 8.3) were studied using Me6TREN as a base at room temperature. When the polymerization was conducted with achiral thiourea TU8/Me6TREN, atactic PLA was produced (suggested by DSC analysis) where no melting peak was observed, and the catalyst showed no preference towards L-LA or D-LA (Figure 8.4b), indicating that the chiral moiety is an essential component for its activity as a stereoselective catalyst. Catalyst TU9 and TU10 was synthesized bearing slightly different chiral moieties. In the presence of TU9 rac-LA was rapidly converted to polymer, but the transformation lack stereoselectivity. TU10 carrying a proline group produced highly isotactic PLA, However, the transformation was much slower than with TU1 under similar reaction conditions. When TU2 with a strong electron withdrawing (EW) group on para position was studied (Table 8.1, entry 10), the reaction reached 90% monomer conversion at a slightly slower rate than TU1 and exhibited less stereo control. Polymerization conducted with TU3 where the EW CF3 groups were replaced by electron donating (ED) CH3 groups, there was no observed conversion in 24 h, this phenomenon is likely a consequence of the diminished polarity of thiourea making it a poor H-bond donor.15,20 Making use of these observations we then designed catalyst TU4 with EW groups closer to the chiral center and ED groups away from the chiral center with the intention of forcing the monomer to selectively bind to the thiourea moiety close to the 340 sterically hindered chiral center. In the presence of TU4 23 % of monomer was converted to polymer in 14 h and the reaction reached equilibrium thereafter. RESULTS AND DISCUSSIONS In the presence of TU5 an analogous structure to TU4 the polymerization reaches 42 % conversion in 18 h and no further conversion was observed. We believe the reason for low conversion with TU4 & TU5 is lack of internal activation of the thiourea moiety making it function more like a mono thiourea rather than a bis thiourea suggested by NMR binding studies (figure 8.9). The difference in conversion with TU4 vs TU5 could be a consequence of position of the bulky group on the catalyst. To confirm this observation chiral mono thiourea TU7 was synthesized and studied with Me6TREN as a base, the polymerization reaches equilibrium at 20% monomer conversion in 24 h corroborating our observation. With the aim of further enhancing the rate by manipulating the thiourea catalyst and taking advantage of a recent discovery by our group where the 5 atom linkers between thiourea moieties showed the fastest rates in bis-thiourea systems,15 we synthesized TU6, a leucine functionalized chiral thiourea with only 5 atoms between the thiourea moieties. As anticipated in the presence of TU6/Me6TREN, fastest rates were observed (Table 8.1, entry 13), yet the polymerization lack stereo control making it unfit to function as a stereoselective catalyst for this transformation. Understanding the mode of monomer activation is critical in comprehending the mechanism of stereoselectivity. Thiourea catalyst in the presence of weak bases are known to catalyze ROP of LA via neutral H-bonding mechanism, where LA was activated by the internally activated bis-thiourea, and the chain end by the base (scheme 8.1).12 Our group revealed that even in the presence of weak bases ROP of LA can proceed through an 341 imidate H-bonding mechanism depending on the acidity of the thiourea.15 Therefore, grasping the mechanism of catalyst activation is crucial to explain the rate enhancement by some thioureas, and why in the presence of others no enhancement of rates or stereoselectivity was observed. Replicating an NMR experiment to determine imidate or neutral H-bonding mechanism, thioureas TU1-TU5, with and without base was compared.3 TU1 and TU2, both show convergence of the aromatic and NH peaks (Figures 8.13) suggesting an imidate nature.3,20 Lack of polymerization in the presence of TU3, corroborates with the NMR experiment, where no to minimal peak shifts were observed in the downfield region, suggesting minimum interaction between the base and TU3. RESULTS AND DISCUSSIONS In the presence of Me6TREN, TU4 and TU5, exhibit broadening of NH peaks associated with the thiourea moiety and show a slight downfield shift of the ortho protons indicating neutral H-bonding mechanism.20 Accordingly, our observations revealed that bis-thiourea chiral catalyst that follow imidate mediated H-bonding ROP show enhanced rates and better stereo-control than the slower catalysts that follow neutral H-bonding mechanism. To further understand the mechanism of stereoselectivity, ratio between the peak intensities of the methine region of the 1H homonuclear decoupled NMR was determined. Previous reports indicate that if the polymerization followed CEC mechanism relative tetrad intensities would be [mrm] : [mrm] : [rmm] 1:1:1 and if it followed ESC [rmr] : [mmr] : [rmm] : [mrm] 1:1:1:2.2 However, after close analysis of the polymer produced in the presence of TU1/Me6TREN (Table 8.1, entry 1) at room temperature, the relative tetrad intensities (Figure 8.8) do not follow either one of the above patterns, suggesting both mechanisms prevail in the polymerization under the conditions reported here in. However, a higher inclination towards ESC mechanism is seen when we associate the Pm values with 342 the Tm obtained for PLA and by considering the kinetic resolution seen when D-LA and L- LA was polymerized using TU1. Apart from the reduced stereo-seletivity of the transformation, a major contributor of stereo-error during the course of polymerization is epimerization of rac-LA that can occur in the presence of bases.2 Close monitoring of the ROP of rac-LA at room temperature at 90% monomer conversion revealed no to minimum epimerization, suggesting negligible contribution towards depressing of the Pm. 343 CONCLUSION In summary, the bis-chiral thiourea bearing a tert-butyl leucine derivative (TU1) has been shown to be a highly effective enantio-selective catalysts that promotes fast stereo-selective ROP of rac-LA at room temperature in the presence of Me6TREN and at cold temperatures (-15 oC) with BEMP. The co-catalysts system not only shows high enantio-selectivity but also exhibits living behavior and produces polymers with precise control of molecular weight and narrow molecular weight distribution. Even though the catalysts system doesn’t produce perfectly isotactic PLA, to our knowledge the system reported herein show faster rates and comparable stereoselectivity at milder conditions compared to other organo-catalysts that have been reported thus far. We also believe there is potential for further manipulation of the catalysts design to improve the stereo-selectivity of these bis- chiral thiourea catalysts systems and build a library of chiral catalysts for the polymerization of other cyclic monomers bearing chiral functionality. 344 LIST OF REFERENCES (1) Stanford, M. J.; Dove, A. P. Stereocontrolled Ring-Opening Polymerisation of Lactide. Chem. Soc. Rev. 2010, 39 (2), 486–494. (2) Orhan, B.; Tschan, M. J. L.; Wirotius, A. L.; Dove, A. P.; Coulembier, O.; Taton, D. Isoselective Ring-Opening Polymerization of Rac-Lactide from Chiral Takemoto’s Organocatalysts: Elucidation of Stereocontrol. ACS Macro Lett. 2018, 7 (12), 1413–1419. (3) Dharmaratne, N. U.; Pothupitiya, J. U.; Bannin, T. J.; Kazakov, O. I.; Kiesewetter, M. K. Triclocarban: Commercial Antibacterial and Highly Effective H-Bond Donating Catalyst for Ring-Opening Polymerization. ACS Macro Lett. 2017, 6 (4), 421–425. (4) Fastnacht, K. V.; Spink, S. S.; Dharmaratne, N. U.; Pothupitiya, J. U.; Datta, P. P.; Kiesewetter, E. T.; Kiesewetter, M. K. Bis- and Tris-Urea H-Bond Donors for Ring-Opening Polymerization: Unprecedented Activity and Control from an Organocatalyst. ACS Macro Lett. 2016, 5 (8), 982–986. (5) Pothupitiya, J. U.; Dharmaratne, N. U.; Jouaneh, T. M. M.; Fastnacht, K. V.; Coderre, D. N.; Kiesewetter, M. K. H-Bonding Organocatalysts for the Living, Solvent-Free Ring-Opening Polymerization of Lactones: Toward an All-Lactones, All-Conditions Approach. Macromolecules 2017, 50 (22), 8948–8954. (6) Zhong, Z.; Dijkstra, P. J.; Feijen, J. Controlled and Stereoselective Polymerization of Lactide: Kinetics, Selectivity, and Microstructures. J. Am. Chem. Soc. 2003, 125 (37), 11291–11298. (6) 345 (7) Kan, Z.; Luo, W.; Shi, T.; Wei, C.; Han, B.; Zheng, D.; Liu, S. Facile Preparation of Stereoblock PLA from Ring-Opening Polymerization of Rac-Lactide by a Synergetic Binary Catalytic System Containing Ureas and Alkoxides. Front. Chem. 2018, 6 (NOV), 1–9. (8) Zhang, L.; Nederberg, F.; Messman, J. M.; Pratt, R. C.; Hedrick, J. L.; Wade, C. G. Organocatalytic Stereoselective Ring-Opening Polymerization of Lactide with Dimeric Phosphazene Bases. J. Am. Chem. Soc. 2007, 129 (42), 12610–12611. (9) Dove, A. P.; Li, H.; Pratt, R. C.; Lohmeijer, B. G. G.; Culkin, D. A.; Waymouth, R. M.; Hedrick, J. L. Stereoselective Polymerization of Rac- and Meso-Lactide Catalyzed by Sterically Encumbered N-Heterocyclic Carbenes. Chem. Commun. 2006, No. 27, 2881–2883. (10) Dove, A. P.; Pratt, R. C.; Lohmeijer, B. G. G.; Waymouth, R. M.; Hedrick, J. L. Thiourea-Based Bifunctional Organocatalysis: Supramolecular Recognition for Living Polymerization. J. Am. Chem. Soc. 2005, 127 (40), 13798–13799. (11) Makiguchi, K.; Yamanaka, T.; Kakuchi, T.; Terada, M.; Satoh, T. Binaphthol- Derived Phosphoric Acids as Efficient Chiral Organocatalysts for the Enantiomer- Selective Polymerization of Rac-Lactide. Chem. Commun. 2014, 50 (23), 2883– 2885. (12) Spink, S. S.; Kazakov, O. I.; Kiesewetter, E. T.; Kiesewetter, M. K. (13) Zhu, J. B.; Chen, E. Y. X. From Meso-Lactide to Isotactic Polylactide: (13) Zhu, J. B.; Chen, E. Y. X. From Meso-Lactide to Isotactic Polylactide: Bond Donating Catalysts for Ring-Opening Polymerization: Mechanistic Insights via (Non)Linear Free Energy Relationships. Macromolecules 2018, 51 (8), 3203– 3211. LIST OF REFERENCES Rate Accelerated Organocatalytic Ring-Opening Polymerization of l -Lactide via the Application of a Bis(Thiourea) H-Bond Donating Cocatalyst. Macromolecules 2015, 48 (17), 6127–6131. (13) Zhu, J. B.; Chen, E. Y. X. From Meso-Lactide to Isotactic Polylactide: (13) Zhu, J. B.; Chen, E. Y. X. From Meso-Lactide to Isotactic Polylactide: 346 Epimerization by B/N Lewis Pairs and Kinetic Resolution by Organic Catalysts. J. Am. Chem. Soc. 2015, 137 (39), 12506–12509. Epimerization by B/N Lewis Pairs and Kinetic Resolution by Organic Catalysts. J. Am. Chem. Soc. 2015, 137 (39), 12506–12509. Epimerization by B/N Lewis Pairs and Kinetic Resolution by Organic Catalysts. J. (14) Dharmaratne, N. U.; Pothupitiya, J. U.; Kiesewetter, M. K. The Mechanistic Duality of (Thio)Urea Organocatalysts for Ring-Opening Polymerization. Org. Biomol. Chem. 2019, 17 (13), 3305–3313. (15) Hewawasam, R. S.; Kalana, U. L. D. I.; Dharmaratne, N. U.; Wright, T. J.; Bannin, T. J.; Kiesewetter, E. T.; Kiesewetter, M. K. Bisurea and Bisthiourea H-Bonding Organocatalysts for Ring-Opening Polymerization: Cues for the Catalyst Design. Macromolecules 2019, 52 (23), 9232–9237. (16) Coudane, J.; Ustariz-Peyret, C.; Schwach, G.; Vert, M. More about the Stereodependence of DD and LL Pair Linkages during the Ring-Opening Polymerization of Racemic Lactide. J. Polym. Sci. Part A Polym. Chem. 1997, 35 (9), 1651–1658. (17) Belleney, J.; Wisniewski, M.; Le Borgne, A. Influence of the Nature of the Ligand on the Microstructure of Poly D,L-Lactides Prepared with Organoaluminum Initiators. Eur. Polym. J. 2004, 40 (3), 523–530. (18) Reisman, S. E.; Doyle, A. G.; Jacobsen, E. N. Enantioselective Thiourea- Catalyzed Additions to Oxocarbenium Ions. J. Am. Chem. Soc. 2008, 130 (23), 7198–7199. (19) Liu, S.; Li, H.; Zhao, N.; Li, Z. Stereoselective Ring-Opening Polymerization of Rac-Lactide Using Organocatalytic Cyclic Trimeric Phosphazene Base. ACS Macro Lett. 2018, 7 (6), 624–628. (20) Pothupitiya, J. U.; Hewawasam, R. S.; Kiesewetter, M. K. Urea and Thiourea H- 347 Bond Donating Catalysts for Ring-Opening Polymerization: Mechanistic Insights via (Non)Linear Free Energy Relationships. Macromolecules 2018, 51 (8), 3203– 3211. 348 entry TU/base [LA]/[BnOH]/[cat]/[base] Time (min) Conv. LIST OF REFERENCES % Mn (g/mol) Mw/Mn Pm Tm ESC CEC 1 TU1/Me6TREN 100/1/2/2 281 90 17,500 1.05 0.87 0.80 168 2 TU1/Me6TREN 100/1/2/1 354 90 17,600 1.05 0.87 0.81 168 3 TU1/Me6TREN 100/1/3/1 200 90 18,500 1.05 0.89 0.83 169 4 TU1/Me6TREN 100/2/2/2 90 91 9,500 1.04 0.86 0.80 161 5 TU1/Me6TREN 100/0.5/2/2 605 90 3,8000 1.04 0.88 0.81 159 6 TU1/t-TACN 100/1/2/2 165 90 19,000 1.04 0.85 0.77 156 7 TU1/PMDTA 100/1/2/2 1020 75 15,000 1.05 0.84 0.76 151 8 TU9/ Me6TREN 100/1/2/2 120 90 16,800 1.04 0.82 0.73 140 9 TU10/ Me6TREN 100/1/2/2 1140 84 0.88 0.80 10 TU2/Me6TREN 100/1/2/2 315 90 17,700 1.06 0.86 0.78 151 11 TU4/Me6TREN 100/1/2/2 840 23 6,000 1.34 0.85 0.77 - 12 TU5/Me6TREN 100/1/2/2 1080 42 0.81 0.72 13 TU6/Me6TREN 100/1/2/2 175 90 19,700 1.06 0.83 0.74 132 Table 8.1. Polymerization of rac-LA at room temperature. Reaction conditions: rac-LA (0.475 mmol, 0.95M), benzyl alcohol (2 mol%), DCM. monomer conversion was determined via 1H NMR. Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. Pm determined by 1H homonuclear decoupled NMR at 50 oC. Tm determined by DSC a explained in SI. entry TU/base [LA]/[BnOH]/[cat]/[base] Time (min) Conv. % Mn (g/mol) Mw/Mn Pm Tm ESC CEC 1 TU1/Me6TREN 100/1/2/2 480 90 17,700 1.06 0.92 0.87 184 2 TU1/t-TACN 100/1/2/2 240 89 17,600 1.05 0.92 0.88 183 3 TU1/BEMP 100/1/2/2 90 90 16,700 1.07 0.91 0.86 186 4 TU1/MTBD 100/1/2/2 60 91 16,600 1.08 0.90 0.86 184 5 BEMP 100/1/2/2 5 90 17,200 1.06 0.79 0.74 - Table 8.2. Polymerization of rac-LA at -15o C. Reaction conditions: rac-LA (0.475 mmol, Table 8.2. Polymerization of rac-LA at -15o C. Reaction conditions: rac-LA (0.475 mmol, 0.95M), benzyl alcohol (2 mol%), DCM. monomer conversion was determined via 1H NMR. Mn and Mw/Mn were determined by GPC (CH2Cl2) versus polystyrene standards. Pm determined by 1H homonuclear decoupled NMR at 50 oC. Tm determined by DSC a explained in SI. 349 Scheme 8.1. a. Neutral H-bonding mediated ROP of LA. b. LIST OF REFERENCES Imidtae H-bonding mediated ROP of LA O O O O + OH O O O O S N H R R N H R R HN S N H O H O O O O O H n O O O O S N R R N H R R HN S N H O H BASE BASE H O O O O S N R R N H R R HN S N H O H BASE H O O O O + OH O O O O O H n BASE TU TU BASE a b O O O O + OH O O O O S N H R R N H R R HN S N H O H O O O O O H n BASE TU BASE a Scheme 8.1. a. Neutral H-bonding mediated ROP of LA. b. Imidtae H-bonding mediated ROP f LA O O O O S N R R N H R R HN S N H O H BASE H O O O O S N R R N H R R HN S N H O H BASE H O O O O + OH O O O O O H n BASE TU b O O O O S N R R N H R R HN S N H O H BASE H O O O O S N R R N H R R HN S N H O H BASE H O O O O O H n SE U Scheme 8.1. a. Neutral H-bonding mediated ROP of LA. b. Imidtae H-bonding mediated ROP of LA Figure 8.1. LIST OF REFERENCES Microstructures of PLA formed by the polymerization of rac-LA O O O O O O O O S S R R L-LA D-LA I O O O O O O O O O O O O O O O OH R R R R S S S S Isotactic stereoblock PLA I O O O O O O O O O O O O O O O OH R R S S R R S S Heterotactic PLA I O O O O O O O OH R R R R Isotactic poly D-LA I O O O O O O O OH S S S S Isotactic poly L-LA I O O O O O O O O O O O O O O O OH R R S S S S R R Atactic PLA rac LA Isotactic stereoblock PLA Figure 8.1. Microstructures of PLA formed by the polymerization of rac-LA 350 a. Chain end-control mechanism Figure 8.2. Mechanisms of stereocontrolled ROP O O O O O O O O random process O O O O OH Bulky catalyst I I Bulky catalyst O O O O O I Bulky catalyst O O O O O O O O O O O O OH Chiral catalyst I I Chiral catalyst O O O O I Chiral catalyst Chiral catalyst Chiral catalyst Chiral catalyst Initiator Initiator selective process O b. Enantio-site control mechanism a. Chain end control mechanism O O O O O O O O random process O O O O OH Bulky catalyst I I Bulky catalyst O O O O O I Bulky catalyst Initiator b. Enantio-site control mechanism b. Enantio-site control mechanism Chiral catalyst Figure 8.2. Mechanisms of stereocontrolled ROP Figure 8.3. Thioureas and bases used in this study. LIST OF REFERENCES H N O tBu N H F3C CF3 N H S TU7 N H S N H N H O tBu H N R2 R2 H N R1 R1 S ’R R’ TU1 : R1 = R2 = CF3 R’ = H TU2 : R1 = R2 = H R’ = NO2 TU3 : R1 = R2 = CH3 R’ = H TU4 : R1 = CH3 R2 = CF3 R’ = H TU5 : R1 = CF3 R2 = CH3 R’ = H S F3C CF3 N H CF3 CF3 N H tBu O H N N H S N H TU6 S F3C CF3 N H CF3 CF3 N H N H S N H TU8 S CF3 CF3 N H CF3 F3C N H O H N H N S H N TU9 TU10 Bases N N N N N N N N N N N N N N tBu N N P N tTACN Me6TREN PMDTA MTBD BEMP S CF3 CF3 HN CF3 F3C N O N H N H S N H N N N N Me6TREN Figure 8.3. Thioureas and bases used in this study. 351 Figure 8.4. a. First order plots for the polymerization of D-LA, rac-LA and L-LA respectively with TU1/ Me6TREN. b. First order plots for the polymerization of L-LA, D- LA and rac-LA respectively with TU8/ Me6TREN. LIST OF REFERENCES y = 0.0691x - 0.0485 y = 0.0753x + 0.0011 y = 0.0726x - 0.0092 0 0.5 1 1.5 2 2.5 3 3.5 4 0 5 10 15 20 25 30 35 ln[LA]o/[LA] time (min) rac-LA+ BisTU (3C) L-LA+ BisTU (3C) D-LA+ BisTU (3C) y = 0.025x + 0.0853 y = 0.0091x + 0.0805 y = 0.0062x + 0.4187 y = 0.0074x + 0.084 0 0.5 1 1.5 2 2.5 0 50 100 150 200 250 300 350 ln([LA]o/[LA]) time (min) D-LA+ TU1 rac-LA+ TU1 rac-LA+ TU1 L-LA+ TU1 y = 0.0691x - 0.0485 y = 0.0753x + 0.0011 y = 0.0726x - 0.0092 0 0.5 1 1.5 2 2.5 3 3.5 4 0 5 10 15 20 25 30 35 ln[LA]o/[LA] time (min) rac-LA+ BisTU (3C) L-LA+ BisTU (3C) D-LA+ BisTU (3C) y = 0.025x + 0.0853 y = 0.0091x + 0.0805 y = 0.0062x + 0.4187 y = 0.0074x + 0.084 0 0.5 1 1.5 2 2.5 0 50 100 150 200 250 300 350 ln([LA]o/[LA]) time (min) D-LA+ TU1 rac-LA+ TU1 rac-LA+ TU1 L-LA+ TU1 Figure 8.4. a. First order plots for the polymerization of D-LA, rac-LA and L-LA respectively with TU1/ Me6TREN. b. First order plots for the polymerization of L-LA, D- LA and rac-LA respectively with TU8/ Me6TREN. Figure 8.5. First order plots for the polymerization of D-LA, rac-LA and L-LA respectively with TU1/ BEMP. y = 0.1918x - 0.002 y = 0.1806x + 0.0313 y = 0.0249x + 0.113 0 0.5 1 1.5 2 2.5 3 3.5 0 20 40 60 80 100 Time (min) ln([LA]o/[LA]) L-LA D-LA rac-LA Figure 8.5. First order plots for the polymerization of D-LA, rac-LA and L-LA respectively with TU1/ BEMP. 352 Figure 8.6. RI and UV GPC traces of the PLA initiated by 1-pyrenebutanol. Conditions: rac-LA (0.95 M, 0.475 mmol), 1-pyrenebutanol (2mol%, 0.02mmol), TU1 (2mol%, 0.05 mmol), Me6TREN (5 mol%, 0.05 mmol) in CH2Cl2. 17 17.5 18 18.5 19 19.5 20 20.5 21 21.5 22 RI UV Figure 8.6. RI and UV GPC traces of the PLA initiated by 1-pyrenebutanol. Conditions: rac-LA (0.95 M, 0.475 mmol), 1-pyrenebutanol (2mol%, 0.02mmol), TU1 (2mol%, 0.05 mmol), Me6TREN (5 mol%, 0.05 mmol) in CH2Cl2. Me6TREN (5 mol%, 0.05 mmol) in CH2Cl2. Figure 8.7. Mn (blue) and Mw/Mn (orange) catalyzed ring-opening polymerization of rac-LA. LIST OF REFERENCES Conditions: rac-LA (0.95 M, 0.95 mmol), benzyl alcohol (1mol%, 0.0095 mmol), TU1 (2mol%, 0.019 mmol), Me6TREN (2mol%, 0.019 mmol) in CH2Cl2. 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 100 Mw/Mn Mn (g/mol) conversion % 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 100 Mw/Mn Mn (g/mol) conversion % Figure 8.7. Mn (blue) and Mw/Mn (orange) catalyzed ring-opening polymerization of rac-LA. igure 8.7. Mn (blue) and Mw/Mn (orange) catalyzed ring-opening polymerization of rac-LA Conditions: rac-LA (0.95 M, 0.95 mmol), benzyl alcohol (1mol%, 0.0095 mmol), TU1 (2mol%, 0.019 mmol), Me6TREN (2mol%, 0.019 mmol) in CH2Cl2. 353 Figure 8.8. MALDI-TOF of the PLA resulting from TU1/Me6TREN cocatalyzed ROP of rac-LA. The peaks represent the whole repeat units m/z = (Na+ + benzyl alcohol + nLA). [M 14+Na] + [M 15+Na] + [M 16+Na] + Figure 8.8. MALDI-TOF of the PLA resulting from TU1/Me6TREN cocatalyzed ROP of rac-LA. Figure 8.8. MALDI-TOF of the PLA resulting from TU1/Me6TREN cocatalyzed ROP of rac-LA. Th k t th h l t it / (N + + b l l h l + LA) The peaks represent the whole repeat units m/z = (Na+ + benzyl alcohol + nLA). The peaks represent the whole repeat units m/z = (Na+ + benzyl alcohol + nLA). Figure 8.9. Homonuclear decoupled 1H NMR spectrum (400 MHz, CDCl3) of the methine region of PLA obtained from TU1 at 50 °C NMR (Table 8.1, entry 1) Tetrad Fitted Areas Relative Fitted Areas mmm 11086 0.739 mmr 721 0.048 rmm 1468 0.098 rmr 362 0.024 mrm 1358 0.091 Sum 14995 1.000 Figure 8.9. Homonuclear decoupled 1H NMR spectrum (400 MHz, CDCl3) of the methine region of PLA obtained from TU1 at 50 °C NMR (Table 8.1, entry 1) 354 Figure 8.10. DSC thermograms of PLA obtained at a heating and cooling rate of 5°C/min (2nd scan after annealing sample at 170 °C for 15 h), PLA produced by ROP at r.t (Table 8.1, entry 1). Figure 8.10. LIST OF REFERENCES DSC thermograms of PLA obtained at a heating and cooling rate of 5°C/min (2nd scan after annealing sample at 170 °C for 15 h), PLA produced by ROP at r.t (Table 8.1, entry 1). Figure 8.10. DSC thermograms of PLA obtained at a heating and cooling rate of 5°C/min (2nd scan after annealing sample at 170 °C for 15 h), PLA produced by ROP at r.t (Table 8.1, entry 1). Figure 8.11. DSC thermograms of PLA obtained at a heating and cooling rate of 5°C/min (2nd scan after annealing sample at 170 °C for 15 h), PLA produced by ROP at -15oC (Table 8.2, entry 1). Figure 8.11. DSC thermograms of PLA obtained at a heating and cooling rate of 5°C/min (2nd scan after annealing sample at 170 °C for 15 h), PLA produced by ROP at -15oC (Table 8.2, entry 1). 355 Figure 8.12. HPLC chromatograms of (run a ) L-LA, (run b) D-LA (run c ), rac-LA as a reference (run d) and the unreacted monomer at 47 % monomer conversion determined using a UV (254 nm) detector (Flow rate, 0.5 mL min-1; eluent, hexane/isopropanol = /3 2 C ) 50.2 % 49.7 % 59.2 % 40.7 % 50.2 % 49.7 % 59.2 % 40.7 % Figure 8.12. HPLC chromatograms of (run a ) L-LA, (run b) D-LA (run c ), rac-LA as a reference (run d) and the unreacted monomer at 47 % monomer conversion determined using a UV (254 nm) detector (Flow rate, 0.5 mL min-1; eluent, hexane/isopropanol = 7/3; temperature; 25 oC.). 7/3; temperature; 25 oC.). 7/3; temperature; 25 oC.). 356 Figure 8.13. Downfield portion of 1 H NMR spectra (400 MHz, ppm) of TU1-TU4 with and without Me6TREN in CH2Cl2 using a DMSO-d6 capillary. !! "#$%& !! !! "#$%& !! !! "#$%& !! !! "#$%& !! ' !! '"#$%& !! LIST OF REFERENCES !! "#$%& !! !! "#$%& !! !! "#$%& !! !! "#$%& !! ' !! '"#$%& !! Figure 8.13. Downfield portion of 1 H NMR spectra (400 MHz, ppm) of TU1-TU4 with and without Me6TREN in CH2Cl2 using a DMSO-d6 capillary. 357 357 Figure 8.14. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU1 ! "#$%#&' (()'+ N H S N H N H O tBu H N CF3 CF3 H N CF3 F3C S ! LIST OF REFERENCES "#$%#&' (()'+ N H S N H N H O tBu H N CF3 CF3 H N CF3 F3C S Figure 8.14. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU1 358 Figure 8.15. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU2. N H S N H N H O tBu H N H N S NO2 O2N N H S N H N H O tBu H N H N S NO2 O2N Figure 8.15. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU2. 1 13 Figure 8.15. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU2. Figure 8.15. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU2. d6 100 MHz, ppm) spectrum of TU2. 359 Figure 8.16. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d 100 MH ) t f TU3 N H S N H N H O tBu H N CH3 CH3 H N CH3 H3C S N H S N H N H O tBu H N CH3 CH3 H N CH3 H3C S Figure 8.16. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- Figure 8.16. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU3. Figure 8.16. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU3. Figure 8.16. (Upper) 1H NMR (DMSO-d6, 400 MHz, ppm), (Lower) 13C NMR (DMSO- d6 100 MHz, ppm) spectrum of TU3. 360 Figure 8.17. LIST OF REFERENCES (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 1 MHz, ppm) spectrum of TU5 !"#"$"%%" &'(( ))* +),-. /"-012 1 N H S N H N H O tBu H N CF3 CF3 H N CH3 H3C S !"#"$"%%" &&'&&()&+, -"+./(0/ !"#"$"%%" &'(( ))* +),-. /"-012 1 N H S N H N H O tBu H N CF3 CF3 H N CH3 H3C S !"#"$"%%" &&'&&()&+, -"+./(0/ Figure 8.17. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU5 361 Figure 8.18. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 6. LIST OF REFERENCES !"# $%%&%%" '()) + ,-./"0%".#12+3 2" H N O tBu N H F3C CF3 N H S !"# $%%&%%" '()) + ,-./"0%".#12+3 2" H N O tBu N H F3C CF3 N H S Figure 8.18. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 6. Figure 8.18. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 6. MHz, ppm) spectrum of TU 6. 362 Figure 8.19. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 7. LIST OF REFERENCES ! "#$#%#&&# '()) + ,-./ !0# .1+2 S F3C CF3 N H CF3 CF3 N H tBu O H N N H S N H ! "#$#%#&&# ' ( )+, !-# +./(0 ! "#$#%#&&# '()) + ,-./ !0# .1+2 S F3C CF3 N H CF3 CF3 N H tBu O H N N H S N H ! "#$#%#&&# S F3C CF3 N H CF3 CF3 N H tBu O H N N H S N H Figure 8.19. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 7. Figure 8.19. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 7. Figure 8.19. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 7. 363 Figure 8.20. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 9. LIST OF REFERENCES !"#"$"%%" &'(( )) +),-. /"-012 1 S CF3 CF3 N H CF3 F3C N H O H N H N S H N !"#"$"%%" &&' &&( )&+, -"+./(0 / !"#"$"%%" &'(( ))* +),-. /"-012 1 S CF3 CF3 N H CF3 F3C N H O H N H N S H N Figure 8.20. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 9. !"#"$"%%" &&' &&( )&+, -"+./(0 / Figure 8.20. (Upper) 1H NMR (CDCl3, 400 MHz, ppm), (Lower) 13C NMR (CDCl3 100 MHz, ppm) spectrum of TU 9. MHz, ppm) spectrum of TU 9. 364
https://openalex.org/W2969656698	https://www.scielo.br/j/clin/a/79Vm9kKGknPc7WmxNMQdnSt/?lang=en&format=pdf	English	null	Turn down - turn up: a simple and low-cost protocol for preparing platelet-rich plasma	Clinics	2,019	cc-by	5,107	KEYWORDS: Platelet-Rich Plasma; Regenerative Medicine; Orthopedics. KEYWORDS: Platelet-Rich Plasma; Regenerative Medicine; Orthopedics. Turn down - turn up: a simple and low-cost protocol for preparing platelet-rich plasma I Servico de Dor e Cuidados Paliativos, Grupo Hospitalar Conceicao, Hospital Nossa Senhora da Conceicao, Porto Alegre, RS, BR. II Centro de Pesquisa e Gerenciamento da Dor, Clinica Univida, Canoas, RS, BR. III Laboratorio de Analises Clinicas, Grupo Hospitalar Conceicao, Hospital Nossa Senhora da Conceicao, Porto Alegre, RS, BR. IV Public Health Institute, Medical School, Universidade do Porto, Porto, Portugal. V Servico de Reumatologia, Grupo Hospitalar Conceicao, Hospital Nossa Senhora da Conceicao, Porto Alegre, RS, BR. Machado ES, Leite R, Santos CC, Artuso GL, Gluszczak F, Jesus LG, et al. Turn down - turn up: a simple and low-cost protocol for preparing platelet-rich plasma. Clinics. 2019;74:e1132 s CC, Artuso GL, Gluszczak F, Jesus LG, et al. Turn down - turn up: a simple and low-cost protocol for preparing platelet-r 32 chado ES, Leite R, Santos CC, Artuso GL, Gluszczak F, Jesus LG, et al. Turn down - turn up: a simple and low-cost protocol sma. Clinics. 2019;74:e1132 Corresponding author. E-mail: ortospine@gmail.com Corresponding author. E-mail: ortospine@gmail.com OBJECTIVE: To describe and analyze a new protocol for the extraction of platelet-rich plasma (PRP) for use in clinical practice and compare this technique with methods that have been previously described in the medical literature. METHODS: We extracted PRP from 20 volunteers using four different protocols (single spin at 1600 g, single spin at 600 g, double spin at 300 and 700 g, and double spin at 600 and 900 g). In another group of 12 individuals, we extracted PRP with our new technique (named ‘turn down-turn up’) consisting of a double spin (200 g and 1600 g) closed system using standard laboratory equipment (including an ordinary benchtop centrifuge), where the blood remained in the same tube during all processes, reducing the risk of contamination. Platelet counts adjusted to baseline values were compared using analysis of covariance (ANCOVA). RESULTS: Using the four previously described protocols (mentioned above), we obtained concentrations of platelets that were 1.15-, 2.07-, 2.18-, and 3.19-fold greater than the baseline concentration, respectively. With the turn down-turn up technique, we obtained a platelet count that was 4.17-fold (95% conﬁdence interval (CI): 3.09 to 5.25) greater than the baseline platelet count (p=0.063 compared with the double spin at 600 and 900 g method). The total cost of the disposable materials used in the extraction process was less than US$10.00 per individual. ORIGINAL ARTICLE ORIGINAL ARTICLE Turn down - turn up: a simple and low-cost protocol for preparing platelet-rich plasma CONCLUSION: In the present study, we described a simple and safe method for obtaining P devices. Received for publication on December 28, 2018. Accepted for publi- cation on May 10, 2019 Box 1 Turn Down-Turn Up PRP Protocol - Double Spin - Closed System: 1. Collect the desired volume (8.5 ml) of blood through peripheral venous access directly into a vacuum tube with acid citrate dextrose (ACD) (1.5 ml). 2. Equalize the remaining vacuum in the tube. 3. Centrifuge the tube at 200 g for 15 minutes with the tube cap facing down (Figure 1). 4. Carefully remove the tube from the centrifuge and maintain the tube in the downward position without turning the tube. 5. Under aseptic conditions, aspirate 3.5 ml of the hematic layer through the rubber cap (Figure 2). 6. Turn the tube to an upright position (cap facing up). 7. Centrifuge the tube at 1600 g for 10 minutes with the lid facing up. 8. Under aseptic conditions, aspirate 3.5 ml of the superior portion of the material (platelet-poor plasma, PPP). 9. Aspirate the desired amount of PRP (1-2 ml) from the lower portion of the tube (Figure 3). The first step to carry out a clinical study on PRP is the standardization of the process for collecting and preparing the concentrate. Several previous studies have reported different methods of obtaining PRP using different centrifu- gation speeds and numbers of spins (11). Appropriate care during collection and manipulation and control of tempera- ture are also important to ensure that the early activation of platelets does not occur (19). Some investigators have pro- posed standard protocols for evaluating the quality of PRP, mainly for reproducibility purposes and for the inclusion of studies in systematic reviews and meta-analyses (20). In general, PRP preparation in clinical studies involves com- plex and expensive protocols, and the majority of studies do not provide sufficient information to allow for adequate reproducibility. A detailed, precise, and stepwise description of the PRP preparation protocol is required to allow com- parisons between studies. In the present study, we aimed to describe and evaluate a simple, inexpensive, and reproducible protocol for obtaining PRP for use in clinical practice. Figure 1 - The tube must be inserted with the cap facing down. ’ INTRODUCTION Protocols 3 and 4, which were double spin methods, consisted of centrifugation at 600 g (24,25) and 300 g (16,26,27) for 5 minutes. After careful collection of 2 ml of the lower portion of the plasma while avoiding aspirating the buffy coat, the material was again centri- fuged at 900 g for 15 minutes (protocol 3) or at 700 g for 17 minutes (protocol 4). demonstration of its potential tissue repair ability. Growth factors present in platelets are responsible for a regenerative stimulus. Vascular endothelial growth factor (VEGF), plate- let-derived growth factor AB (PDGF-AB), and transforming growth factor beta 1 (TGF-b1) are several growth factors released by platelets that are responsible for tissue repair (14). Cell proliferation, angiogenesis, and cell migration are stimulated by these growth factors, resulting in tissue repair and regeneration (15,16). The number of platelets per volume of blood varies between individuals, and the expression of growth factors is also not uniform (17). There is still no absolute definition of the number of platelets that is required for a regenerative stimulus. Ideally, the platelet concentra- tion should be close to 1,000,000 platelets/mL, and studies showing the best results in bone healing reported a plate- let concentration that is 3 to 8 times greater than the basal concentration (18). (21,22), and the samples were centrifuged at 1600 g for 3 minutes, while samples obtained by protocol 2 (23) were centrifuged at 600 g for 5 minutes. Both protocols involved single spin methods. Protocols 3 and 4, which were double spin methods, consisted of centrifugation at 600 g (24,25) and 300 g (16,26,27) for 5 minutes. After careful collection of 2 ml of the lower portion of the plasma while avoiding aspirating the buffy coat, the material was again centri- fuged at 900 g for 15 minutes (protocol 3) or at 700 g for 17 minutes (protocol 4). In the second part of the study, we investigated our protocol (‘turn down-turn up’) for obtaining PRP. The procedures are described in Box 1. The estimated time for preparation of PRP using this protocol was approximately 35 minutes. ’ INTRODUCTION minimally invasive therapies for orthopedics has led to the use of biological therapies, especially platelet-rich plasma (PRP), to optimize clinical outcomes and improve tissue recovery (5). Studies on the therapeutic applications of PRP have been conducted in several fields, such as maxillofacial surgery (6), cardiovascular surgery (7), and plastic surgery (8). In orthopedics, the clinical indications include hip and knee osteoarthritis, rotator cuff pathologies (8), epicondylitis and tendinitis (9), fracture healing (10), and back pain (11). A recent meta-analysis pooled the results of clinical trials of PRP injection in patients with knee osteoarthritis and demon- strated that PRP is superior to intra-articular injections of corticosteroids and hyaluronic acid (12). The incidence of articular degenerative diseases is increas- ing (1), resulting in a substantial economic and social burden (2). Even with advancements in understanding the biological mechanisms of joint diseases, a therapy that controls or reverses the degenerative process is not yet available in clinical practice. Surgical procedures (arthroplasties and arthrodesis) that are associated with high costs and a high rate of complications remain the definitive options for these types of lesions (3,4). The development of low-cost and PRP is defined as an autologous preparation obtained from the centrifuged peripheral blood of patients with a concentration that is 3-5 times greater than the basal con- centration (13). Through the separation of blood components (plasma, platelets, leukocytes and red blood cells) by centri- fugation, platelets are located in the lower portion of the plasma layer next to the hematic region. There has been increasing interest in the therapeutic effects of PRP upon the Copyright & 2019 CLINICS – This is an Open Access article distributed under the terms of the Creative Commons License (http://creativecommons.org/licenses/by/ 4.0/) which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original work is properly cited. No potential conﬂict of interest was reported. 1 A low-cost protocol for preparing platelet-rich plasma Machado ES et al. A low-cost protocol for preparing platelet-rich plasma Machado ES et al. A low-cost protocol for preparing platelet-rich plasma Machado ES et al. CLINICS 2019;74:e1132 CLINICS 2019;74:e1132 (21,22), and the samples were centrifuged at 1600 g for 3 minutes, while samples obtained by protocol 2 (23) were centrifuged at 600 g for 5 minutes. Both protocols involved single spin methods. Turn Down-Turn Up PRP Protocol - Double Spin - Closed System: ’ MATERIALS AND METHODS The first part of the study involved the reproduction of four different protocols based on a literature review. Healthy blood donors were selected from a hospital blood bank. After explaining the research objectives to patients and obtaining signed informed consent forms from the patients, 42.5 ml of peripheral venous blood was obtained from each patient. For coagulation analysis, blood was collected using blood col- lection tubes (BD Vacutainer ACD Solution A Ref. 364606) containing 3.2% sodium citrate at a volumetric ratio of 1.5 ml of anticoagulant to 8.5 ml of blood. A Sorologic Kasvi centrifuge (model K14-0815A) with a fixed angle rotor was used. The blood samples were analyzed on a Sysmex XE5000 hematology counter prior to the centrifugation procedures to establish the baseline values of platelets, red blood cells, and leukocytes in each patient. After obtaining the concentrate, a new count was obtained to enable the evaluation of the platelet concentration factor (the ratio between platelet counts in the concentrate and the original blood sample). The extraction of PRP was performed following four different protocols. Protocol 1 was used to obtain samples based on commercial kits mentioned in previous studies Figure 1 - The tube must be inserted with the cap facing down. 2 Figure 2 - Aspiration of the hematic layer with the cap facing down. Figure 3 - The ﬁnal appearance of the tube after the second spin. CLINICS 2019;74:e1132 A low-cost protocol for preparing platelet-rich plasma Machado ES et al. A low-cost protocol for preparing platelet-rich plasma Machado ES et al. A low-cost protocol for preparing platelet-rich plasma Machado ES et al. CLINICS 2019;74:e1132 Figure 3 - The ﬁnal appearance of the tube after the second spin. A low cost protocol for preparing platelet rich plasma Machado ES et al. Figure 2 - Aspiration of the hematic layer with the cap facing down. Figure 2 - Aspiration of the hematic layer with the cap facing down. Figure 3 - The ﬁnal appearance of the tube after the second spin. Figure 3 - The ﬁnal appearance of the tube after the second spin. +700 g protocol yielded a platelet concentration that was 2.21 times greater than the basal concentration. A higher concentration was obtained with the 600 g+900 g pro- tocol that was 3.09 times greater than the basal concentration (Table 1). ’ RESULTS For the first part of the study (the investigation of pre- viously described protocols), 20 male donors aged 26 to 54 years (mean age: 39.9 years, standard deviation (SD): 9.9 years) were selected. The baseline counts of the blood components were within the normal range of values (Table 1). The PRP extraction results are described in Table 1. For the single spin protocols, the mean final concentration of platelets was 1.17 times greater than the basal concentration with the 1600 g protocol and 2.15 times greater than the basal concentration with the 600 g protocol. The 300 g Ethics The present study was approved by the Institutional Review Board of our institution (Hospital Nossa Senhora da Conceic¸ão, Grupo Hospitalar Conceic¸ão) and registered at Plataforma Brasil (the Brazilian government’s registry of scientific studies) under CAAE (‘Certificado de Apresen- tac¸ão de Apreciac¸ão Ética’, Certificate of Presentation of Ethic Appreciation) number 53166416.1.0000.5530. Statistical analysis The platelet counts were compared with Student’s t-test and analysis of covariance (adjusted for baseline platelet counts) using IBM SPSS Statistics for Windows, Version 20.0 (IBM Corporation, Armonk, NY, USA). The 95% confidence intervals (CIs) were estimated for the ratio between the PRP and baseline concentrations. ( ) In the second part of the study (the investigation of the turn down-turn up PRP protocol developed by our team), another group of donors (10 males and 2 females) was selec- ted (mean age: 38.7 years, SD: 9.9 years, range: 28-54 years). Our protocol achieved a mean platelet concentration that was 4.07 times greater than the basal concentration with a mean platelet count of 749x103 platelets/mL (Table 1). This platelet concentration tended to be greater than that obtained with the 600 g+900 g protocol (Student’s t-test, p=0.059; ANCOVA adjusted for the baseline platelet count, p=0.063). Analysis of the PPP revealed negligible values (values ranging from 1 to 16 x 103 cells/uL) of red and white blood cells. The number of platelets and erythrocytes lost at the end of the first centrifugation was also evaluated, and the median count was 48.75 x 103 platelets/mL (26.4% of the baseline platelet count), which varied from 29 to 66 x 103 platelets/mL. Therefore, an average of 73.6% of the total number of plate- lets was recovered by our PRP preparation protocol. ’ DISCUSSION The use of PRP in various medical fields has resulted in the exponential growth of the number of studies using PRP therapy. In orthopedics, PRP serves as an interme- diate between conservative and surgical treatments (28). 3 CLINICS 2019;74:e1132 Table 1 - Results of the platelet counts using the different PRP extraction techniques that were investigated in this study. First part of the study (the investigation of previously described protocols; N=20). Baseline 1600 g (Protocol 1) 600 g (Protocol 2) 300+700 g (Protocol 3) 600+900 g (Protocol 4) Mean platelet count±SD (range of values)* 188.8±28.9 (152 to 248) 225.4±162.8 (17 to 524) 394.0±87.8 (202 to 524) 412.5±83.0 (249 to 594) 599.9±41.5 (355 to 1081) Fold change relative to the baseline value±SD (95% CI) ------------ 1.15±0.77 (0.813 to 1.49) 2.07±0.30 (1.94 to 2.2) 2.18±0.29 (2.05 to 2.31) 3.19±0.91 (2.79 to 3.59) Second part of the study (the investigation of the turn down-turn up technique); N=12. Baseline Turn down-turn up Mean Platelet count±SD (range of values)* 185.4±27.9 (152 to 225) 749.0±307.0 (426 to 1395) Fold change relative to the baseline value±SD (95% CI) ---------- 4.17±1.9 (3.09 to 5.25) * values represent thousands of platelets per mL. CI: conﬁdence interval. First part of the study (the investigation of previously described protocols; N=20). * values represent thousands of platelets per mL. CI: conﬁdence interval. single and double spin protocols The protocols using double spins were also investigated in this study but yielded heterogeneous results. Centrifugation at 300 g followed by 700 g achieved an intermediate platelet concentration (2.18 times greater than the baseline concentration). How- ever, in the original study conducted by Amable et al. (16), who obtained platelets at concentrations up to 7 times greater than the baseline concentration, all steps were performed in a refrigerated centrifuge with temperatures ranging from 8 to 19oC. The use of a refrigerated centrifuge is justified because the plasma density and viscosity differ at lower and higher temperatures (16). Since we aimed to identify a reproducible protocol for easy clinical use, the experiments that we did were conducted at room tempera- ture (22oC), and the centrifuges were not refrigerated. The diversity of protocols causes difficulties in comparing and/or reproducing results (13). Chahla et al. ’ DISCUSSION conducted a systematic review of studies investigating the use of PRP for the treatment of orthopedic pathologies and indicated that only 10% of these studies provided comprehensive reporting that included a clear description of the preparation protocol. They concluded that a detailed, precise, and stepwise description of the PRP preparation protocol is required to allow comparisons between studies and enable reproduci- bility (11). The need for consensus on the minimum reporting requirements for studies evaluating biologic treatments has been suggested by the American Academy of Orthopedic Surgeons. After a think tank symposium, a consensus of experts was published listing the minimum data that should be reported in studies to favor standardization of the analyses (29). The minimum information for clinical studies evaluating biologics in orthopedics (MIBO statement) is available at the following web address: www.mibo-state- ment.org. Some authors proposed methods of PRP classifica- tion. A more comprehensive and comparative investigative strategy has been proposed by Lana et al. (30): the MARSPILL classification describes the method (custom-made or commer- cial kit), activation, presence of red and white blood cells, number of spins, platelet concentration and number, light activation and image guidance of the application. This new classification focuses on mononuclear cells, which are as important as the platelet content due to their action on neovasculogenesis and cellular proliferation. g g Among the previously established protocols investigated in this study, the highest concentrations of platelets were obtained using the method reported by Jo et al. (24). In their original study, they achieved 633 x 103 platelets/uL, which was 4.3 times greater than the basal concentration of platelets, with a nonrefrigerated centrifuge. Using this method, we obtained a platelet concentration that was 3.09 times greater than the basal concentration. This difference may be explained by several factors, which we briefly discuss here. The results may have been influenced by the design of the centrifuge. Rotors with a fixed/nonfixed angle and the distance to the center of the rotor may result in different forces, leading to different concentration values, and the size of the blood tube may also influence the results (25). In addition to the lack of standardization of the methods to obtain PRP, the frequency and number of PRP injections remain controversial. In previous studies, the protocols for injections are highly variable, with up to 6 weekly applications in the lumbar muscles for the treatment of back pain (11). A low-cost protocol for preparing platelet-rich plasma Machado ES et al. follows the principles of low-speed centrifugation to pre- serve the maximum volume of platelets and growth factors (34,35) and uses discontinuous centrifugation to modulate and control platelet recovery. According to Perez et al., the centrifugal acceleration, time, size of the rotor, volume of processed blood, and prevention of platelet aggregation are the most relevant factors that must be controlled in the centrifugation step for the preparation of PRP (27). Their maximum platelet recovery was observed after centrifuga- tion at 100 g for 10 minutes in 3.5 ml tubes. Therefore, we adapted the first spin at 200 g for 15 minutes considering that the size of the tube used by us was 10 mL. We analyzed the volume of platelets lost in the hematic portion, and the values were within an acceptable range according to pre- vious studies. The very low number of platelets and white blood cells found in the PPP after the second centrifugation confirmed the high degree of platelet recovery of our tech- nique. We estimated that our cost of PRP preparation in a hospital facility is less than US$10.00 without considering the expenses of equipment and laboratory personnel. Various other centrifuge-based closed systems, such as PRP kits, are available commercially with a price ranging between US $300.00 to $1,500.00 (20). Our technique has the advantages of a closed system but with less cost and a low risk of contamination. 2. Dibonaventura Md, Gupta S, McDonald M, Sadosky A. Evaluating the health and economic impact of osteoarthritis pain in the workforce: results from the National Health and Wellness Survey. BMC Musculoskelet Disord. 2011;12:83. https://doi.org/10.1186/1471-2474-12-83 p g 3. Murphy NJ, Eyles JP, Hunter DJ. Hip Osteoarthritis: Etiopathogene- sis and Implications for Management. Adv Ther. 2016;33(11):1921-46. https://doi.org/10.1007/s12325-016-0409-3 p g 4. Mino DE, Munterich JE, Castel LD. Lumbar fusion surgery for degen- erative conditions is associated with significant resource and narcotic use 2 years postoperatively in the commercially insured: a medical and pharmacy claims study. J Spine Surg. 2017;3(2):141-8. https://doi.org/ 10.21037/jss.2017.04.02 j 5. LaPrade RF, Dragoo JL, Koh JL, Murray IR, Geeslin AG, Chu CR. AAOS Research Symposium Updates and Consensus: Biologic Treatment of Orthopaedic Injuries. J Am Acad Orthop Surg. 2016;24(7):e62-78. https://doi.org/10.5435/JAAOS-D-16-00086 p g 6. Annunziata M, Guida L, Nastri L, Piccirillo A, Sommese L, Napoli C. The Role of Autologous Platelet Concentrates in Alveolar Socket Preserva- tion: A Systematic Review. Transfus Med Hemother. 2018;45(3):195-203. 1. Park J, Mendy A, Vieira ER. Various Types of Arthritis in the United States: Prevalence and Age-Related Trends From 1999 to 2014. Am J Public Health. 2018;108(2):256-8. https://doi.org/10.2105/AJPH.2017.304179 A low-cost protocol for preparing platelet-rich plasma Machado ES et al. https://doi.org/10.1159/000488061 p g 7. Kirmani BH, Jones SG, Datta S, Mclaughlin EK, Hoschtitzky AJ. A meta- analysis of platelet gel for prevention of sternal wound infections fol- lowing cardiac surgery. Blood Transfus. 2017;15(1):57-65. https://doi. org/10.2450/2016.0231-15 8. Sommeling CE, Heyneman A, Hoeksema H, Verbelen J, Stillaert FB, Monstrey S. The use of platelet-rich plasma in plastic surgery: a systema- tic review. J Plast Reconstr Aesthetic Surg. 2013;66(3):301-11. https://doi. org/10.1016/j.bjps.2012.11.009 g j jp 9. Hurley ET, Lim Fat D, Moran CJ, Mullett H. The Efficacy of Platelet-Rich Plasma and Platelet-Rich Fibrin in Arthroscopic Rotator Cuff Repair: A Meta-analysis of Randomized Controlled Trials. Am J Sports Med. 2019; 47(3):753-61. https://doi.org/10.1177/0363546517751397 p g 10. Mi B, Liu G, Zhou W, Lv H, Liu Y, Wu Q, et al. Platelet rich plasma versus steroid on lateral epicondylitis: meta-analysis of randomized clinical trials. Phys Sportsmed. 2017;45(2):97-104. https://doi.org/10.1080/00913 847.2017.1297670 ’ CONCLUSIONS 11. Chahla J, Cinque ME, Piuzzi NS, Mannava S, Geeslin AG, Murray IR, et al. A Call for Standardization in Platelet-Rich Plasma Preparation Protocols and Composition Reporting: A Systematic Review of the Clin- ical Orthopaedic Literature. J Bone Joint Surg Am. 2017;99(20):1769-79. https://doi.org/10.2106/JBJS.16.01374 In the present study, we described a simple and safe method of obtaining PRP using low-cost devices. Our results were consistent with current standards of quality for PRP preparation for clinical trials suggested by the American Academy of Orthopedic Surgeons working group, who presented the MIBO statement (3). Compared with other currently employed techniques, our method may have signi- ficant advantages. The proposed technique does not demand expensive laboratory kits or a refrigerated centrifuge and has a low risk of contamination considering the use of a closed system. 12. Hussein M, Hussein T. Effect of autologous platelet leukocyte rich plasma injections on atrophied lumbar multifidus muscle in low back pain patients with monosegmental degenerative disc disease. SICOT J. 2016;2:12. https://doi.org/10.1051/sicotj/2016002 p g j 13. Campbell KA, Saltzman BM, Mascarenhas R, Khair MM, Verma NN, Bach BR Jr, et al. Does Intra-articular Platelet-Rich Plasma Injection Provide Clinically Superior Outcomes Compared with Other Therapies in the Treatment of Knee Osteoarthritis? A Systematic Review of Overlapping Meta-analyses. Arthroscopy. 2015;31(11):2213-21. https://doi.org/10.1016/ j.arthro.2015.03.041 j 14. Dhurat R, Sukesh M. Principles and Methods of Preparation of Platelet- Rich Plasma: A Review and Author’s Perspective. J Cutan Aesthet Surg. 2014;7(4):189-97. https://doi.org/10.4103/0974-2077.150734 CLINICS 2019;74:e1132 CLINICS 2019;74:e1132 ’ AUTHOR CONTRIBUTIONS p g 15. Sánchez-González DJ, Méndez-Bolaina E, Trejo-Bahena NI. Platelet-rich plasma peptides: key for regeneration. Int J Pept. 2012;2012:532519. https://doi.org/10.1155/2012/532519 Machado ES provided substantial contribution to the concept and design of the study, data acquisition, analysis and interpretation, manuscript drafting, critical revision for important intellectual content and ﬁnal approval. Leite R, Santos CC, Artuso GL provided substantial contribution to the concept and design of the study, data acquisition, manuscript critical revision for important intellectual content and ﬁnal approval. Caldas JM provided substantial contribution to the concept and design of the study, manuscript critical revision for important intellectual content and ﬁnal approval. Gluszczak F provided substantial contribution to data acquisition, analysis and interpretation, manuscript critical revision for important intellectual content and ﬁnal approval. Jesus LG provided substantial contribution to the data analysis and interpretation, manuscript drafting and ﬁnal approval. Bredemeier M provided substantial contribution to the data analysis and interpretation, manuscript drafting, critical revision for important intellec- tual content and ﬁnal approval. All of the authors take public responsibility for appropriate portions of the content. p g 16. Amable PR, Carias RB, Teixeira MV, da Cruz Pacheco I, Corrêa do Amaral RJ, Granjeiro JM, et al. Platelet-rich plasma preparation for regenerative medicine: optimization and quantification of cytokines and growth fac- tors. Stem Cell Res Ther. 2013;4(3):67. https://doi.org/10.1186/scrt218 p g 17. Sundman EA, Cole BJ, Fortier LA. Growth factor and catabolic cytokine concentrations are influenced by the cellular composition of platelet-rich plasma. Am J Sports Med. 2011;39(10):2135-40. https://doi.org/10.1177/ 0363546511417792 18. Weibrich G, Hansen T, Kleis W, Buch R, Hitzler WE. Effect of platelet concentration in platelet-rich plasma on peri-implant bone regeneration. Bone. 2004;34(4):665-71. https://doi.org/10.1016/j.bone.2003.12.010 p g j 19. Everts PA, Knape JT, Weibrich G, Schönberger JP, Hoffmann J, Overdevest EP, et al. Platelet-rich plasma and platelet gel: a review. J Extra Corpor Technol. 2006;38(2):174-87. 20. Chahla J, Cinque ME, Piuzzi NS, Mannava S, Geeslin AG, Murray IR, et al. A Call for Standardization in Platelet-Rich Plasma Preparation Protocols and Composition Reporting: A Systematic Review of the Clini- cal Orthopaedic Literature. J Bone Joint Surg Am. 2017;99(20):1769-79. https://doi.org/10.2106/JBJS.16.01374 ’ DISCUSSION Furthermore, there is evidence that a single injection in an arthritic knee could be enough (31). In Brazil, the Federal Medical Council currently recognizes PRP as a promising therapy, but its use is officially permitted only in clinical protocols that are duly evaluated and approved by the CEP/ CONEP system. The main argument is the lack of standardi- zation of PRP preparation and lack of information regarding the optimal dose and frequency of applications (32). There is still no consensus or scientific basis for the concentration of platelets that is necessary to induce tissue repair. However, there is evidence that even relatively low platelet concentrations may be effective. Sanchez et al. found that, compared with hyaluronic acid injection in the knee, using the single spin technique, which produced a concen- tration of platelets that was two-fold greater than the baseline concentration, showed a superior effect (33). Furthermore, most investigators suggest concentrations that are 4-5 times greater than the baseline concentration. A study conducted by Sugaya et al. (34) comparing PRP and bone marrow aspi- rate concentrate (BMAC) showed a wide range of basal platelet and growth factor concentrations between research volunteers. In the present study, we performed PRP extraction using four methods adapted from previously reported protocols. We obtained a low volume of platelet concentrate using single spin protocols in accordance with previous reports. Cavallo et al. (31) compared different formulations of PRP extraction and found a significant difference between the After reviewing the current literature and investigating four different techniques of PRP extraction, we developed a reproducible and low-cost closed protocol, which we named the turn down-turn up technique. This two-spin method 4 CLINICS 2019;74:e1132 22. Fitzpatrick J, Bulsara MK, McCrory PR, Richardson MD, Zheng MH. Analysis of Platelet-Rich Plasma Extraction: Variations in Platelet and Blood Components Between 4 Common Commercial Kits. Orthop J Sports Med. 2017;5(1):2325967116675272. https://doi.org/10.1177/232 5967116675272 29. Murray IR, Geeslin AG, Goudie EB, Petrigliano FA, LaPrade RF. Mini- mum Information for Studies Evaluating Biologics in Orthopaedics (MIBO): Platelet-Rich Plasma and Mesenchymal Stem Cells. J Bone Joint Surg Am. 2017;99(10):809-19. https://doi.org/10.2106/JBJS.16.00 793 23. Anitua E, Aguirre JJ, Algorta J, Ayerdi E, Cabezas AI, Orive G, et al. Effectiveness of autologous preparation rich in growth factors for the treatment of chronic cutaneous ulcers. J Biomed Mater Res B Appl Biomater. 2008;84(2):415-21. https://doi.org/10.1002/jbm.b.30886 30. Lana JFSD, Purita J, Paulus C, Huber SC, Rodrigues BL, Rodrigues AA, et al. Contributions for classification of platelet rich plasma - proposal of a new classification: MARSPILL. Regen Med. 2017;12(5):565-74. https:// doi.org/10.2217/rme-2017-0042 g 31. Cavallo C, Filardo G, Mariani E, Kon E, Marcacci M, Pereira Ruiz MT, et al. Comparison of platelet-rich plasma formulations for cartilage heal- ing: an in vitro study. J Bone Joint Surg Am. 2014;96(5):423-9. https://doi. org/10.2106/JBJS.M.00726 p g j 24. Jo CH, Roh YH, Kim JE, Shin S, Yoon KS. Optimizing platelet-rich plasma gel formation by varying time and gravitational forces during cen- trifugation. J Oral Implantol. 2013;39(5):525-32. https://doi.org/10.1563/ AAID-JOI-D-10-00155 g J J 32. Resoluc¸ão CFM 2128/2015. Available at: https://sistemas.cfm.org.br/ normas/visualizar/resolucoes/BR/2015/2128 J 25. Akhundov K, Pietramaggiori G, Waselle L, Darwiche S, Guerid S, Scaletta C, et al. Development of a cost-effective method for platelet-rich plasma (PRP) preparation for topical wound healing. Ann Burns Fire Disasters. 2012;25(4):207-13. 33. Sánchez M, Fiz N, Azofra J, Usabiaga J, Aduriz Recalde E, Garcia Gutierrez A, et al. A randomized clinical trial evaluating plasma rich in growth factors (PRGF-Endoret) versus hyaluronic acid in the short-term treatment of symptomatic knee osteoarthritis. Arthroscopy. 2012;28(8): 1070-8. https://doi.org/10.1016/j.arthro.2012.05.011 ( ) 26. Perez AG, Lana JF, Rodrigues AA, Luzo AC, Belangero WD, Santana MH. Relevant aspects of centrifugation step in the preparation of platelet-rich plasma. ISRN Hematol. 2014;2014:176060. https://doi.org/10.1155/2014/ 176060 p g j 34. Sugaya H, Yoshioka T, Kato T, Taniguchi Y, Kumagai H, Hyodo K, et al. Comparative Analysis of Cellular and Growth Factor Composi- tion in Bone Marrow Aspirate Concentrate and Platelet-Rich Plasma. Bone Marrow Res. 2018;2018:1549826. https://doi.org/10.1155/2018/1549826 27. Perez AG, Lichy R, Lana JF, Rodrigues AA, Luzo AC, Belangero WD, et al. ’ REFERENCES p g 21. Degen RM, Bernard JA, Oliver KS, Dines JS. Commercial Separation Systems Designed for Preparation of Platelet-Rich Plasma Yield Differ- ences in Cellular Composition. HSS J. 2017;13(1):75-80. https://doi.org/ 10.1007/s11420-016-9519-3 5 A low-cost protocol for preparing platelet-rich plasma Machado ES et al. CLINICS 2019;74:e1132 CLINICS 2019;74:e1132 Prediction and modulation of platelet recovery by discontinuous centrifugation of whole blood for the preparation of pure platelet-rich plasma. Biores Open Access. 2013;2(4):307-14. https://doi.org/10.1089/ biores.2013.0015 35. Choukroun J, Ghanaati S. Reduction of relative centrifugation force within injectable platelet-rich-fibrin (PRF) concentrates advances patients’ own inflammatory cells, platelets and growth factors: the first introduc- tion to the low speed centrifugation concept. Eur J Trauma Emerg Surg. 2018;44(1):87-95. https://doi.org/10.1007/s00068-017-0767-9 28. Navani A, Li G, Chrystal J. Platelet Rich Plasma in Musculoskeletal Pathology: A Necessary Rescue or a Lost Cause? Pain Physician. 2017; 20(3):E345-E356. 6 6
W2791704659.txt	https://zenodo.org/records/1124819/files/10.21307_ane-2017-059.pdf	en		Marble burying as compulsive behaviors in male and female mice	Acta Neurobiologiae Experimentalis	2,017	cc-by	4,938	Research paper Acta Neurobiol Exp 2017, 77: 254–260 Marble burying as compulsive behaviors in male and female mice George T. Taylor1,2, Sandra Lerch1, and Sabine Chourbaji1 1 Interfaculty Biomedical Research Facility (IBF), Heidelberg Universität, Heidelberg, Germany, 2 Behavioral Neuroscience, University of Missouri, St. Louis, USA, * Email: geot@UMSL.edu Marble burying is considered an, albeit controversial, animal model of the compulsive like behaviors of obsessive‑compulsive disorder (OCD). Hallmark features of OCD patients are similarities and, more prominent, differences from anxiety disorders, e.g., the absence of sex differences and resistance to spontaneous remission. We report an experiment on marble burying by male and female C57/BL6/N mice. Animals were administered either the classic anxiolytic drug, diazepam, that targets the GABA receptor or a “pure” inhibitor of the serotonin transporter, escitalopram, that has been reported to be particularly effective in OCD. A burying paradigm that more precisely mimics the human condition was used, e.g., testing in the home environment, chronic drug exposure and acknowledging individual differences by pre‑selecting for high marble burying. Results were that there were no sex differences in groups treated with drugs or in control mice. Both diazepam and escitalopram decreased numbers of marbles buried compared to vehicle‑only controls in the absence of correlated changes in anxiety. Diazepam, however, was more effective than escitalopram in suppressing MB. The conclusion is that along with serotonin, GABA is involved in regulating compulsive behaviors. The marble burying paradigm may prove more useful for pharmacological drugs tests of impulsivity or attention deficit because of the involvement of serotonin and GABA in both disorders. Key words: animal model, obsessive‑compulsive disorder, anxiety, OCD, sex differences, GABA, serotonin © 2017 by Acta Neurobiologiae Experimentalis INTRODUCTION Obsessive‑compulsive disorder, (OCD) is character‑ ized by unwanted, intrusive thoughts and images (ob‑ sessions) and repetitive, ritualistic behaviors (behavior‑ al compulsions). The latter presumably serves to reduce anxiety caused by the obsessions. Yet, OCD now is sepa‑ rated from anxiety disorders in DSM‑V, largely based on the obsessional component (American Psychiatric Asso‑ ciation 2013). Our research interests are in developing and assessing reliable and valid animal models for psychiatric condi‑ tions. OCD has proven to be a particularly difficult condi‑ tion to model. The task of animal modelers is to develop behavioral measures that isolate compulsions from anxi‑ ety measures. Spontaneous burying of marbles in rodents has been suggested as a compulsive‑like behavior (Broekkamp et al. 1986, Deacon 2006, Gyertyan 1995). Marble burying (MB), however, has been criticized on both conceptual and empirical grounds for its ability to serve as a unique benchmark of OCD (Albelda and Joel 2012, Wolmarans de Received 7 June 2017, accepted 27 June 2017 7_1143 Taylor_v5.indd 254 et al. 2016). Indeed, MB has also appeared in the literature as a measure for autism, motivation or general anxiety disorder (GAD) (Ene et al. 2016, Jury et al. 2015, Silverman et al. 2015). Despite the criticisms, MB continues to ap‑ pear regularly in the literature as a measure of compul‑ sion (Gawali et al. 2016, Kudryashov et al. 2016, Nichols et al. 2016, Satta et al. 2016). We designed an experiment with mice to assess MB in relation to unique features in OCD patients. 1) In con‑ trast to the notably higher frequencies of most anxiety disorders in women, incidences of OCD have no reliable sex differences (Martin 2003). Our experiment used both males and female mice. 2) There are clearly individu‑ al differences in compulsive behaviors among people. Our experiment used pre‑tests to select the mice who were most likely to bury marbles. 3) Unlike the anxiety disorders, untreated OCD frequently fails to remit with the passage of time (Taylor et al. 2011). Our experiment tested the selected mice repeatedly to determine spon‑ taneous reductions as marbles became familiar. 4) The consensus is that the neural circuits for OCD and anxiety differ (Burguiere et al. 2015, Hoffman 2011). Neurotrans‑ mitters underlying OCD are the monoamines, mostly Correspondence should be addressed to: George T. Taylor Email: geot@UMSL.edu 23/10/17 21:48 Marble Burying Animal Model 255 Acta Neurobiol Exp 2017, 77: 254–260 serotonin and, likely, glutamate (Bokor and Anderson 2014, Egashira et al. 2008). GABA remains the primary transmitter thought to underlie most anxieties. Our study compared a classic benzodiazepine, diazepam, and a SSRI, escitalopram, that has proven particularly effec‑ tive for OCD (Shim et al. 2011, Zohar 2008). Finally, all animals were examined in the open field apparatus as a general measure of anxiety. Hypotheses for the study included no sex differences in MB, consistency over time for individual MB habits in un‑ treated animals and greatest effectiveness of escitalopram in reducing MB. METHODS Animals A total of 35 male and 35 female C57/BL6/N mice, 40 days of age obtained from Charles River (Sulzfeld,Ger‑ many), were acclimatized for 2 weeks before pretests were conducted. After pretesting for spontaneous MB, 42 mice equally divided between sexes were selected as subjects for the experiment. All mice were individually housed in flat bottom plastic Macrolon type II cages measuring 360 cm2 (Tecniplast, Italy) under SPF conditions. Standard lab diet (Rod16A, LASvendi, Soest) and water were avail‑ able ad libitum. The colony room lighting was a 12:12 h reversed light/dark cycle with lights off at 9am; room temperature (20–22 °C) and relative humidity (50%) are controlled automatically. The Institutional Animal Care and Use Committee and the local authorities (Regierung‑ spräsidium Karlsruhe) approved the experimental proto‑ col (permit number: G‑37/15). Materials All behavioral sessions were conducted during the nocturnal light cycle and under dim illumination. The open field apparatus (50×50×50 cm) was constructed of black Plexiglas. Movement of the animal in the open field was measured automatically by the Ethovision 4.0 track‑ ing system (Noldus Information Technology, Wageningen, The Netherlands) via an overhead, infrared camera (Ike‑ gami Digital). The tracking system can record locomotor movement in the various quadrants of the open field. The innermost quadrant (25×25cm) was designated as the center arena. After each test, the apparatus was cleaned with 70% ethanol. Marbles used were multi‑colored and approximately 16 mm diameter. The home cage of the an‑ imal was used for measurement of MB. Bedding consisted of aspen wooden chips (ABEDD LTE‑001, Lab & Vet Service, Vienna, Austria) that was approximately 5 cm deep. Bed‑ 7_1143 Taylor_v5.indd 255 ding was changed weekly but never on the days before a behavioral test. Diazepam ampules were purchased (Boeringer, Mann‑ heim, Germany). Escitalopram was purchased (Sigma‑Al‑ drich Chemical Company, St. Louis, MO, USA) and solubi‑ lized in a 0.9% saline solution for injection. Experimental Design Following an initial test to identify the tendency of each mouse to bury marbles, males (N=21) and females (N=21) were selected that had buried at least 6 marbles in the pre‑test and randomly assigned to groups. There were 3 groups of each sex (n=7 per grp) that were s.c. injected daily for 11 days with either 0.9% saline (Fresenius Kubi, Bad Homburg) vehicle (Veh only), 2 mg/kg bwt diazepam (Diaz) or 2 mg/kg bwt escitalopram (Esc). These drug dosages are in the low to mid ranges of those employed in the rodent literature (Erfanparast and Tamaddonfard 2015, Nicolas et al. 2006, Pandey et al. 2009, Schneider and Popik 2007). Procedures Pre‑selection tests of animals for the experiment were completed over 2 days. The pre‑selection MB trial was treated as a pre‑test, i.e., prior to drug treatments. Using the procedure described below for MB, the 35 fe‑ males and 35 females were tested for spontaneous MB in their home cages. The 21 mice of each sex burying the most marbles were retained for the experiment, and the other mice were removed to another animal hous‑ ing room. Drug administrations began at the beginning of behavioral tests of the animals. Injections were done 1 hr before tests. Behavioral sessions were conducted over 11 days in the open field and in the home cages. Order of tests were counterbalanced between and within groups. In addition to the pretests, experimental subjects were given 2 tests in both apparatus for a total of 4 tests separated by at least 2 days. Test 1 behaviors were conducted during the initial 4 days and Test 2 was conducted during the last 4 days of the 11 days of testing. Animals were injected on all days, including “off” days. For a session in the open field, a mouse was re‑ moved from the colony room to an adjacent experi‑ mental room. The animal was placed in the center of the open field apparatus and movement was record‑ ed over the 30 min session. Also recorded was time spent in the center quadrant of the apparatus. The open field is a marker of activity changes under drug influences and time in the central area relative to the 23/10/17 21:48 256 GT. Taylor et al. other areas adjacent to the walls serves as a measure of anxiety (Archer et al. 1987, Benatti et al. 2014, Ene et al. 2016). The procedure used for MB followed the paradigm used commonly in the literature (Deacon 2006, Gawali et al. 2016, Witkin 2008), except that we conducted the tests in the home cages rather than in novel cages. Log‑ ic was to mimic human OCD in which compulsive be‑ haviors occur with unsettling disturbances of a famil‑ iar environment. For a test, the home cage was moved from the cage rack to a nearby table and the mouse removed from its home cage to a holding cage for ap‑ proximately 1 min. During that time, 12 marbles were distributed equally around the perimeter of the home cage at least 2 cm from the walls. Marbles were placed on top of the approximately 5 cm‑deep wood chip bed‑ ding of the home cage. The mouse was returned to its home cage that was placed back into its normal place in the cage rack. After 30 min, the cage was again moved to the table, and the animal placed in the holding cage while the numbers of marbles buried were counted. Although some marbles were buried out of sight, most often the marbles were buried only partially. We counted the marble buried if it was covered half or more by the bedding. Statistical Analyses Assessment of behavioral differences among groups was accomplished with 3‑way analyses of variance Fig. 1. Numbers of marbles buried by male and female mice during a pre‑drug test and over two trials during 11 daily administrations of either vehicle only (Veh), escitalopram (Esc) or diazepam (Diaz). An asterisk () indicates significant differences (p<0.05) of drug groups from veh controls. The double asterisk () indicates Diaz groups differed significantly (p<0.05) from the Esc groups. 7_1143 Taylor_v5.indd 256 Acta Neurobiol Exp 2017, 77: 254–260 (ANOVAs). The first 3 x 2 x 3 ANOVA on numbers of marbles buried had main factors of Drug (diazepam, es‑ citalopram or vehicle only) x Sex with Trial (Pre‑test, Trial 1 and Trial 2) as a repeated measure. Open field activity used a similar 3×2×2 arrangement, except with 2 trials as repeated measure. Numbers of marbles bur‑ ied were the primary measure of compulsive behavior. Time in the center arena of the open field served as the measure of anxiety. Distance traveled, in cm, was a measure of general activity. Post‑hoc Tukey‑Kram‑ er tests were used for pair‑wise comparisons of mean group differences. The p<0.05 confidence value was used for all analyses. RESULTS Numbers of marbles buried in the home cage of the different groups of mice over the three trials (Pre‑test and 2 trials with drug exposure) are depicted in Fig. 1. Results of the ANOVA indicated a statistically significant difference for the main effects for drug F(2,35)=11.45, P<0.001, partial eta squared (η2)=0.396, and for the re‑ peated factor of trials, F(2,35)=25.41, P<0.001, η2=0.417. The only 2‑way interaction that was statistically reli‑ able was trial x drug F(4,70)=10.59, P<0.001, η2=0.377, al‑ though trial x sex approached significance, F(2,70)=2.82, P=0.067, η2=0.074. The 3‑way interaction failed to achieve significance. Posthoc group comparisons revealed statistically significant differences (p<0.05) between and within groups. The groups did not differ on the pre‑test data. However, between group differences showed the mice administered diazepam burying the fewest marbles on Fig. 2. Distance traveled by male and female mice in the open field during a pre‑drug test and over trials during exposure either vehicle only (Veh), escitalopram (Esc) or diazepam (Diaz). There were no differences between groups although the combination of female groups differed significantly (p<0.05) the combinations of male groups. 23/10/17 21:48 Acta Neurobiol Exp 2017, 77: 254–260 Trials 1 and 2 and the control animals the most, with the escitalopram mice differing from both groups on both trials. Within group differences indicate that the control animals did not change burying over the pre‑trial and two drug trials, but the drug groups did change over that period. Diazepam animals reduced burying from Pre‑trial to Trial 1 but no further reduc‑ tions for Trial 2. The escitalopram group had yet a third pattern, decreasing burying from the Pre‑trial to Trial 1 and then increasing again for Trial 2. There were no reliable differences between male and female mice in any of the MB comparisons. Results for the distance traveled measure of general locomotor activity (Fig. 2) revealed no between group differences and only the single main factor of sex was statistically reliable, F(1,35)=7.02, P=0.012, η2=0.167. The females were more active than the male mice, indepen‑ dent of drug treatments. None of the interactions with the sex factor was statistically significant. Examination of time spent in the center arena of the open field was conducted as a measure of anxiety. Results are in Fig. 3. The 2×3×2 factorial analysis indicated the sex × drug interaction was statistically reliable, F(1,35)=3.61, P=0.037, partial eta squared (η2)=0.171. Further analyses revealed that the females in the diazepam group spent the most time in the center arena, and, surprisingly, fe‑ males of the escitalopram group had the least time in the center area. All other groups were statistically similar. The 3‑way interaction nor any of the other 2‑way inter‑ actions achieved statistical significance. Within group results were that only the Trials main effect was significant, F(1,35)=10.30, P=0.003. Overall, both sexes spent more time in the center arena during their second trial than the first trial. Fig. 3. Time in a central arena of the open field by male and female mice over two trials during 11 daily administrations of either vehicle only (Veh), escitalopram (Esc) or diazepam (Diaz). An asterisk () indicates significant differences (p<0.05) from the other groups. 7_1143 Taylor_v5.indd 257 Marble Burying Animal Model 257 DISCUSSION Results of the experiment included no sex differences in control and drug groups of mice selected for high MB. Males and females administered only vehicle showed consistent levels of burying behaviors while both sexes administered diazepam and escitalopram reduced their MB. Diazepam was more effective in eliminating the be‑ havior than escitalopram. Indeed, the influence of escit‑ alopram on burying appeared to weaken over time. The implication is that, along with the serotonergic system, GABA neurotransmission is critically involved in MB. The literature places emphasis on serotonin in both patients and animal models of compulsive behaviors. Findings of certain SSRIs being the first line treatment for OCD and that those same SSRIs reduce MB have re‑ inforced the serotonin hypothesis (Egashira et al. 2008). Escitalopram is a notable example (Stein et al. 2008, Wol‑ marans de et al. 2016). However, longer durations and higher doses of the SSRIs often are needed to treat OCD patients compared to other psychiatric disorders (Bokor and Anderson 2014). More telling, SSRIs fail to reduce symptomology in 40‑60% of OCD patients (Pallanti and Quercioli 2006), although benzodiazepines were even less effective (Goddard et al. 2008). The clear implication is that OCD is a complex disorder that involves multi‑ ple systems rather the current emphasis on the sero‑ tonin receptor (Egashira et al. 2008, Marazziti et al. 2010, Takeuchi et al. 2002). The value of MB as an animal model for OCD remains an open question (Albelda and Joel 2012). Nonetheless, there are empirical reasons indicating that rodents do not bury objects simply because they are anxious. MB has features not observed in other anxiety paradigms. For example, burying will occur in the safety of the home cage, in the absence of obvious fear or stressful stimuli and burying fails to habituate over test sessions (Chotiwat and Harris 2006, Greene‑Schloesser et al. 2011, Thomas et al. 2009). Our findings provide additional evidence. The most common behaviors in OCD patients are compulsive check‑ ing, involving the performance of routines related to se‑ curity, orderliness, and accuracy but without resolution (Taylor et al. 2011). Our findings confirm that the absence of a reduction in compulsive burying of the control an‑ imals. Over time the control animals revealed a similar resistance to extinction of burying despite becoming fa‑ miliar with the marbles. Finally, there was no obvious re‑ lation in our study between MB and time in the center arena of our open field, a measure of anxiety (Benatti et al. 2014). That measure indicated females administered diazepam were least anxious, but escitalopram females were the most anxious. All other group comparisons were not significantly different. The data for general locomo‑ 23/10/17 21:48 258 GT. Taylor et al. tor activity indicated that, over all groups, females were more active. Untreated female rodents often are found to more active than untreated males (Blizard et al. 1975, Palanza et al. 2001, Taylor et al. 2011). A combination of methodological features makes our experiment a unique contribution to this literature. An‑ imals were tested in their familiar home cages rather than in a novel, neutral apparatus (Witkin 2008). We ac‑ knowledged the fact that there are individual differences in compulsive behaviors by pre‑selecting mice that bur‑ ied marbles (Fineberg et al. 2015, Wirth‑Dzięciołowska et al. 2005). Whereas we examined burying by both sexes, almost all previous reports have used males, even when testing the influence of ovarian steroids (Gomez et al. 2002, Umathe et al. 2009). We compared diazepam, a clas‑ sic benzodiazepine and a GABA agonist (Nicolas et al. 2006), with escitalopram, an SSRI that has been described as a pure inhibitor of the serotonin transporter (Stahl 2013). And animals were chronically exposed to the drugs as opposed to acute treatments in most reports in the lit‑ erature (Jimenez‑Gomez et al. 2011). Yet, diazepam proved more effective (Joel et al. 2004) than escitalopram in decreasing MB. The present experi‑ ment, essentially, failed to resolve MB as an animal model capable of dissociating compulsive and anxious behaviors (Albelda and Joel 2012). This may not be a failing so much as empirical sup‑ port for OCD and some forms of anxiety being inseparable (Schneier et al. 2008), despite the newest DSM moving OCD from anxiety categories (American Psychiatric Association 2013). Notably, both disorders share neural pathology of the cortico‑striatal‑thalamic‑cortical circuitry (Milad and Rauch 2007, Stahl 2013). Generalized anxiety disorder and OCD both show frontal/striatal hyperactivity. Social anxi‑ ety and OCD both show similar anterior cingulate dysfunc‑ tion (Kim and Gorman 2005). We remain convinced the MB paradigm has import‑ ant potential as an animal model for psychiatric disor‑ ders. The puzzle is that it is not clear what is actually being measured. Perhaps it is too narrow of a perspec‑ tive to focus on anxiety or compulsion as the only pos‑ sibilities. It is entirely possible that there are other di‑ mensions being measured, for example, impulsivity that is neither specifically anxiety or specifically compulsive. Indeed, it has been proposed that individual differenc‑ es in patients suggest a continuum of compulsivity to impulsivity (Allen et al. 2003, Geller 2006). DSM‑V indi‑ cates that impulsivity may be performed in patients for pleasure or gratification rather than relief of tension or anxiety. MB could prove useful for innovative pharma‑ cological treatments for impulsive‑control, attention deficit and related psychiatric disorders. These are con‑ ditions for which both serotonin and GABA, as well as dopamine, glutamate and neurosteroids, have been im‑ 7_1143 Taylor_v5.indd 258 Acta Neurobiol Exp 2017, 77: 254–260 plicated (Hoffman 2011, Perry et al. 2011, Schule et al. 2011, Yates et al. 2012). We believe a fresh look at the MB paradigm is warranted. CONCLUSIONS Our results suggest the conclusion that, independent of sex, marble burying can be suppressed with therapeutic drugs used to treat both anxious and compulsive patients. However, MB cannot clearly distinguish compulsions from anxiety. MB remains an intriguing animal model partly because burying objects appears to be an inherent trait of some, but not all, mice. Moreover, burying has the virtues of ease, reliability and sensitivity to drug treatments, A broader perspective, thinking “outside the box,” may reveal MB as useful for pharmacological drugs tests of im‑ pulsivity, attention deficit disorder or other psychiatric disorders that have proven difficult to model in animals. ACKNOWLEDGEMENTS This research was supported in part by a grant to GTT from the Alexander von Humboldt Foundation, Bonn, Ger‑ many. We thank the animal caretakers of the clinical and experimental facility of the IBF for their support. CONFLICT OF INTERESTS The authors report no conflicts of interest. REFERENCES Albelda N, Joel D (2012) Current animal models of obsessive compulsive disorder: an update. Neuroscience 211: 83–106. Allen A, King A, Hollander E (2003) Obsessive‑compulsive spectrum disorders. Dialog Clin Neurosci 5: 259–271. Archer T, Fredriksson A, Lewander T, Soderberg U (1987) Marble burying and spontaneous motor activity in mice: interactions over days and the effect of diazepam. Scand J Psychol 28: 242–249. American Psychiatric Association (2013) Diagnostic and Statistical Manual of Mental Disorders. 5th edn. Am Psychiatr Assn: Washington, D.C. Benatti C, Alboni S, Blom JM, Gandolfi F, Mendlewicz J, Brunello N, Tascedda F (2014) Behavioural and transcriptional effects of escitalopram in the chronic escape deficit model of depression. Behav Brain Res 272: 121–130. Blizard DA, Lippman HR, Chen JJ (1975) Sex differences in open‑field behavior in the rat: The inductive and activational role of gonadal hormones. Physiol Behav 14: 601–608. Bokor G, Anderson PD (2014) Obsessive‑compulsive disorder. J Pharm Pract 27: 116–130. Broekkamp CL, Rijk HW, Joly‑Gelouin D, Lloyd KL (1986) Major tranquillizers can be distinguished from minor tranquillizers on the basis of effects on 23/10/17 21:48 Acta Neurobiol Exp 2017, 77: 254–260 marble burying and swim‑induced grooming in mice. Eur J Pharmacol 126: 223–229. Burguiere E, Monteiro P, Mallet L, Feng G, Graybiel AM (2015) Striatal circuits, habits, and implications for obsessive‑compulsive disorder. Curr Opin Neurobiol 30: 59–65. Chotiwat C, Harris RB (2006) Increased anxiety‑like behavior during the post‑stress period in mice exposed to repeated restraint stress. Horm Behav 50: 489–495. Deacon RM (2006) Digging and marble burying in mice: simple methods for in vivo identification of biological impacts. Nature Protocol 1: 122–124. Egashira N, Okuno R, Matsushita M, Abe M, Mishima K, Iwasaki K, Nishimura R, Oishi R, Fujiwara M (2008) Aripiprazole inhibits marble‑burying behavior via 5‑hydroxytryptamine (5‑HT)1A receptor‑independent mechanisms. Eur J Pharmacol 592: 103–108. Ene HM, Kara NZ, Barak N, Reshef Ben‑Mordechai T, Einat H (2016) Effects of repeated asenapine in a battery of tests for anxiety‑like behaviours in mice. Acta Neuropsychiat 28: 85–91. Erfanparast A, Tamaddonfard E (2015) Effects of intracortical microinjection of vitamin B12 on penicillin‑induced epileptiform activity in rats. Acta Neurobiol Exp 75: 200–207. Fineberg NA, Reghunandanan S, Simpson HB, Phillips KA, Richter MA, Matthews K, Stein DJ, Sareen J, Brown A, Sookman D (2015) Obsessive‑compulsive disorder (OCD): Practical strategies for pharmacological and somatic treatment in adults. Psychiat Res 227: 114–125. Gawali NB, Chowdhury AA, Kothavade PS, Bulani VD, Nagmoti DM, Juvekar AR (2016) Involvement of nitric oxide in anticompulsive‑like effect of agmatine on marble‑burying behaviour in mice. Eur J Pharmacol 770: 165–171. Geller D (2006) Obsessive‑compulsive and spectrum disorders in children and adolescents. Psychiatr Clin North Am 29: 353–370. Goddard AW, Shekhar A, Whiteman AF, McDougle CJ (2008) Serotoninergic mechanisms in the treatment of obsessive‑compulsive disorder. Drug Discov Today 13: 325–332. Gomez C, Saldivar‑Gonzalez A, Delgado G, Rodriguez R (2002) Rapid anxiolytic activity of progesterone and pregnanolone in male rats. Pharmacol Biochem Behav 72: 543–550. Greene‑Schloesser DM, Van der Zee EA, Sheppard DK, Castillo MR, Gregg KA, Burrow T, Foltz H, Slater M, Bult‑Ito A (2011) Predictive validity of a non‑induced mouse model of compulsive‑like behavior. Behav Brain Res 221: 55–62. Gyertyan I (1995) Analysis of the marble burying response: marbles serve to measure digging rather than evoke burying. Behav Pharmacol 6: 24–31. Hoffman KL (2011) Animal models of obsessive compulsive disorder: recent findings and future directions. Expert Opin Drug Discov 6: 725–737. Jimenez‑Gomez C, Osentoski A, Woods JH (2011) Pharmacological evaluation of the adequacy of marble burying as an animal model of compulsion and/or anxiety. Behav Pharmacol 22: 711–713. Joel D, Ben‑Amir E, Doljansky J, Flaisher S (2004) “Compulsive” lever‑ ‑pressing in rats is attenuated by the serotonin re‑uptake inhibitors paroxetine and fluvoxamine but not by the tricyclic antidepressant desipramine or the anxiolytic diazepam. Behav Pharmacol 15: 241–252. Jury NJ, McCormick BA, Horseman ND, Benoit SC, Gregerson KA (2015) Enhanced responsiveness to selective serotonin reuptake inhibitors during lactation. PLoS One. Kim J, Gorman J (2005) The psychobiology of anxiety. Clin Neurosci Res 4: 335–347. Kudryashov NV, Kalinina TS, Zhmurenko LA, Voronina TA (2016) Anticompulsive activity of a new pyrazolo[C]pyridine derivative GIZh‑72 under conditions of unpredictable chronic mild stress. Bull Exp Biol Med 161: 377–380. 7_1143 Taylor_v5.indd 259 Marble Burying Animal Model 259 Marazziti D, Consoli G, Baroni S, Catena Dell’Osso M (2010) Past, present and future drugs for the treatment of obsessive‑compulsive disorder. Curr Med Chem 17: 3410–3421. Martin P (2003) The epidemiology of anxiety disorders: a review. Dialog Clin Neurosci 5: 282–298. Milad MR, Rauch SL (2007) The role of the orbitofrontal cortex in anxiety disorders. Ann NY Acad Sci 1121: 546–561. Nichols JN, Deshane AS, Niedzielko TL, Smith CD, Floyd CL (2016) Greater neurobehavioral deficits occur in adult mice after repeated, as compared to single, mild traumatic brain injury (mTBI) Eur J Pharmacol 547: 106–115. Nicolas LB, Kolb Y, Prinssen EP (2006) A combined marble burying‑ ‑locomotor activity test in mice: a practical screening test with sensitivity to different classes of anxiolytics and antidepressants. Eur J Pharmacol 547: 106–115. Palanza P, Morley‑Fletcher S, Laviola G (2001) Novelty seeking in periadolescent mice: sex differences and influence of intrauterine position. Physiol Behav 72: 255–262. Pallanti S, Quercioli L (2006) Treatment‑refractory obsessive‑compulsive disorder: Methodological issues, operational definitions and therapeutic lines. Prog Neuropsychopharmacol Biol Psych 30: 400–412. Pandey DK, Yadav SK, Mahesh R, Rajkumar R (2009) Depression‑like and anxiety‑like behavioural aftermaths of impact accelerated traumatic brain injury in rats: A model of comorbid depression and anxiety? Behav Brain Res 205: 436–442. Perry JL, Joseph JE, Jiang Y, Zimmerman RS, Kelly TH, Darna M, Huettl P, Dwoskin LP, Bardo MT (2011) Prefrontal cortex and drug abuse vulnerability: Translation to prevention and treatment interventions. Brain Res Rev 65: 124–149. Satta V, Scherma M, Giunti E, Collu R, Fattore L, Fratta W, Fadda P (2016) Emotional profile of female rats showing binge eating behavior. Physiol Behav 163: 136–143. Schneider T, Popik P (2007) Attenuation of estrous cycle‑dependent marble burying in female rats by acute treatment with progesterone and antidepressants. Psychoneuroendocrinology 32: 651–659. Schneier FR, Martinez D, Anissa Abi‑Dargham A, Zea‑Ponce Y, Simpson HB, Liebowitz MR, Liebowitz MR, Laruelle M (2008) Striatal dopamine D2 receptor availability in OCD with and without comorbid social anxiety disorder: preliminary findings. Depress Anx 25: 1–7. Schule C, Eser D, Baghai TC, Nothdurfter C, Kessler JS, Rupprecht R (2011) Neuroactive steroids in affective disorders: target for novel antidepressant or anxiolytic drugs? Prog Neurobiol 113: 79–87. Shim G, Park HY, Jang JH, Kim E, Park HY, Hwang JY, Kim SN, Jang GE, Kwon JS (2011) What is the optimal dose of escitalopram for the treatment of obsessive‑compulsive disorder? A naturalistic open‑label study. Int Clin Psychopharmacol 26: 284–290. Silverman JL, Pride MC, Hayes JE, Puhger KR, Butler‑Struben HM, Baker S, Crawley JN (2015) GABAB receptor agonist R‑baclofen reverses social deficits and reduces repetitive behavior in two mouse models of autism. Neuropsychopharmacology 40: 2228–2239. Stahl SM (2013) Stahl’s Essential Psychopharmacology: Neuroscientific basis and practical applications. 4th edn. Cambridge University Press: New York. Stein DJ, Carey PD, Lochner C, Seedat S, Fineberg N, Andersen EW (2008) Escitalopram in obsessive‑compulsive disorder: response of symptom dimensions to pharmacotherapy. CNS Spectrums 13: 492–498. Takeuchi H, Yatsugi S, Yamaguchi T (2002) Effect of YM992, a novel antidepressant with selective serotonin re‑uptake inhibitory and 5‑HT 2A receptor antagonistic activity, on a marble‑burying behavior test as an obsessive‑compulsive disorder model. Jpn J Pharmacol 90: 197–200. Taylor GT, Boggiano J, Cabrera O, Kroutil K (2011) Steroidal influences on anxiety disorders in childhood and their animal models. Curr Topics Steroid Res 8: 47–64. 23/10/17 21:48 260 GT. Taylor et al. Thomas A, Burant A, Nghiem B, Graham D, Yuva‑Paylor LA, Paylor P (2009) Marble burying reflects a repetitive and perseverative behavior more than novelty‑induced anxiety. Psychopharmacology (Berl) 204: 361–373. Umathe SN, Vaghasiya JM, Jain NA, Dixit PV (2009) Neurosteroids modulate compulsive and persistent behavior in rodents: implications for obsessive‑compulsive disorder. Prog Neuropsychopharmacol Biol Psych 33: 1161–1166. Wirth‑Dzięciołowska E, Lipska A, Węsierska M (2005) Selection for body weight induces differences in exploratory behavior and learning in mice. Acta Neurobiol Exp 65: 243–253. 7_1143 Taylor_v5.indd 260 Acta Neurobiol Exp 2017, 77: 254–260 Witkin JM (2008) Animal models of obsessive‑compulsive disorder. Curr Prot Neurosci Unit 9: 30. Wolmarans de W, Stein DJ, Harvey BH (2016) Of mice and marbles: Novel perspectives on burying behavior as a screening test for psychiatric illness. Cognitive & Affective Behavioral Neuroscience 16: 551–560. Yates JR, Marusich JA, Gipson CD, Beckmann JS, Bardo MT (2012) High impulsivity in rats predicts amphetamine conditioned place preference? Pharm Biochem Behav 100: 370–376. Zohar J (2008) Escitalopram in the treatment of obsessive‑compulsive disorder. Expert Rev Neurother 8: 339–349. 23/10/17 21:48
https://openalex.org/W4387069425	https://revistas.unal.edu.co/index.php/cal/article/download/92451/89127	es		Uso y manejo tradicional de la fauna silvestre por nahuas de Santa Catarina, Tepoztlán, Morelos, México	Caldasia	2,022	cc-by	8,409	CALDASIA Caldasia 45(1):36-48 \| Enero-abril 2023 http://www.revistas.unal.edu.co/index.php/cal Fundada en 1940 ISSN 0366-5232 (impreso) ISSN 2357-3759 (en línea) ETNOZOOLOGÍA Uso y manejo tradicional de la fauna silvestre por nahuas de Santa Catarina, Tepoztlán, Morelos, México Traditional use and management of wildlife by nahuas from Santa Catarina, Tepoztlán, Morelos, Mexico María Inés Ayala Enríquez José Manuel Pino Moreno , Alejandro García Flores 1, Erika Román Montes De Oca , Feliciano García Lara 4 1 , 2 3 • Recibido: 13/Ene/2021 • Aceptado: 23/Mar/2022 • Publicación en línea: 09/Jun/2022 Citación: Ayala Enríquez MI, García Flores A, Román Montes de Oca E, Pino Moreno JM, García Lara F. 2023. Uso y manejo tradicional de la fauna silvestre por nahuas de Santa Catarina, Tepoztlán, Morelos, México. Caldasia 45(1):36–48. doi: https://doi.org/10.15446/caldasia.v45n1.92451 ABSTRACT This research characterized the traditional knowledge on the use and management of wildlife in the Nahuatl community of Santa Catarina, Morelos, Mexico. A qualitative methodology was used, applying ethnozoological research techniques such as semi-structured interviews (n = 55), ethnographic observation, and guided tours with local experts. In addition, the Cultural Importance Index (IIC) was calculated for each species used. A total of 41 species from different phyla were recorded, which are part of the tangible and intangible cultural manifestations of the Nahuas. Eleven categories of use are reported, of which nutritional (n = 57.5 %), forecast (n = 25 %), medicinal (n = 22.5 %) and legends (n = 12.5 %) recorded the higher percentage of species. The wild animals with the highest IIC were the white-tailed deer (Odocoileus virginianus; 16.8), the field rabbit (Sylvilagus cunicularius; 13.6), the huilota (Zenaida macroura; 10.9), the rattlesnake (Crotalus culminatus; 12.8) and the winged ant (gynes and males of Atta mexicana; 10.8). Obtaining wildlife resources is done through individual or group gathering and hunting with firearms and traditional weapons in forest areas and agroecosystems within their territory. Wild fauna is a renewable natural resource that has a socio-cultural importance for the Nahua inhabitants because it is part of their subsistence strategies and cosmology: it provides them with raw materials for material needs, but also has symbolic representations associated with animals in traditional medicine, the practice of hunting and in the narratives of oral tradition. Keywords: traditional knowledge, worldview, Nahuatl culture, ethnozoology, subsistence strategies, El Tepozteco National Park. Universidad Autónoma del Estado de Morelos, Centro de Investigaciones Biológicas. Av. Universidad 1001, Col. Chamilpa, C. P. 62209. Cuernavaca, Morelos, México, ines.ayala@uaem.mx; alejandro.garcia@uaem.mx 2 Universidad Autónoma del Estado de Morelos, Facultad de Ciencias Agropecuarias. Av. Universidad 1001, Col. Chamilpa, C. P. 62209. Cuernavaca, Morelos, México, erika.romanm@uaem.edu.mx 3 Universidad Nacional Autónoma de México, Instituto de Biología. Cto. Zona Deportiva s/n. C. U. Coyoacán 04510, Ciudad Universitaria jpino@ib.unam.mx 4 Universidad Autónoma del Estado de Morelos, Centro de Lengas Extranjeras. Av. Universidad 1001, Col. Chamilpa, C. P. 62209. Cuernavaca, Morelos, México, feliciano.garcia@uaem.mx Autor para correspondencia. 1 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 RESUMEN Se caracterizó el conocimiento tradicional sobre el uso y manejo de la fauna silvestre en la comunidad Náhuatl de Santa Catarina, Morelos, México. Se utilizó una metodología cualitativa, aplicando técnicas de investigación etnozoológicas como entrevistas semiestructuradas (n=55), observación etnográfica y recorridos guiados con expertos locales. Además, se calculó el Índice de Importancia Cultural (IIC) de cada especie utilizada. Se registraron 41 especies de diversos phyla, los cuales forman parte de las manifestaciones culturales tangibles e intangibles de los nahuas. Se reportan once categorías de uso, de las cuales la alimentaria (n=57,5 %), pronosticadora (n=25 %), medicinal (n=22,5 %) y leyendas (n=12,5 %), registraron el mayor porcentaje de especies. Los animales silvestres con mayor IIC fueron el venado cola blanca (Odocoileus virginianus; 16,8), el conejo de campo (Sylvilagus cunicularius; 13,6), la huilota (Zenaida macroura; 10,9), la víbora de cascabel (Crotalus culminatus; 12,8) y la “hormiga con alas” (reproductores de Atta mexicana; 10,8). La obtención de los recursos faunísticos se realiza mediante la recolección y la cacería individual o grupal con armas de fuego y métodos tradicionales usados en áreas forestales y agroecosistemas dentro de su territorio. La fauna silvestre es un recurso natural renovable que tiene una importancia sociocultural para los habitantes nahuas porque forma parte de sus estrategias de subsistencia y cosmovisión, al proveerlos de materias primas para necesidades materiales, pero también por poseer representaciones simbólicas asociadas a los animales en la medicina tradicional, la práctica de la cacería y en narrativas de la tradición oral. Palabras clave: conocimiento tradicional, cosmovisión, cultura Náhuatl, etnozoología, estrategia de subsistencia, Parque Nacional El Tepozteco. INTRODUCCIÓN ideas y representaciones simbólicas asociadas a los animales en la tradición oral como creencias, mitos, leyen- La relación histórica de las comunidades indígenas con das, así como en las prácticas rituales (Osorio-López et al. la naturaleza de acuerdo con su propia cosmovisión, les 2017). El aprovechamiento de la fauna se diversifica más ha permitido elaborar conocimientos tradicionales sobre allá de la alimentación, porque también se emplean pro- el uso y manejo de los recursos naturales de su territo- ductos animales en la medicina tradicional, como materia rio (Herrera-Flores et al. 2019). La fauna silvestre es un prima para la elaboración de artesanías, adornos, herra- recurso significativo de su realidad (Argueta-Villamar mientas, utensilios y como animales de compañía, como 2008). La forma en que cada grupo originario ha concebi- lo evidencia la diversidad de mamíferos, aves, anfibios y do su interacción con los animales y las categorías de uso reptiles con categoría de uso registrados en comunidades que le asigna varían de acuerdo con el contexto geográfi- de México (Alonso-Castro 2014, Ávila-Nájera et al. 2018a, co, histórico y cultural (Pérez-Gil et al. 1996). En la región Ávila-Nájera et al. 2018b, Zarazúa-Carbajal et al. 2020). cultural de Mesoamérica y particularmente en México, los animales silvestres han sido un recurso importante Los grupos humanos realizan diversas prácticas de manejo para la subsistencia de comunidades indígenas desde la de la fauna silvestre. Entre estas se encuentran las que tie- época precolombina hasta la actualidad, principalmente nen como objetivo la obtención de estos recursos, las que se por el aporte de proteína animal como un complemento enfocan a acciones concretas para aumentar o mantener la de la dieta (Ramírez-Barajas y Calme 2015). En regiones disponibilidad de los animales silvestres, y otras para fines del trópico húmedo mexicano se ha estimado que la car- antropogénicos adicionales (Zarazúa-Carbajal et al. 2020). ne proveniente de la fauna silvestre puede aportar hasta el En el contexto mesoamericano, la interacción humano- 50 % de la proteína en la dieta de pobladores originarios fauna silvestre ha incluido una gama de prácticas de mane- (Tejeda-Cruz et al. 2014). La fauna además es un elemento jo como la extracción oportunista o planeada de animales de la cosmovisión indígena mesoamericana al encontrar mediante la caza o la recolección; la cría en cautiverio, la 37 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 domesticación e hibridación; las enfocadas al manejo del hábitat como el cuidado de plantas hospederas y alimentarias para la fauna, o la milpa-comedero-trampa para atraer especies de interés; hasta aquellas prácticas de manejo dirigidas hacia los usuarios para regular el acceso a este recurso (Valadez-Azúa y Rodríguez-Galicia 2013, Santos-Fita et al. 2013, Valadez-Azúa 2016, Zarazúa-Carbajal et al. 2020). El Náhuatl es la lengua originaria predominante en el estado de Morelos, México. Aproximadamente son 35 comunidades de este origen cultural, concentrando la mayor población en los pueblos de Hueyapan, Tetelcingo, Cuentepec, Xoxocotla y Santa Catarina (INPI 2020). La comunidad de Santa Catarina se localiza en el norte del estado. Sus habitantes preservan actividades tradicionales de aprovechamiento de los recursos naturales que forman parte de su identidad cultural como la milpa, la recolección de hongos y plantas silvestres, además la cacería (AyalaEnríquez et al. 2019). Sin embargo, las parcelas agrícolas y la superficie forestal han disminuido por los cambios de uso de suelo, el crecimiento urbano y la fragmentación del territorio, lo que tiene impacto en la disponibilidad de sus recursos naturales (Ayala-Enríquez et al. 2019). En dicho contexto, entender como los grupos culturales aprovechan sus recursos faunísticos y las formas de manejo a nivel comunitario, constituye información base para la toma de decisiones que permita favorecer su conservación y aprovechamiento sostenible (Puc y Retana 2012). Ante estas premisas, el objetivo de este trabajo fue caracterizar, por primera vez para esta comunidad, el conocimiento tradicional sobre el uso y manejo de la fauna silvestre entre nahuas de Santa Catarina, Morelos, México. MATERIAL Y MÉTODOS Área y población de estudio La comunidad Náhuatl de Santa Catarina se localiza en el municipio de Tepoztlán, Morelos, centro de México (Fig. 1), a una altitud de 1620 m. Los antecedentes históricos de los pueblos nahuatlacas muestran que fue un grupo cultural que se desarrolló en la Cuenca de México y valles centrales colindantes. En Morelos, los tlahuicas y xochimilcas fueron los pueblos nahuas que se asentaron en el CentroSur y en los Altos respectivamente (Morayta 2011). Actualmente, en Morelos el Náhuatl es la lengua indígena predominante. La distribución actual de los pueblos na38 huas es heterogénea, ubicándose en la región de los Altos, Centro y Sur. La comunidad de estudio se localiza dentro de la región de los Altos en la transición del bosque templado y la selva baja caducifolia entre los 1300 y 2000 m (Morayta 2011). Su territorio se ubica en el Parque Nacional El Tepozteco, área natural protegida que integra parte del Corredor Biológico Chichinautzin (Fig. 1). El clima es semicálido con lluvias en verano (Taboada et al. 2009). En Santa Catarina se distribuyen la selva baja caducifolia, el bosque de pino y encino, el matorral crasicaule y las “huizacheras” que representan vegetación secundaria de selva baja por ser terrenos agrícolas abandonados (AyalaEnríquez et al. 2019). De acuerdo con la SEDESOL (c2015), la población es de 4521 habitantes, de los cuales 894 aún mantiene el náhuatl o mexicano como su idioma. La población campesina de esta comunidad basa su reproducción social en las prácticas agrícolas, pecuarias y forestales. La actividad agrícola se establece alrededor de la milpa en la que se cultivan maíces criollos (Zea mays L.) como las variedades amarillas, el rojo, el pinto, el ancho, el pitillo, el picado y el azul. Además, se cultiva calabaza dulce (Cucurbita pepo L.), frijol yepatlach o cuaresmero (Phaseolus lunatus L.) y se recolectan plantas arvenses. Paralelamente, algunos campesinos se han especializado en monocultivos de jitomate (Solanum lycopersicum L.) o pepino (Cucumis sativus L.) destinados para la comercialización (Ayala-Enríquez et al. 2019). Otras estrategias de subsistencia campesina son la recolección de hongos (localmente conocidos como hongueros), la práctica de la cacería (conocidos como tiradores), picadores de piedra (conocidos como pedreros) y en recolección de plantas útiles, extracción de leña y vendedores de tierra de hoja. Recopilación de información El trabajo de campo se llevó acabo de enero a diciembre de 2019. En su realización se consideró el Código de Ética de la Sociedad Latinoamericana de Etnobiología, al obtener el permiso de la autoridad local para realizar la investigación, el consentimiento libre e informado de habitantes para su participación en ella y respetar su anonimato bajo el principio de confidencialidad (Cano-Contreras et al. 2014). Se utilizó una metodología cualitativa que permitió obtener datos descriptivos a partir de las propias palabras de las personas y sus conductas observables mediante la utilización de técnicas de investigación social aplicadas en la etnozoología como la entrevista, observación etnográfica y recorridos guiados (Taylor y Bogdan 1987, Santos-Rodrigues 2009). Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 Figura 1. Ubicación de la comunidad de Santa Catarina, Tepoztlán, Morelos, México. En el desarrollo del trabajo de campo se contó con el como organización, armas, técnicas y sitios de cacería, y acompañamiento de un habitante originario de Santa Ca- prácticas de manejo enfocadas al hábitat o los usuarios. tarina. Con el apoyo de este contacto en la comunidad, la Las entrevistas fueron grabadas digitalmente, siempre información se obtuvo mediante la identificación de un que el informante lo permitiera y posteriormente fue- primer grupo de personas claves con conocimientos sobre ron transcritas. Para una interacción más cercana con la el tema de investigación (Patton 2002). Posteriormen- realidad y el contexto social de la población de estudio se te la selección se realizó mediante la técnica de Bola de realizó observación etnográfica (Taylor y Bogdan 1987), al nieve, la cual consistió en la identificación de habitantes estar presente en actividades de la milpa, en fiestas de la conocedores del tema con base en la recomendación de comunidad, la colocación de la ofrenda en el día de muer- los entrevistados anteriores (Martín-Crespo y Salamanca- tos, en actividades de manejo del traspatio, recolección de Castro 2007). Otro criterio que se consideró para la se- plantas útiles y en la preparación de alimentos. lección fue la disponibilidad de tiempo y diálogo. Se usaron 55 entrevistas semiestructuradas a partir de un guion Para la identificación taxonómica de las especies animales preestablecido (Santos-Rodrigues 2009), la cual se reali- referidas en las entrevistas, se realizaron recorridos guia- zó en español y de forma individual en los hogares de los dos con los informantes locales en los sistemas producti- entrevistados. Se recabó información sobre la ocupación, vos agrícolas y forestales. Esta actividad se realizó median- edad, animales conocidos, formas de uso, creencias rela- te técnicas directas e indirectas y con el apoyo de guías de cionadas con la fauna silvestre, además de las prácticas campo (Castro-Franco y Bustos-Zagal 2006, Aranda-Sán- de manejo para la obtención de los recursos faunísticos, chez 2012, Gaviño 2015). Para los insectos se utilizaron las 39 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 claves taxonómicas propuestas en Blandin (2007), Ward (2010), DGSV (2016), Mariño-Pérez y Song (2018) y Song et al. (2018). El análisis de las entrevistas fue inductivo, en el cual a partir de la revisión detallada línea por línea de los textos y los datos se construyeron las categorías y los patrones que describen y explican el fenómeno bajo estudio (Patton 2002). La categorización de los usos en la fauna se realizó a partir de las descripciones de los informantes y la revisión de literatura (Pérez-Gil et al. 1996, Santos-Fita et al. 2012, Chontal-Chagala et al. 2019, Nahuat-Cervera et al. 2021). La información se consolidó en una base de datos para realizar análisis de estadística descriptiva. A partir del recuento de datos se obtuvo el número de informantes por género, la distribución porcentual de las ocupaciones, promedio de edad, la riqueza de especies reconocidas, la riqueza de especies con categoría de uso, la frecuencia de mención por categoría de uso y especies, la distribución porcentual de las especies por categoría de uso y la distribución porcentual de especies con categoría de uso por grupo taxonómico. Se aplicó el Índice de Importancia Cultural para cuantificar la relevancia local de las especies de fauna a partir de la ecuación propuesta por Ávila-Nájera et al. (2018b): IICz = ∑ (Iuz + Fmz + Vutz) Donde: IICz: Índice de Importancia Cultural de una especie determinada. Iuz: número de usos de la especie z de todos los informantes dividido entre el número total de usos de todas las especies y de todos los informantes, multiplicado por 100. Fmz: número de menciones de la especie z de todos los usos y de todos los informantes dividido entre el número total de menciones de todas las especies para todos los usos y de todos los informantes, multiplicado por 100. Vutz: número de menciones de la especie z para un uso de todos los informantes dividido entre el número total de menciones de todas las especies para un uso x y de todos los informantes, multiplicado por 100. 40 RESULTADOS Y DISCUSIÓN Se aplicaron entrevistas a 31 hombres y 24 mujeres. La edad promedio fue de 80 años. Las ocupaciones de los entrevistados fueron: ama de casa (44 %), campesino (25,2 %), pedrero (10,9 %), albañil (10,9 %), comerciante (7,2 %) y peón (1,8 %). El alto porcentaje de amas de casa fue debido a su disponibilidad de tiempo, pero también porque son quienes llegan a preparar los alimentos con carne de animales silvestres y en algunos casos también son quien deciden que se caza y se come, mientras que los hombres son los encargados de obtener los animales silvestres (tiradores). Fauna silvestre y categorías de uso Los habitantes entrevistados mencionaron 45 especies, que taxonómicamente se clasifican en 31 familias (Anexo 1); el 84 % son nombradas en Náhuatl. De las especies reconocidas, 39 están incluidas en alguna categoría de uso, las cuales se distribuyen en los siguientes grupos taxonómicos: mamíferos (28,8 %), aves (22,2 %), reptiles (15,5 %) e insectos (20 %). Se registraron once categorías de uso, siendo la alimentaria (57,5 %), pronosticador (25 %), medicinal (22,5 %) y animales con leyendas (12,5 %) las que suman el mayor porcentaje (Fig. 2); así mismo fueron las categorías con mayor número de menciones (Anexo 2). En el Parque Nacional El Tepozteco se han registrado 469 especies de vertebrados de las cuales 67 son mamíferos, 301 aves y 74 reptiles (González y Contreras-MacBeath 2020), lo que equivale al 19,4 %, 3,32 % y 9,46 % de las especies registradas con categoría de uso respectivamente. Así mismo, el total de vertebrados numerados con uso representa el 6,4 % de la riqueza total registrada en el área. Históricamente, los animales silvestres han sido un recurso utilizado por los pueblos nahuas para diversos fines materiales y significados simbólicos desde la época prehispánica como lo evidencian documentos históricos como el Códice Florentino (Sahagún 2005). Dicho aprovechamiento diversificado de la fauna es una práctica con continuidad histórica que igualmente se registra en otras comunidades nahuas del centro de México (Martínez 2014, GutiérrezSantillán et al. 2019). De las categorías de uso reportadas en el presente trabajo, sobresale la alimentaria que agrupa más de la mitad de los animales y tiene la mayor frecuencia de mención de las especies (n=582) (Anexo 2). Dicha prevalencia de uso de la fauna también se registra en otras comunidades del Parque Nacional El Tepozteco y del Corredor Biológico Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 Figura 2. Número de especies por categorías de uso entre los nahuas de Santa Catarina. Chichinautzin, donde más del 50 % de las especies están ubicadas en esa categoría (Monroy et al. 2011, García-Flores et al. 2014). El consumo de partes musculares o vísceras de animales silvestres con frecuencia es la principal categoría de uso en comunidades de origen étnico o mestizo de Latinoamérica (Guerra et al. 2010, Pitriello y Stronza 2019). La ingesta de carne silvestre es un complemento de la dieta en la población rural, lo cual no solo está relacionado con satisfacer el hambre, también con el deseo y gusto por el sabor de esta carne o porque es considerada saludable y nutritiva (Alcalá y De la Riva 2016, Zavala-Sánchez et al. 2018, Pitriello y Stronza 2019). A pesar de que una de las principales necesidades que se busca satisfacer con el uso de la fauna silvestre en México es la alimentación, los animales silvestres pueden aprovecharse integralmente para obtener productos ornamentales, amuletos, herramientas, además de tener animales de compañía (Naranjo et al., 2010). Este aprovechamiento se repite entre pobladores nahuas del área de estudio, quienes obtienen bienes de la fauna con fines ornamentales, enseres domésticos o agrícolas, y abono. Alimento En esta categoría se registraron 23 especies distribuidas en once mamíferos, seis aves, tres reptiles y tres insectos, que forman parte de la cultura alimentaria al prepararse en diferentes platillos de la cocina tradicional como salsas, mixiotes, barbacoa, carne frita o asada. Las especies animales con mayor frecuencia de mención en esta categoría fueron Sylvilagus cunicularius (n=54), Zenaida macroura (n=54), Odocoileus virginianus (n=51), Dasypus novemcinctus (n=34), Sylvilagus floridanus (n=32), Sphenarium purpurascens (n=33) y Columbina inca (n=31). La fauna silvestre sigue siendo parte del menú en la cultura alimentaria de los habitantes de comunidades rurales en regiones tropicales como la de Santa Catarina, a pesar de entornos cada vez más cambiantes por la globalización, el mayor acceso a alimentos industrializados, la tercerización productiva, la degradación ambiental y el cambio cultural (Alves y van Vliet 2018). Las especies de vertebrados con mayor número de menciones como los conejos, el venado, el armadillo y las palomas silvestres son de amplia distribución geográfica, por lo que forman parte de la cultura alimentaria de diversas comunidades originarias y campesinas de México (Valdez y Ortega 2014, Ávila-Nájera et al. 2018b). De los grupos taxonómicos registrados, los mamíferos aportan la mayor riqueza de especies con fines comestibles y también son el grupo con mayor frecuencia de mención en esta categoría (n=287) (Anexo 2), similar a lo que se registra en otras regiones de México (Ávila-Nájera et al. 2018b). En el caso de los insectos, aunque el número de especies fue menor al registrar solo el consumo de chapulines del género Sphenarium y hormigas del género Atta, muestra la vigencia de la entomofagia en la comunidad, una práctica con tradición prehispánica (Sahagún 2005). Por otra parte, es necesario estimar aspectos como la extracción de fauna y carne consumida, variables que permitirán contextualizar con mayor precisión el aporte alimentario de los animales silvestres en la dieta de los habitantes locales. La evaluación cuantitativa del aprovechamiento de la fauna y su aporte a la alimentación es escasa en grupos originarios del centro de país en comparación con el sureste mexicano, donde se ha evidenciado que a pesar de que existe una amplia diversidad de animales comestibles registrada por medio de técnicas de investigación cualitativa como las entrevistas, la obten- 41 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 ción de carne silvestre habitualmente se centra en ciertas especies de talla mediana y grande (León y Montiel 2008, Naranjo et al. 2010, Santos-Fita et al. 2012, Tejeda-Cruz et al. 2014, Ramírez-Barajas y Calme 2015). Pronosticador Se identificaron diez animales clasificados como pronosticadores de enfermedad, muerte, de mal agüero e indicadores de fenómenos meteorológicos y estado del tiempo. Por ejemplo, es frecuente que los campesinos encuentren a ehecatotolin (Geococcyx velox) en su recorrido a la milpa, pero no se arriesgan a pasar junto a ellos porque si llega a atravesarse en su camino les indica que alguien del pueblo o un familiar va a morir. Lo mismo sucede con el tecolotl (Glaucidium brasilianum) y la lechuza (Tyto alba), cuyo ulular se relaciona con presagios de muerte, por lo cual los ahuyentan de las viviendas. Entre los nahuas de Santa Catarina se registran insectos indicadores de lluvia como las chicharras (Cicada orni), especie que estridula en las mañanas del mes de mayo hasta que empiezan las primeras lluvias. Los moyotes (Cotinis mutabilis) hacen vuelos nocturnos entre mayo y junio anunciando el inicio de las lluvias. Mientras las hormigas con alas (reproductores de A. mexicana) hacen su arribo en las primeras lluvias del temporal. La mariposa atolera (Morpho polyphemus) aparece por las viviendas en mayo-junio y al final de los temporales, entre octubre-noviembre, por lo que los nahuas la consideran un animal indicador de inicio y término de las lluvias. En las aves, las golondrinas (Hirundo rustica) con su vuelo pronostican la llegada de las lluvias. Las tortolitas (C. inca y C. passerina) también son reconocidas como indicadoras del estado del tiempo, cuando la temperatura del ambiente aumenta las aves bajan al suelo en busca de humedad. En el contexto cultural de los pueblos con tradición mesoamericana como los nahuas de Santa Catarina, es frecuente la identificación de especies animales asociadas con el presagio de diferentes sucesos que influyen en la vida de las personas, como de forma similar se documenta en comunidades tzotziles, tojolabales o mayas de la península de Yucatán en México, donde invertebrados y vertebrados silvestres anuncian cambios en el clima y sucesos que pueden ser perjudiciales o benéficos (Serrano-Gonzáles et al. 2011, Nahuat-Cervera et al. 2021). En el caso de la cosmovisión prehispánica náhuatl, la palabra tetzáhuitl se relacionaba a una “cosa escandalosa, espantosa o cosa de agüero” traducido al español como augurio o presagio, un fenómeno considerado una señal que anuncia un hecho futuro, que no necesariamente es un acontecimiento negativo, también 42 de buena fortuna y no significaban hechos definitivos. Estos presagios se pueden manifestar por medio de objetos, fenómenos, comportamientos o también por la presencia de animales como se documenta en el presente trabajo (Flores-Rodríguez 2021, Pastrana-Flores 2021). Con relación al conocimiento tradicional sobre los fenómenos meteorológicos y cambio en el estado del tiempo, este se encuentra relacionado con características reproductivas y de comportamiento en insectos y aves. De acuerdo con Alves y Barboza (2017), la predicción de eventos meteorológicos a través de los animales ha persistido en numerosas culturas humanas a lo largo de generaciones. Las personas observan aspectos de comportamiento, fisiológicos y reproductivos de la fauna para las predicciones climáticas, siendo las aves, insectos y mamíferos los grupos más utilizados (Alves y Barboza 2017). Medicina Esta categoría incluye nueve especies: tres mamíferos, tres reptiles, dos aves y un insecto, para tratar padecimientos o enfermedades de origen fisiológico y de filiación cultural. Entre las partes utilizadas de los animales se encuentran la carne, sangre, colmillos, piel o los animales completos. Por ejemplo, se consume la carne guisada del yepatl (Conepatus leuconotus) para quitar los granos y sarpullido. La ingesta de la sangre fresca del mazatl (O. virginianus), ayuda a fortalecer y aumentar la sangre. Los colmillos de tectli (Crotalus culminatus) se remojan en alcohol por un año, posteriormente son utilizados para el dolor de cabeza y reumas; mientras su piel se utiliza en forma de cataplasma para tratar la picadura de alacrán. La sangre de la akuekuespalin (Ctenosaura pectinata) es recomendada para la debilidad, ceguera, y aumentar las plaquetas, para lo cual se mezcla con refresco de cola y se bebe. En el caso de la ayotl (Kinosternon integrum) se utiliza para el tratamiento de la enfermedad cultural conocida con el nombre tortuga, la cual se caracteriza por un incremento en la temperatura y porque el cuello no tiene fuerza para detener la cabeza del niño. Para el tratamiento la sangre del cuello de la tortuga se coloca en las articulaciones y la espalda del enfermo, después preparan un caldo con la carne y se le da de comer. La carne guisada del ratoncito de tierra tlalquiquimichi (Peromyscus melanophrys) se emplea para la enfermedad de infantes conocida como chachahualistli. Esta se caracteriza por falta de apetito, mejillas inflamadas y los niños almacenan su comida en la boca sin ingerirla o solo juegan con ella. Mientras a los niños que ya están en edad de hablar y no pueden, los alimentan con las ciruelas Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 picadas por las aves, principalmente de las tórtolas C. inca y C. passerina para su tratamiento. Otro remedio para la falta de habla consiste en poner una chikilishtli (Cicada orni) dentro de la boca para que cante y le otorgue el habla al niño. La riqueza de especies medicinales reportada entre los nahuas representa el 1,54 % de la fauna medicinal registrada en Latinoamérica (n = 584) (Alves y Alves 2011) y el 5,52 % de la registrada en la medicina tradicional mexicana (n = 163) (Alonso-Castro 2014). El zorrillo, la víbora de cascabel, el venado cola blanca y la iguana representan especies con una amplia tradición de uso desde la época prehispánica y que se mantiene en la medicina tradicional mexicana contemporánea (Alonso-Castro 2014). Los animales no solo son utilizados para el tratamiento de enfermedades de origen fisiológico, también de aquellas que son propias de un contexto cultural como el padecimiento de “tortuga” y chachahualistli, este último padecimiento también se registra en otros grupos nahuas del país de acuerdo con fuentes etnohistóricas y etnográficas (Acosta 2013, Good 2013, Romero 2020). Leyenda Los nahuas de Santa Catarina conservan en su tradición oral leyendas que se asocian con cinco especies animales. Los campesinos tienen respeto a las hormigas tzicame (Atta spp.), porque simbólicamente son relacionadas con las víboras, con el viento, el Tepozteco y la lluvia. En el caso de la víbora tzikacoatl (Micrurus laticollaris Peters, 1870) se menciona que es un animal venenoso que habita en los hormigueros y es la mamá de las hormigas, las protege y éstas le consiguen alimento. En el caso de la víbora tlilcoatl, (Drymarchon melanurus Duméril, Bibron & Duméril, 1854), la tradición menciona que hipnotiza a las mujeres que están amamantando para tomarse la leche y para que los niños no lloren les coloca su cola como chupón. Mientras que los murciélagos son asociados con vampiros que se alimentan con la sangre de las personas y la oscuridad. Las hormigas del género Atta son uno de los elementos clave en el sistema simbólico de los pueblos mesoamericanos porque aparecen en narrativas relacionadas con el viento y la lluvia, fenómenos atmosféricos esenciales en el ciclo agrícola para el crecimiento del maíz, alimento básico en la subsistencia en Mesoamérica, o incluso representan el papel de héroe civilizador al dar este grano al hombre (Katz 2005). En el área de estudio, el significado de estas hormigas en la cosmovisión de los campesinos evita que se apliquen insecticidas en los hormigueros de las parcelas a pesar del daño que pueden ocasionar al maíz, utilizando para su control diferentes plantas distractoras de las que se alimentan (Ayala-Enríquez et al. 2018). Las serpientes probablemente representan uno de los grupos de reptiles que tienen mayor presencia en el imaginario colectivo debido a que son relacionadas con diferentes narrativas como mitos, leyendas y creencias (Casas 2000). La narración sobre una serpiente reconocida como “madre de las hormigas” y que vive en los hormigueros ya se reporta por Sahagún y Hernández en el siglo XVI, quienes probablemente describan un coralillo del género Micrurus, que se alimentan de otros ofidios que consumen termitas y hormigas (Casas 2004, Katz 2005). Mientras que las serpientes que maman los pechos de las madres es una narrativa común en diferentes regiones de la República Mexicana, donde además del tilcuate (Drymarchon corais Holbrook, 1842), también se registra a Pituophis deppei Duméril, 1853 y Spilotes pullatus Linnaeus, 1758 (Casas 2000). En la cosmovisión mesoamericana los murciélagos han sido asociados con la oscuridad, el inframundo y la muerte, pero no en el contexto negativo (Flores-Monter et al. 2017). Control biológico Los zorrillos (C. leuconotus) se alimentan de nextikuilli (Phyllophaga obsoleta Blanchard, 1851) por lo que se tolera su presencia en los patios de las casas y la milpa, así como la nexhua (Coluber mentovarius Duméril, Bibron & Duméril, 1854) porque controlan la población de chachahua (ratoncitos de tierra) que ocasionan daños al maíz. Otra especie es la víbora de cascabel cuya dieta está constituida por roedores que pueden ser perjudiciales para la actividad agrícola. El reconocimiento de los beneficios de vertebrados silvestres en los agroecosistemas es de interés para llevar a cabo acciones conservacionistas, especialmente de aquellas especies como las serpientes cuyas percepciones negativas terminan con su eliminación (Gatica-Colima y Jiménez-Castro 2009, Leyte-Manrique et al. 2016, García-López et al. 2017). Amuleto Los testículos de los conejos (S. cunicularius y S. floridanus) son de buena suerte en la práctica de la cacería. Para esto se colocan en algodón dentro de un frasco con alcohol. Del venado, los tiradores utilizan su pata o cola como amuleto para la fortuna en la cacería. En estudios sobre la cacería en pueblos de origen mesoamericano, se ha documentado que dicha actividad puede estar enmarcada en un contexto simbólico al registrar elementos como 43 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 prácticas rituales, creencias, tabúes y el uso de amuletos que otorgan fortuna o establece el número de presas permitidas para cazar (Olivier 2015, Osorio-López et al. 2017, Herrera-Flores et al. 2018). Enseres de trabajo Las patas de conejo y venado se incrustan en la pared de las casas para utilizarlas como percheros. Además, las puntas de las astas del venado se utilizan como agujas para coser los costales en donde se transporta la mazorca del maíz (Z. mays). Abono El guano del murciélago (Artibeus lituratus) se utiliza para abonar la milpa, el cual se extrae de cuevas ubicadas en el Parque Estatal El Texcal. El aprovechamiento del guano de murciélago para la actividad agrícola también se registra en otras comunidades del Parque Nacional del Tepozteco (García-Flores et al. 2014). Sin embargo, en la comunidad de Santa Catarina probablemente se trate de una categoría de uso ya no vigente, pero si en la memoria de las personas. Esto porque en un estudio reciente sobre la caracterización de la milpa en Santa Catarina (AyalaEnríquez et al. 2019) se describe que en las actividades del ciclo agrícola se utiliza abono químico y que la extracción de guano para la fertilización de parcelas se realizaba hace 30 años, de acuerdo con campesinos de mayor edad. Uso lúdico y ornamental El uso lúdico se refiere a los animales que utilizan los niños para jugar. El moyotl (Cotinis mutabilis), es recolectado por los niños al inicio del verano (junio-julio), ellos amarran un hilo delgado en su cabeza y lo hacen volar. La piel de mazacoatl (Boa constrictor) y de mazatl se usa para adornar las casas de los tiradores y el caparazón del ayotochtli (Dasypus novemcinctus) es utilizado como instrumento musical. Índice de Importancia Cultural El venado cola blanca O. virginianus presentó el IIC más alto (16,8) después de S. cunicularius (13,6), C. culminatus (12,8), Z. macroura (10,9) y A. mexicana (10,8) (Anexo 1). El venado cola blanca registró el mayor número de categorías de uso, cuatro, con tres destacadas como la alimentación, medicinal o materia prima para enseres de trabajo; pero también el mayor número de menciones (n = 129), obteniendo la mayor frecuencia por su aporte de carne (n = 51). El conejo de campo y la víbora de cascabel registran tres categorías de uso, siendo las de mayor frecuencia la 44 alimentaria (n = 54) y la medicinal (n = 44) respectivamente. La huilota solo se registra en la categoría alimentaria, sin embargo, fue la especie con mayor frecuencia de mención en esta categoría (n = 54). La hormiga con alas es mencionada por ser un insecto parte de la cultura alimentaria local, pero de acuerdo con la frecuencia de mención es mayormente relacionada con un significado simbólico al ser pronosticadora de lluvias (n = 44). El venado cola blanca es uno de los vertebrados silvestres con mayor relevancia sociocultural en comunidades campesinas e indígenas de México (Ávila-Nájera et al. 2018b), como se reporta en esta investigación. Dicha relevancia podría estar en relación a que es una de las presas predilectas y apreciadas en la cacería de subsistencia (Ávila-Nájera et al. 2018b), pero también porque se aprovecha integralmente al proveer otros productos (Retana-Guiascón y LorenzoMonterrubio 2016, Ávila-Nájera et al. 2018b). Manejo de la fauna Organización, armas, técnicas y sitios de cacería Los recursos faunísticos son conseguidos mediante la extracción al practicar la cacería y en el caso de los insectos mediante la recolección. Las armas utilizadas en la práctica de la cacería fueron: 1) onda, este instrumento está hecho a base del ixtle (Agave angustifolia Haw). 2) Varas, consiste en una rama de árbol o arbusto de Dodonea viscosa (L.) Jacq), Tecoma stans (L.) Juss. ex. Kunth) o Bursera fagaroides (Kunth) Engl). 3) Resortera para cazar aves, ésta se fabrica con un hule, fragmento de cuero y una base en forma de horqueta, esta se obtiene de una madera dura de Psidium guajava L. o de Vitex molli Kunth. 4) Machete, es adquirido en el mercado. 5) Escopeta calibre 16 con postas, se emplea para cazar venados y munición para el tochtli, huilotl y el tlacuatzin. 6) Rifle de diábolos, para cazar a la cuapioton (Ortalis poliocephala). Las técnicas de cacería registradas fueron: la arreada, consiste en perseguir a la presa hasta acorralarla, se recurre a ella cuando van más de cinco tiradores e incluso emplean perros. Cuando deciden ir en busca de un venado se organizan por familia o amigos, dicen “vamos a la arreada”, juntan 24 personas, entre ellos doce arreadores con perros. Los tiradores entrenan a los chichime (Canis lupus familiaris) para ayudar a cazar conejos, armadillos, ardillas y venados. La espiada consiste en observar a animales en los caminos, así como huellas, comederos y abrevaderos. Los adultos la utilizan para cazar tectli y cuando observan o escuchan sonidos de animales mayores con interés como el mazatl. En la imitación, los adultos y niños aprenden a remedar los sonidos, Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 principalmente de las aves para que se acerquen a ellos y porque les gusta el canto, otros imitan las peleas del venado con ayuda de cuernos. La técnica de la carnada consiste en colocar en las cuevas frutas o flores que saben le gusta al animal, para acorralarlo cuando se acerque. Los tiradores ubican áreas donde han visto que llega la fauna, como abrevaderos, madrigueras, cuevas, cañadas y los árboles predilectos de los animales. Por ejemplo, el mazatl se alimenta del cazahuatl (Ipomoea spp.), pochotl (Ceiba aesculifolia (Kunth) Britt & Baker f.) y xiloxochitl (Pseudobombax ellipticum (Kunth) Dugand). Sus astas las afilan en árboles de encino (Quercus spp.), tepehuaxi (Lysiloma acapulcense (Kunth) Benth) y palo dulce (Eysenhardtia polystachya (Ortega) Sarg). Persiste el uso de armas tradicionales como la onda o la resortera, en un contexto latinoamericano donde los cazadores nahuas poseen sobre su territorio y las especies tales como hábitos alimentarios, distribución, rastros, cantos o sitios de refugio, los cuales son básicos para elegir donde cazar. Las prácticas de manejo registradas se enfocan en la obtención del recurso faunístico mediante la extracción por cacería y recolección en ambientes naturales y antropizados, no se registra ninguna enfocada al manejo de los usuarios, del hábitat o de las especies animales con la finalidad de aumentar o mantener el número de individuos. En este contexto, es pertinente evaluar integralmente el uso de la fauna silvestre, que además de considerar aspectos socioculturales también integre ecológicos como tasas de captura, biomasa extraída y estado de las poblaciones animales. Aunque la cacería de subsistencia puede representar una menor presión para las poblaciones animales en comparación con la comercial, la compleja interacción de factores como la pérdida y fragmentación del hábitat, una de las principales problemáticas en la parte baja del Parque Nacional El Tepozteco (González y ContrerasMacBeath 2020), así como la sobreexplotación ocasionada por la extracción excesiva, pueden impactar la fauna, lo que pone en peligro la conservación de la diversidad faunística, sus funciones ecológicas y los bienes que aportan a las poblaciones humanas (Naranjo et al. 2010, Ávila-Nájera et al. 2018b). tegias de subsistencia y representaciones simbólicas. De acuerdo a su conocimiento tradicional reconocen 45 especies, de las cuales el 84 % son nombradas en Náhuatl. Se registraron 41 especies entre vertebrados e invertebrados que forman parte de manifestaciones culturales tangibles e intangibles. Las principales categorías de uso fueron la alimentaria (57,5 %), pronosticador (25 %), medicinal (22,5 %) y animales con leyendas (12,5 %). El venado cola blanca obtuvo el mayor Índice de Importancia Cultural al registrar el mayor número de categorías de uso y mayor frecuencia de mención, sobresaliendo en su aporte de carne para la alimentación. La obtención de los recursos faunísticos se realiza mediante la recolección y la cacería individual o grupal en ambientes naturales o antropizados, mediante armas de fuego y tradicionales. Es necesario realizar estudios integrales que además de considerar aspectos socioculturales, también incluyan aspectos ecológicos para obtener una mayor profundidad en el análisis de la relación humano-fauna entre los habitantes nahuas de Santa Catarina, Morelos, México. PARTICIPACIÓN DE LOS AUTORES MIAE: Concepción teórica, diseño de la investigación, recolección de datos en campo e identificación de material botánico. AGF: Diseño de la investigación, análisis de datos y escritura del artículo. ERM: Toma y análisis de datos en campo. JMPM: Identificación de insectos. FGL: Toma y análisis de datos en campo e identificación de material botánico. AGRADECIMIENTOS A los habitantes de Santa Catarina por su participación y apoyo para la realización de la presente investigación. CONFLICTO DE INTERESES Los autores declaran no tener conflicto de interés. LITERATURA CITADA CONCLUSIONES La fauna silvestre es un recurso con importancia sociocultural para los pobladores nahuas de Santa Catarina, cuyo uso es parte de sus formas de vida al figurar en sus estra- Acosta E. 2013. La relación del itonal con el chikawalistli en la constitución y deterioro del cuerpo entre nahuas de Pahuatlán, Puebla. Dimensión Antropológica, 58:115-148. nahuas de Pahuatlán, 58:115-148. Puebla. Dimensión Antropológica, 45 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 Alcalá S, De la Riva G. 2016. Uso tradicional de fauna silvestre en las serranías del occidente del estado Aguascalientes, México. Etnobiol. 14(2):20-36. Ayala-Enríquez MI, Román-Montes de Oca E, García-Lara F. 2018. Caracterización del sistema milpa en Santa Catarina, Tepoztlán, Morelos, México. Acta agric. Pecuaria 5(1): 11-23. Alonso-Castro AJ. 2014. Use of medicinal fauna in Mexican traditional medicine. J. Ethnopharmacol. 152(1):53-70. doi: https:// doi.org/10.1016/j.jep.2014.01.005 Blandin P. 2007. The systematics of the genus Morpho, Fabricius, 1807: (Lepidoptera Nymphalidae, Morphinae). Canterbury, UK: Hillside Books. Alves RR, Alves HN. 2011. The faunal drugstore: Animalbased remedies used in traditional medicines in Latin America. J. Ethnobiol. Ethnomed. 7(9). doi: https://doi. org/10.1186/1746-4269-7-9 Cano-Contreras EJ, Medinaceli A, Sanabria-Diago OL, ArguetaVillamar A. 2014. Código de Ética para la investigación, la investigación-acción y la colaboración etnocientífica en América Latina. Etnobiol. 12(4):5-6. Alves RRN, Barboza RRD. 2017. Animals as Ethnozooindicators of Weather and Climate. En: Alves RRN, Albuquerque U, editores. Ethnozoology: Animals In Our Lives. Cambridge, Massachusetts: Academic Press. p. 383-420. doi: https://doi. org/10.1016/B978-0-12-809913-1.00021-1 Casas G. 2000. Mitos, leyendas y realidades de los reptiles en México. Cienc. Ergo Sum. 7:286−291. Alves RRN, Mendonça LET, Confessor MVA, Vieira WLS, Lopez LCS. 2009. Hunting strategies used in the semi-arid region of northeastern Brazil. J. Ethnobiol. Ethnomed. 5(12). doi: http:// dx.doi.org/10.1186/1746-4269-5-12 Alves, RRN, van Vliet N. 2018. Wild fauna on the menu. En: Alves RRN, Albuquerque UP, editors. Ethnozoology Animals in Our Lives. Oxford, UK: Elsevier. p. 167-194. doi: https://doi. org/10.1016/B978-0-12-809913-1.00010-7 Aranda Sánchez MJ. 2012. Manual para el rastreo de mamíferos silvestres de México. México: CONABIO, INECOL. doi: https:// doi.org/10.5962/bhl.title.113211 Argueta-Villamar A. 2008. Los saberes P’urhépecha. Los animales y el diálogo con la naturaleza. Morelia, México: UNAM, UMSNH, PNUMA, JP, Gobierno del Estado de Michoacán y Universidad Intercultural. Ávila-Nájera DM, David Mendoza G, Villarreal O, Serna-Lagunes R. 2018a. Uso y valor cultural de la herpetofauna en México: Una revisión de las últimas dos décadas (1997-2017). Acta Zool. Mex. Nueva serie. 34(1):1-15. doi: https://doi.org/10.21829/ azm.2018.3412126 Ávila-Nájera DM, Naranjo EJ, Tigar BJ, Villarreal OA, Mendoza GD. 2018b. An Evaluation of the Contemporary Uses and Cultural Significance of Mammals in Mexico. Ethnobiol. Lett. 9(2):124-135. doi: https://doi.org/10.14237/ebl.9.2.2018.1106 Ayala-Enríquez MI, Cuevas-Salgado MI, Montes de Oca ER. 2018. Conocimiento tradicional de plantas distractoras para fauna nociva en milpas de Santa Catarina, Tepoztlán, Morelos. Rev. Geograf. Agríc. 60:67-79. doi: https://doi.org/10.5154/r. rga.2018.60.002 Ayala-Enríquez MI, García-Lara F, Román-Montes de Oca E. 2019. La apropiación de los recursos naturales, silvestres y cultivados. Estrategia alimentaria de los pueblos originarios. En: Román-Montes de Oca E, editora. Prácticas agropecuarias como estrategias de seguridad alimentaria. Cuernavaca, México: Universidad Autónoma del Estado de Morelos. p. 70-102. 46 Casas G. 2004. Nuevas interpretaciones y adiciones a los anfibios y reptiles en la obra del naturalista Francisco Hernández (15171584). Cienc. Ergo Sum. 11(3):308-312. Castro-Franco R, Bustos-Zagal MG. 2006. Herpetofauna de las áreas naturales protegidas, Corredor Biológico Chichinautzin y la Sierra de Huautla, Morelos, México. 1ª Edición. Comisión Nacional para el Conocimiento y Uso de la Biodiversidad. México: Universidad Autónoma del Estado de Morelos. Chontal-Chagala Y, Aguilar-Cordero W, Chablé-Santos J. 2019. Estudio etnozoológico del uso tradicional de aves silvestres en la comunidad maya de Xanláh, Yucatán, México. Ethnoscientia 4. doi: https://doi.org/10.22276/ethnoscientia.v4i1.256 [DGSV] Dirección General de Sanidad Vegetal. 2016. Chapulines de Importancia Económica en México en cultivo de frijol: Brachystola magna, B. mexicana, Melanoplus differentialis, Sphenarium purpurascens y Taeniopoda spp. México: SEDAGRO, SENASICA. Flores-Monter YM, Reyna-Trujillo T, López-Wilchis R, NavarijoOrnelas L. 2017. Aproximación a la percepción sobre los murciélagos en la población de la mixteca poblano-oaxaqueña, México. Etnobiol. 15(2):16-31. Flores-Rodríguez MG. c2021. Tetzáhuitl en la cosmovisión indígena nahua. [Revisada en: 15 jul 2021]. https://www.gob.mx/inpi/es/ articulos/tetzahuitl-en-la-cosmovision-indigena-nahua?idiom=es García-Flores A, Lozano-García MA, Ortiz-Villaseñor AL, Monroy-Martínez R. 2014. Uso de mamíferos silvestres por habitantes del Parque Nacional El Tepozteco, Morelos, México. Etnobiol. 12(3):57-67. García-López R, Villegas A, Pacheco-Coronel N, Gómez-Álvarez G. 2017. Uso tradicional y percepción de serpientes por los nahuas de Cuetzalan del Progreso, Puebla, México. J. Ethnobiol. Ethnomed. 13(6). doi: https://doi.org/10.1186/s13002-016-0134-7 Gatica-Colima A, Jiménez-Castro JA. 2009. Serpientes de cascabel: percepción por algunos pobladores del desierto chihuahuense en el estado Chihuahua. Rev. Latinoame. Rec. Nat. 5(3):198-204. Gaviño G. 2015. Aves de Morelos. Estudio general, guía de campo y recopilación bibliográfica. México: CONABIO. Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 González L, Contreras-MacBeath T. 2020. Áreas naturales protegidas. En: Comisión Nacional para el Conocimiento y Uso de la Biodiversidad, editor. La biodiversidad del estado de Morelos. Estudio de Estado 2. Vol. II. México: CONABIO. p. 255-279. Good C. 2013. Formas de organización familiar náhuatl y sus implicaciones teóricas. La Ventana 37:9-40. Guerra R, Calme S, Gallina TS, Naranjo E. 2010. Uso y manejo de fauna silvestre en el norte de Mesoamérica. México: Instituto Nacional de Ecología. Gutiérrez-Santillán T, Moreno-Fuentes A, Sánchez-González A, Sánchez-Rojas G. 2019. Knowledge and use of biocultural diversity by Nahua in the Huasteca region of Hidalgo, Mexico. Ethnobiol. Conserv. 8(7):1-31. doi: https://doi.org/10.15451/ ec2019-06-8.07-1-31 Herrera-Flores B, Santos-Fita D, Naranjo EJ, Hernández-Betancourt S. 2019. Importancia cultural de la fauna silvestre en comunidades rurales del norte de Yucatán, México. Península 14(2):27-55. Herrera-Flores B, Santos-Fita D, Naranjo EJ, Hernández-Betancourt SF. 2018. Creencias y prácticas rituales en torno a la cacería de subsistencia en comunidades del norte de Yucatán, México. Etnobiol. 16(1):5-18. [INPI] Instituto Nacional de los Pueblos Indígenas. c2020. Etnografía de los nahuas de Morelos. [Revisada en: 30 may 2021]. https://www.gob.mx/inpi/articulos/ etnografia-de-los-nahuas-de-morelos Katz E. 2005. Las hormigas, el maíz y la lluvia. An. Antrop. 39:215-229. León P, Montiel S. (2008). Wild Meat Use and Traditional Hunting Practices in a Rural Mayan Community of the Yucatan Peninsula, Mexico. Hum. Ecol. 36:249-257. doi: https://doi. org/10.1007/s10745-007-9139-0 Leyte-Manrique A, Gutiérrez-Álvarez N, Hernández-Navarro E. 2016. Percepción cultural de la herpetofuana en tres comunidades rurales del municipio de Irapuato, Guanajuato, México. Etnobiol. 14(1):73-84. Mariño-Pérez PR, Song H. 2018. Phylogeny of the grasshopper family Pyrgomorphidae (Caelifera, Orthoptera) based on morphology. Syst. Entomol. 43(1):90-108. doi: https://doi. org/10.1111/syen.12251 Martín-Crespo BMC, Salamanca-Castro AB. 2007. El muestreo en la investigación cualitativa. Nure Inv. 27(1):1-4. Martínez M. 2014. Un estudio etnozoológico: la fauna en la cuenca del Balsas, Guerrero. Suplemento cultural el tlacuache INAH 626:3-4. Monroy R, Pino Moreno JM, Lozano García MA, García Flores A. 2011. Estudio etnomastozoológico en el Corredor Biológico Chichinautzin (COBIO), Morelos, México. Sitientibus, Sér. Ciênc. Biol. 11(1):16-23. doi: https://doi.org/10.13102/scb101 Morayta LM. 2011. Los pueblos nahuas de Morelos: atlas etnográfico. México: Gobierno del Estado de Morelos, Instituto Nacional de Antropología e Historia. Nahuat-Cervera P, Estrada-Riaño I, Peraza-Romero F, UitzilCollí M, Basora-Dorantes R, Buenfil-Morales S. 2021. Conocimiento y aprovechamiento tradicional de vertebrados silvestres en la comunidad maya de Zavala, municipio de Sotuta, Yucatán, México. Estud. Cult. Maya 57:275-304. doi: https://doi. org/10.19130/iifl.ecm.57.2021.18660 Naranjo EJ, López-Acosta JC, Dirzo R. 2010. La cacería en México. Biodiversitas 91:6-10. Olivier G. 2015. Cacería, sacrificio y poder en Mesoamérica. Tras las huellas de Mixcóatl, “Serpiente de Nube”. México: FCE, UNAM, CEMC. Osorio-López D, Mariaca-Méndez R, Santos-Fita D, Nazar-Beutelspacher DA, Huicochea-Gómez L. 2017. Cacería y cosmovisión en una comunidad ayuuk en San José El Paraíso, Oaxaca, México. Etnobiol. 15:54–66. Pastrana-Flores M. 2021. Presagios, augurios y portentos de las conquistas de Mesoamérica. Una perspectiva comparada. Rev. Hist. Cien. Soc. 1(1):15-40. doi: https://doi.org/10.22136/ korpus2120213 Patton MQ. 2002. Qualitative research and evaluation methods. Thousand Oaks, CA: Sage Publications. Pérez-Gil R, Jaramillo F, Muñiz A, Torres M. 1996. Importancia económica de los vertebrados silvestres en México. México: PG7 Consultores y CONABIO. Pitriello MA, Stronza AL. 2019. Campesino hunting and conservation in Latin America. Conserv. Biol. 34(2):338-353. doi: https://doi.org/10.1111/cobi.13396 Puc R, Retana O. 2012. Uso de la fauna silvestre en la comunidad maya Villa de Guadalupe, Campeche, México. Etnobiol. 10(2):1-11. Ramírez-Barajas JP, Calmé S. 2015. Subsistence Hunting and Conservation. En: Islebe G, Calme S, Leon J, Schmook B., editores. Biodiversity and Conservation of the Yucatán Peninsula. New York: Springer International Publishing. p. 333-354. doi: https://doi.org/10.1007/978-3-319-06529-8_13 Retana-Guiascón O, Lorenzo-Monterrubio C. 2016. Valor Cinegético y Cultural del Venado Cola Blanca en México. Etnobiol. 14(3):60-70. Romero L. 2020. Niñez, transformación y consumo de maíz entre los nahuas de la Sierra Negra de Puebla. Trace 78:5-19. doi: https://doi.org/10.22134/trace.78.2020.739 Sahagún B. 2005. Fauna de la Nueva España. Ciudad de México: Fondo de Cultura Económica. Santos-Fita D, Naranjo EJ, Rangel-Salazar JL. 2012. Wildlife uses and hunting patterns in rural communities of the Yucatan Peninsula, Mexico. J. Ethnobiol. Ethnomed. 8(38). doi: https://doi. org/10.1186/1746-4269-8-38 Santos-Fita D, Naranjo-Pinera Eduardo J, Bello-Baltazar E, Estrada-Lugo E, Mariaca-Méndez R, Macario-Mendoza, P. 2013. La milpa comedero-trampa como una estrategia de cacería tradicional maya. Estud. Cult. Maya 42:87-118. doi: https://doi. org/10.1016/S0185-2574(13)71387-X 47 Ayala Enríquez et al., 2023. Caldasia 45(1):36-48 Santos-Rodríguez A. 2009. Metodología de la investigación etnozoológica. En: Costa-Neto E, Santos-Fita D, Vargas M, editores. Manual de Etnozoología. España: Tundra ediciones. p. 253-272. SEDESOL Secretaría de Desarrollo Social. c2015. Catálogo de localidades de Tepoztlán. [Revisada en: 01 dic 2020]. http:// www.microregiones.gob.mx Serrano-González R, Guerrero-Martínez F, Serrano-Velázquez R. 2011. Animales medicinales y agoreros entre tzotziles y tojolabales. Estudios Mesoamericanos 11:29-42. Song H, Mariño-Pérez P, Woller DA, Cigliano MM. 2018. Evolution, diversification, and biogeography of grasshoppers (Orthoptera: Acrididae). Insect. Syst. Divers. 2(4):3. doi: https://doi. org/10.1093/isd/ixy008 Taboada M, Granjeno A, Oliver R. 2009. Normales climatológicas (temperatura y precipitación) del estado de Morelos. Cuernavaca, México: Universidad Autónoma del Estado de Morelos. Taylor SJ, Bogdan R. 1987. Introducción a los métodos cualitativos de investigación. Barcelona: Paydos. Tejeda-Cruz C, Naranjo-Piñera EJ, Medina-Sanson LM, Guevara-Hernández F. 2014. Cacería de subsistencia en comunidades rurales de la selva lacandona, Chiapas, México. Quehacer Científico en Chiapas 9(1):59-73. 48 Valadez-Azúa R, Rodríguez-Galicia B. 2013. Uso de la fauna, estudios arqueozoológicos y tendencias alimentarias en culturas prehispánicas del centro de México. An. Antrop. 48:139-166. doi: https://doi.org/10.1016/S0185-1225(14)70493-X Valadez-Azúa R. 2016: Importancia de la domesticación animal en la antigua ciudad de Teotihuacán. Antrop. Am. (1)1:103-125. Valdez R, Ortega S. 2014. Ecología y manejo de fauna silvestre en México. México: Colegio de Posgraduados. Ward PS. 2010. Taxonomy, phylogenetics, and evolution. En: Lach L, Parr C, Abbott K, editores. Ant Ecology. Oxford: Oxford University Press. p. 3-17. Zarazúa-Carbajal M, Chávez-Gutiérrez M, Romero-Bautista Y, Rangel-Landa S, Moreno-Calles AI, Alvarado-Ramos LF, Smith SE, Blancas J, Del Val E, Arizmendi M, Casas A. 2020. Use and management of wild fauna by people of the Tehuacán-Cuicatlán Valley and surrounding areas, Mexico. J. Ethnobiol. Ethnomed. 16(4). doi: https://doi.org/10.1186/s13002-020-0354-8 Zavala-Sánchez Z, Segura-Pacheco HR, Ávila-Nájera DM, Herrera-Castro ND, Barrera-Catalán E, Sarabia-Ruiz G. 2018. Valoración cultural y uso de la fauna silvestre en San Vicente de Benítez, Guerrero, México. Etnobiol. 16(3):78-92.
https://openalex.org/W2768605762	https://europepmc.org/articles/pmc5717252?pdf=render	English	null	Learning the value of information and reward over time when solving exploration-exploitation problems	Scientific reports	2,017	cc-by	13,011	Learning the value of information and reward over time when solving exploration-exploitation problems Received: 7 August 2017 Accepted: 22 November 2017 Published: xx xx xxxx Irene Cogliati Dezza1, Angela J. Yu3, Axel Cleeremans1 & William Alexander2 To flexibly adapt to the demands of their environment, animals are constantly exposed to the conflict resulting from having to choose between predictably rewarding familiar options (exploitation) and risky novel options, the value of which essentially consists of obtaining new information about the space of possible rewards (exploration). Despite extensive research, the mechanisms that subtend the manner in which animals solve this exploitation-exploration dilemma are still poorly understood. Here, we investigate human decision-making in a gambling task in which the informational value of each trial and the reward potential were separately manipulated. To better characterize the mechanisms that underlined the observed behavioural choices, we introduce a computational model that augments the standard reward-based reinforcement learning formulation by associating a value to information. We find that both reward and information gained during learning influence the balance between exploitation and exploration, and that this influence was dependent on the reward context. Our results shed light on the mechanisms that underpin decision-making under uncertainty, and suggest new approaches for investigating the exploration-exploitation dilemma throughout the animal kingdom. Imagine a nice summer day in Rome. You stop by a gelateria to buy a refreshing ice cream. After entering the shop, you are confronted with dozens of different ice-cream flavours. A good strategy would be to choose your favourite flavour (i.e., vanilla), because the likelihood that you will find it satisfying is very high. However, the shop also sells flavours that seem appealing but that you have never tasted (e.g., ginger and cinnamon). Do you select the flavour that you know you will enjoy, or do you select a flavour you have never tried before, potentially finding either a new favourite, or profound disappointment? The decision problem you are faced with is an exam- ple of the exploration-exploitation dilemma. It reflects a conflict between exploiting a known source of reward (i.e., vanilla) and exploring other less well-known sources (i.e., ginger and cinnamon) that may turn out to be either a more rewarding outcome, or a wasted opportunity.f Humans can resolve the exploration-exploitation dilemma through adopting different exploratory strate- gies, and different theoretical frameworks have offered different accounts of the factors that drive exploration. www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Received: 7 August 2017 Accepted: 22 November 2017 Published: xx xx xxxx Learning the value of information and reward over time when solving exploration-exploitation problems The process is iterated for the length of free-choice task and at the beginning of each game expected reward values are initialized to Q0 and information to zero. Figure 1. Comparison between sRL and kRL. Both models learn reward values Qt+1,j(c) during the forced-choice task using a δ learning rule, but only kRL learns also information I c( ) t j , - as function of the number of previous observations. At the end of the forced-choice task, kRL integrates information and reward in a same choice value Vt,j(c) where the expected reward value Qt+1,j(c) is reduce by the amount of information gained during previous trials and its weighted parameter ω. In order to generate choice probabilities based on choice values, both models use a softmax choice function. However, sRL enters only expected reward values Qt+1,j(c), whereas kRL enters the combined information-reward value Vt,j(c). The process is iterated for the length of free-choice task and at the beginning of each game expected reward values are initialized to Q0 and information to zero. recently developed a new version of the bandit task in which the information participants had about the payoffs of each slot machine was controlled, thereby eliminating the reward and information confound (i.e. information and reward were ‘decorrelated’). Based on this paradigm, Wilson et al. showed that participants indeed adopted both random and directed exploratory strategies: when participants were given unequal amounts of information regarding two options, they more frequently chose the option about which they had less information, even when that option was associated with lower gains.h p g Thus, withholding information about choice reward appears to promote directed exploration and to mod- ulate the manner in which humans resolve the exploration-exploitation dilemma. However, in Wilson et al.’s task, payoffs concerning previous decision outcomes were displayed to participants at the time of the decision. In daily life, however, outcomes concerning previous choices are not always perfectly available; instead, it is their learned values that play the central role during the decision process. Moreover, previously learned outcomes influence future decisions in a way that depends on the type of learning involved9, on the context in which deci- sions are made10 and on the weight that each individual gives to decision outcomes11. Nonetheless, the number of times an outcome is experienced during learning might eventually influence the decision process. Learning the value of information and reward over time when solving exploration-exploitation problems In reinforcement learning (RL) theory, exploration may be random: a decision-maker who learns to maximize a numerical reward signal1 may nevertheless make choices associated with lower reward value (exploration) due to a noisy response-generation process2. Alternatively, in optimal decision-making theories3, exploration is viewed as a directed or intentional process (i.e. directed or uncertainty-driven exploration). Here, the decision-maker is also trying to maximize reward, but he takes a longer view and is willing to sacrifice short-term high-reward choices so as to gain information about more uncertain novel options, ultimately hoping to achieve an overall higher level of reward4. In directed exploration, the absence of information related to the estimation of the reward value of an option constitutes a key factor for exploration: all else being equal (e.g., estimated reward value and reward variance of options), directed exploration minimizes the difference in information amongst options. On the contrary, random exploration is merely the result of a noisy process. y p y y p In previous research, evidence concerning the extent to which human decision-makers engage in directed exploration has been inconsistent. Some studies failed to observe directed exploration in humans5–7, possibly because of using tasks (i.e., “bandit tasks”) that make it difficult to separately identify directed and random exploratory strategies8 due to rise in information/reward confound2. To overcome this limitation, Wilson et al.2 1Centre for Research in Cognition & Neurosciences (CRCN), Université Libre de Bruxelles, Brussels, Belgium. 2Department of Experimental Psychology, Ghent University, Gent, Belgium. 3Department of Cognitive Science, University of California San Diego, La Jolla, CA, United States. Correspondence and requests for materials should be addressed to I.C.D. (email: icogliat@ulb.ac.be) SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 1 www.nature.com/scientificreports/ Figure 1. Comparison between sRL and kRL. Both models learn reward values Qt+1,j(c) during the forced-choice task using a δ learning rule, but only kRL learns also information I c( ) t j , - as function of the number of previous observations. At the end of the forced-choice task, kRL integrates information and reward in a same choice value Vt,j(c) where the expected reward value Qt+1,j(c) is reduce by the amount of information gained during previous trials and its weighted parameter ω. In order to generate choice probabilities based on choice values, both models use a softmax choice function. However, sRL enters only expected reward values Qt+1,j(c), whereas kRL enters the combined information-reward value Vt,j(c). SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w Learning the value of information and reward over time when solving exploration-exploitation problems (a) In the unequal information condition (when information concerning the 3 decks was not always available), sRL (right frames) predicts a higher probability of choosing highly rewarded option (i.e., exploitation), whereas kRL (left frames) predicts a higher probability of choosing the most informative/never experienced option (i.e., 0seen). (b) Both models predict a decrement in directed exploration in high reward contexts (when the generative mean was set to 50 points) compared to low reward contexts (when the generative mean was set to 30 points). However, in the high reward context, kRL still chooses more often 0seen options compared to exploitation (left frames), whereas sRL chooses more often exploitative options (right). (c) In the equal information condition (Baseline + Reward) under different reward contexts, both models predict increased random exploration in the high reward context compared to the low reward context. Figure 2. Model predictions. (a) In the unequal information condition (when information concerning the 3 Figure 2. Model predictions. (a) In the unequal information condition (when information concerning the 3 decks was not always available), sRL (right frames) predicts a higher probability of choosing highly rewarded option (i.e., exploitation), whereas kRL (left frames) predicts a higher probability of choosing the most informative/never experienced option (i.e., 0seen). (b) Both models predict a decrement in directed exploration in high reward contexts (when the generative mean was set to 50 points) compared to low reward contexts (when the generative mean was set to 30 points). However, in the high reward context, kRL still chooses more often 0seen options compared to exploitation (left frames), whereas sRL chooses more often exploitative options (right). (c) In the equal information condition (Baseline + Reward) under different reward contexts, both models predict increased random exploration in the high reward context compared to the low reward context. information2 to a dynamic and trial-dependent computation adaptable to more complex scenarios. Specifically, we computed information as discrete quantities, which accumulate over time as the number of observations of an option increases. Both learning models make different predictions regarding how humans solve the exploration and exploitation dilemma when information concerning competing options is not always available (e.g., not inte- grating information tends to facilitate exploitative strategies), and when it is presented under different learning contexts (e.g., higher reward context reduces exploration of most informative options in a way that differs when the algorithm integrates or not integrates information) (Fig. 2). Learning the value of information and reward over time when solving exploration-exploitation problems For example, increasing the number of times an option is observed decreases its uncertainty3, and the value of an option also depends on the level of uncertainty12. Accordingly, information might accumulate during learning, affecting the decision process itself. Consequently, learning information and reward over time might influence the resolution of the exploration-exploitation dilemma. Following this hypothesis, standard RL theories need to be augmented with optimal decision-making approaches specifying how reward and information interact to influence choices. Moreover, the context in which learning is achieved might influence this reward/information integration. Indeed, a vast literature has highlighted the crucial role of context in influencing decision strategies10,13,14. Experiencing richer contexts, for example, might speed up exploratory process as predicted by optimal foraging theory15. p g p p p y p p y p g g y In this study, we compare two possible computational solutions to the learning problem faced by humans during the exploration-exploitation dilemma: (1) a standard Reinforcement Learning (sRL) model where only rewards are considered during the learning and decision process; (2) and a novel reinforcement learning model in which quantities related both to reward and information are learned over time and combine to derive choice values (knowledge Reinforcement Learning- kRL; Fig. 1). Previous models2 have suggested how information and reward might be integrated in driving choice behaviour, but have not included trial-by-trial learning dynam- ics, thus potentially limiting the scope of their predictive power. Moreover, we extended previous measures of SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 2 www.nature.com/scientificreports/ Figure 2. Model predictions. (a) In the unequal information condition (when information concerning the 3 decks was not always available), sRL (right frames) predicts a higher probability of choosing highly rewarded option (i.e., exploitation), whereas kRL (left frames) predicts a higher probability of choosing the most informative/never experienced option (i.e., 0seen). (b) Both models predict a decrement in directed exploration in high reward contexts (when the generative mean was set to 50 points) compared to low reward contexts (when the generative mean was set to 30 points). However, in the high reward context, kRL still chooses more often 0seen options compared to exploitation (left frames), whereas sRL chooses more often exploitative options (right). (c) In the equal information condition (Baseline + Reward) under different reward contexts, both models predict increased random exploration in the high reward context compared to the low reward context. Figure 2. Model predictions. Learning the value of information and reward over time when solving exploration-exploitation problems To assess the models’ predictions, we developed a novel version of the Wilson et al. paradigm in which participants make repeated choices about three decks of cards in such a way that previous deck-selection influences choices through learning (i.e., outcomes disappeared from the screen after 300 ms of deck selection - Fig. 3). The use of three options allows us to investigate partici- pants’ behaviour in a more complex and realistic scenario. By comparing the fits of the two models to trial-by-trial participant responses, our aim was to establish which model better describes participants’ behaviour while per- forming our task. In the following sections, we present the main results whereas the mathematical description of the two models and their predictions, derived from model simulations, can be found in the last section of this manuscript. information2 to a dynamic and trial-dependent computation adaptable to more complex scenarios. Specifically, we computed information as discrete quantities, which accumulate over time as the number of observations of an option increases. Both learning models make different predictions regarding how humans solve the exploration and exploitation dilemma when information concerning competing options is not always available (e.g., not inte- grating information tends to facilitate exploitative strategies), and when it is presented under different learning contexts (e.g., higher reward context reduces exploration of most informative options in a way that differs when the algorithm integrates or not integrates information) (Fig. 2). To assess the models’ predictions, we developed a novel version of the Wilson et al. paradigm in which participants make repeated choices about three decks of cards in such a way that previous deck-selection influences choices through learning (i.e., outcomes disappeared from the screen after 300 ms of deck selection - Fig. 3). The use of three options allows us to investigate partici- pants’ behaviour in a more complex and realistic scenario. By comparing the fits of the two models to trial-by-trial participant responses, our aim was to establish which model better describes participants’ behaviour while per- forming our task. In the following sections, we present the main results whereas the mathematical description of the two models and their predictions, derived from model simulations, can be found in the last section of this manuscript. SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 3 www.nature.com/scientificreports/ Figure 3. Behavioral paradigm. Learning the value of information and reward over time when solving exploration-exploitation problems (d) During a game, participants faced 6 consecutive trials of the forced-choice task and between 1 and 6 trials of the free-choice task. The number of free-choice trials was exponentially distributed such that a higher proportion of games allowed subjects to make 6 free choices. Analysis in the Results section mostly focuses on choices made by participants during the first trial of the free-choice task (in yellow) being the only trial where reward and information were uncorrelated2. Learning the value of information and reward over time when solving exploration-exploitation problems The number of times each deck was played varied throughout the experiment: 2 times each deck - equal information condition; 0, 2, 4 times - unequal information condition. (b) During the free-choice task, participants made their own decisions among the same three decks of cards displayed during the forced-choice task. Participants were instructed to attempt to maximize the total points earned at the end of the experiment. After each trial, the points displayed on the screen were added to the participants’ total score. (c) Participants indicated their choices using the forefinger, middle finger and ring finger and pressing the keyboard keys ‘1’, ‘2’ and ‘3’, respectively. (d) During a game, participants faced 6 consecutive trials of the forced-choice task and between 1 and 6 trials of the free-choice task. The number of free-choice trials was exponentially distributed such that a higher proportion of games allowed subjects to make 6 free choices. Analysis in the Results section mostly focuses on choices made by participants during the first trial of the free-choice task (in yellow) being the only trial where reward and information were uncorrelated2. Figure 3. Behavioral paradigm. A variant of Horizon task2, in which participants chose between options in two different phases: forced-choice task and free-choice task. (a) During the forced-choice task, three decks of cards were displayed on the screen (a blue, a red and green deck) and participants were forced to choose a preselected deck outlined in blue. After selecting the deck, the card turned and revealed the points associated with the selected option, between 1 and 100 points. At this stage, the points displayed to participants were not added to their total score. To allow learning, feedbacks of previous trials did not remain visible to participants. The number of times each deck was played varied throughout the experiment: 2 times each deck - equal information condition; 0, 2, 4 times - unequal information condition. (b) During the free-choice task, participants made their own decisions among the same three decks of cards displayed during the forced-choice task. Participants were instructed to attempt to maximize the total points earned at the end of the experiment. After each trial, the points displayed on the screen were added to the participants’ total score. (c) Participants indicated their choices using the forefinger, middle finger and ring finger and pressing the keyboard keys ‘1’, ‘2’ and ‘3’, respectively. Learning the value of information and reward over time when solving exploration-exploitation problems A variant of Horizon task2, in which participants chose between options in two different phases: forced-choice task and free-choice task. (a) During the forced-choice task, three decks of cards were displayed on the screen (a blue, a red and green deck) and participants were forced to choose a preselected deck outlined in blue. After selecting the deck, the card turned and revealed the points associated with the selected option, between 1 and 100 points. At this stage, the points displayed to participants were not added to their total score. To allow learning, feedbacks of previous trials did not remain visible to participants. The number of times each deck was played varied throughout the experiment: 2 times each deck - equal information condition; 0, 2, 4 times - unequal information condition. (b) During the free-choice task, participants made their own decisions among the same three decks of cards displayed during the forced-choice task. Participants were instructed to attempt to maximize the total points earned at the end of the experiment. After each trial, the points displayed on the screen were added to the participants’ total score. (c) Participants indicated their choices using the forefinger, middle finger and ring finger and pressing the keyboard keys ‘1’, ‘2’ and ‘3’, respectively. (d) During a game, participants faced 6 consecutive trials of the forced-choice task and between 1 and 6 trials of the free-choice task. The number of free-choice trials was exponentially distributed such that a higher proportion of games allowed subjects to make 6 free choices. Analysis in the Results section mostly focuses on choices made by participants during the first trial of the free-choice task (in yellow) being the only trial where reward and information were uncorrelated2. Figure 3. Behavioral paradigm. A variant of Horizon task2, in which participants chose between options in two different phases: forced-choice task and free-choice task. (a) During the forced-choice task, three decks of cards were displayed on the screen (a blue, a red and green deck) and participants were forced to choose a preselected deck outlined in blue. After selecting the deck, the card turned and revealed the points associated with the selected option, between 1 and 100 points. At this stage, the points displayed to participants were not added to their total score. To allow learning, feedbacks of previous trials did not remain visible to participants. Results In this section, we first describe the analysis related to participants’ behaviour investigating the presence of behavioural patterns in participants’ data predicted by each model (qualitative model comparison analysis). Subsequently, we present a quantitative analysis to investigate which model better explained participants’ choices. Because all participants accurately performed the gambling-task, playing strategically and learning reward out- comes throughout the experiment (for details refers to supplementary material), we did not exclude any partici- pants from our analyses. Qualitative model comparison analysis. Prediction 1: Directed Exploration vs. Exploitation. The first prediction of kRL and sRL concerns the information manipulation condition, in which decks are sampled une- venly during the forced-choice task (Fig. 2a). Under this scenario, the kRL model predicts more frequent selection of the never-experienced deck (0seen) whereas sRL predicts to more frequent selection of options associated with highest expected reward (exploitation). To investigate how information manipulation affected participants’ strategy selection, we computed directed exploration, random exploration and exploitation in the first free choice trials of the Information condition (where information was manipulated but mean reward was equal among decks; see Methods). We conducted the analysis on the first trial of each game, it being the only trial where reward and information confound were uncorrelated2. Trials were classified as directed exploratory when par- ticipants chose the option that was never selected during forced-choice trials (0seen deck), exploitative when participants chose the experienced deck with the highest average of points (regardless of the number of times that deck had been selected during the forced-choice task) and random exploratory when the classification did SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 4 www.nature.com/scientificreports/ Figure 4. Human behaviour. (a) When information concerning the 3 decks was not always available – unequal information – participants were more likely to directed explore in the first trial of the free-choice task compare to exploit and random explore as predicted by kRL. The graph reports results related with Information condition only (when mean reward were equals among the decks). (b) When unequal amounts of information were available from the decks (left), the probability of exploitation and random exploration increases under the High Reward context while the probability to directed explore decreases as predicted by both models. However, participants showed higher preferences toward directed exploratory options compared to exploitative ones as predicted by kRL. Results In the equal information condition under different reward contexts (right), participants were more likely to engage in random exploration in the High Reward context and to engage in exploitation in the Low Reward context as predicted by both RL models. Figure 4. Human behaviour. (a) When information concerning the 3 decks was not always available – unequal information – participants were more likely to directed explore in the first trial of the free-choice task compare to exploit and random explore as predicted by kRL. The graph reports results related with Information condition only (when mean reward were equals among the decks). (b) When unequal amounts of information were available from the decks (left), the probability of exploitation and random exploration increases under the High Reward context while the probability to directed explore decreases as predicted by both models. However, participants showed higher preferences toward directed exploratory options compared to exploitative ones as predicted by kRL. In the equal information condition under different reward contexts (right), participants were more likely to engage in random exploration in the High Reward context and to engage in exploitation in the Low Reward context as predicted by both RL models. not meet the previous criteria. Averaged values were entered into single factor repeated measure ANOVA. Results revealed an effect of information on decision strategies F(1,40) = 80.65, p < 10−14 (Fig. 4a). Pairwise comparison using paired t-test showed significant differences between all comparisons with all p values <10−4, indicating that under the manipulation of information participants adopted a significantly more directed-exploratory strat- egy (M = 0.682, SD = 0.168) to the detriment of random exploration (M = 0.086, SD = 0.1069) and exploitation (M = 0.232, SD = 0.132). This analysis is in line with the first kRL prediction where the probability of engaging in directed exploration is higher compared to exploitation (for details refer to method section). Similar results were obtained investigating the decision strategies overall in the unequal information condition (Information + Mixed condition) and in the Mixed condition alone (where both information and reward were manipulated- see sup- plementary material). Prediction 2: reward context. The second prediction concerns the effect of reward context on decision strate- gies (Fig. 2b). Both the sRL and kRL models predict reward context-dependent changes in selecting the 0seen deck (for details see method section). Results To investigate the models’ predictions, we computed random exploration and exploitation in the first free choice trials of the Information condition under High and Low Reward con- text. We conducted a 2 (context: Low Reward, High Reward) by 3 (strategies: exploitation, random exploration, directed exploration) non-parametric ANOVA. The test allows the use of two-way repeated measure ANOVA in a non-parametric setting using aligned rank transformation (e.g., ARTool package in R, http://depts.washington. edu/madlab/proj/art/)16. Results showed an effect of strategy F(2,100) = 161.9, p < 10−15, and an effect of context X strategy F(2,100) = 26.359, p < 10−9 (Fig. 4b), whereas we did not find a general effect of context, p = 0.926. Post-hoc comparisons showed an increase in exploitation in High Reward context (M = 0.329, SD = 0.171) compared to the Low Reward (M = 0.134, SD = 0.128), p < 10−5; a decrease in directed exploration in the High Reward context (M = 0.561, SD = 0.201) compared to the Low Reward context (M = 0.803, SD = 0.149), p < 10−7; and an increment in random exploration in the High Reward context (M = 0.11, SD = 0.092) compared to the Low Reward context (M = 0.062, SD = 0.074), p = 0.042. The above results are in line with both models’ predic- tions where the probability to be engaged in directed exploration was reduced in High Reward context. However, the two models draw different predictions concerning the degree by which the 0seen deck is selected. As a result of information integration in choice value, kRL predicts higher probability of selecting the 0seen deck in both reward contexts, whereas sRL predicts to select most often exploitative options in High Reward context. Post-hoc comparisons in the High Reward context showed that participants performed directed exploration significantly more compared to exploitation, p = 0.0104 as predicted by the kRL model. p p p p y To examine the models’ predictions in the equal information condition (both models predict an increase in random exploration in the High Reward context- Fig. 2c), we computed decision strategies under High and Low Reward context. Here, the first trial of each game was labelled as exploitation when participants selected the deck with the highest average number of points gained during the forced-choice task, and as random explo- ration otherwise. Results Instead, in order to disentangle directed exploration from random exploration, we used a version of the task in which subjects were faced with 3 options. Given our version of the task did not include a specific horizon manipulation, it is possible that our interpretation of choices of the 0seen option as evidence of directed exploration may be confounded with “uncertainty preference”17. As discussed in that study, uncertainty preference arises when, in choosing the least observed option, the potential use of future information is irrelevant (in contrast with directed exploration where potential use of future information is the driving factor of information seeking). These authors observed an increase in uncertainty preference in a loss context compared to gain context17. However, our use of 3 options allows us to rule out uncertainty preference as the sole explanation for the observed decrease in selec- tion of the 0seen option in high reward contexts relative to low reward contexts. Because uncertainty preference should act on both 0seen and 2seen options, if uncertainty preference was the driving factor in our task, we should also expect to find higher selection of the 2seen option relative to the 4seen option during the Low Reward context compared to the High Reward context. We therefore tested this by computing the probability of choosing each of the non-zero options (2seen and 4seen) in High and Low Reward context of the Information condition. A 2 (option: 2seen, 4seen) by 2 (context: High Reward, Low Reward) repeated measure ANOVA did not show any effect of option p = 0.289, any effect of context p = 0.584, or interaction effect between context and option p = 0.919, ruling out uncertainty preference as a factor underlying differences in choice behaviour between high and low reward contexts (Fig. 5). Quantitative model comparison analysis. Our analyses of behavioural data provide support for the kRL account of participants’ behaviour. We additionally conducted a quantitative model comparison analysis to investigate which model best predicted participants’ decision process. We first investigated the goodness-of-fit by comparing kRL and sRL model responses to participants’ responses toward 0seen options in the Information con- dition. R-squared computation revealed that kRL was able to explain approximately the 90% of variance in par- ticipants’ data, R2 (19) = 0.8975 p < 10−11, whereas sRL explained approximately 10%, R2 (19) = 0.073 p = 0.236. Results A 2 (context: Low Reward, High Reward) by 2 (strategy: exploitation, random exploration) non-parametric ANOVA was adopted. The results showed an effect of strategy F(1,60) = 216.4, p < 10−15, and effect of condition X strategy F(1,60) = 26.9, p < 10−5) (Fig. 4b), whereas a general effect of context did not SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 5 www.nature.com/scientificreports/ Figure 5. Uncertainty preference and ‘seen analysis’. Probability to choose in the first free-choice trials the option experienced 2 times (2seen) and 4 times (4seen) during the forced-choice task under different reward conditions. 2seen and 4seen decks were chosen at the same rate within and between reward conditions suggesting that choices labelled as directed exploration where indeed guided by an exploratory drive and not by participants’ preference toward uncertainty. Figure 5. Uncertainty preference and ‘seen analysis’. Probability to choose in the first free-choice trials the option experienced 2 times (2seen) and 4 times (4seen) during the forced-choice task under different reward conditions. 2seen and 4seen decks were chosen at the same rate within and between reward conditions suggesting that choices labelled as directed exploration where indeed guided by an exploratory drive and not by participants’ preference toward uncertainty. reach significance, p = 1. Post-hoc comparison indicated an increment in random exploration in High Reward (M = 0.331, SD = 0.139) compared to Low Reward context (M = 0.186, SD = 0.115), p < 10−4 as predicted by both sRL and kRL model. In the supplementary material, we show that this increase in random exploration in the High Reward context might be due to an effect of utility vs. cost induced by reward context (Fig. S2). Horizon, Uncertainty Preference, and Directed Exploration. In Wilson et al.2, directed exploration was quanti- fied as an increase in exploratory decisions on the first free-choice trial for conditions in which participants were allowed multiple free choices vs. conditions in which they were allowed a single choice. In our version of the task, we do not include a specific horizon manipulation: the number of free-choice trials was selected randomly from an exponential distribution such that, in a plurality of games, subjects were allowed 6 free-choices. However, subjects were not informed as to the number of free-choice trials they would experience during a game. Results Subsequently, we graphically investigated the two models in terms of negative log likelihood. Only one participant did not benefit from one of the two models laying on the identity line (Fig. 6a). However, because R-squared and likelihood tend to favour complex models (i.e., models with more explanatory variables) we also compared the likelihoods of the data given the model while accounting for their different numbers of free parameters (i.e., model complexity). To do so, we computed BIC and AIC18,19 values as reported in the methods section. We conducted a classical frequentist analysis with BIC values of the two models entered into the Wilcoxon Signed Rank Test. Results showed that kRL model (BICkRL = 163.978) best represented participants’ data com- pared to sRL (BICsRL = 241.838), Z = 3.945, p < 10−6 (Fig. 6b). Results for AIC values were similar to those for BIC values (Table 1). Fitting to all free-choice trials did not change model comparison results, and simulating both models with the best-fit parameters reproduced behavioural patterns observed during simulation with random SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 6 www.nature.com/scientificreports/ Figure 6. Comparative fit of the kRL and sRL. (a) The comparison of the fit is based on the negative log likelihood (i.e., probability of observing the data given the model) of both models. Each point is one participant. The kRL fits better when the point is below the 45° line. Except for one participant who is lying on the identity line, all participants’ behaviour is better represented by kRL model. (b) BIC of both models (i.e., the likelihoods of the data given the model while accounting for their different numbers of free parameters). kRL had a significantly lower BIC compared to sRL. Model with lower BIC values are better able to explain the variability in the data. Figure 6. Comparative fit of the kRL and sRL. (a) The comparison of the fit is based on the negative log likelihood (i.e., probability of observing the data given the model) of both models. Each point is one participant. The kRL fits better when the point is below the 45° line. Except for one participant who is lying on the identity line, all participants’ behaviour is better represented by kRL model. (b) BIC of both models (i.e., the likelihoods of the data given the model while accounting for their different numbers of free parameters). Results kRL had a significantly lower BIC compared to sRL. Model with lower BIC values are better able to explain the variability in the data. sRL (1stchoices) kRL (1stchoices) gkRL (1stchoices) sRL (All Trials) kRL (All Trials) gkRL (All Trials) BIC 241.838 (32.1) 163.978 (50.603) 163.141 (52.469) 809.064 (98.152) 791.092 (97.723) 683.505 (143.789) AIC 236.135 (32.099) 155.424 (50.603) 151.736 (52.47) 800.554 (98.153) 778.326 (97.724) 666.484 (144.791) Neg. Log-Lik 116.068 (16.05) 74.712 (25.302) 71.868 (26.235) 398.277 (49.077) 386.163 (48.862) 329.242 (72.395) Table 1. Model Comparison. BIC, AIC, negative log-likelihood group average and standard deviation reported for sRL, kRL, and gkRL models fitted to both the 1st and all free-choice trials. Models that best scored during the model comparison are underlined. sRL (1stchoices) kRL (1stchoices) gkRL (1stchoices) sRL (All Trials) kRL (All Trials) gkRL (All Trials) BIC 241.838 (32.1) 163.978 (50.603) 163.141 (52.469) 809.064 (98.152) 791.092 (97.723) 683.505 (143.789) AIC 236.135 (32.099) 155.424 (50.603) 151.736 (52.47) 800.554 (98.153) 778.326 (97.724) 666.484 (144.791) Neg. Log-Lik 116.068 (16.05) 74.712 (25.302) 71.868 (26.235) 398.277 (49.077) 386.163 (48.862) 329.242 (72.395) Table 1. Model Comparison. BIC, AIC, negative log-likelihood group average and standard deviation reported Table 1. Model Comparison. BIC, AIC, negative log-likelihood group average and standard deviation reported for sRL, kRL, and gkRL models fitted to both the 1st and all free-choice trials. Models that best scored during the model comparison are underlined. parameters (details for both analysis in the supplementary). In sum, both our qualitative and quantitative model analyses indicate that participants’ choice behaviour integrated both reward and information into a choice value. parameters (details for both analysis in the supplementary). In sum, both our qualitative and quantitative model analyses indicate that participants’ choice behaviour integrated both reward and information into a choice value. Non-linear information gain. Although kRL was better able to explain the participants’ behaviour using a para- metric integration of information, our assumption, building from the binary coding of information suggested by Wilson et al.2, that information scales linearly with the number of samples (equation 4) may be incorrect. It may be the case that information gain is subject to diminishing returns: each additional observation of an option may yield progressively less additional information regarding that option. Results Alternately, it may be the case that, at least for situations (as in the present task) in which only a few samples of each option are available, additional observa- tions may provide a greater-than-linear amount of information. To better investigate the nature of the scaling, we evaluated the kRL model against a version (the gamma kRL, or gkRL model) that allows information to increase sub- or super-linearly with the number of observations: ∑ =       =    ≠ = γ I c i c i c choice c choice c ( ) ( ) where, ( ) 0, 1, (1) t j t t j t j , 1 , , (1) Here information was computed by including an exponential term, gamma, for the sum of observations It,j(c) and then integrating this computation in the choice value Vt,j(c) as in equation (5). Gamma defines the degree of non-linearity in the amount of information obtained as a function of the number of observations, it,j Choices were made as in equation 6. As with the kRL model, we fit the gkRL model to both first free-choice trials and all free-choice trials and conducted model comparisons. A paired t-test on BIC values showed no difference between SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 7 www.nature.com/scientificreports/ gkRL model (BICgkRL = 163.141) and kRL (BICkRL = 163.978), t(20) = −0.595, p = 0.559, when fit to first-free choices only, whereas a paired t-test on AIC values revealed a lower value in gkRL model (AICgkRL = 151.736) and kRL (AICkRL = 155.424), t(20) = −2.62, p = 0.016. However, when fit to all free-choices, Wilcoxon Signed Rank Test showed gkRL model (BICgkRL = 683.505) best represented participants’ data compared to kRL (BICkRL = 791.092), Z = 6.054, p < 10−4. The discordant results obtained when fitting only the first free-choice trials might be due to the bias toward simple models for BIC and to complex models for AIC20. While the gkRL model was not decisively better than the kRL model when fitted to first free-choice trials, the BIC and AIC scores from all trials suggest at least equivocal evidence in favour of the gkRL model. Analysis of the gamma parameter estimates suggest that, to the extent the gkRL model explains our data better than the kRL model, it is due to a sub-linear influence of observations on information. Discussion Th l The exploration-exploitation dilemma is a fundamental conflict faced by humans when adapting to the demands of their environment. The resolution of the dilemma is achieved following either a random or a directed strategy. However, in previous work directed strategies were mostly confounded by random strategies making difficult to investigate whether information processing influenced the resolution of the dilemma. In this study, we imple- mented a version of a previous paradigm that is able to separately identify directed and random strategies2 and we examined how learning about reward and information over time influences the resolution of dilemma. We mod- ified the previous version of the paradigm to make learning a requisite for good performance (see also21), and we compare two computational solutions to investigate the learning problem. Our results suggest that humans learn a combined reward/information value when facing sequential exploration-exploitation problems. Moreover, the degree to which reward and information are integrated depends on the learning context in which decisions are made (i.e., the number of times options are selected and the overall level of reward associated with the options). Together, these findings shed additional light on the mechanisms underlying decision-making under uncertainty and suggest new approaches for investigating the exploration-exploitation dilemma. In the following, we discuss the implications of our main results. Our data showed that participants played strategically during the games, learning about reward outcomes throughout each game. Moreover, participant behaviour was better described by a δ learning rule compared to estimating an average of observed outcomes (supplementary material), as suggested by previous works2. Furthermore, our results showed that choices made by participants reflected the integration of both reward and information over time, replicating findings that information-seeking behaviour intervenes during the resolution of the exploration-exploitation dilemma2,22,23. When making decisions among alternatives that differ in their informative value, participants preferentially sample from options that will provide more information, inhibiting their default tendency to choose highly rewarded options5 in favour of more uncertain, but potentially better, alternatives. Under previous formulation, information gain is prospective, and prospective gains in information contribute positive value to the expected value (in points) of choosing an uncertain option2. Conversely, in kRL’s formulation, information is considered retrospective - the amount of information acquired during the past- and it is added as negative value: directed exploration is realized by reducing the probability of selecting well-known alternatives. Results The mean gamma parameter estimate when fit only to the first free-choice trials was 0.411, significantly lower than 1 (linear), p < 10−4. This sub-linear influence is more pronounced in fits of the gkRL model to all free-choice trials: here the mean gamma estimate is 0.103, signifi- cantly lower than the estimated gamma parameter for fits only to the first free-choice trials (p = 0.002). These results suggest that, as the number observations increases (in the free-choice trials), the sub-linear information gain becomes more relevant, while in the first free-choice trial, the effective level of information gained from 4 observations, while less than double that gained from 2 observations, is still behaviourally relevant. SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w Discussion Th l Our results showed that participants integrated information as parametric function of the number of observations of one option compared to a simple binary coding scheme signifying greater or lesser information (supplementary material) as previous models would predict2. Our results also suggest that information integra- tion is sub-linearly related to number of observations, consistent with decreases in uncertainty in the Bayesian posterior (quantified by either entropy or variance). While both retrospective and prospective formulations can account for exploratory behaviours in humans, the exact manner in which information is computed and is inte- grated in choice values remains an open question and should be the subject of future research.i While our study replicates prior findings of the use of information during the resolution of the dilemma, the evidence concerning directed exploration has is not been consistent in the literature. Several studies have failed to find information integration during exploration5,6. For example, in Daw et al.5, a model that considered only reward as driving factor for its choices was better able to explain participants’ performance on a bandit task com- pared to a model that also integrated information. However, the particular instantiation of bandit tasks in those studies might explain this failure8. Generally, in bandit tasks, reward and information are confounded24, hence making it difficult to discern random from directed exploration2. Additionally, bandit tasks often involve drifts in the rate of payoffs associated with each bandit5, leading to the possibility that participants cannot accurately track reward mean or variance as predicted by the model. Therefore, even if subjects did utilize directed exploration, the model would not be able to uncover it due to mis-modeling of subjects’ probabilistic beliefs and the reward distributions. Additionally, our data indicated that the reward context (i.e., the overall level of reward associated with the options) may influence how humans integrate reward and information when solving the exploration-exploitation dilemma. In higher reward contexts, participants showed an increment in their preference for highly rewarded decks and a decrement in their preference for the most informative option. This effect was not due to a context-dependent decrease in uncertainty preference, but it was instead a specific effect of directed exploration. These results are in line with previous findings where learning contexts appeared to influence decision-making strategies13,14. Methods P i i Participants. Twenty-one young healthy adults participated in this study (12 women; aged 19–29 years, mean age = 23.24). The inclusion criteria were as follows: normal color vision, right-handed, and absence of psycho- active treatment. The entire group belonged to the Belgian Flemish-speaking community. The experiment was approved by the UZGent ethics committee and conducted according to the Declaration of Helsinki. Informed consent was obtained from all participants prior to the experiment. Procedure. Participants played 128 games of a gambling-task, where on each game repeated choices were made among 3 decks of cards (Fig. 3). The task is a variant of the Horizon task2, in which participants chose between two options in two different phases: forced-choice task and free-choice task. Our experimental task differs in four ways from the Wilson et al. task. First, participants chose among three options, depicted as decks of cards (Fig. 3). Second, information regarding the points earned following a choice did not remain visible following feedback, allowing learning during the task. Third, while the forced-choice task lasted 6 trials in all the games (constant number as in Wilson et al.2), the length of the free-choice task ranged from 1 to 6 trials. The number of free-choice trials was exponentially distributed such that a higher proportion of games allowed subjects to make 6 free choices. Fourth, the length of the free-choice task (i.e., horizon) was not cued to participants, so that in our version participants were not aware of the total number of free-choice trials they would have on any trial.ht The decks of cards were arranged on the left, centre and right side of the computer screen (Fig. 3) and partic- ipants indicated their choices using the keyboard keys ‘1’, ‘2’ and ‘3’, respectively. After each choice, the chosen card was flipped to reveal a number indicating the points earned by the participant for selecting that deck. Each option paid off between 1 and 100 points, and the number of points earned for selecting a deck was sampled from a Gaussian distribution with standard deviation of 8 points. On each game, the generative mean of each deck was set to 30 (Low Reward condition) or 50 (High Reward condition). Discussion Th l Moreover, the framing effect10 can explain the decrement in directed exploration observed in our SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 8 www.nature.com/scientificreports/ study during high reward context: when experiencing gains, humans tend to engage in uncertainty-averse behav- iours17,25. Furthermore, framing the dilemma in terms of losses has also been associated with exploratory strat- egies26. Exploratory behaviour thus seems to depend both on the amount of information regarding available options as well as the context in which the decision is made. However, the results of our study concerning random exploration nuance this view; we found that random exploration increased under high reward context both in the equal and unequal information conditions. Our results suggest that the decrease in random exploration in low reward context might be explained by an increase in cost or by a decrease in utility of switching among alter- natives in this condition (supplementary material). These findings are in line with predictions made by optimal foraging theory15: in a rich habitat, exploration among patches may increase because searching further is more likely to yield better outcomes, whereas in poorer environments, patch exploitation increases because it will be more difficult and costly to find better outcomes27. Thus, two different processes appear to underlie the resolution of the exploration-exploitation dilemma under different reward contexts, and their characterization depends on the type of exploratory strategies adopted. Our results are in line with recent findings where random and directed exploration showed distinct underlying mechanisms21, however our results also confirm that both random and directed exploration show a common exploratory drive (see supplementary).i p p y ( pp y) Although our study replicated findings of the use of information by humans during the resolution of the dilemma bringing additional knowledge on the learning mechanism involved, there are nonetheless limitations that may influence the scope of our results. These limitations relate primarily to the formulation of our task as a series of discrete games in which task-relevant information (global mean reward value, information about options) is reset prior to the start of each game. Discussion Th l In typical bandit tasks, which have been well-studied from a cog- nitive and machine-learning perspective, the value of options evolves over the course of an experiment, leading to the incorporation of mechanisms by which the influence of previous observations is discounted as a function of time28, estimates of uncertainty related to ongoing experience with the task5,22, and continuous adjustments of value predictions for options. In order to adapt the kRL model to these more traditional tasks, similar mecha- nisms may need to be included in future iterations of the model. y Despite these limitations, a new view of the exploration-exploitation dilemma is suggested, considering explo- ration and exploitation trade-off as a class of problems spanning different scales29,30. We showed that depend- ing on the scale adopted (information, reward, experience, context) the resolution of the dilemma might differ as well as its hidden mechanisms. Future studies should take into account this perspective to better explain exploration-exploitation phenomenon throughout the animal kingdom. SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w www.nature.com/scientificreports/ option should be closer together and around chance. Conversely, if the change in information maximization is due to decreased directed exploration, thus not affecting the level of noise in participants’ choices, the probabili- ties of choosing the other two options should not change from one context to another. g p g During the forced-choice task, we manipulated the information about the decks of cards acquired by par- ticipants (i.e., the number of times each deck of cards was played). The information manipulation we adopted differed from Wilson et al. Here, on each game, participants were forced to either choose each deck 2 times (equal information condition), or to choose one deck 4 times, another deck 2 times, and 0 times for the remaining deck (unequal information condition). We decided to adopt this manipulation so as to maintain the differences of infor- mation among decks (4 times – 2 times and 2 times – 0 times) to be two, as in Wilson et al. In 50% of the games, participants played with the equal information condition. The order of card selection was randomized in both information conditions. Participants were instructed to choose among the three decks of cards in the free-choice task so as to maximize their final score (i.e., the amount of points earned throughout the experiment). The total score was converted to a monetary payoff at the end of the experiment (i.e., 0.01 euros every 60 points). Participants were told that during the forced-choice task of certain games they could sample options at different rate, and that the decks of cards did not change during the same game, but were replaced by new decks at the beginning of each new game. However, they were not informed of the reward manipulation and the underlying generative distribution we adopted.f y p y g g p Crossing the information and reward manipulations, we therefore obtained 4 different conditions: Reward, Information, Baseline, and Mixed. Reward and information were the main conditions, whereas Baseline and Mixed were the control conditions. The main conditions were used to test our hypotheses. Each of the main con- ditions consisted of 42 games each. In the Reward condition, reward was manipulated (unequal reward), whereas information was equal (equal information). In the Information condition the generative means of each option were equal (equal reward) and information varied (unequal information). www.nature.com/scientificreports/ The sRL model uses a simple δ learning rule31 to compute the expected reward value Q(c) or each deck of cards c (=Left, Central or Right) on each trial, using the following equation: α δ = + × + Q c Q c ( ) ( ) (2) t j t j t j 1, , , α δ = + × + Q c Q c ( ) ( ) t j t j t j 1, , , (2) where Qt,j(c) is the expected reward value for trial t and game j. δ = − R c Q c ( ) ( ) t j t j t j , , , is the prediction error, which quantifies the discrepancy between the previous predicted outcome and the actual outcome obtained at trial t and game j. The prediction error enables the model to adjust its predictions on a trial-by-trial basis, adapt- ing reward expectations to new information. The degree by which new outcomes are integrated with the expected reward value Qt+1,j(c) depends on the learning rate α. With small α the model slowly updates it estimate of Q in response to new information (i.e., points obtained after a trial), whereas with higher values the model integrates new information more rapidly. In other words, with lower learning rate the model expects lower rate of change of the environment32. The expected reward Qt+1,j(c) is updated using the above rule only if an outcome from the deck c is observed, otherwise Qt+1,j(c) keeps the value of the previous trial. In order to generate choice probabili- ties based on expected reward values, the model uses a softmax choice function5,33,34. The softmax rule is expressed as: β β \| = × ∑ × + + + ( ) ( ) ( ) P c Q c Q c Q c ( ) exp ( ) exp exp ( ) (3) t j i t j i t j i 1, 1, 1, (3) where β is a free parameter known in reinforcement learning as the inverse temperature. β determines the degree to which choices are directed toward the highest rewarded option. With higher β the model is mainly exploitative, whereas with lower β the model chooses the deck more randomly. Considering that participants were told that games were independent from one another, Q0 is initialized at the beginning of each game35. www.nature.com/scientificreports/ The control conditions consisted of 20 Baseline games and 22 Mixed games. In the Baseline condition, both reward and information for each deck were equal (i.e. the generative means and information were equal among the 3 decks, and each deck was seen twice during the forced-choice phase), whereas in the Mixed condition both manipulations were introduced (unequal reward/unequal information). In each condition, reward context was manipulated by setting the mean of the Gaussian to 30 in 50% of the games and to 50 in the rest of the games to which the noise of +/− 0, 4, 12, 20 points per deck was added for each specific game. For each participant, the order of appearance of the 4 conditions was selected randomly, so each participant experienced a different game sequence. Computational models. To achieve good performance during our task, participants needed to learn reward outcomes associated with each deck of cards on a trial-by-trial basis. Tracking this information helps them max- imize their final score and earn more money during the experiment. Because our task involved learning reward outcomes over time, we implemented two computational models based on reinforcement learning. In order to quantify how reward and information may influence choice behaviour, we compared two alternative RL models to determine which model could potentially explain participants’ behaviour during our task. The first model is a standard Reinforcement Learning model (sRL) that focuses its learning on reward outcomes only, and uses reward expectations as the driving factor of choice. The second model is a knowledge Reinforcement Learning that learns reward outcomes (as in sRL), and tracks the information gained from experienced decks. At the time of the decision, kRL integrates reward and information into an overall choice value. A comparison between the features of sRL and kRL is shown in Fig. 1. In the following sections, we present the mathematical description of the two models as well as their predictions, derived from model simulations. tandard RL model. Methods P i i To avoid the influence of learned meta-information regarding the task (specifically, the possibility that participants may learn that, on each game, the points for each deck are generated from a specific mean of determinable value, and thus could estimate the value of a deck after a single trial), deck means were additionally adjusted by +/− 0, 4, 12, & 20 points on each game (generative Gaussian ranges from 10 to 70 points). The mean and standard deviation of the generative Gaussian were stable within a game and varied between games. The 3 decks of cards had the same generative means in 50% of the games (equal reward) and different means in the rest of the games (unequal reward). In the unequal reward condition, the generative means differed so that two options had higher means compared to the third one in 25% of the total games (High Reward Context) and had lower means in the remaining games of the unequal reward condition (Low Reward Context).hf The use of three options allows us to dissociate effects related to directed exploration from random exploration between different reward contexts. For example, if choices that maximize information gain decrease from one context to another, this decrease could be a consequence either of decreased directed exploration (participants become less information seeker), or increased random exploration (participants’ choices become noisier). If the change is due to an increase in random exploration, then, for trials in which the subject chose one of the remain- ing available options in which information gain is not maximized, the probabilities of the subject choosing either SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 9 www.nature.com/scientificreports/ www.nature.com/scientificreports/ Information accumulated during the forced-choice task scales values Vt,j(c) so that increasing the number of observations of one option decreases its final value. In other words, when one option is over-selected It,j(c) becomes larger resulting in lower Vt,j(c) On the contrary, if one option is never-selected It,j(c) is zero, and = + V c Q c ( ) ( ) t j t j , 1, . ω is a weighting parameter that determines the degree by which the model integrates information in choice value Vt,j(c).t ,j Choices are made probabilistically using the softmax rule where options with higher values Vt,j(c) result in higher probability to be selected: β β = × ∑ × ( ) ( ) ( ) P c V c V c V c / ( ) exp ( ) exp exp ( ) (6) t j i t j i t j i , , , (6) Model predictions. In order to derive behavioural predictions, we simulated the sRL and kRL models performing our sequential gambling-task (forced-choice task + free-choice task). For the purpose of the simulation, we drew random values for model parameters α]0,1], β]0,20], ω]0,20] at every simulation. Both models were simulated 80 times. As the difference between the two models related specifically to information computation, we examined their predictions prin- cipally in the Information condition, even though the model was simulated in all experimental conditions. Predictions for the equal information conditions are also reported considering their relevance for the behavioural analysis. Model predictions. In order to derive behavioural predictions, we simulated the sRL and kRL models performing our sequential gambling-task (forced-choice task + free-choice task). For the purpose of the simulation, we drew random values for model parameters α]0,1], β]0,20], ω]0,20] at every simulation. Both models were simulated 80 times. As the difference between the two models related specifically to information computation, we examined their predictions prin- cipally in the Information condition, even though the model was simulated in all experimental conditions. Predictions for the equal information conditions are also reported considering their relevance for the behavioural analysis. Prediction 1: Directed Exploration vs. Exploitation. In the information manipulation condition, in which decks are sampled unevenly during the forced-choice task, the kRL model predicts more frequent selection of the 0seen deck compared to the other two decks. www.nature.com/scientificreports/ However, despite the independence among games, we assumed that participants could learn the long-run average of all decks over where β is a free parameter known in reinforcement learning as the inverse temperature. β determines the degree to which choices are directed toward the highest rewarded option. With higher β the model is mainly exploitative, whereas with lower β the model chooses the deck more randomly. Considering that participants were told that games were independent from one another, Q0 is initialized at the beginning of each game35. However, despite the independence among games, we assumed that participants could learn the long-run average of all decks over SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 10 www.nature.com/scientificreports/ the course of the experiment; that is, participants did not begin each game as if starting the experiment with no knowledge, but with an estimate of the possible average value of each deck. In other words, prior reward expec- tations Q0 are set as the value of the mean between the two-main generative Gaussian distributions. To obtain a global estimate of the expected reward values for each deck, we simulated the δ learning rule on 1200 forced-trials ×128 games. For each game, Q values were initialized to 0 and updated over all 1200 trials. The final Q values for each game were recorded, and the average of the final Q values (~40 points for each option) for each of the 128 games were used as the value of Q0 (equation 2). the course of the experiment; that is, participants did not begin each game as if starting the experiment with no knowledge, but with an estimate of the possible average value of each deck. In other words, prior reward expec- tations Q0 are set as the value of the mean between the two-main generative Gaussian distributions. To obtain a global estimate of the expected reward values for each deck, we simulated the δ learning rule on 1200 forced-trials ×128 games. For each game, Q values were initialized to 0 and updated over all 1200 trials. The final Q values for each game were recorded, and the average of the final Q values (~40 points for each option) for each of the 128 games were used as the value of Q0 (equation 2). Knowledge RL model. www.nature.com/scientificreports/ To account for potential influence of information on choice values, we implemented a version of the sRL that incorporates a mechanism reflecting the knowledge gained about each deck during a game - the kRL model. As in the sRL, the kRL model computes expected reward values using the δ learning rule described in equation 1, and Q0is initialized using the global estimate of reward values (Q0). Additionally, kRL tracks information gained from each deck based on how often it is selected, as follows: ∑ = =   ≠ − I c I c I c choice c choice c ( ) ( ) where, ( ) 0, 1, (4) t j t t j t j , l , , (4) It,j(c) is the amount of information associated with the deck c at trial t and game j. After six trial of forced choice task, if one option has never been selected It,j(c) has value zero, whereas in the case one option is selected 4 times, It,j(c) has the value 4. This dynamic and trial-dependent measure of information has been developed as an extension of the previous ‘nothing-or-all’ information bonus discussed in2. As in previous work2, information is assumed to interact with reward in influencing choice behaviour as an additional factor in determining value, and this interaction may positive (more information regarding a deck increases the subjective value of that deck) or negative (more information regarding a deck decreases that value of selecting that deck). In the context of the current task, on the first free choice trial, acquired information is thought to have a negative drive- it is beneficial for the subject to explore additional options that they never or less experienced during the forced- choice task. Therefore, in the kRL model, we model the impact of information as negative: after each forced-choice task, kRL subtracts the information gained It,j(c) to the expected reward value Qt+1,j(c): ω = − ∗ + V c Q c I c ( ) ( ) ( ) (5) t j t j t j , 1, , ω = − ∗ + V c Q c I c ( ) ( ) ( ) t j t j t j , 1, , (5) Vt,j(c) is the final value of the deck c. ω is a weighting parameter that determines the degree by which the model integrates information in choice value Vt,j(c). SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w www.nature.com/scientificreports/ www.nature.com/scientificreports/ Prediction 2: Reward Context. Both the sRL and kRL models predict reward context-dependent changes in directed exploration (Fig. 2 b, both p < 10−3). In the high-reward context (generative mean of all decks is 50) the expected reward Qt+1,j(c) from experienced decks is higher than the prior expectation of the decks Q0. The long-run value of Q0 is around the mean of the two Gaussian distributions (~40 points). In the High Reward context, Qt+1,j(c) increases for decks that are sampled during the forced-choice component of the task, while the expected value of the 0seen decks is not updated, = + Q seen Q (0 ) t j 1, 0 . p p + Q Q ( ) t j 1, 0 As a consequence, the expected value of the 0seen deck is lower than the expected value of the other two decks, and the probability to select the 0seen deck decreases. Conversely, in the Low Reward context, the Q values for decks sampled during the forced-choice task decrease below the prior Q value, with the result that the Q value for the 0seen deck tends to be higher, resulting in an increased likelihood of selecting the 0seen option. However, the two models draw different predictions concerning the degree by which the probability of selecting the 0seen deck. Because kRL integrates information with Q values during choice, the choice value Vt,j(c) the probability of selecting the 0seen deck is higher relative to the 2seen and 4seen options in both reward contexts (Fig. 2b, all p < 10−3). In contrast, while the sRL predicts a higher probability of selecting the 0seen option in the Low Reward context (similar to kRL), it also predicts a higher probability of selecting the option with the highest Q value in the High Reward Context (Fig. 2b, both p < 10−3). Additionally, we also looked at the effect of reward contexts on ran- dom exploration in the equal information condition. Both models predict participants to be more likely to engage in random exploration in the High Reward context compared to Low Reward context (Fig. 2c, both p < 10−3). Model fitting and model comparison. To decide which model better explains participants’ behaviour during the task, we collected trial-by-trial participants’ choices to fit the free parameters α, β, (sRL & kRL) and ω (kRL) of each model (Table S1). www.nature.com/scientificreports/ Parameters were fit using the first free-choice trials, as in2 and all free-choice trials as reported in the supplementary material. During the fitting procedure, the negative log likelihood - ∑= = P c log( ( )) j j 1 1 128 for each participant under each model was computed. The negative log likelihood was then minimized to obtain an estimate of the parameters for each participant under both models using MATLAB and Statistics Toolbox Release 2015b unconstrained optimization function fminsearch. The resulting negative-log like- lihoods were used to compute the model evidence (or the log model evidence - the probability of obtaining the observed data given a particular model). We adopted an approximation to the (log) model evidence, namely Bayesian Information Criterion (BIC18) and Akaike Information Criterion (AIC19). Model evidence was used as a comparing estimate of the accuracy of the two models in describing participants’ performance on our task. Statistical analysis. Statistical analysis was performed using RStudio (https://www.rstudio.com/), functions and packages adopted are reported in the results section. To determine how reward and information affected strategy selection, we conducted repeated measure ANOVA. When violations of parametric tests were indi- cated, non-parametric tests were performed. All statistical tests were two-tailed. Multiple tests were corrected by Bonferroni correction and a p-value of <0.05 was considered significant. Data availability. The data that support the findings of this study are available from the corresponding author upon request. References 1. Sutton, R. S. & Barto, A. G. Reinforcement learning: An introduction. (MIT Press, 1998). 1. Sutton, R. S. & Barto, A. G. Reinforcement learning: An introduction. (MIT Press, 1998). 1. Sutton, R. S. & Barto, A. G. Reinforcement learning: An introduction. (MIT Press, 1998). f g 2. Wilson, R. C., Geana, A., White, J. M., Ludvig, E. A. & Cohen, J. D. Humans use directed and random exploration to solve the explore-exploit dilemma. Journal of experimental psychology. General 143, 2074–2081, https://doi.org/10.1037/a0038199 (2014).h explore exploit dilemma. Journal of experimental psychology. General 143, 2074 2081, https://doi.org/10.1037/a0038199 (2014). 3. Gittins, J. & Jones, D. In Progress in statisics (ed J. Gans) 241–266 (The Netherlands: North-Holland, 1974).h p p f p p y gy p g 3. Gittins, J. & Jones, D. In Progress in statisics (ed J. Gans) 241–266 (The Netherlands: North-Holland, 1974). , J J , g ( J ) (h , ) 4. Averbeck, B. B. Theory of choice in bandit, information sampling and foraging tasks. PLoS computational biology 11, e1004164 https://doi.org/10.1371/journal.pcbi.1004164 (2015). g j 5. Daw, N. D., O’Doherty, J. P., Dayan, P., Seymour, B. & Dolan, R. J. Cortical substrates for exploratory decisions in humans. Nature 441, 876–879, https://doi.org/10.1038/nature04766 (2006). 6. Payzan-LeNestour, E. & Bossaerts, P. Risk, unexpected uncertainty,and estimation uncertainty: Bayesian learning in unstable settings. PLoS Computational Biology, e1001048, https://doi.org/10.1371/journal.pcbi.1001048 (2011). 7. Allais, M. Le Comportement de l’Homme Rationnel devant le Risque: Critique des Postulats et Axiomes de l′Ecole Americaine Econometrica 21, 503–546 (1953). 8. Cavanagh, J. F., Figueroa, C. M., Cohen, M. X. & Frank, M. J. Frontal theta reflects uncertainty and unexpectedness during exploration and exploitation. Cereb Cortex 22, 2575–2586, https://doi.org/10.1093/cercor/bhr332 (2012). p p p g 9. Daw, N. D., Niv, Y. & Dayan, P. Uncertainty-based competition between prefrontal an control. Nat Neurosci 8, 1704–1711, https://doi.org/10.1038/nn1560 (2005). 9. Daw, N. D., Niv, Y. & Dayan, P. Uncertainty-based competition between prefrontal and dorsolateral striatal systems for behavioral control. Nat Neurosci 8, 1704–1711, https://doi.org/10.1038/nn1560 (2005). 9. Daw, N. D., Niv, Y. & Dayan, P. Uncertainty based competition between prefrontal and dorsolateral striatal systems for behaviora control. Nat Neurosci 8, 1704–1711, https://doi.org/10.1038/nn1560 (2005). y y t Neurosci 8, 1704–1711, https://doi.org/10.1038/nn1560 (2005). p g 10. Kahneman, D. & Tversky, A. Prospect theory: An analysis of decision under risk. Econometrica 47, 263–291 (1979). 11. Behrens, T. E., Woolrich, M. W., Walton, M. E. & Rushworth, M. F. www.nature.com/scientificreports/ Because kRL integrates information into a composite value estimate, the proba- bility of choosing the 0seen deck is higher compared to exploitation (i.e., choosing the deck with the highest average of points obtained in the forced-choice task) or random exploration (i.e., choosing the experienced deck with lower point average) (Fig. 2a) (both p < 10−3). Under kRL, the value of an option is adjusted by the amount of information gained during the forced-choice task. Increasing the number of observations of a deck decreases its value Vt,j(c) and thus its probability of being selected. Because the 0seen deck has not been observed, there is no penalty associated with it, resulting in a higher Vt,j(c) and higher probability of selecting the 0seen over the 4seen and 2seen options.h g g ,j g p y g p The sRL model, on the other hand, predicts a higher probability of selecting the option with the highest expected reward (exploitation), compared to the other two options (Fig. 2a, both p < 10−2). Because in sRL infor- mation does not influence the value of observed options, the tendency to select 0seen is reduced and, the prob- ability of selecting highly-rewarded and well-known options increases. Thus, under the kRL model, humans are expected to preferentially select options in order to gain more information, while the sRL model suggests that humans should select the option with the highest expected payoff. SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 11 SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w www.nature.com/scientificreports/ www.nature.com/scientificreports/ 1. Zajkowski, W. K., Kossut, M. & Wilson, R. C. A causal role for right frontopolar cortex in directed, but not random, exploration. Elife 6, https://doi.org/10.7554/eLife.27430 (2017).f Elife p g 2. Frank, M. J., Doll, B. B., Oas-Terpstra, J. & Moreno, F. Prefrontal and striatal dopaminergic genes predict individual differences in exploration and exploitation. Nat Neurosci 12, 1062–1068, https://doi.org/10.1038/nn.2342 (2009). p p p g ( ) 23. Payzan-LeNestour, E. & Bossaerts, P. Do not bet on the unknown versus try to find out more: Estimation uncertainty and “unexpected uncertainty” both modulate exploration. Frontiers in Neuroscience, 150, https://doi.org/10.3389/fnins.2012.00150 (2012). 4. Hertwig, R. & Erev, I. The description-experience gap in risky choice. Trends Cogn Sci 13, 517–523, https://doi.org/10.1016/j tics.2009.09.004 (2009). 5. Platt, M. L. & Huettel, S. A. Risky business: the neuroeconomics of decision making under uncertainty. Nat Neurosci 11, 398–403 https://doi.org/10.1038/nn2062 (2008).f p g 26. Cooper, J. A., Blanco, N. J. & Maddox, W. T. Framing matters: Effects of framing on older adults’ exploratory decision-making. Psychol Aging 32, 60–68, https://doi.org/10.1037/pag0000146 (2017). y g g p g p g 7. Stephens, D. W., Couzin, I. & Giraldeau, L. In Cognitive search: evolution, algorithms, and the brain (eds Todd, P. M., Hills, T. T. & Robbins, T.W.) (2012). 8. Garivier, A. & Moulines, E. In ALT (ed J. Kivinen, Szepesvári, C., Ukkonen, E., Zeugmann, T.) 174–188 (Springer, Berlin, Heidelberg (2011). ( ) 29. Cohen, J. D., McClure, S. M. & Yu, A. J. Should I stay or should I go? How the human brain manages the trade-off between exploitation and exploration. Philos Trans R Soc Lond B Biol Sci 362, 933–942, https://doi.org/10.1098/rstb.2007.2098 (2007).f 30. Mehlhorn, K. et al. Unpacking the exploration–exploitation tradeoff: A synthesis of human and animal literatures. Decision 2, 191–215 (2015). 1. Rescorla, R. A. & Wagner, A. R. A theory of Pavlovian conditioning: Variations in the effectiveness of reinforcement and nonreinforcement. Classical conditioning: Current research and theory 64–99 (1972). 32. O’Reilly, J. X. Making predictions in a changing world-inference, uncertainty, and learning. Front Neurosci 7, 105, https://doi. org/10.3389/fnins.2013.00105 (2013). g 33. Wilson, R. C. & Niv, Y. Inferring relevance in a changing world. Front Hum Neurosci 5, 189, https://doi.org/10.3389/ fnhum.2011.00189 (2011).f 4. Humphries, M. D., Khamassi, M. & Gurney, K. Dopaminergic Control of the Exploration-Exploitation Trade-Off via the Basa Ganglia. Front Neurosci 6, 9, https://doi.org/10.3389/fnins.2012.00009 (2012). 35. Khamassi, M., Enel, P., Dominey, P. F. & Procyk, E. www.nature.com/scientificreports/ Medial prefrontal cortex and the adaptive regulation of reinforcement learning parameters. Prog Brain Res 202, 441–464, https://doi.org/10.1016/B978-0-444-62604-2.00022-8 (2013). Author Contributions I.C.D. and W.A. designed and carried out the experiment and discussed the computational modelling. I.C.D. performed the analysis of the data and the model analysis. I.C.D., A.J.Y. and W.A. discussed and interpreted the data. I.C.D., A.J.Y., A.C. and W.A. wrote the manuscript. SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w References Learn Neurosci 10, 1214–1221, https://doi.org/10.1038/nn1954 (2007).f 11. Behrens, T. E., Woolrich, M. W., Walton, M. E. & Rushworth, M. F. Learning the value of information in an uncertain world. Nat Neurosci 10, 1214–1221, https://doi.org/10.1038/nn1954 (2007).f Neurosci 10, 1214–1221, https://doi.org/10.1038/nn1954 (2007).f p g 2. Bossaerts, P., Preuschoff, K. & Hsu, M. In Neuroeconomics: Decision Making and the Brain (eds Glimcher, P. W., Camerer, C. F., Fehr E. & Poldrack, R. A.) (Academic Press, 2008). p g 12. Bossaerts, P., Preuschoff, K. & Hsu, M. In Neuroeconomics: Dec f E. & Poldrack, R. A.) (Academic Press, 2008). ) ( ) 3. Seymour, B. & McClure, S. M. Anchors, scales and the relative coding of value in the brain. Curr Opin Neurobiol 18, 173–178, https:/ doi.org/10.1016/j.conb.2008.07.010 (2008). g j ( ) 4. Green, L. & Myerson, J. A discounting framework for choice with delayed and probabilistic rewards. Psychol Bull 130, 769–792 https://doi.org/10.1037/0033-2909.130.5.769 (2004).h p g 15. Stephens, D. W. & Krebs, J. R. Foraging Theory. (Princeton University Press, 1986). h 16. Conover, W. J. & Iman, R. L. Rank transformations as a bridge between parametric and nonparametric statistics. Ame Statistician 35, 124–129 (1981). , ( ) 17. Krueger, P. M. Strategies for exploration in the domain of losses. Judgement and Decision Making 12, 104–117 (2017). 17. Krueger, P. M. Strategies for exploration in the domain of losses. Judgement and Decision Making 12, 104–117 (2017). g 19. Akaike, H. A new look at the statistical model identification. IEEE Trans. Automatic Control, 716–723 (1974).t i 20. Kass, R. E. & Raftery, A. E. Bayes factors and model uncertainty. (University of Washington, 1993). SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 12 Acknowledgements g ICD is a researcher supported by F.R.S.-FNRS grant (Belgium). This work was in part funded by NSF CRCNS grant (BCS-1309346) to AJY. AC is a research director with the F.R.S.-FNRS (Belgium). WA was supported in part by FWO-Flanders Odysseus II Award #G.OC44.13N. Additional Information Additional Information Supplementary information accompanies this paper at https://doi.org/10.1038/s41598-017-17237-w. Competing Interests: The authors declare that they have no competing interests. Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2017 SCIENtIfIC Reports \| 7: 16919 \| DOI:10.1038/s41598-017-17237-w 13
https://openalex.org/W4299914177	http://kig.kartografija.hr/index.php/kig/article/download/253/463	Bosnian	null	11th Annual Assembly of the Croatian Cartographic Society	DOAJ (DOAJ: Directory of Open Access Journals)	2,013	cc-by-sa	4,835	11th Annual Assembly of the Croatian Cartographic Society March 1 5, 201 3, Zagreb March 1 5, 201 3, Zagreb The 11th Annual Assembly of the Croatian Cartographic Society (CCS) took place at the Hall of the Faculty of Architecture, Geodesy and Civil Engin- eering, Kačićeva 26 in Zagreb on March 15, 2013. The assembly was led by Prof. Dr. Miljenko Lapaine, the president of CCS. knowledge gained during the training atthemilitaryacademy. promotion of the monograph Topo- graphic maps on the territory of Croatia at Croatian State Archive in cooperation with State Geodetic Administration (SGA). After the lecture there was a dis- cussion during which Dr. Pahernik gladly answered to all given questions. Afterwards, the president of CCS awar- ded the colonel a monograph Topo- graphic maps on the territory of Croatia. Then the president of CCS started the official part of the Assembly with a proposalofagenda: On May 23, 2012 at the Ceremonial hall of the University of Zadar, there was a commemorative event on 500 years of Gerard Mercator’s birth. The organizers were Geography and His- tory departments of the University of Zadar, Croatian Cartographic Society and Croatian Geographic Society Zadar. There were four lectures held at the Event. M. Lapaine held the lecture on Mercator’sprojection. At the beginning Assist. Prof. Dr. Dražen Tutić, the Vice-Dean for educa- tion and students greeted everyone in behalf of the Faculty of Geodesy. After- wards, in the first unofficial part, Col- onel Dr. Mladen Pahernik held keynote lecture Military Cartography at Croatian MilitaryAcademy. 1. Approvalofagenda l f h bl 2. Approval of the previous assembly minutes 3. President's report about work in 2012 The lecture consisted of three seg- ments: on Croatian Military Academy Petar Zrinski, programmes connected to information technology and carto- graphy on different levels of Military Academy studies and CAX simulation. The Assembly participants learnt about the history of Croatian Military Aca- demy Petar Zrinski and all its levels, through a short film during the first segment of the lecture. The second seg- ment was linked to a programme inter- connection of geoscientific subjects on all academy levels, passed by soldiers through every training level and in all fields, from inventorying of space at the lowest level, through space analysis, to space management at the highest level of training. Geosubjects are con- nected to geodesy, geoinformation, GIS and geography. FROM PROFESSIONAL SOCIETIES 11th Annual Assembly of the Croatian Cartographic Society March 1 5, 201 3, Zagreb FROM PROFESSIONAL SOCIETIES 11th Annual Assembly of the Croatian Cartographic Society March 1 5, 201 3, Zagreb FROM PROFESSIONAL SOCIETIES 11. godišnja skupština Hrvatskoga kartografskog društva 1 5. ožujka 201 3., Zagreb Jedanaesta godišnja skupština Hr- vatskoga kartografskog društva (HKD) održana je 15. ožujka 2013. s početkom u 12 sati u Vijećnici AGG fakulteta, Ka- čićeva 26 u Zagrebu. Skupštinu je vodio prof. dr. sc. Miljenko Lapaine, predsjed- nikHKD-a. Nakon predavanja razvila se ras- prava u kojoj je dr. Pahernik rado odgo- varao na postavljena pitanja. Nakon rasprave predsjednik HKD-a je u znak zahvalnosti pukovniku uručio mono- grafiju Topografske karte na području Hr- vatske. Zatim je predsjednik HKD-a otvorio službeni dio godišnje skupštine prijedlogomdnevnogreda: Najprije je doc. dr. sc. Dražen Tutić, prodekan za nastavu i studente u ime Geodetskog fakulteta pozdravio sve na- zočne. Zatim je u prvom, neslužbenom dijelu, pukovnik dr. sc. Mladen Pahernik održao pozvano predavanje Vojna karto- grafijanaHrvatskomvojnomučilištu. 1. Utvrđivanjednevnogareda 2. Ovjera zapisnika s prethodne Skup- štine 3. Izvještaj predsjednika o radu u 2012.godini 4. Financijskiizvještajza2012.godinu 5. IzvještajNadzornogodbora 6. Planradaza2013.godinu 7. Financijskiplanza2013.godinu 8. DodjelaZahvalnicaHKD-a 9. Različito 1. Utvrđivanjednevnogareda 2. Ovjera zapisnika s prethodne Skup- štine 3. Izvještaj predsjednika o radu u 2012.godini 4. Financijskiizvještajza2012.godinu 5. IzvještajNadzornogodbora 6. Planradaza2013.godinu 7. Financijskiplanza2013.godinu 8. DodjelaZahvalnicaHKD-a 9. Različito 1. Utvrđivanjednevnogareda 2. Ovjera zapisnika s prethodne Skup- štine 3. Izvještaj predsjednika o radu u 2012.godini Predavanje je bilo strukturirano u tricjeline:oHrvatskomvojnomučilištu Petar Zrinski, programima koji su ve- zani uz informatiku i kartografiju na različitim razinama studija vojnog uči- lišta te o simulaciji CAX. Kroz kratki film u prvom dijeli predavanja, sudi- oniciSkupštinesaznalisuopovijesnom razvoju Hrvatskog vojnog učilišta Petar Zrinski i svim razinama od kojih se uči- lište sastoji. Drugi dio predavanja bio je vezan uz programsku povezanost ge- oznanstvenih predmeta na svim razi- nama učilišta, koji vojnici prolaze kroz sve stupnjeve školovanja i u svim stru- kama, od inventarizacije prostora na najnižoj razini, preko analize prostora pa sve do upravljanja prostorom na najvišoj razini školovanja. Geopredmeti su vezani uz geodeziju, geoinformacije, GIS i geografiju. Kroz CAX simulaciju prikazanajeutrećemdijelupredavanja stožernavježbakojunakrajustudijana vojnom učilištu polaznici polažu, a po- vezuje sva teorijska znanja koju su voj- nici stekli tijekom školovanja na voj- nomučilištu. Sjednice Predsjedništva održane su uskladusaStatutomčetiriputaugodini. akademija u povodu 500. obljetnice ro- đenja Gerarda Mercatora. Organizatori su bili Odjel za geografiju i Odjel za po- vijest Sveučilišta u Zadru, Hrvatsko kartografsko društvo i Hrvatsko geo- grafsko društvo – Zadar. Na svečanoj akademiji održana su četiri predavanja. OMercatorovojprojekcijipredavanjeje održaoM.Lapaine. U 2012. objavljeni su brojevi 17 i 18 našeg časopisa Kartografija i geoinfor- macije. 11th Annual Assembly of the Croatian Cartographic Society March 1 5, 201 3, Zagreb In the third segment, CAX simulation showed a staff exercise which is taken by military academy attendants at the end of the pro- gramme, and it connects all theoretical OnJuly9,2012CCSorganizedatour Zagreb–Varaždin–Borl–Jeruzalem–Za- greb on the occasion of 250 years of Vi- enna meridian survey and a visit to the oldest trigonometry point in Croatia. 25 people participated in the tour. That tour should have been the first step in attempt to mark out, with a memorial or a monument, the final point of Vi- enna meridian survey (the Varaždin Cathedraltower). 4. Financialreportfor2012 5. ReportoftheSupervisoryBoard 6. Workplanfor2013 8. AwardingCCSacknowledgments 9. Miscellaneous The chairmanship meetings had been held four times in a year in ac- cordancewiththeStatute. Arrangements and negotiations with Croatian State Archive (CSA) about keeping the SDI Days started on December 15, 2011 with the idea of keepingthejoint4thNSDIandINSPIRE Day and the 8th Cartography and Geoinformation Conference. The ne- gotiations lasted until May 14, 2012 when the president of CCS was invited to a meeting with the head of SGA, Dr. Danko Marinković. At that meeting it was agreed that CCS and SGA will or- ganize the conference together, in September 2012. SDI Days was the five In 2012 two issues of our journal, Cartography and Geoinformation, num- bers 17 and 18, were published. To pre- parethetextlayoutandtherelease,the open source software Scribus was star- ted to be used. The appearance of the journal has been slightly altered in ac- cordancetographicdesignerremarks. On March 5, 2012 there was a 500 year anniversary of birth of Gerard Mercator, officially celebrated with a symposium and an appropriate exhib- itionattheFacultyofGeodesy. O A il 18 2012 CCS i d In 2012 two issues of our journal, Cartography and Geoinformation, num- bers 17 and 18, were published. To pre- parethetextlayoutandtherelease,the open source software Scribus was star- ted to be used. The appearance of the journal has been slightly altered in ac- cordancetographicdesignerremarks. On March 5, 2012 there was a 500 year anniversary of birth of Gerard Mercator, officially celebrated with a symposium and an appropriate exhib- itionattheFacultyofGeodesy. On April 18 2012, CCS organized a KiG No. 19, Vol. 12, 2013 KiG No. 19, Vol. 12, 2013 182 182 IZ DRUŠTAVA KiG Br. 19, Vol. 12, 2013 11. godišnja skupština Hrvatskoga kartografskog društva Za pripremu časopisa za prije- lom teksta i za tisak prešlo se na slobo- dan (open source) softver Scribus. Izgled časopisa je malo promijenjen u skladu s primjedbamagrafičkogdizajnera. Dana 9. 6. 2012. HKD je organizirao stručni obilazak Zagreb–Varaždin–Borl –Jeruzalem–Zagreb u povodu 250. go- dišnjice izmjere Bečkog meridijana i posjetu najstarijoj trigonometrijskoj točkiuHrvatskoj.Nastručnomobilasku sudjelovalo je 25 osoba. Taj bi stručni obilazak trebao biti početni korak u nas- tojanju da se završna točka izmjere Beč- kog meridijana (toranj varaždinske Dana5.3.2012.svečanojeobilježena 500. godišnjica rođenja Gerarda Merca- tora znanstvenim skupom i prigodnom izložbomnaGeodetskomfakultetu. Dana 18. 4. 2012. HKD je organizirao promociju monografije Topografske karte na području Hrvatske u Hrvatskom dr- žavnomarhivuuzsuradnjuDGU-a. Dana23.5.2012.usvečanojdvorani Sveučilišta u Zadru, održana je svečana KiG Br. 19, Vol. 12, 2013 KiG Br. 19, Vol. 12, 2013 KiG Br. 19, Vol. 12, 2013 183 183 183 FROM PROFESSIONAL SOCIETIES a two-hour lecture New Cartographic projection HTRS96/TM. In the name of CCS he greeted the participants of the Assembly at the ceremonial opening and the special item of agenda was an introductionofCCS. was invited to the Annual assembly of Association of surveyors of Osijek- Baranja County in Osijek, and in the name of CCS he greeted the assembly participantsattheceremonialopening. On October 10, 2012 we celebrated the 11th birthday of CCS. On that occa- sion, at the Faculty of Geodesy, Uni- versityofZagreb,Prof.Dr.MarijanHerak from the Geophysics Department of the Faculty of Science, University of Zagreb held a lecture Croatian Map of Earthquake Risk–howdoesitstartandwhatnext? On December 18, 2012 the presid- ent of CCS was invited to a festive re- ception organized by Croatian Cham- berofcharteredengineersofgeodesy. Especiallygoodnews: On November 15, 2012 Aldo Šošić defended the dissertation Kartografija Istre s posebnim poudarkom na razvoju kartografskega izražanja (Cartography of Istria with special emphasis on carto- graphic expression) at the Faculty of Engineering and Geodesy, University of Ljubljana. On October 16, 2012 the president, in the name of the Faculty of Geodesy, was awarded with the letter of thanks, for successful cooperation with Croa- tian Standards Institute on the occa- sion of World Standards Day. This cooperation was realized due to tem- perature, wind and snow maps, made thisyearbyMarinaRajaković,Miljenko LapaineandDraženTutić. Boris Blagonić finished doctoral studies at the Faculty of Geodesy, Uni- versity of Zagreb on July 7, 2012 by de- fending a doctoral thesis Katastar vodova u lokalnoj infrastrukturi prostornih podataka (Utility Cadastre in a Local SpatialDataInfrastructure). days conference in Zagreb, at the In- ternational Hotel. KiG No. 19, Vol. 12, 2013 11. godišnja skupština Hrvatskoga kartografskog društva The first two days was the 2nd Regional INSPIRE Training with participating only with an invita- tion. On September 27, there was the First Regional INSPIRATION Forum and 4th Croatian NSDI and INSPIRE Day, on September 28 the 8th Cartography and Geoinformation Conference and on September 29 there was a tour to Hr- vatskoZagorje. On October 19, 2012 the president of CCS was invited to the 5th symposi- um of Croatian chartered engineers of geodesy, in Opatija, where he held a keynote speech in the name of CCS at theceremonialopening. Mario Mađer, on November 5, 2012 at the Faculty of Geodesy, University of Zagreb defended a doctoral thesis Mod- el povezivanja katastra sa srodnim upis- nicima (A model for Linking Cadaster withRelatedRegisters). On November 14, 2012 CCS marked the World GIS Day organizing the con- ference Cartography and Eurocodes held in the Hall of the Faculty of Architec- ture, Civil Engineering and Geodesy, in Zagreb, Kačićeva 26. Co-organizers we- re the Faculty of Geodesy, University of Zagreb, Meteorological and Hydrolo- gical Service, Croatian Meteorological Society and Croatian Standards Insti- tute. Five lectures were held after ope- ning words and musical number on a clarinetbyJosipMimica. Prof. Dr. Mithad Kozličić was elec- ted a member of Academy of Science and Arts of Bosnia and Herzegovina (a member outside office – citizen of Bos- niaandHerzegovinawholivesabroad). SDI Days included these events as well: • Announcement of Competition of ideasofINSPIRATIONproject • Visiting an exhibition Zagreb’s special data infrastructure in Za- grebForum At the end of his report the presid- ent of CCS thanked for successful co- operation the Secretary Martina Tri- plat Horvat, MEng, web-page editor Marina Rajaković, MEng, Irena Iveko- vić, who has been an accountant for many years (Irvi Ltd), the main editor of the journal Cartography and Geoin- formation Assist. Prof. Dr. Dražen Tutić and Technical editor Ana Kuveždić Di- vjak,MEng. • Visiting the Collection of maps and atlases in National and University LibraryinZagreb On November 15, 2012 the presid- ent was invited to participate at the scientific convention Geography in Sci- ence and Education, which took place at the Antunović Hotel in Zagreb where he greeted participants in the name of CCSattheceremonialopening. • GeneralassemblyofAGISEE • GeneralassemblyofAGISEE At SDI Days there were 200 parti- cipantsfrom15countries. On September 26, 2012 CCS ap- peared at the Faculty of Geodesy Day where it got a letter of thanks for suc- cessful cooperation with the Faculty of Geodesy,theUniversityofZagreb. OnNovember30,2012thepresident was invited to participate at the Annual assemblyoftheAssociationofsurveyors of Istarska County in Novigrad. He held Total income in 2012 was 187.858,57 HRK and expenses 227.072,31 HRK which means that in 2012 we operated with a loss of 39.213,74 HRK. Account On October 5, 2012 the president KiG No. 19, Vol. 12, 2013 184 IZ DRUŠTAVA katedrale) obilježi spomenikom ili ba- remspomenpločom. Dogovori i pregovori s Državnom geodetskom upravom (DGU) o održa- vanju Dana infrastrukture prostornih podataka (Dani IPP-a / SDI Days) počeli su 15. 12. 2011. s idejom o zajedničkom održavanju 4. NIPP i INSPIRE dana i 8. savjetovanja kartografija i geoinforma- cije. Pregovori su trajali do 14. 5. 2012. kada je predsjednik HKD-a pozvan na sastanak s ravnateljem DGU-a, dr. sc. Dankom Markovinovićem. Na tom je sastanku dogovoreno da će DGU i HKD zajednički organizirati konferenciju u rujnu 2012. U Zagrebu u hotelu Inter- national održani su od 25. do 29. rujna 2012. Dani infrastrukture prostornih podataka(SDIDays):prvadvadana2nd Regional INSPIRE Training sa sudjelo- vanjem samo po pozivu, zatim 27. 9. First Regional INSPIRATION Forum i 4th Croatian NSDI and INSPIRE Day, 28. 9. bila je 8th Cartography and Geoin- formation Conference i 29. 9. stručni obilazakHrvatskimzagorjem. društva. Tom je prigodom na Geodet- skom fakultetu Sveučilišta u Zagrebu održano predavanje prof. dr. sc. Mari- jana Heraka s Geofizičkog odsjeka PMF- a Sveučilišta u Zagrebu pod naslovom Hrvatska karta potresne opasnosti – kako nastajeikakodalje? i Hrvatski zavod za norme. Nakon poz- dravnih riječi i glazbenih točaka u iz- vođenju Josipa Mimice na klarinetu održanojepetpredavanja. Dani IPP-a uključili su i ova događa- nja: • Najavu natječaja Nadmetanje ideja projektaINSPIRATION Dana 15. 11. 2012. predsjednik je po pozivu sudjelovao na znanstvenom skupu Geografija u sustavu znanosti i od- gojno-obrazovnom sustavu, održanom u hotelu Antunović u Zagrebu i u ime HKD-a pozdravio učesnike na sveča- nomotvaranju. • Posjet izložbi Zagrebačka infrastru- ktura prostornih podataka u Zagreb Forumu Dana 16. 10. 2012. predsjednik je primio u ime Geodetskog fakulteta Za- hvalnicu za uspješnu suradnju s Hrvat- skim zavodom za norme u povodu Svjetskoga dana norma. KiG Br. 19, Vol. 12, 2013 • GeneralassemblyofAGISEE Ta je suradnja ostvarena na kartama temperatura, vjetra i snijega što su ih ove godine iz- radili Marina Rajaković, Miljenko La- paineiDraženTutić. • Posjet Zbirci karata i atlasa Nacio- nalne i sveučilišne knjižnice u Za- grebu Dana 30. 11. 2012. predsjednik je po pozivu sudjelovao na Godišnjoj skup- štini Udruge geodeta Istarske županije u Novigradu. Održao je dvosatno pre- davanje Nova kartografska projekcija HTRS96/TM.UimeHKD-apozdravioje učesnike skupštine na svečanom otva- ranju, a u posebnoj točki dnevnoga re- dabilojepredstavljanjeHKD-a. • GeneralnuskupštinuAGISEE-ja. Na Danima IPP-a bilo je 200 učesni- kaiz15zemalja. Dana 19. 10. 2012. predsjednik HKD- a bio je pozvan na 5. simpozij ovlašte- nih inženjera geodezije, u Opatiju, gdje je u ime HKD-a na svečanom otvaranju održaopozdravnigovor. Dana26.9.2012.HKDjenastupiona Danu Geodetskog fakulteta, tom prigo- dom HKD je dobio Zahvalnicu za us- pješnu suradnju s Geodetskim fakul- tetomSveučilištauZagrebu. Dana 18. 12. 2012. predsjednik HKD- a bio je pozvan na svečani domjenak u organizacijiHKOIG-a. Dana 5. 10. 2012. predsjednik je po pozivu sudjelovao na Godišnjoj skup- štiniUdrugegeodetaOsječko-baranjske županije u Osijeku, i u ime HKD-a poz- dravioučesnikeskupštinenasvečanom otvaranju. Dana 14. 11. 2012. HKD je obilježio Svjetski dan GIS-a znanstveno-struč- nim skupom Kartografija i eurokodovi održanim u Vijećnici AGG fakulteta, u Zagrebu, Kačićeva 26. Suorganizatori su bili Geodetski fakultet Sveučilišta u Zagrebu, Državni hidrometeorološki zavod, Hrvatsko meteorološko društvo EuropeanGeodesyandGeoinformation DayonMarch22,2013 5. To intensify the sending of the jou- rnal to more important libraries and anexchangewithrelatedjournals. EuropeanGeodesyandGeoinformation DayonMarch22,2013 3. To broaden the circle of people and institutions which financially back uptheworkoftheSociety Marking of European Geodesy and Geoinformation Day was arranged at the meeting held on March 22, 2013 at SGA. The organizers were SGA, the Fac- ulty of Geodesy and Croatian Chamber of chartered engineers of geodesy. CCS participated with two pages in a pub- licationwhichwillbeissued. 6. To start publishing the News every threemonths 4. To attract more young people and students 4. To attract more young people and students 7. To design an award for young carto- graphers, nominate honorary mem- bersandplanlettersofthanks. TheCCSmembershipmeetings l TheCCSmembershipmeetings 8. To establish a co-operation with the journalists and to inform about the Society’sactivitiesontime. 1. To organize expert lectures once a month 2. Toorganizeanexcursiononceayear 2. Toorganizeanexcursiononceayear CartographyandArtonApril15,2013 On the proposal and the initiative of the CCS president, the Faculty of Ge- odesy and the Academy of Fine Arts, University of Zagreb together with the Department of Geography, University of Zadar are preparing a symposium under a name Cartography and Art. Sym- posiumwillbeheldonApril15,2013on a birthday of famous Leonardo da Vinci who was a versatile scientist and artist, andlessknownasacartographer.Atthe same time in the Gallery on the first 9. To expand the international co-op- eration with the aim of promoting the Society in international com- munity(ICA,EUROGI,AGISEEetc.) Promotion 1. To constantly modernize web-pages oftheSociety The Assembly has accepted the fol- lowingworkplanfor2013: The Assembly has accepted the fol- lowingworkplanfor2013: 2. To refresh an e-list of members and periodicallysendnotifications Membershipandfinance 3. To involve secondary schools of geo- desyandtherelatedsocieties 1. Toorderlyconductthemembership recordsandmembershipfees 4. To publish the journal Cartography andGeoinformationregularly 2. To take care of the contracts with institutionswhichhelptheSociety Legislation 1. To take part in activities at making and passing the legislation for the field of cartography and geoinfor- mation RegulationsonCCSAwards incartography 3. ToparticipateatSciencefestivals 4. To organize lectures and workshops forallages 5. To involve secondary schools of geo- desy Posebnodobrevijesti: AldoSošićobranioje15.11.2012.na Fakultetu za građevinarstvo i geodeziju Sveučilišta u Ljubljani disertaciju Kar- tografija Istre s posebnim poudarkom na Dana 10. 10. 2012. proslavili smo 11. rođendan Hrvatskoga kartografskog KiG Br. 19, Vol. 12, 2013 185 185 FROM PROFESSIONAL SOCIETIES status on December 31, 2012 was 206.779,0 HRK. Therefore CCS was not economicallysuccessfulintheprevious year. The main reason was a drastic re- ductioninhelpgivenbySGAforthelast 17 years and a lack of financial help by the Ministry of Science, Education and Sport which has not been asked for since the deadlines to apply ended due to prolonged negotiations with SGA. Financial state of CCS is not alarming sincewesuccessfullyoperatedprevious years but it is obvious that we cannot expectfurtherhelpfromSGAexceptin co-financing the journal based on the contractfromDecember2011. RegulationsonCCSAwards At the last presidency session of CCS the proposal of the Regulations on CCS Awards was discussed. Accord- ingly, the award would be given to a personundertheageof35whomadea noticeable personal scientific or expert progress or accomplished contribution tothecartography.Afterthediscussion it was concluded that CCS could assign 3 awards: to a student, to a senior per- son or an institution and for the best map. That proposition will be further developed and proposed for its adop- tionatnextGeneralAssembly. incartography 3 ToparticipateatSciencefestivals Projects Projects 1. To propose scientific-technical pro- jects which will improve Croatian cartography (e.g. The future of to- pographiccartography) 1. To propose scientific-technical pro- jects which will improve Croatian cartography (e.g. The future of to- pographiccartography) 2. To propose an instalment of SGA’s co-financing the making of national reports for the International Carto- graphicAssociation. Furtherco-operationwithSGA At the meeting held on November 7, 2012 the president gave the Director ofSGAamemoinwhichSGAisaskedto help financially by paying the annual membership in ICA, the participation at conventions in Vienna, Helsinki and Dresden and the organization of the 9th conference of Cartography and Geoinformation. The response has not arrived until today therefore we can conclude that SGA does not intend to helpanyplannedactivityofCCS. 3. To suggest reviewing and establish- ing the quality of cartographic and geoinformation products to the ap- propriateinstitutions. Dr. Dubravka Mlinarić, the Head of Supervisory Board, reported that CCS acted in accordance with the Society Act, Society Statute and it worked ma- terially and financially valid. The Su- pervisory Board report was accepted unanimously. Članstvoifinanciranje Članstvoifinanciranje SusretičlanovaDruštva SusretičlanovaDruštva SusretičlanovaDruštva 3. Predložiti odgovarajućim instituci- jama recenziranje i utvrđivanje kvalitete njihovih kartografskih i geoinformacijskihproizvoda. 1. U pravilu jednom mjesečno organi- ziratistručnapredavanja. 2. Jednomnagodinuorganiziratiizlet. Zakonodavstvo Promidžba Projekti 1. Voditi uredno evidenciju članstva i plaćanjačlanarine 1. Predložiti znanstveno-stručne pro- jekte kojima bi se unaprijedila hr- vatska kartografija (npr. Budućnost topografskekartografije). 2. Voditi brigu o ugovorima s instituci- jamakojepomažuradDruštva 3. Proširiti krug osoba i institucija koje financijskipodupiruradDruštva 2. Državnoj geodetskoj upravi predlo- žiti nastavak sufinanciranja izrade nacionalnih izvještaja za Međuna- rodnokartografskodruštvo. 4. Privućivišemladihosobaistudenata Conferences 1. To continue with organizing con- ferences on cartography, geoinfor- mationandtheirinfrastructure. Education 1. To actively participate in promoting knowledge and skills in cartography ateverylevelofeducation 2. The preparations can start a year in advance. 2. To organize a competition for the best children’s drawing on a theme 3. To persist on the international level andco-operation. KiG No. 19, Vol. 12, 2013 186 IZ DRUŠTAVA IZ DRUŠTAVA Izvještaj Nadzornog odbora je prihva- ćenjednoglasno. 2. Organizirati natječaj za najbolje dje- čjecrteženakartografsketeme. razvoju kartografskega izražanja (Karto- grafija Istre s posebnim naglaskom na kartografskoizražavanje). Boris Blagonić završio je doktorski studij na Geodetskom fakultetu Sve- učilišta u Zagrebu 13. 7. 2012. obranom doktorskog rada pod naslovom Katastar vodova u lokalnoj infrastrukturi prostornih podataka. Mario Mađer obranio je 5. studeno- ga 2012. na Geodetskom fakultetu Sve- učilištauZagrebudoktorskudisertaciju Model povezivanja katastra sa srodnim upisnicima. Prof.dr.sc.MithadKozličić,izabran je za člana ANU BiH ("član van radnog sastava" – državljanin BiH koji živi u inostranstvu/inozemstvu). Na kraju svojeg izvještaja predsjed- nik HKD-a zahvalio je na uspješnoj su- radnji tajnici Martini Triplat Horvat, dipl. ing., urednici web-stranica Marini Rajaković, dipl. ing., gospođi Ireni Ive- ković koja već više godina vodi računo- vodstvo (Irvi d.o.o.), glavnom uredniku časopisa Kartografija i geoinformacije doc. dr. sc. Draženu Tutiću i tehničkoj uredniciAniKuveždićDivjak,dipl.ing. 2. Organizirati natječaj za najbolje dje- čjecrteženakartografsketeme. 2. Organizirati natječaj za najbolje dje- čjecrteženakartografsketeme. 2. Organizirati natječaj za najbolje dje- čjecrteženakartografsketeme. Izvještaj Nadzornog odbora je prihva- ćenjednoglasno. Izvještaj Nadzornog odbora je prihva- ćenjednoglasno. 3. SudjelovatinaFestivalimaznanosti. Skupština je prihvatila sljedeći plan radaza2013.godinu: Skupština je prihvatila sljedeći plan radaza2013.godinu: 4. Organizirati predavanja i radionice zasveuzraste. Ukupni prihodi u 2012. godini iz- nosili su 187.858,57 kn, a rashodi 227.072,31 kn. Znači da smo u 2012. poslovali s gubitkom od 39.213,74 kn. Stanje računa na dan 31. prosinca 2012. bilo je 206.779,0 kn. Odatle slijedi da HKD nije uspješno poslovao u prethod- noj godini. Osnovni razlog tome je drastično smanjenje pomoći koju je u proteklih 17 godina pružao DGU i u iz- ostanku financijske pomoći MZOŠ-a koja nije ni zatražena jer su zbog oteza- nja pregovora s DGU-om prošli rokovi za prijavu na natječaj. Financijsko sta- nje HKD-a nije alarmantno jer smo prethodnih godina uspješno poslovali, ali je očito da se do daljnjega od DGU vi- še ne može očekivati pomoć, osim sufi- nanciranjačasopisanatemeljuugovora izprosinca2011.godine. 5. Uključitisrednjegeodetskeškole. 5. Uključitisrednjegeodetskeškole. Promidžba 1. Uključiti se u aktivnosti pri izradi i donošenju zakonske regulative za područje kartografije i geoinforma- cija. 1. Stalno osuvremenjivati web-strani- ceDruštva. Dr. sc. Dubravka Mlinarić, pred- sjednicaNadzornogodborapodnijelaje izvještaj u kojem navodi da je Društvo djelovalo u skladu sa Zakonom o udru- gama, Statutom Društva te da je mate- rijalno i financijski poslovalo uredno. 2. Osuvremeniti e-popis članova i po- vremenoslatiobavijesti. Obrazovanje Obrazovanje 3. Uključiti srednje geodetske škole i srodnadruštva. 1. Aktivno sudjelovati na promicanju kartografskih znanja i vještina na svimrazinamaobrazovanja. 4. Redovito izdavati časopis Kartogra- fijaigeoinformacije KiG Br. 19, Vol. 12, 2013 187 FROM PROFESSIONAL SOCIETIES floor of the Faculty of Geodesy there will be an exhibition of Vlado Martek called kARTe mARTek. In co-operation with the Academy of Fine Arts, Univer- sity of Zagreb there will be an art work- shop for primary school children as well. Pupils and students, inspired by maps, will paint imaginary parts on Earth,maybeevenknownones,butina waytheywanttoseethem. floor of the Faculty of Geodesy there will be an exhibition of Vlado Martek called kARTe mARTek. In co-operation with the Academy of Fine Arts, Univer- sity of Zagreb there will be an art work- shop for primary school children as well. Pupils and students, inspired by maps, will paint imaginary parts on Earth,maybeevenknownones,butina waytheywanttoseethem. It is planned in Zadar in co-operati- on with the Department for Geography, UniversityofZadar. All preparatory activities have been doneontime.32institutionsresponded with the overall of 171 works. The jury Assist. Prof. Dr. I. Kljajić, Assoc. Prof. Ines Krasić, Prof. Dr. Miljenko Lapaine, Igor Čabraja and Ana Kuveždić Divjak, chose six drawings for the exhibition in Dre- sden. Furthermore, CCS will especially commend another 10 drawings which enteredtheshortlist. The proposed work plan for 2013 wasunanimouslyadopted. Theassemblyunanimouslyadopted a proposition to award Assist. Prof. Dr. IvkaKljajićwiththeletterofthanksfrom CCS for perennial successful organizati- on and management of best children’s drawing competition that represented Croatia at international exhibitions for BarbaraPetchenikaward. The public has been called to parti- cipateattheinternationalexhibitionof maps by an e-mail address and a web- pageofCCS.Detailedpropositionshave been given together with the invitati- on. The deadline is March 23, 2013. The finalselectionofcartographicworksfor this international exhibition will be made by the Presidency of CCS on Mar- ch27,2013. ICC2013,Dresden,August25to30,2013 There have been ten abstracts/arti- cles sent from Croatia. Reviewers were, amongst others D. Tutić and M. Lapaine. The Committee for Map Projections of ICAwillholdasessiononAugust24,2013 from 5 pm to 7 pm in the lecture hall at TU Dresden, Bergstraße 64, 01069 Dre- sden.Everyoneisinvited! Promidžba Theassemblyunanimouslyadopted a proposition to award Ana Kuveždić Divjak, MEng with the letter of thanks fromCCSforperennialeffortandagreat dedication in making and promoting a visual identity of CCS, especially for tec- hnical editing of the journal Cartograp- hyandGeoinformation. A. Kuveždić Divjak and I. Kljajić are creditable for the selection and exhibi- tion of children’s drawings from Croatia. WorldGISDayand9thCartography and Geoinformation Conference, No- vember20to22,2013 MartinaTriplatHorvat MartinaTriplatHorvat Kartografijaiumjetnost15.4.2013. Na prijedlog i inicijativu predsjedni- ka Hrvatsko kartografsko društvo, Ge- odetski fakultet i Akademija likovnih umjetnosti Sveučilišta u Zagrebu zajed- nosOdjelomzageografijuSveučilištau Zadru pripremaju simpozij pod naslo- vom Kartografija i umjetnost. Simpozij će se održati 15. travnja 2013. na rođendan slavnoga Leonarda da Vincija koji je bio svestraniznanstvenikiumjetnik,amalo je poznato da je bio i kartograf. Istodob- nouGalerijina1.katubitćepostavljena izložba slika Vlade Marteka pod naslo- vom kARTe mARTek. Istodobno, a u su- radnji s Akademijom likovnih umjetno- stiSveučilištauZagrebu(ALU)održatće se likovna radionica s učenicima osnov- ne škole. Učenici i studenti inspirirani kartamaoslikatćezamišljenekrajevena Zemlji, a možda i one poznate, ali onako kakobiihoniželjelividjeti. 1. Nastaviti s organiziranjem konfe- rencija o kartografiji, geoinformaci- jamainjihovojinfrastrukturi. 2. Pripreme početi godinu dana una- prijed. SvjetskidanGIS-ai9.savjetovanjeKarto- grafijaigeoinformacije20–22.11.2013. 3. Ustrajati na međunarodnoj razini i suradnji. PlaniranojeuZadruusuorganizaciji s Geografskim odjelom Sveučilišta u Za- dru. Daneuropskegeodezijeigeoinformacija 22.3.2013. 8. Uspostaviti suradnju s novinarima i pravodobno informirati javnost o aktivnostima Društva preko svih medija. Nasastankuodržanom22.1.2013.u DGU-u dogovoreno je obilježavanje Dana europske geodezije i geoinforma- cija. Organizatori su DGU, Geodetski fa- kultet i HKOIG. HKD sudjeluje s dvije stranice u publikaciji koja će se izdati timpovodom. 9. Proširiti međunarodnu suradnju s ciljem promoviranja Društva u me- đunarodnoj zajednici (ICA, EUROGI, AGISEEidr.). Poziv za sudjelovanje na međuna- rodnojizložbikarataupućenjejavnosti putem e-adresa i web-stranica HKD-a. Uz poziv su dane i detaljne propozicije. Rok za prijavu je 23. 3. 2013. Konačni iz- bor kartografskih djela za tu međuna- rodnu izložbu obavit će Predsjedništvo HKD-a27.3.2013. Savjetovanja PravilnikonagradamaHKD-a Na posljednjoj sjednici Predsjed- ništva HKD-a razmatrao se prijedlog Pravilnika o nagradama HKD-a. Prema tom prijedlogu nagrada bi se dodijelila osobi mlađoj od 35 godina koja je os- tvarila zamjetan osobni znanstveni ili stručni napredak ili postigla zapaženi doprinos kartografiji. Nakon rasprave zaključeno je da HKD može dodijeliti 3 nagrade: jednu studentu, drugu "stari- joj"osobiiliinstitucijiitrećunagraduza najbolju kartu. Taj će se prijedlog dalje razraditi i na sljedećoj skupštini pred- ložitizausvajanje. Predloženi plan rada za 2013. godi- nuusvojenjejednoglasno. Skupština je jednoglasno prihvatila prijedlog da se doc. dr. sc. Ivki Kljajić dodijeli Zahvalnica HKD-a za višego- dišnje vrlo uspješno organiziranje i vo- đenje natječaja za najbolje dječje crteže koji su predstavljali Hrvatsku na među- narodnim izložbama za nagradu Bar- barePetchenik. Skupština je jednoglasno prihvatila prijedlog da se Ani Kuveždić Divjak, dipl. ing. geod. dodijeli Zahvalnica HKD- azavišegodišnjitrudivelikupredanost na izradi i promicanju vizualnog iden- titeta Hrvatskoga kartografskog druš- tva, a posebno za tehničko uređivanje časopisaKartografijaigeoinformacije. IZ DRUŠTAVA KiG Br. 19, Vol. 12, 2013 IZ DRUŠTAVA 5. Pojačati slanje časopisa u važnije knjižnice i razmjenu sa srodnim ča- sopisima. 5. Pojačati slanje časopisa u važnije knjižnice i razmjenu sa srodnim ča- sopisima. Helsinkiju i Dresdenu te organiziranje 9. savjetovanja Kartografija i geoinforma- cije. Do danas odgovor nije stigao pa mo- žemo zaključiti da DGU ne namjerava pomagatiplaniraneaktivnostiHKD-a. Za izbor i izložbu dječjih crteža iz Hrvatske posebno su zaslužne A. Kuve- ždić Divjak i I. Kljajić. Sve pripremne aktivnosti obavljene su na vrijeme. Na poziv su se odazvale 32 ustanove s ukupno 171 radom. Žiri u sastavu doc. dr.sc.I.Kljajić,prof.InesKrasić,prof.dr. sc. Miljenko Lapaine, Igor Čabraja i Ana Kuveždić Divjak izabrao je za izložbu u Dresdenu šest dječjih radova. Osim to- ga, HKD će posebno pohvaliti još 10 dječjihcrtežakojisuušliuužiizbor. 6. Pokrenuti izdavanje Novosti svaka trimjeseca. 6. Pokrenuti izdavanje Novosti svaka trimjeseca. 7. Osmisliti nagradu mladim karto- grafima, nominirati počasne člano- ve,planiratizahvalnice. Daneuropskegeodezijeigeoinformacija 22.3.2013. ICC2013,Dresden,25–30.8.2013. Iz Hrvatske je poslano 10 sažeta- ka/članaka. Recenzenti su, između os- talih, bili D. Tutić i M. Lapaine. Povje- renstvo za kartografske projekcije ICA-e održat će u Dresdenu sjednicu 24. 8. 2013. od 17 do 19 sati u predavaonici na TU Dresden, Bergstraße 64, 01069 Dre- sden.Svisudobrodošli! DaljnjasuradnjasDržavnomgeodetskom upravom DaljnjasuradnjasDržavnomgeodetskom upravom Na sastanku održanom 7. 11. 2012. predsjednik je predao ravnatelju DGU-a dopis u kojem moli da DGU financijski pomogne plaćanje godišnje članarine u ICA-i, sudjelovanje na skupovima u Beču, MartinaTriplatHorvat KiG Br. 19, Vol. 12, 2013 189
https://openalex.org/W2012324443	https://archimer.ifremer.fr/doc/00372/48356/48545.pdf	English	null	Emulsifying Activity and Stability of a Non-Toxic Bioemulsifier Synthesized by Microbacterium sp. MC3B-10	International journal of molecular sciences	2,013	cc-by	6,550	Printed Edition of the Special Issue Published in IJMS Biofilms: Extracellular Bastions of Bacteria Edited by Alan W. Decho Alan W. Decho (Ed.) 1. Edition 2014 MDPI • Basel • Beijing Biofilms: Extracellular Bastions of Bacteria This book is a reprint of the special issue that appeared in the online open access journal International Journal of Molecular Sciences (ISSN 1422-0067) in 2013 (available at: http://www.mdpi.com/journal/ijms/special_issues/extracellular-bastions-of-bacteria). Guest Editor Alan W. Decho Department of Environmental Health Sciences Arnold School of Public Health, University of South Carolina Columbia, SC. 29208 USA Guest Editor Alan W. Decho Department of Environmental Health Sciences Arnold School of Public Health, University of South Carolina Columbia, SC. 29208 USA Editorial Office MDPI AG Klybeckstrasse 64 Basel, Switzerland 1. Edition 2014 MDPI • Basel • Beijing Juan Carlos Camacho-Chab, Jean Guézennec, Manuel Jesús Chan-Bacab, Elvira Ríos-Leal, Corinne Sinquin, Raquel Muñiz-Salazar, Susana del C. De la Rosa-García, Manuela Reyes-Estebanez and Benjamín Otto Ortega-Morales Abstract: A previously reported bacterial bioemulsifier, here termed microbactan, was further analyzed to characterize its lipid component, molecular weight, ionic character and toxicity, along with its bioemulsifying potential for hydrophobic substrates at a range of temperatures, salinities and pH values. Analyses showed that microbactan is a high molecular weight (700 kDa), non-ionic molecule. Gas chromatography of the lipid fraction revealed the presence of palmitic, stearic, and oleic acids; thus microbactan may be considered a glycolipoprotein. Microbactan emulsified aromatic hydrocarbons and oils to various extents; the highest emulsification index was recorded against motor oil (96%). The stability of the microbactan-motor oil emulsion model reached its highest level (94%) at 50 °C, pH 10 and 3.5% NaCl content. It was not toxic to Artemia salina nauplii. Microbactan is, therefore, a non-toxic and non-ionic bioemulsifier of high molecular weight with affinity for a range of oily substrates. Comparative phylogenetic assessment of the 16S rDNA gene of Microbacterium sp. MC3B-10 with genes derived from other marine Microbacterium species suggested that this genus is well represented in coastal zones. The chemical nature and stability of the bioemulsifier suggest its potential application in bioremediation of marine environments and in cosmetics. Reprinted from Int. J. Mol. Sci. Cite as: Camacho-Chab, J.C.; Guézennec, J.; Chan-Bacab, M.J.; Ríos-Leal, E.; Sinquin, C.; Muñiz-Salazar, R.; del C. De la Rosa-García, S.; Reyes-Estebanez, M.; Ortega-Morales, B.O. Emulsifying Activity and Stability of a Non-Toxic Bioemulsifier Synthesized by Microbacterium sp. MC3B-10. Int. J. Mol. Sci. 2013, 14, 18959-18972. Reprinted from Int. J. Mol. Sci. Cite as: Camacho-Chab, J.C.; Guézennec, J.; Chan-Bacab, M.J.; Ríos-Leal, E.; Sinquin, C.; Muñiz-Salazar, R.; del C. De la Rosa-García, S.; Reyes-Estebanez, M.; Ortega-Morales, B.O. Emulsifying Activity and Stability of a Non-Toxic Bioemulsifier Synthesized by Microbacterium sp. MC3B-10. Int. J. Mol. Sci. 2013, 14, 18959-18972. ISBN 978-3-906980-95-9 © 2014 by the authors; licensee MDPI, Basel, Switzerland. All articles in this volume are Open Access distributed under the Creative Commons Attribution 3.0 license (http://creativecommons.org/licenses/by/3.0/), which allows users to download, copy and build upon published articles even for commercial purposes, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. However, the dissemination and distribution of copies of this book as a whole is restricted to MDPI, Basel, Switzerland. © 2014 by the authors; licensee MDPI, Basel, Switzerland. All articles in this volume are Open Access distributed under the Creative Commons Attribution 3.0 license (http://creativecommons.org/licenses/by/3.0/), which allows users to download, copy and build upon published articles even for commercial purposes, as long as the author and publisher are properly credited, which ensures maximum dissemination and a wider impact of our publications. However, the dissemination and distribution of copies of this book as a whole is restricted to MDPI, Basel, Switzerland. 217 Reprinted from Int. J. Mol. Sci. Cite as: Camacho-Chab, J.C.; Guézennec, J.; Chan-Bacab, M.J.; Ríos-Leal, E.; Sinquin, C.; Muñiz-Salazar, R.; del C. De la Rosa-García, S.; Reyes-Estebanez, M.; Ortega-Morales, B.O. Emulsifying Activity and Stability of a Non-Toxic Bioemulsifier Synthesized by Microbacterium sp. MC3B-10. Int. J. Mol. Sci. 2013, 14, 18959-18972. 3. Biotechnological Aspects Emulsifying Activity and Stability of a Non-Toxic Bioemulsifier Synthesized by Microbacterium sp. MC3B-10 Juan Carlos Camacho-Chab, Jean Guézennec, Manuel Jesús Chan-Bacab, Elvira Río Corinne Sinquin, Raquel Muñiz-Salazar, Susana del C. De la Rosa-García, Manuela Reyes-Estebanez and Benjamín Otto Ortega-Morales 1. Introduction Microorganisms produce a wide variety of high and low molecular weight biosurfactants. These active molecules include proteins, polysaccharides, lipopeptides, glycolipids, flavolipids, phospholipids and lipopolysaccharides [1]. High molecular weight biosurfactants, usually referred to as bioemulsifiers, can form and stabilize oil-in-water or water-in-oil emulsions, work at low concentrations and exhibit considerable substrate specificity [2]. The chemical diversity and functional properties of these compounds result in a broad spectrum of potential applications in sectors as diverse as agriculture, cosmetics, environmental, food, leather, paper, pharmaceutical and textile industries [2,3]. Biologically derived surface-active molecules have advantages over their synthetic counterparts, 218 including biodegradability, lower toxicity, novel structural chemistry and high stability at extreme temperature, salinity and pH [2,4]. Some bioemulsifiers, e.g., emulsan [5], have been extensively characterized and reached commercial applications, chiefly in the bioremediation sector. Bioprospection of various habitats, in particular marine environments [6,7], has yielded many novel biomolecules produced by microorganisms. An earlier report showed that the extracellular biopolymer produced by Microbacterium sp. MC3B-10 emulsified hexane, decane and hexadecane at higher efficiencies than commercial surfactants Triton® X-100 and Tween® 80 [8]. This extracellular biopolymer will be referred to as microbactan after its bacterial producer. The goal of the present study was to characterize in greater detail this extracellular biopolymer and determine the emulsifying stability as a function of key variables. 2.1. Characterization of Microbactan Microbactan was produced by fermentation in shaker flasks; no attempt was made at this time to optimize its production using a bioreactor, where more strict control of variables can be achieved. The reproducibility of production conditions employed in this study was confirmed by the yield and the primary chemical profile of the biopolymer, which was comparable to the previous report [8]. Our previous work showed that microbactan was dominated by carbohydrates and proteins, and was preliminarily considered a glycoprotein. The present study also revealed the presence of lipids, shown by the peak at 2927 cmí1 in the FT-IR analysis, which represents the asymmetric stretch (C–H) of –CH2 groups combined with that of –CH3 groups in lipids [9]. Quantitative assessment by absorbance showed that the lipid signature detected by FT-IR corresponded to a lipid content of 8% ± 0.5%. The gas chromatograph analysis of the lipid fraction revealed the presence of palmitic (C-16), stearic (C-18) and oleic (C-18:1) acids. Other fatty acids, found at lower levels, were myristoleic (C-14:1) and linoleic (C-18:2). The lipid content and the previously reported chemical composition of ~90% carbohydrates and proteins [8] represented almost 100% of the total mass of the microbactan on a dry weight basis, suggesting that this biopolymer is a class of glycolipoprotein. The fatty acids detected (palmitic, stearic and oleic) have been reported in bioemulsifiers produced by Yarrowia lipolytica [10] and Penicillium sp. [11]. However, the presence of other substituents of extracellular biopolymers such as pyruvil, succinyl and sulfates, which often occur at minor levels [4,9], was not investigated. High performance size exclusion chromatography showed that microbactan had a molecular weight (MW) approaching 700 kDa and a polydispersity index (Mw/number average molecular weight) of 1.3; a value close to 1 denotes a polymer with homogeneous monomer grouping. These results indicate that microbactan is a homogeneous high molecular weight glycolipoprotein, a finding that is consistent with previous reports showing that, in general, bioemulsifiers are high molecular weight polymers [1,12]. The commercial bioemulsifier alasan® has a molecular weight of approximately 1000 kDa [13]. Most bacterial bioemulsifiers are polymers of either carbohydrates or proteins, or even 219 glycoproteins [1,6]. Thavasi and colleagues report marine-derived glycolipoprotein emulsifiers from Corynebacterium kutcheri and Bacillus megaterium [14,15]. The putative glycolipoprotein nature of microbactan is, therefore, to some extent unusual among marine bioemulsifiers. Microbactan was shown by the modified double diffusion test to be a non-ionic emulsifier. 2.1. Characterization of Microbactan The ionic character of polymers is one of the features that contribute strongly to the functional properties of emulsifiers; this property is often reported for synthetic emulsifiers or novel emulsifying formulations used for a range of applications. Despite the importance of this feature, it is rarely reported in newly discovered bioemulsifiers. The non-ionic nature of emulsifiers greatly contributes to emulsion stability, generating a number of short-range repulsive forces, such as steric, hydration, and thermal fluctuation interactions, which prevent the droplets from getting too close together [16]. a Results are expressed as percentages of the total height occupied by the emulsion; values are means of at least three determinations. a Results are expressed as percentages of the total height occupied by the emulsion; values are means of at le three determinations lts are expressed as percentages of the total height occupied by the emulsion; values are means of at least 2.2. Emulsifying Potential Tables 1 and 2 show the emulsifying activity of microbactan and control commercial emulsifiers tested against different substrates at different times. Statistical analysis showed significant differences between microbactan and these controls (F = 84.24; p < 0.05). Synthetic surfactants Triton® X-100 and Tween® 80 were more efficient than microbactan and the commercial biopolymers, reaching emulsification efficiencies of 100% against oils (with the exception of motor oil), irrespective of time of incubation (24 and 96 h). Microbactan showed stronger emulsifying activities than gum arabic but was comparable to the emulsifying efficiency of xanthan gum, another bacterially-produced biopolymer. The emulsifying activity of microbactan was dependent on the type of substrate (F = 16.2; p < 0.05); this is consistent with other reports on substrate-specific hydrocarbon metabolism by marine bacteria [17,18]. Microbactan emulsions with the tested substrates were rather stable over time (Tables 1 and 2), varying only slightly between 24 and 96 h. Table 1. Emulsifying activity of microbactan, commercial synthetic surfactants and natural biopolymer emulsifiers on various hydrophobic substrates after 24 h of evaluation. Table 1. Emulsifying activity of microbactan, commercial synthetic surfactants and n biopolymer emulsifiers on various hydrophobic substrates after 24 h of evaluation. Hydrophobic substrate Microbactan a Synthetic surfactants a Biopolymers a Tween 80 Triton-X-100 Gum arabic Xanthan gum Benzene 76.9 ± 2.4 98.7 ± 0.6 54.3 ± 2.0 80.7 ± 3.0 64.7 ± 1.3 Xylene 81.5 ± 3.4 94.7 ± 6.1 98.2 ± 0.4 72.6 ± 3.4 76.8 ± 1.7 Crude oil 76.5 ± 2.1 82.9 ± 0.6 100 54.7 ± 3.7 95.0 ± 0.9 Motor oil 96.3 ± 0.1 54.9 ± 1.4 68.5 ± 6 95.2 ± 1.1 89.5 ± 3.3 Sunflower oil 84.4 ± 5.4 100 100 77.8 ± 1.5 62.4 ± 2.4 Corn oil 81.1 ± 1.6 100 100 0 90.6 ± 1.9 Olive oil 76.1 ± 0.3 100 100 96.6 ± 0.1 76.6 ± 5.4 Mineral oil 0 95.9 ± 5.6 100 60.3 ± 2.8 86.5 ± 3.8 a Results are expressed as percentages of the total height occupied by the emulsion; values are means of at least 220 Table 2. Emulsifying activity of microbactan, commercial synthetic surfactants and natural biopolymer emulsifiers on various hydrophobic substrates after 96 h of evaluation. 2.2. Emulsifying Potential Hydrophobic substrate Microbactan a Synthetic surfactants a Biopolymers a Tween 80 Triton-X-100 Gum arabic Xanthan gum Benzene 75.3 ± 3.8 93.2 ± 5 54.3 ± 2.1 78.9 ± 0 61.7 ± 1.4 Xylene 75.4 ± 0.5 94.1 ± 5 97.9 ± 0.5 63.7 ± 3 75 ± 1.6 Crude oil 72 ± 4.8 85.5 ± 5 100 ± 0 53.2 ± 2.4 93.8 ± 2.5 Motor oil 96.3 ± 0.1 53.5 ± 2.3 59 ± 4.2 92.8 ± 1.8 83.7 ± 3.2 Sunflower oil 83.8 ± 3.4 100 100 77.8 ± 2.1 62.4 ± 2.4 Corn oil 81.1 ± 1.6 100 100 0 91.5 ± 0.8 Olive oil 76.1 ± 0.5 100 100 96.6 ± 0.6 78.2 ± 2.6 Mineral oil 0 91.2 ± 2.3 100 ± 0 56.8 ± 2.3 82.9 ± 2 a Results are expressed as percentages of the total height occupied by the emulsion; values are means of at least three determinations. a Results are expressed as percentages of the total height occupied by the emulsion; values are means of at least three determinations. An established criterion for emulsion-stabilizing capacity is the ability of an emulsifier to maintain at least 50% of the original volume of the emulsion for 24 h [4,19]. Microbactan emulsions remained stable for several months, showing no sign of droplet coalescence after standing at room temperature (28 °C). This extended stability has previously been observed for glycoprotein bioemulsifiers produced by a marine Antarctobacter [20]. Emulsifying and surfactant activities are together responsible for important functional properties of bacterial exopolymers. Our analyses revealed that microbactan has emulsifying but not surfactant activity. Surface-active biomolecules are classified as surfactants when they lower the interfacial or surface tension and emulsifiers when they form stable emulsions [19,21]. Overall, our results indicate that microbactan is a true bioemulsifier. 2.3. Effect of Temperature, pH and NaCl on Emulsifying Activity The effect of these three key variables that influence emulsifying activity was assessed at 24, 48, 72 and 96 h. As expected, they influenced the emulsifying activity of microbactan, but their effects occurred at different levels. A time-dependent decrease of emulsifying activity was observed at 100 °C (F = 38.13; p < 0.005). This was not the case for activities at 50 °C and 5 °C. The highest level of activity was found at 50 °C (emulsifying activity of 95.7% ± 2.5%) irrespective of time of incubation (Figure 1). The loss of emulsifying activity at 100 °C can be explained by denaturation of the protein fraction of microbactan during heating [22], as seen with other microbial biosurfactants [23]. Similarly, the activity associated with the upper limit of the range of salinities tested (3.5%, 5%, and 10%) declined as a function of time to level off at 72 and 96 h. The highest levels of activity were observed at 3.5% (Figure 2), not dissimilar to the biosurfactant produced by Aeromonas spp., which maintained emulsifying activity up to 5% NaCl [24]. On the other hand, no significant effect on activity was observed as a function of time with the pH values tested (Figure 3). However, slightly higher levels of emulsifying activity were recorded at acid and alkaline pH values, suggesting the ionization of 221 functional groups that resulted in the activation of less surface-active species within the bioemulsifier matrix [25]. In comparison, biodispersan from Acinetobacter calcoaceticus A2 had an optimum functional pH value in the range of 9 to 12 for limestone-dispersing activity [26]. It has been shown that the emulsifying activity of certain polymers is modified at different extents when temperature and pH covariate, this is probably due to the synergistic influence of these factors on surface-active proteins, whose conformation and functional groups are influenced as a function of these factors. Figure 1. Emulsifying activity of microbactan on motor oil at different temperatures up to 96 h. 5 °C (Ɣ), 50 °C (Ŷ) and 100 °C (Ÿ). Values represent means ± SD (n = 3). Figure 2. Emulsifying activity of microemulsan on motor oil at various salinities up to 96 h. 3.5% (Ɣ), 5% (Ŷ) and 10% (Ÿ) NaCl concentration. Values represent means ± SD (n = 3). Figure 1. Emulsifying activity of microbactan on motor oil at different temperatures up to 96 h. 2.3. Effect of Temperature, pH and NaCl on Emulsifying Activity 5 °C (Ɣ), 50 °C (Ŷ) and 100 °C (Ÿ). Values represent means ± SD (n = 3). Figure 1. Emulsifying activity of microbactan on motor oil at different temperatures up to 96 h. 5 °C (Ɣ), 50 °C (Ŷ) and 100 °C (Ÿ). Values represent means ± SD (n = 3). Figure 1. Emulsifying activity of microbactan on motor oil at different temperatures up 96 h. 5 °C (Ɣ), 50 °C (Ŷ) and 100 °C (Ÿ). Values represent means ± SD (n = 3). Figure 2. Emulsifying activity of microemulsan on motor oil at various salinities up to 96 h. 3.5% (Ɣ), 5% (Ŷ) and 10% (Ÿ) NaCl concentration. Values represent means ± SD (n = 3). 222 Figure 3. Emulsifying activity of microbactan on motor oil at different pH values. pH 4 (Ɣ), pH 7 (Ŷ) and pH 10 (Ÿ).Values represent means ± SD (n = 3). These results suggest that microbactan could find application in environmental marine processe uch as enhanced oil recovery, cleaning of oil reservoirs and enhancement of biodegradation rates o pilled oils [2]. Bioemulsifiers have a wide diversity of composition and structure and are haracterized by improved functionality and stability. Their potential applications include: the oil and petroleum industries, water and soil bioremediation, metal treatment and processing, detergents and aundry supplies, agriculture, textile manufacturing, pulp and paper processing, paints, cosmetics pharmaceuticals, personal care products and food processing [4,6]. Microbactan has higher emulsifying Figure 3. Emulsifying activity of microbactan on motor oil at different pH values. pH 4 (Ɣ), pH 7 (Ŷ) and pH 10 (Ÿ).Values represent means ± SD (n = 3). Figure 3. Emulsifying activity of microbactan on motor oil at different pH values. pH 4 (Ɣ), pH 7 (Ŷ) and pH 10 (Ÿ).Values represent means ± SD (n = 3). h l h i b ld fi d li i i i l i These results suggest that microbactan could find application in environmental marine processes such as enhanced oil recovery, cleaning of oil reservoirs and enhancement of biodegradation rates of spilled oils [2]. Bioemulsifiers have a wide diversity of composition and structure and are characterized by improved functionality and stability. Their potential applications include: the oil and petroleum industries, water and soil bioremediation, metal treatment and processing, detergents and laundry supplies, agriculture, textile manufacturing, pulp and paper processing, paints, cosmetics, pharmaceuticals, personal care products and food processing [4,6]. 2.3. Effect of Temperature, pH and NaCl on Emulsifying Activity Microbactan has higher emulsifying activity at 50 °C and 3.5% NaCl, conditions typical of intertidal environments [8]. If the molecule remains stable under these conditions for long periods, it may prove specifically useful for bioremediation of polluted intertidal habitats [27]. The stability of the emulsions under diverse conditions, such as temperature, pressure, pH and ionic strength, makes this biopolymer a versatile emulsifier for use in many food and pharmaceutical formulations. 2.4. Toxicity of Microbactan Based solely on the Artemia salina toxicity test, microbactan proved innocuous, as expected for a biologically derived surface-active agent; this class of compound is generally biodegradable and non-toxic [2]. Although synthetic surfactants exhibited the highest emulsifying activity in this study, Triton® X-100 (also a non-ionic surface-active), at least, proved to be toxic in our Artemia bioassay (Table 3). Polyoxyethylene octyl phenols (the Triton X series) are known to be highly cytotoxic, solubilizing the membrane lipid bilayer [28]. In fact, these synthetic surfactants can actually inhibit aromatic hydrocarbon biodegradation via toxic interactions, making them less suitable for bioremediation purposes [27,29]. The Artemia salina toxicity bioassay is a reliable primary screen, given the sensitivity of this crustacean to a wide range of biologically active compounds of diverse chemistries, including pesticide residues, mycotoxins, stream pollutants, anaesthetics, dinoflagellate toxins, 223 morphine-like compounds and oil dispersants [30]. In addition, this bioassay has demonstrated good correlation with other cell-based tests such as tumor cell lines (e.g., KB, P-388, 388, L5178Y and L1210) and mammalian systems [31,32]. Reinforcing this finding, an experimental study on marine biofilm colonization on surfaces coated with microbactan showed that this biopolymer did not affect biofilm formation (unpublished results). Table 3. Anti-crustacean activity of surfactants against Artemia salina nauplii. Table 3. Anti-crustacean activity of surfactants against Artemia salina nauplii. Surfactant LC50 (μg/mL) Microbactan >1000 Triton X-100 100.3 ± 3.8 Tween 80 >1000 2.5. Phylogenetic Reassessment of Microbacterium sp. MC3B-10 2.5. Phylogenetic Reassessment of Microbacterium sp. MC3B-10 The reassessment of the phylogenetic position of Microbacterium sp. MC3B-10 based on its 16S rDNA gene, using a more robust bioinformatics approach, confirmed that its closest relative was Microbacterium trichothecenolyticum, with a similarity of 99.2% (underline, Figure 4). Stackebrandt and Ebers [33] established that a cutoff of 98.7% 16S rDNA gene homology is appropriate for species differentiation within a genus. Following this criterion and given the fact that Microbacterium is a very tight genus with respect to 16S rDNA gene homology between valid species [34], Microbacterium sp. MC3B-10 could be classified as M. trichothecenolyticum. However, definitive identification of this bacterial isolate requires a polyphasic approach including biochemical, physiological, chemotaxonomic, and nucleic acid-based methods, along with a range of microscopies, as previously shown for newly described Microbacterium species [35]. 2.4. Toxicity of Microbactan The in-silica analysis revealed novel strains of the genus Microbacterium (MC24 and MC60) that had been isolated from an intertidal environment in Brazil and were capable of synthesizing bioemulsifiers [36]. These isolates were also related to Microbacterium sp. MC3B-10, but at lower similarity levels (97%). It should be noted that, except for M. resistens, M. hominis, and M. paraoxydans, all microbacteria are considered environmental bacteria [34]. Certain Microbacterium species, such as M. thalassium, M. halophilum, and M. phyllosphaerae, occur in coastal habitats [35,37]. However, the biotechnological potential of this genus as a marine bioemulsifier producer has only been reported recently [8,36]. The occurrence of bioemulsifier-producing Microbacterium species in intertidal environments is not surprising, given the metabolic versatility exhibited by this genus; some of them can metabolize hydrocarbons, presumably through bioemulsifier synthesis [38,39]. In addition, the availability of hydrophobic substrates in coastal marine environments may select for bacteria capable of synthesizing surface-active molecules to enhance nutrient uptake [1]. 224 Figure 4. Neighbor-joining phylogenetic tree, based on 16S rDNA gene sequences, showing the positions of strain MC3B-10 (AY833570) relative to all known Microbacterium species. Accession numbers of 16S rDNA gene sequences of reference organisms are shown in parentheses. Bootstrap values (1000 replicates, >50%) are shown above the nodes. The bar indicates the relative sequence divergence (0.01 nucleotide substitutions per site). Curtobacterium luteum and C. michiganense were used as outgroups. M. sp. MC3B 10 (AY833570) M. trichothecenolyticum (AB167383) M. ketosireducens IFO 14548 (AB004724) M. terrae IFO 15300 (AB004720) M. sp. Mc24 (DQ512484) M. sp. Mc60 (DQ512485) M. kitamiense kitami (AB013919) M. laevaniformans IFO 14471T (D21344) M. dextranolyticum IFO 14592T (D21341) M. aquimaris JS63 1 (AM778450) M. aquimaris JS54 2 (AM778449) M. resistens AGP4 3 (AY277553) M. nematophilum CBX102 (AF319539) M.arabinogalactanolyticum DSM8611 Y17228 M. terregens IFO 12961 (AB004721) M. koreense JS53 5 (AY962575) M. schleiferi DSM 20489 (Y17237) M. sp. 1 (DQ512483) M. oleivorans DSM 16091 (AJ698725) M. lacticum IFO 14135T (D21343) M. aurum IFO 15204T (D21340) M. keratanolyticum IFO13309 (AB004717) M. phyllosphaerae DSM 13468 (AJ277840) M. phyllosphaerae SAFR 012 (AY167852) M. aerolatum V 73 (AJ309929) M. hydrocarbonoxydans DSM16089T AJ698726 M. paraoxydans C57 33 (AJ581908) M. luteolum DSM 20143 (Y17235) M. maritypicum DSM 12512 (AJ853910) M. oxydans CV71a (AJ717358) M. oxydans AC94 (AJ717356) M. oxydans S15 M5 (AM234159) M. sp. ZD M2 (DQ417926) M. sp. JYC17 (EU036699) M. sp. F10a (EU196564) M. ulmi (AY062021) M. arborescens IFO 3750 (AB007421) M. 3.3. High Performance Size Exclusion Chromatography The molecular weight of microbactan was determined using an HPLC system Prominence Shimadzu™, a PL aquagel-OH mixed, 8 μm (Varian) guard column (U 7.5 mm × L50 mm), and a PL aquagel-OH mixed (Varian, Palo Alto, CA, USA) separation column (U 7.5 × 300 mm, operating range 102–107 g/mol). Elution was performed at 1 mL/min with 0.1 M ammonium acetate containing 0.03% NaN3, and the eluate filtered through a 0.1 μm membrane (Durapore Membrane, PVDF, Hydrophilic type VVLP, Millipore®, Saint Quentin en Yvelines, France). A differential refractive index (RI) detector (L2490, VWR Hitachi, Fontenay sous bois, France) and a multi-angle light scattering detector (Dawn Heleos™, Wyatt, Toulouse, France) were coupled on-line and data computed with Astra software for absolute molar mass determination. 3.1. Production of Microbactan Microbacterium sp. MC3B-10 was originally isolated from pristine rocky intertidal shores in the state of Campeche, southern Gulf of Mexico, Mexico. Microbactan was produced following a batch shake flask fermentation of an overnight culture (50 mL) of Microbacterium sp. MC3B-10 as previously reported [8]. The exopolymeric material was extracted from the fermented broth, redissolved in a small volume of distilled water and then dialyzed (molecular weight cutoff 12,000–14,000 Da, Spectrum®, CA, USA) for 72 h. The resulting material was lyophilized and kept in the dark before analysis. 2.4. Toxicity of Microbactan imperiale IFO 12610T (D21342) A. globiformis (M23411) C. luteum DSM 20542 (X77437) C. michiganensis DSM 46364 (X77435) 99 99 99 99 99 99 98 97 96 54 93 83 82 71 92 90 82 81 76 65 55 52 99 57 225 3.2. Lipid Analysis and Fatty Acid Composition Lipids were determined using Fourier-transform infrared spectroscopy (FT-IR) [8] and spectrophotometry, using triolein as standard lipid [40]. Fatty acid composition was investigated as follows. After mild acid hydrolysis of the exopolymer (0.7064 g) at 80 °C for 30 min, the hydrolysate was extracted with CHCl3:CH3OH:H2O (3:2:1) and the aqueous phase extracted two more times with CHCl3 (1:1). The CHCl3 fractions were combined and evaporated under reduced pressure [41]. Before GC-MS analysis, the sample was pre-treated as described previously [42]. The methylation of fatty acids was carried out with 5 mL of HCl-methanol at 80 °C for 30 min. The fatty acid methyl ethers were extracted with hexane and subjected to analysis. GC-MS was performed using helium as carrier gas on a Perkin Elmer Clarus 580 GC equipped with a Clarus SQ 8S mass spectrometer, equipped with an Elite-5 capillary column (30 m × 0.32 mm i.d., 0.25 μm film thickness). 3.6. Effects of Temperature, Salinity and pH on Emulsifying Activity The effect of temperature, NaCl content and pH on emulsifying activity of microbactan was tested using motor oil because this substrate yielded the highest levels of emulsification (~96%). The influence of temperature was determined by pretreating at desired temperatures (5, 50 and 100 °C) microbactan solutions that were either heated for 15 min in a water bath or cooled for 10 min in a freezer, before being assayed [23]. The effect of NaCl was investigated at three concentrations (3.5, 5 and 10% w/v) and pH values of 4, 7 and 10 were studied. Emulsification with motor oil was carried out at room temperature as in Item 3.5 [22,24]. 3.4. Ionic Charge The ionic charge of microbactan was assessed using a modified double diffusion technique [43]. Briefly, two 12 mm-diameter plugs were taken from Petri dishes containing 1% agar. One of the 226 resulting wells was filled with 200 μL of microbactan solution at 1 mg/mL and the other with commercial anionic (sodium dodecyl sulphate SDS and Teepol® 610S Sigma-Aldrich Química, S. de R.L. de C.V., Toluca, Mexico State, Mexico) or cationic (barium chloride and cetyltrimethylammonium bromide) surfactants supplied by Sigma-Aldrich. These surfactants were used at 20 mmol, except for barium chloride (50 mmol). The appearance of precipitation lines between the wells was indicative of the ionic character of microbactan. Plates were incubated at 28 °C and monitored daily. 3.5. Emulsifying Activity The emulsifying activity of microbactan was assessed against aromatic hydrocarbons (benzene and xylene), vegetable oils (corn, olive and sunflower) and mineral, motor and crude oils, all at 1%. Oils were acquired from local commercial suppliers and the hydrocarbons were of analytical grade (J.T. Baker, Center Valley, PA, USA; E.M. Sciences, Hatfield, PA, USA). The emulsifying activity was measured by combining equal volumes of microbactan solution (1% w/v) and the target hydrophobic substrates in 12-mm-diameter glass tubes, as previously described [8]. Controls included Triton® X-100 and Tween® 80 (Research Organics, Cleveland, OH, USA), a plant-derived polysaccharide (gum arabic from Spectrum®) and xanthan gum, a bacterial polysaccharide (ICN Biomedicals, Inc, Irvine, CA, USA). All evaluations were performed in triplicate. 3.6. Effects of Temperature, Salinity and pH on Emulsifying Activity 3.9. Phylogenetic Reassessment The purpose of this analysis was twofold: first, to corroborate the identification of the species of Microbacterium MC3B-10 based on its 16S rDNA gene (AY833570) using a more robust bioinformatics approach (see details below) with the sequences reported in [8], and second to prove if recently reported species of Microbacterium displaying bioemulsifying activity or originally isolated from polluted environments were phylogenetically close to Microbacterium sp. MC3B-10. The following additional sequences were obtained from the GenBank database, Microbacterium sp. Mc1 DQ512483, Mc24 DQ512484, Mc60 DQ512485, M. aquimaris JS54-2 AM778449, M. aquimaris JS63-1 AM778450, Microbacterium sp. F10a EU196564, M. oleivorans DSM 16091 AJ698726 and M. sp. ZD-M2 DQ417926. In total, 41 sequences (1418 bp) were aligned using the CLUSTALW algorithm implemented in MEGA 4.3 software [46]. Phylogenetic analyses were performed using the Neighbor Joining (NJ) algorithm: phylogenetic analysis using parsimony V4.0 beta 10, 2002. For NJ analyses, the evolutionary model selected was the GTR + I + G (General Time Reversible Model + Invariable sites + Gamma distribution) inferred from the program Modeltest V.3.06 [47]. Support for nodes of the NJ tree was determined by calculating bootstrap proportion values [48] based on 1000 resamplings of neighbor-joining searches. The 16S rDNA sequences of Curtobacterium luteum and C. michiganense were used as outgroup. 3.8. Statistical Analysis All the experiments were run in triplicate. For the emulsification assays, a two-way analysis of variance (ANOVA) was used to assess the effect of biopolymer and controls on the hydrophobic substrates. When there was a significant difference (p 0.05) between microemulsion and controls, we applied the Tukey test aposteriori. These tests were performed with Sigma Stat software version 2.0 (1997). 3.7. Toxicity Test Brine shrimp eggs (Salt Creek™, Salt Lake City, UT, USA) were hatched at 27 °C under continuous aeration and illumination in seawater prepared with sea salts (Coralife®, Rhinelander, WI, USA) at 38 g/L and supplemented with 6 mg/L of dried yeast [44]. Solutions of Triton® X-100 and Tween® 80 (Research Organics, Cleveland, OH, USA) and microbactan were dissolved in seawater at concentrations of 1000, 500, 100, 50 and 10 μg/mL [45]. The 50% Lethal Concentration (LC50) was determined by counting the dead nauplii after an incubation period of 24 h. Data were analyzed with the Finney computer program as described previously [27]. 227 4. Conclusions This study demonstrated that the novel Microbacterium sp. MC3B-10, probably indigenous to the marine intertidal zone, synthesizes an environmentally-friendly exopolymeric non-ionic glycolipoprotein capable of emulsifying aromatic hydrocarbons and oils. The functional stability of this bioemulsifier was retained for long periods of time and at a range of temperature, NaCl concentration and pH. The chemistry, activity and stability of microbactan make it useful for environmental and personal care applications. Overall, our results suggest that bioemulsifier-producing bacteria warrant intensified bioprospection in the intertidal zones. Additionally, this study corroborates the still untapped resource represented by marine microorganisms for new biosurfactants, bioemulsifiers and biopolymers. This work thus strengthens the notion that marine microbial diversity is a non-exhausted source of novel biomolecules and particularly emphasizes the biotechnological significance of Microbacterium species derived from the intertidal environment. 228 Acknowledgments This research was supported by a grant Marine biofilm inhibition by bacterial exopolymeric substance-based coatings and characterization of producing isolates (F/3617-2F) 2008-2011, from the International Foundation for Science (IFS) F/3617-1, Stockholm, Sweden; Organization for the Prohibition of Chemical Weapons, The Hague, The Netherlands. Conflicts of Interest The authors declare no conflict of interest. The authors declare no conflict of interest. References E.; Rosenberg, E. Natural roles of biosurfactants. Environ. Microbiol. 2001, 3, 229–236. 1. Ron, E.; Rosenberg, E. Natural roles of biosurfactants. Environ. Microbiol. 2001, 2. Banat, I.M.; Franzetti, A.; Gandolfi, I.; Bestetti, G.; Martinotti, M.; Fracchia, L.; Smyth, T.; Marchant, R. Microbial biosurfactants, production, applications and future potential. Appl. Microbiol. Biotechnol. 2010, 87, 427–444. 2. Banat, I.M.; Franzetti, A.; Gandolfi, I.; Bestetti, G.; Martinotti, M.; Fracchia, L.; Smyth, T.; Marchant, R. Microbial biosurfactants, production, applications and future potential. Appl. Microbiol. Biotechnol. 2010, 87, 427–444. 3. Freitas, F.; Alves, V.; Reis, M. Advances in bacterial exopolysaccharides: From production to biotechnological applications. Trends Biotechnol. 2011, 29, 388–398. 3. Freitas, F.; Alves, V.; Reis, M. Advances in bacterial exopolysaccharides: From production to biotechnological applications. Trends Biotechnol. 2011, 29, 388–398. 4. Freitas, F.; Alves, V.; Carvalheira, M.; Costa, N.; Oliveira, R.; Reis, M. Emulsifying behaviour and rheological properties of the extracellular polysaccharide produced by Pseudomonas oleovorans grown on glycerol byproduct. Carbohyd. Polym. 2009, 78, 549–556. 5. Satpute, S.; Banat, I.; Dhakephalkar, P.; Banpurkar, A.; Chopade, B. Biosurfactants, bioemulsifiers and exopolysaccharides from marine microorganisms. Biotechnol. Adv. 2010, 28, 436–450. 6. Gutiérrez, T.; Leo, V.; Walker, G.; Green, D. Emulsifying properties of a glycoprotein extract produced by a marine Flexibacter species strain TG382. Enzyme Microb. Technol. 2009, 45, 53–57. 7. Ortega-Morales, B.O.; Chan-Bacab, M.J.; de la Rosa-García, S.C.; Camacho-Chab, J.C. Valuable processes and products from marine intertidal microbial communities. Curr. Opin. Biotechnol. 2010, 21, 346–352. 8. Ortega-Morales, B.O.; Santiago, J.; Chan-Bacab, M.J.; Miranda, E.; Fardeau, M.; Carrero, J.; Bartolo, P.; Valadéz, A.; Guezennec, J. Characterization of extracellular polymers synthesized by tropical intertidal biofilm bacteria. J. Appl. Microbiol. 2007, 102, 254–264. 9. Beech, I.; Hanjagsit, L.; Kalaji, M.; Neal, A.L.; Zinkevich, V. Chemical and structural characterization of exopolymers produced by Pseudomonas sp. NCIMB 2021 in continuous culture. Microbiology 1999, 145, 1491–1497. 10. Amaral, P.; Silva, J.; Lehocky, M.; Barros-Timmons, A.; Coelho, M.; Marrucho, I.; Coutinho, J.A.P. Production and characterization of a bioemulsiftier from Yarrowia lipolytica. Process Biochem. 2006, 41, 1894–1898. 229 11. Luna-Velasco, M.A.; Esparza-García, F.; Cansares-Villanueva, R.O.; Rodríguez-Vázquez, R. Production and properties of a bioemulsifier synthesized by phenanthrene-degrading Penicillum sp. Process Biochem. 2007, 42, 310–314. 12. Maier, R.M. Biosurfactants: Evolution and diversity in bacteria. Adv. Appl. Microbiol. 2003, 52, 101–121. 13. Navon-Venezia, S.; Zosim, Z.; Gottlieb, A.; Legmann, R.; Carmeli, S.; Ron, E.Z.; Rosenberg, E. Alasan, a new bioemulsifier from Acinetobacter radioresistens. Appl. Environ. Microb. References 1995, 61, 3240–3244. 14. Thavasi, R.; Jayalakshmi, S.; Balasubramanian, T.; Banat, I. Biosurfactant production by Corynebacterium kutscheri from waste motor lubricant oil and peanut oil cake. Lett. Appl. Microbiol. 2007, 45, 686–691. 15. Thavasi, R.; Jayalakshmi, S.; Balasubramanian, T.; Banat, I. Production and characterization of a glycolipid biosurfactant from Bacillus megaterium using economically cheaper sources. World J. Microbiol. Biotechnol. 2007, 24, 917–925. 16. Whitman, W. Food Emulsifiers: Chemistry, Technology, Functional Properties and Applications; Ten Alps Publishing: London, UK, 1990. 17. Kim, S.; Ravichandran, D.; Khan, S.; Kim, Y. Prospective of the cosmeceuticals derived from marine organisms. Biotechnol. Bioprocess Eng. 2008, 13, 511–523. 18. Kumar, A.; Mody, K.; Jha, B. Evaluation of biosurfactant/bioemulsifier production by a marine bacterium. Bull. Environ. Contam. Toxicol. 2007, 79, 617–621. 19. Willumsen, P.A.; Karlson, U. Screening of bacteria, isolated from PAH-contaminated soils, for production of biosurfactants and bioemulsifiers. Biodegradation 1997, 7, 415–423. 20. Gutiérrez, T.; Mulloy, B.; Bavington. C.; Black, K.; Green, D. Partial purification and chemical characterization of a glycoprotein (putative hydrocolloid) emulsifier produced by a marine bacterium Antarctobacter. Appl. Microbiol. Biotechnol. 2007, 76, 1017–1026. 21. Perfumo, A.; Smyth, T.J.; Marchant, R.; Banat, I.M. Production and Roles of Biosurfactants and Bioemulsifiers in Accessing Hydrophobic Substrates. In Handbook of Hydrocarbon and Lipid Microbiology; Timmis, K.N., Ed.; Springer-Verlag: Berlin, Germany, 2010; pp. 1502–1510. 22. Sarubbo, L.; Moura, J.; Campos-Takaki, G. Production and stability studies of the bioemulsifier obtained from a new strain of Candida glabrata UCP 1002. Electron. J. Biotechnol. 2006, 9, 400–406. 23. Rufino, R.; Sarubbo, A.; Campos, G. Enhancement of stability of biosurfactant produced by Candida lipolytica using industrial residue as substrate. World J. Microb. Biotechnol. 2007, 23, 729–734. 24. Ilori, M.O.; Amobi, C.; Odocha, A. Factors affecting biosurfactant production by oil degrading Aeromonas spp. isolated from a tropical environment. Chemosphere 2005, 61, 985–992. 25. Sarubbo, L.; Farias, C.; Campos-Takaki, G. Co-utilization of canola oil and glucose on the production of a surfactant by Candida lipolytica. Curr. Microbiol. 2007, 54, 68–73. 230 26. Rosenberg, E.; Rubinovitz, C.; Gottlieb, A.; Rosenhak, S.; Ron, E.Z. Production of biodispersan by Acinetobacter calcoaceticus A2. Appl. Environ. Microb. 1988, 54, 317–322. 27. Pelletier, E.; Delille, D.; Delille, B. Crude oil bioremediation in sub-Antarctic intertidal sediments: Chemistry and toxicity of oiled residues. Mar. Environ. Res. 2004, 57, 311–327. 28. Weyermann, J.; Lochmann, D.; Zimmer, A. A practical note on the use of cytotoxicity assays. Int. J. Pharm. 2005, 288, 369–376. 29. 40. Izard, J.; Limberger, R. Rapid screening method for quantification of bacterial cell lipids from whole cells. J. Microbiol. Meth. 2003, 55, 411–418. References Meyer, B.N.; Ferrigni, N.R.; Putnam, J.E.; Jacobsen, L.B.; Nichols, D.E.; McLauglin, J.L. Brine shrimp: A convenient general bioassay for active plant constituents. Planta. Med. 1982, 45, 31–34. 30. McLaughlin, J.L. Crown-gall Tumors in Potato Discs and Brine Shrimp Lethality: Two Simple Bioassays for Higher Plant Screening and Fractionation. In Methods in Plant Biochemistry; Hostettmann, K., Ed.; Academic Press: London, UK, 1991; Volume 6, pp. 1–32. 31. Birndorf, H.C.; D’Alossio, J.; Bagshaw, J.C. DNA-dependant RNA-polymerases from Artemia, embryos. Characterization of polymerases I and II from nauplius larvae. Dev. Biol. 1975, 45, 34–43. 32. Perfeito, J.P.; Santos, M.L.; López, K.S.; Paula, J.E.; Silveira, D. Characterization and biological properties of Pouteria torta extracts. Rev. Bras. Farmacogn. 2005, 15, 183–186. 33. Stackebrandt, E.; Ebers, J. Taxonomic parameters revisited: Tarnished gold standards. Microbiol. Today 2006, 33, 152–155. 34. Gneiding, K.; Frodl, R.; Funke, G. Identities of Microbacterium spp. encountered in human clinical specimens. J. Clin. Microbiol. 2008, 46, 3646–3652. 35. Takeuchi, M.; Hatano, K. Proposal of six new species in the genus Microbacterium and transfer of Flavobacterium marinotypicum ZoBell and Upham to the genus Microbacterium as Microbacterium maritypicum comb. nov. Int. J. Syst. Bacteriol. 1998, 48, 973–982. 36. Aniszewski, E.; Silva, R.; Faria, F.; Gomes, S.; Soares, A. Bioemulsifier production by Microbacterium sp. strains isolated from mangrove and their application to remove cadmiun and zinc from hazardous industrial residue. Braz. J. Microbiol. 2010, 41, 235–245. 37. Burmølle, M.; Webb, J.; Rao, D.; Hansen, L.; Sørensen, S.; Kjelleberg, S. Enhanced biofilm formation and increased resistance to antimicrobial agents and bacterial invasion are caused by synergistic interactions in multispecies biofilms. Appl. Environ. Microb. 2006, 72, 3916–3925. 38. Schippers, A.; Bosecker, K.; Sproer, C.; Schumann, P. Microbacterium oleivorans sp. nov. and Microbacterium hydrocarbonoxydans sp. nov., novel crude-oil-degrading Gram-positive bacteria. Int. J. Syst. Evol. Microbiol. 2005, 55, 655–660. 39. Franzetti, A.; Gandolfi, I.; Bertolini, V.; Raimondi, C.; Piscitello, M.; Papacchini, M.; Bestetti, G. Phylogenetic characterization of bioemulsifier-producing bacteria. Int. Biodeterior. Biodegrad. 2011, 65, 1095–1099. 40. Izard, J.; Limberger, R. Rapid screening method for quantification of bacterial cell lipids from whole cells. J. Microbiol. Meth. 2003, 55, 411–418. 231 41. Ganfield, M.C.; Pieringer, R.A. Phosphatidylkojibiosyl diglyceride. The covalently linked lipid constituent of the membrane lipoteichoic acid from Streptococcus faecalis (faecium) ATCC 9790. J. Biol. Chem. 1975, 250, 702–709. 42. Rezanka T.; Vokoun, J.; Slavicek, J.; Podojil, M. Determination of fatty acids in algae by capillary gas chromatography-mass spectrometry. J. Chromatogr. References 1983, 268, 71–78. 43. Van Oss, C.J. Specifically impermeable precipitate membranes formed through double diffusion in gels: Behavior with complex forming and with simple systems. J. Colloid Interf. Sci. 1968, 27, 684–690. 44. Ortega-Morales, B.O.; Chan-Bacab, M.J.; Miranda-Tello, E.; Fardeau M.L.; Carrero J.C.; Stein, T. Antifouling activity of sessile bacilli derived from marine surface. J. Ind. Microbiol. Biotechnol. 2008, 35, 9–15. 45. Solís, P.; Wright, C.; Anderson, M.; Gupta, M.; Phillipson, D.A. A Microwell cytotoxic using Artemia salina (brine shrimp). Planta Med. 1993, 59, 250–252. 46. Tamura, K.; Dudley, J.; Nei, M.; Kumar, S. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 2007, 24, 1596–1599. 47. Posada, D.; Crandall, K.A. Modeltest: Testing the model of DNA substitution. Bioinformatics 1998, 14, 817–818. 48. Felsenstein, J. Confidence limits on phylogenies: An approach using the bootstrap. Evolution 1985, 39, 783–791.
https://openalex.org/W4299615395	https://periodicos.ufpi.br/index.php/pet/article/download/2153/1977	Portuguese	null	AMPLIANDO AS FRONTEIRAS DO PET FILOSOFIA-UFPI	Cadernos do PET Filosofia	2,010	cc-by	353	1 Professor-Tutor do PET Filosofia/UFPI; Professor-Adjunto do Departamento de Filosofia/UFPI. Editorial Editorial Editorial Cadernos do PET Filosofia – Volume 1, Nº 1, 2010 Editorial 0 Editorial 0 Carvalho, Helder B. A. de Ampliando as fronteiras do PET Filosofia-UFPI Helder Buenos Aires de Carvalho1 Este é o primeiro número da CADERNOS DO PET FILOSOFIA, revista eletrônica que nasceu dos objetivos de integrar ensino, pesquisa e extensão no âmbito do grupo PET Filosofia do Programa de Educação Tutorial na Universidade Federal do Piauí. Sua publicação representa a concretização de um esforço coletivo, por docentes e discentes do Curso de Filosofia da UFPI, bem como de nossos colegas professores de outras IES do país, em busca de um espaço de discussão crítica e coletiva que permita a partilha genuína de reflexões, preocupações e trabalho teórico no esclarecimento e inovação das questões filosóficas e do seu ensino nos mais diferentes níveis, dessa forma, ampliando as fronteiras do PET Filosofia para além dos muros de nossa UFPI. A revista está voltada para o incentivo à produção filosófica de alunos de graduação e pós-graduação em filosofia no país, bem como aos pesquisadores das mais diferentes áreas da filosofia, seja sobre temáticas filosóficas específicas, seja sobre temáticas relativas ao ensino de filosofia. Publicamos também resenhas e notas bibliográficas, bem como materiais didáticos voltados para o ensino de filosofia, seja tradução de textos filosóficos, seja trabalhos com propostas didáticas para o ensino de filosofia. Na constituição desse primeiro número, além dos colegas professores que analisaram os trabalhos publicados e dos nossos estudantes do PET Filosofia, gostaria de agradecer especialmente o apoio do Prof. Dr. Luizir de Oliveira, Coordenador do Curso de Filosofia/UFPI. O apoio do Prof. Dr. Ricardo Alaggio, da Editora da UFPI, foi também essencial. Estou particularmente feliz, como Editor-Chefe, por alcançar mais um dos objetivos do PET Filosofia, e agradeço aos colegas do Departamento de Filosofia pelo apoio e suporte nessa trajetória. Vida longa à Cadernos do PET Filosofia em sua trajetória de busca do diálogo filosófico na internet! Teresina, PI Julho/2010 Teresina, PI Julho/2010 Cadernos do PET Filosofia – Volume 1, Nº 1, 2010 ISSN 2178-5880 ISSN 2178-5880
https://openalex.org/W3046916171	https://www.scielo.br/j/jatm/a/gYPdM5sZK4NYSNjHqjBnyjP/?lang=en&format=pdf	English	null	Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness	Journal of Aerospace Technology and Management	2,020	cc-by	7,172	How to cite R. K. Naren Shankar and N. Dilip Raja (2020) Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness. J Aerosp Technol Manag, 18: e2520 https://doi. org/10.5028/jatm.v12.1122 R. K. Naren Shankar and N. Dilip Raja (2020) Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness. J Aerosp Technol Manag, 18: e2520 https://doi. org/10.5028/jatm.v12.1122 R. K. Naren Shankar https://orcid.org/0000-0002-3363-4478 N. Dilip Raja https://orcid.org/0000-0001-9825-6503 ABSTRACT: Effect of Mach number on coflowing jet at lip thickness of 0.2 Dp, 1.0 Dp and 1.5 Dp (where Dp is primary nozzle exit diameter, 10 mm) at Mach numbers 1.0, 0.8 and 0.6 were studied experimentally. It was found that an increase in Mach number does not have any profound effect on axial total and static pressure variation for 0.2 Dp. Decreasing the mean diameter is due to the geometrical constraints. In this study, the primary nozzle dimension and secondary duct is maintained constant for comparison. For the case of 0.2 Dp, static pressure is almost equal to atmospheric pressure for all Mach numbers. Whereas for other two lip thickness, increase in Mach number marginally influences axial total pressure and profoundly varies static pressure. It is noted that it varies considerably up to 11.1% in the axial direction and up to 17% in the radial direction for Mach number 1.0. For lower Mach numbers, such variation is not observed. Increase in Mach number increases static pressure variation in the coflowing jet flow field with lip thickness 1.0 Dp and 1.5 Dp. KEYWORDS: Coflowing jet; Lip thickness; Static pressure; Subsonic jet; Mixing enhancement; Mixing inhibition. WORDS: Coflowing jet; Lip thickness; Static pressure; Subsonic jet; Mixing enhancement; Mixing inhibition 1 1 1. Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology –Department of Aeronautical Engineering – Chennai/Tamil Nadu – India 2. Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology –Department of Mechanical Engineering – Chennai/Tamil Nadu – India. Corresponding author: ndilipraja@veltech.edu.in Received: 23 Sep 2019 \| Accepted: 17 Dec 2019 Section Editor: T. John Tharakan J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 1 https://doi.org/10.5028/jatm.v12.1122 1 https://doi.org/10.5028/jatm.v12.1122 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Naren Shankar Radha Krishnan1, Dilip Raja Narayana2, INTRODUCTION Mixing in coflowing jet (CFJ) suppresses noise and reduces infrared (IR) signature in commercial aircraft turbofan engines. To reduce IR signature, jet potential core volume and surface area should be reduced. Modification of coflowing nozzle can alter mixing in CFJ (Zaman and Papamoschou 2000; Papamoschou 2000; Shupe 2007). Such modifications do not require additional power source and that are called as passive control in CFJ. Some of the factors that are associated with CFJ include: lip thickness (LT), wake dominance, angular jet and mixing enhancement. Lip thickness (LP) is the distance separating primary nozzle and secondary duct. For low subsonic jets, it is well proven that increase in lip thickness enhances mixing because of wake dominance (Buresti et al. 1994). Figure 1 shows the sketch of the coflowing jet profile. It reveals various regions within the jet recirculation zone become dominant under the influence of lip thickness and bypass ratio. Wake dominance creates a recirculation zone that deflects the secondary jet towards primary jet. This increases interaction between the jets and enhances mixing. For high subsonic jets, only few researchers have performed the effect of lip thickness with varying Mach number. When coflow exist at an angle to the horizontal with respect to CFJ, it is called angular coflow. The secondary jets in CFJ that flow at an inclination to the primary jets are called as angular jet. Papamoschou (2000) studied the effect of secondary jet on primary jet on high subsonic and supersonic Mach numbers mixing enhances with high velocity secondary jet, particularly supersonic. A similar work with coannular nozzle was reported by Zaman J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Shankar RKN, Raja ND Shankar RKN, Raja ND and Papamoschou (2000). The mixing enhanced in the surrounding jet around exit Mach number of 0.8 to 1.2, irrespective of the operating conditions of the central jet. A scaled down version of typical turbo fan exhaust has been studied with varying secondary nozzle area (Shupe et al. 2007). They made secondary nozzle as converging-diverging and varied the secondary nozzle exit area to the throat area ratio from 1.0 to 1.6. Centerline velocity survey was taken for exit Mach number 0.6 to 1.0 in the primary jet. INTRODUCTION It was concluded that the effect of varying Mach number of the primary jet is relatively minor, whereas core of primary jet is practically eliminated for the secondary nozzle area ratio above 1.4. Lip Thickness Secondary Duct Primary Nozzle Initial Merging Zone Fully Merging Zone Intermediate Merging Zone Recirculation Zone Inner Potential Core Outer Potential Core Figure 1. Profile of coflowing jet with finite lip thickness (Shankar 2017). Inner Potential Core Recirculation Zone Outer Potential Core Secondary Duct Intermediate Merging Zone Fully Merging Zone Initial Merging Zone Figure 1. Profile of coflowing jet with finite lip thickness (Shankar 2017). In all the above cases the primary nozzle lip thickness was maintained as 0.7 mm and mixing enhancement is achieved only by varying the secondary jet Mach number. Lovaraju and Rathakrishnan (2011) studied the effect of Mach number in CFJ below sonic, correctly expanded sonic and under expanded sonic up to nozzle pressure ratio (NPR) (nozzle exit pressure against back pressure) of NPR 7. They concluded that mixing inhibits with upsurge in primary jet Mach number. At this stage it is worth noting that, increase in LT from 0.7 to 2.65 mm influences the mixing behavior of CFJ. The potential core length of primary jet elongates in the presence of secondary jet. Srinivasarao et al. (2013) analyzed CFJ from orifices with lip thickness 1.5 and 4.5 mm. CFJ with 4.5 mm lip enhanced mixing and the effect of Mach number is found marginal. Similar study on CFJ nozzles with LT 3 and 15 mm was performed for a constant Mach number of 0.6 and NPR 7. But they did not study the effect of varying Mach number in such finite range of lip thicknesses. Few studies were performed by (Shankar et al. 2016; 2017; 2018; 2019; Thanigaiarasu et al. 2018) on the effect of bypass ratio and lip thickness on CFJ. But specific emphasis was not given for the effect of increasing Mach number in CFJ on range of lip thickness. The present study caters the gap in literature. The current work studies the effect of Mach number on characteristics of typical coflowing jet that has varying lip thickness. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 EXPERIMENTAL Experiments were carried out at the High-speed Jet Laboratory, KCG College of Technology, Chennai, India. In this current study, the primary and the secondary jets have a single feed system and hence the Mach number of the secondary jet is not controlled. Due to friction, the secondary jet encounters losses that are tabulated in Table 1. In all the cases of the present study, exit dimensions for primary and secondary jet are maintained constant. The exit diameter primary nozzle (Dp) is 10 mm and the width of secondary duct is 5 mm. Three LTs of 2 mm (0.2 Dp), 10 mm (1.0 Dp) and 15 mm (1.5 Dp) were analyzed for primary jet exit Mach numbers 0.6, 0.8 and 1.0 with NPR 1.28, 1.52 and 1.89 respectively. The designs of coflowing nozzles are shown in Figs. 2, 3 and 4. Free jet facility is used to carry out the experiments as shown in Fig. 5. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness 3 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Table 1. Experimental conditions at CFJ Nozzle exit. CFJ Pos/Pop (%) P0p (atm) P0s (atm) Mp Ms Percentage loss in secondary jet (%) LT 0.2 Dp 90.0 1.276 1.248 0.6 0.57 2.2 90.0 1.524 1.472 0.8 0.76 3.4 90.0 1.893 1.804 1.0 0.95 4.7 LT 1.0 Dp 82.5 1.276 1.227 0.6 0.55 3.8 84.3 1.524 1.442 0.8 0.74 5.4 86.0 1.893 1.768 1.0 0.94 6.6 LT 1.5 Dp 42.1 1.276 1.162 0.6 0.40 13.8 54.0 1.524 1.283 0.8 0.61 14.7 72.2 1.893 1.645 1.0 0.87 15.5 Compressed air can start to expand and spread out through the primary nozzle and secondary duct passage as soon as it is relieved from the settling chamber. Pressure survey is made using a traverse mechanism mounted with pitot-static tube. Total as well as static pressures along the axial and radial directions of the CFJ were recorded. Table 1. Experimental conditions at CFJ Nozzle exit. Compressed air can start to expand and spread out through the primary nozzle and secondary duct passage as soon as it is relieved from the settling chamber. Pressure survey is made using a traverse mechanism mounted with pitot-static tube. Total as well as static pressures along the axial and radial directions of the CFJ were recorded. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Shankar RKN, Raja ND 4 4 4 Section view A-A Scale: 3:4 Front view Scale: 3:4 A 10 A ϕ10 ϕ40 ϕ50 ϕ60 ϕ97 60 5 5 15 10 5 5 5 5 Rear view Figure 4. CFJ model with LT 1.5 Dp. Pressure Regulating Valve Dry Air From Reservoirs Wire-Mesh Screens Settling Chamber Temperature Port Pressure Port Model Holder Figure 5. Open jet facility. Section view A-A Scale: 3:4 Front view Scale: 3:4 A 10 A ϕ10 ϕ40 ϕ50 ϕ60 ϕ97 60 5 5 15 10 5 5 5 5 Rear view Figure 4. CFJ model with LT 1.5 Dp. Section view A-A Scale: 3:4 Front view Scale: 3:4 A 10 A ϕ10 ϕ40 ϕ50 ϕ60 ϕ97 60 5 5 15 10 5 5 5 5 Rear view Figure 4. CFJ model with LT 1.5 Dp. Front view Front view Scale: 3:4 A Rear view Section view A-A Scale: 3:4 10 5 Figure 4. CFJ model with LT 1.5 Dp. Pressure Regulating Valve Dry Air From Reservoirs Wire-Mesh Screens Settling Chamber Temperature Port Pressure Port Model Holder Figure 5. Open jet facility. Pressure Regulating Valve Dry Air From Reservoirs Wire-Mesh Screens Settling Chamber Temperature Port Pressure Port Model Holder Figure 5. Open jet facility. Pressure Regulating Valve Temperature Port Model Holder Temperature Port M Settling Chamber Dry Air From Reservoirs Wire-Mesh Screens Pressure Port Figure 5. Open jet facility. Figure 5. Open jet facility. In the present investigation, the Reynolds numbers of the jet streams coming out of the nozzle are 0.14 × 106, 0.186 × 106, and 0.233 × 106, respectively for the primary jet exit Mach numbers 0.6, 0.8, and 1.0 in the present investigation. Hence, there is not much effect on the Reynolds numbers of the jet by varying lip thickness. Stagnation probe is placed on the axis facing the flow direction to measure the total pressure. Four static probes of 1 mm diameter each were positioned along the stagnation point to measure the static pressure. Separate static pressure runs were conducted close to the nozzle to observe the effect of sharper gradients and locations of pitot pressure data were correlated (Zaman 2007). Figure 6 shows the image of the probe used to measure the static and total pressures. The probe used in this research had an internal diameter of 0.4 mm and a thickness of 0.1 mm. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 EXPERIMENTAL Compressed air can start to expand and spread out through the primary nozzle and secondary duct passage as soon as it is relieved from the settling chamber. Pressure survey is made using a traverse mechanism mounted with pitot-static tube. Total as well as static pressures along the axial and radial directions of the CFJ were recorded. 5 5 5 Rear view Section view A-A Front view A 10 A ϕ10 ϕ14 ϕ34 ϕ52 ϕ92 60 2 5 5 5 5 5 Rear view Figure 2. CFJ model with LT 0.2 Dp. Section view A-A Scale: 3:4 Front view Scale: 3:4 A 10 A ϕ30 ϕ40 ϕ50 ϕ97 60 5 10 10 5 5 5 LT 5 Rear view Dp Figure 3. CFJ model with LT 1.0 Dp. Section view A-A Front view A 10 A ϕ10 ϕ14 ϕ34 ϕ52 ϕ92 60 2 5 5 5 5 5 Rear view Figure 2. CFJ model with LT 0.2 D . Front view Rear view Section view A-A 5 5 5 5 5 ar view 5 Figure 2. CFJ model with LT 0.2 Dp. Section view A-A Scale: 3:4 Front view Scale: 3:4 A 10 A ϕ30 ϕ40 ϕ50 ϕ97 60 5 10 10 5 5 5 LT 5 Rear view Dp Figure 3. CFJ model with LT 1.0 Dp. Figure 2. CFJ model with LT 0.2 Dp. Figure 2. CFJ model with LT 0.2 Dp. Front view Scale: 3:4 A Section view A-A Scale: 3:4 Dp Rear view 10 LT Figure 3. CFJ model with LT 1.0 Dp. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Shankar RKN, Raja ND CENTERLINE TOTAL PRESSURE VARIATION Centerline total pressure decay of the primary jet is an authentic measure to quantify the regions of CFJ (Rathakrishnan 2010). Potential core length (PCL) is the axial extent within which stagnation pressure at nozzle exit prevails. The total pressure was plotted along the jet axis, at intervals of 1 mm, up to 20 Dp. The jet centerline total pressure decay quantifies jet mixing. This indicates the mixing of primary and secondary jets and combined jet with ambient air. To study the jet spread caused by the CFJ, the total pressure contours were analyzed. In order to evaluate the changes in features of decay, the increase in primary jet exit Mach number, for LT 0.2, 1.0 and 1.5 Dp, centerline and pitot pressure variation is plotted against axial distance based on primary jet exit Mach numbers 0.6, 0.8, and 1.0. Figure 7 shows the centerline pitot pressure decay for LT 0.2 Dp at Mach numbers 1.0, 0.8 and 0.6. Error in the pressure plots in this paper is maintained within ± 2%. The PCL for LT 0.2 Dp for Mach 1.0, 0.8 and 0.6 CFJ are X/Dp = 5 uniformly for all three Mach numbers. This may be due to the fact that Mach number effect on coflow is insignificant (Srinivasarao et al. 2013). The effect of inclination due to geometrical constraints is found minimal. This is due to the fact that the included angle is 2.86°, which may not have profound effect on primary jet. When the included angle increases significantly, the potential core of the primary jet might be affected because of the impinging nature of the angular jets, which is not so in the present case. The deviation was low along the characteristic decay region. Hence, it can be concluded that the variation in Mach number does not have any significance in LT 0.2 Dp in case of CFJ. The changes in slope along the axis from X/Dp = 2 to 5 is 0 since the potential core ends only up to X/Dp = 5 for Mach numbers 1.0, 0.8 and 0.6 along primary jet exit in the CFJ. Once the flow is established, the flow becomes steady till the secondary jets losses its strength. When the coflow exist at an angle to the horizontal with respect to CFJ, it is described as angular coflow. Shankar RKN, Raja ND 4 The ratio of the jet cross section area to the projected area of the probe was maintained greater than 64 to neglect the effect of probe blockage. Hence the exit diameter of the coflowing jet nozzle was at least 24 mm. Total pressure was measured by arranging the probe as parallel to the flow. The static pressure was measured by arranging the probe as normal to the flow. Flow ϕ0,6 15 1,52 Figure 6. Measure of total and static pressure using probe. Flow Figure 6. Measure of total and static pressure using probe. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness 5 RESULTS AND DISCUSSION To study the effect of Mach number in subsonic coflowing jets, centerline total pressure, centerline static pressure and radial static pressure are analyzed. Three major regions exist in axial total pressure variation such as potential core, characteristic decay and fully developed region are studied. Axial static pressure was assumed constant in previous studies which becomes inappropriate with finite lip thickness and varying Mach number. Thus, axial static pressure is analyzed in the present work. CENTERLINE TOTAL PRESSURE VARIATION In the present study, CFJ exist at the angle of 2.86° for LT 0.2 Dp. However, for the other two cases, i.e., LT 1.0 Dp and LT 1.5 Dp, the secondary jet flow parallel to the primary jet at the exit of the nozzle. In other words, the incident angle is 0° for the CFJ having finite lip thickness. Mach 0.6 Mach 0.8 Mach 1.0 1.00 0.75 0.50 0.25 0.00 0 5 10 15 20 X/Dp Pt/Po Figure 7. Centerline pitot pressure variation for CFJ with LT 0.2 Dp. p Figure 7. Centerline pitot pressure variation for CFJ with LT 0.2 Dp. Figure 8 shows the centerline pitot pressure decay for LT 1.0 Dp at primary jet exit Mach numbers 0.6, 0.8, and 1.0. The drop in Pt/Po is indicated as characteristic decay length (CDL) in the plot. Potential core length for CFJ with LT 1.0 Dp for primary jet exit Mach numbers 0.6, 0.8, and 1.0 are X/Dp = 2.5, 2.0 and 2.0 respectively. Within the characteristic decay region, due to increase in Mach number, the variation found is very meagre, but unlike LT 0.2 Dp, there exists a minor variation in this region. Between X/Dp = 2 and 5 unlike the previous case, due to increased mixing the slope of the curve increases. The variation in J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Shankar RKN, Raja ND Shankar RKN, Raja ND 6 slope along the axis from X/Dp = 2 to 5 is 0.137, 0.148 and 0.158 for primary jet exit Mach number 1.0, 0.8 and 0.6. It is found to be zero for CFJ with LT 0.2 Dp. Hence the variation in slope increases as LT increases. This is because a dominant wake region that is present in-between primary and secondary jets alters the characteristics in the near field. The variation of slope increases marginally with increase in Mach number for CFJ with LT 1.0 Dp. slope along the axis from X/Dp = 2 to 5 is 0.137, 0.148 and 0.158 for primary jet exit Mach number 1.0, 0.8 and 0.6. It is found to be zero for CFJ with LT 0.2 Dp. Hence the variation in slope increases as LT increases. This is because a dominant wake region that is present in-between primary and secondary jets alters the characteristics in the near field. l g y g p 7 The flow behavior that lead to improvement in mixing when LT increases can be explained as follows: when LT increases, along the intermediate region, there will be strong mixing between the primary and secondary jets. In this region, extra momentum will be exerted over primary jet by the secondary jet. The increased lip thickness can be used not only to trigger but to control the evolution of organized structures. However, it could increase the turbulence in and around the shear layers. So, it can aid the mixing of the two streams of CFJ (Orlu et al. 2008). This gives more practical importance to such flows. The turbulence affects the characteristics in the initial merging region which lies prior to the intermediate region. The vortices generated by higher lip thickness has larger geometry than that generated by its counterpart having lower lip thickness. This can result in the fast entrainment of secondary flow field and also make the entrainment to readily mix with the composition of primary jet. This can shorten the length of jet by disturbing the PCL of primary jet. CENTERLINE TOTAL PRESSURE VARIATION The variation of slope increases marginally with increase in Mach number for CFJ with LT 1.0 Dp. Mach 0.6 Mach 0.8 Mach 1.0 1.00 0.75 0.50 0.25 0.00 0 5 10 15 20 X/Dp Pt/Po CDL Figure 8. Centerline pitot pressure variation for CFJ with LT 1.0 Dp. p Figure 8. Centerline pitot pressure variation for CFJ with LT 1.0 Dp. Figure 9 shows centerline pitot pressure decay for LT 1.5 Dp at primary jet exit Mach numbers 1.0, 0.8 and 0.6. The potential core length for CFJ with LT 1.5 Dp at primary jet exit Mach numbers 0.6, 0.8, and 1.0 are X/Dp = 1.6, 1.6 and 1.2 respectively. The variation in slope in the near field from X/Dp = 2 to 5 is 0.194, 0.195 for primary jet exit Mach numbers 0.6 and 0.8. CDL Mach 0.6 Mach 0.8 Mach 1.0 1.00 0.75 0.50 0.25 0.00 0 5 10 15 20 X/Dp Pt/Po Figure 9. Centerline pitot pressure variation for CFJ with LT 1.5 Dp. Mach 0.6 p Figure 9. Centerline pitot pressure variation for CFJ with LT 1.5 Dp. For primary jet exit Mach number 1.0 the changes in slope along the axis from X/Dp = 1 to 4 is 0.23. Again, the variation in slope keeps increasing as lip thickness increases. The variation of slope increases marginally with increase in Mach number for CFJ with LT1.5 Dp. The above discussed results are summarized in Table 2. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Table 2. PCL in terms of nozzle exit diameter and its reduction for LT 0.2 Dp, LT 1.0 Dp, LT 1.5 Dp at Mp of 0.6, 0.8 and 1.0. Type of jet Mach 0.6 Mach 0.8 Mach 1.0 PCL (X/Dp) DPCL PCL (X/Dp) DPCL PCL (X/Dp) DPCL LT 0.2 Dp 5 ---- 5 ---- 5 ---- LT 1.0 Dp 2.5 50 % 2 60 % 2 60 % LT 1.5 Dp 1.6 68 % 1.6 68 % 1.2 76 % J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Shankar RKN, Raja ND 8 Shankar RKN, Raja ND Shankar RKN, Raja ND 8 Shankar RKN, Raja ND 8 The fall in static pressure below atmospheric pressure (ΔPsfall) from nozzle exit to X/Dp = 0.5 is 3%, 5.9% and 10.1% atm for primary jet exit Mach numbers 1.0, 0.8 and 0.6 respectively. The rise in static pressure above atmospheric pressure (ΔPsrise) from X/Dp = 0.5 to 2.0 is 3.3%, 6.4%, and 11% atm for primary jet exit Mach numbers 0.6, 0.8, and 1.0 respectively. The variation in slope of sinusoidal static pressure curve for primary jet exit Mach number 1.0 is more than that of Mach 0.6 and 0.8. A similar trend is observed for LT 1.5 Dp of CFJ at primary jet exit Mach numbers 0.6, 0.8, and 1.0 as shown in Fig. 12. Hence, there is considerable variation in the static pressure irrespective of the Mach numbers. Mach 0.6 Mach 0.8 Mach 1.0 1.00 0.75 0.50 0.25 0.00 0 2 4 6 8 10 X/Dp Ps/Patm Figure 12. Centerline static pressure variation for CFJ with LT 1.5 Dp. p Figure 12. Centerline static pressure variation for CFJ with LT 1.5 Dp. The variation of Mach number profoundly influences axial static pressure and it is quantitatively tabulated in Table 3 for primary jet exit Mach numbers 0.6, 0.8 and 1.0 respectively for the three different lip thickness. The variation of Mach number profoundly influences axial static pressure and it is quantitatively tabulated in Table 3 for primary jet exit Mach numbers 0.6, 0.8 and 1.0 respectively for the three different lip thickness. Table 3. Comparison of static pressure variation based on primary jet exit Mach number for different lip thickness. Lip Thickness Type of Jet Psmin ΔPsfall Psmax ΔPsrise LT 0.2 Dp Mach 0.6 3.5 2.7 ----- ----- Mach 0.8 4.5 2.9 ----- ----- Mach 1.0 6.0 3.2 ----- ----- LT 1.0 Dp Mach 0.6 0.5 3.0 2.0 3.3 Mach 0.8 0.5 5.9 2.0 6.5 Mach 1.0 0.5 10.1 2.0 11.1 LT 1.5 Dp Mach 0.6 0.7 1.2 2.7 2.3 Mach 0.8 0.7 3.3 2.7 5.1 Mach 1.0 0.7 7.7 2.7 10.8 Figure 13 shows comparison of CFJ with LT 1.0 Dp and bypass ratio 6.4 with Matsumoto et al. (1973), where bypass ratio was very high because area ratio was 363 and velocity ratio was 0.54, giving rise to bypass ratio of 196. In this study the bypass ratio is comparatively low. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 CENTERLINE STATIC PRESSURE VARIATION If Mach number of the primary jet increases, then it can cause a considerable variation in static pressure. Hence centerline static pressure variation is plotted against axial distance based on primary jet exit Mach numbers 0.6, 0.8 and 1.0, for LT 0.2 Dp, 1.0 Dp, and 1.5 Dp as shown in Figs. 9, 10 and 11 respectively. Figure 10 shows the centerline static pressure decay for LT 0.2 Dp at primary jet exit Mach numbers of 1.0, 0.8 and 0.6. Static pressure decreases gradually from the exit of the nozzle and becomes constant after an axial extent as discussed in the previous section. Hence, there is negligible variation in the static pressure irrespective of the Mach numbers. It can be clearly seen from the plot that there is no significant variation found in the near field region. Therefore, the static pressure variation due to the increase in primary jet exit Mach number does not have any noticeable influence on LT 0.2 Dp of CFJ (Matsumoto et al. 1973). Mach 0.6 Mach 0.8 Mach 1.0 1.00 0.75 0.50 0.25 0.00 0 2 4 6 8 10 X/Dp Ps/Patm Figure 10. Centerline static pressure variation for CFJ with LT 0.2 Dp. p Figure 10. Centerline static pressure variation for CFJ with LT 0.2 Dp. Figure 11 shows the centerline static pressure decay for CFJ with lip thickness 1.0 Dp at primary jet exit Mach numbers 1.0, 0.8 and 0.6. In the near field region, as the Mach number increases, rapid variation of static pressure occurs. Hence, there is considerable variation in the static pressure irrespective of the Mach numbers. Mach 0.6 Mach 0.8 Mach 1.0 1.00 0.75 0.50 0.25 0.00 0 2 4 6 8 10 X/Dp Ps/Patm Figure 11. Centerline static pressure variation for CFJ with LT 1.0 Dp. p Figure 11. Centerline static pressure variation for CFJ with LT 1.0 Dp. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Shankar RKN, Raja ND 8 Hence, there is noticeable variation in the static pressure irrespective of the Mach numbers. However, in both studies, static pressure rises and lowers along the axis and becomes constant after a far axial extend. Figure 13 shows comparison of CFJ with LT 1.0 Dp and bypass ratio 6.4 with Matsumoto et al. (1973), where bypass ratio was very high because area ratio was 363 and velocity ratio was 0.54, giving rise to bypass ratio of 196. In this study the bypass ratio is comparatively low. Hence, there is noticeable variation in the static pressure irrespective of the Mach numbers. However, in both studies, static pressure rises and lowers along the axis and becomes constant after a far axial extend. Figure 13 shows comparison of CFJ with LT 1.0 Dp and bypass ratio 6.4 with Matsumoto et al. (1973), where bypass ratio was very high because area ratio was 363 and velocity ratio was 0.54, giving rise to bypass ratio of 196. In this study the bypass ratio is comparatively low. Hence, there is noticeable variation in the static pressure irrespective of the Mach numbers. However, in both studies, static pressure rises and lowers along the axis and becomes constant after a far axial extend. Figure 13 shows comparison of CFJ with LT 1.0 Dp and bypass ratio 6.4 with Matsumoto et al. (1973), where bypass ratio was very high because area ratio was 363 and velocity ratio was 0.54, giving rise to bypass ratio of 196. In this study the bypass ratio is comparatively low. Hence, there is noticeable variation in the static pressure irrespective of the Mach numbers. However, in both studies, static pressure rises and lowers along the axis and becomes constant after a far axial extend. Figure 13 shows comparison of CFJ with LT 1.0 Dp and bypass ratio 6.4 with Matsumoto et al. (1973), where bypass ratio was very high because area ratio was 363 and velocity ratio was 0.54, giving rise to bypass ratio of 196. In this study the bypass ratio is comparatively low. Hence, there is noticeable variation in the static pressure irrespective of the Mach numbers. However, in both studies, static pressure rises and lowers along the axis and becomes constant after a far axial extend. J. Aerosp. Technol. Shankar RKN, Raja ND 8 Manag., São José dos Campos, v12, e2520, 2020 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness 9 Present Study Matsumoto et al. 1.12 1.08 1.04 1.00 0.96 0.92 0.88 0 1 2 3 4 5 6 7 8 X/Dp Ps/Patm Figure 13. Primary jet axial static pressure comparison between present study and Matsumoto et al. (1973). Present Study Matsumoto et al. 1.12 1.08 1.04 1.00 0.96 0.92 0.88 0 1 2 3 4 5 6 7 8 X/Dp Ps/Patm Present Study Matsumoto et al. igure 13. Primary jet axial static pressure comparison between present study and Matsumoto et al. (19 RADIAL STATIC PRESSURE VARIATION Radial near field static pressure variation for CFJ with LT 1.0 Dp, at Mp 1.0. 1.84e+00 1.79e+00 1.74e+00 1.69e+00 1.64e+00 1.59e+00 1.54e+00 1.49e+00 1.44e+00 1.39e+00 1.34e+00 1.29e+00 1.24e+00 1.18e+00 1.13e+00 1.08e+00 1.03e+00 9.83e+01 9.33e+01 8.83e+01 8.32e+01 Figure 18. Contours of static pressure for CFJ with LT 1.0 Dp, at Mp 1.0. 1.26e+00 1.24e+00 1.23e+00 1.21e+00 1.20e+00 1.18e+00 1.16e+00 1.15e+00 1.13e+00 1.12e+00 1.10e+00 1.09e+00 1.07e+00 1.05e+00 1.04e+00 1.02e+01 1.01e+01 9.91e+01 9.76e+01 9.60e+00 9.44e+01 Figure 16. Contours of static pressure for CFJ with LT 1.0 Dp, at Mp 0.6. X/D = 0.1 X/D =0.5 X/D =1.0 X/D =1.5 X/D = 2.0 X/D =2.5 X/D =3.0 X/D =4.0 1.2 1.1 1.0 0.9 0.8 0 5 10 15 20 R/Dp Ps/Patm Figure 17. Radial near field static pressure variation for CFJ with LT 1.0 Dp, at Mp 1.0. 1.26e+00 1.24e+00 1.23e+00 1.21e+00 1.20e+00 1.18e+00 1.16e+00 1.15e+00 1.13e+00 1.12e+00 1.10e+00 1.09e+00 1.07e+00 1.05e+00 1.04e+00 1.02e+01 1.01e+01 9.91e+01 9.76e+01 9.60e+00 9.44e+01 Figure 16. Contours of static pressure for CFJ with LT 1.0 Dp, at Mp 0.6. X/D = 0.1 X/D =0.5 X/D =1.0 X/D =1.5 X/D = 2.0 X/D =2.5 X/D =3.0 X/D =4.0 1.2 1.1 1.0 0.9 0.8 0 5 10 15 20 R/Dp Ps/Patm Figure 17. Radial near field static pressure variation for CFJ with LT 1.0 Dp, at Mp 1.0. Figure 17. Radial near field static pressure variation for CFJ with LT 1.0 Dp, at Mp 1.0. 1.84e+00 1.79e+00 1.74e+00 1.69e+00 1.64e+00 1.59e+00 1.54e+00 1.49e+00 1.44e+00 1.39e+00 1.34e+00 1.29e+00 1.24e+00 1.18e+00 1.13e+00 1.08e+00 1.03e+00 9.83e+01 9.33e+01 8.83e+01 8.32e+01 Figure 18. Contours of static pressure for CFJ with LT 1.0 Dp, at Mp 1.0. Figure 18. Contours of static pressure for CFJ with LT 1.0 Dp, at Mp 1.0. The static pressure is nondimensionalized with atmospheric pressure. It can be clearly seen that static pressure varies along the radial direction in the wake region. There are more variations in static pressure along radial direction which is due to the flow turbulence. The static pressure variation is noted to increase with the Mach number of the CFJ. On the other hand, the static pressure variation is profound only in the near field region up to an axial station of X/Dp = 4.0, for all coflowing jets analyzed in the present study. Beyond this axial station the static pressure variation becomes almost negligible. Static pressure contours obtained from numerical simulation confirms the experimental trend. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 RADIAL STATIC PRESSURE VARIATION Radial total pressure variation does not vary much with varying Mach number (Shankar et al. 2018). Figure 14 shows the experimental and CFD comparison with varying grids. This reveals that CFD work matches well with experimental work. The CFD contours were obtained from Shankar et al. (2016; 2017; 2018). The CFD study had good agreement with experimental study by the same author in the previous studies (Shankar et al. 2016; 2017; 2018). In radial direction the variation is very minimal i.e., within ± 1%. The nondimensionalized static pressure rise along radial direction is shown in Figs. 15 and 17 and corresponding contours are presented in Figs. 16 and 18 respectively. Experiment Fine grid Medium grid Coarse grid 1.00 0.75 0.50 0.25 0.00 0 5 10 15 20 X/Dp Pt/Po Figure 14. Experimental and CFD comparison of CFJ. Experiment Fine grid Medium grid Coarse grid Experiment Fine grid Medium grid Coarse grid 1.00 0.75 0.50 0.25 0.00 0 5 10 15 20 X/Dp Pt/Po Figure 14. Experimental and CFD comparison of CFJ. X/D = 0.1 X/D =0.5 X/D =1.0 X/D =1.5 X/D = 2.0 X/D =2.5 X/D =3.0 X/D =4.0 1.2 1.1 1.0 0.9 0.8 0 5 10 15 20 R/Dp Ps/Patm Figure 15. Radial near field static pressure variation for CFJ with LT 1.0 Dp, at Mp 0.6. p Figure 14. Experimental and CFD comparison of CFJ. X/D = 0.1 X/D =0.5 X/D =1.0 X/D =1.5 X/D = 2.0 X/D =2.5 X/D =3.0 X/D =4.0 1.2 1.1 1.0 0.9 0.8 0 5 10 15 20 R/Dp Ps/Patm 15 R di l fi ld t ti i ti f CFJ ith LT 1 0 D p Figure 15. Radial near field static pressure variation for CFJ with LT 1.0 Dp, at Mp 0.6. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Shankar RKN, Raja ND 10 Shankar RKN, Raja ND 10 N, Raja ND , j 10 1.26e+00 1.24e+00 1.23e+00 1.21e+00 1.20e+00 1.18e+00 1.16e+00 1.15e+00 1.13e+00 1.12e+00 1.10e+00 1.09e+00 1.07e+00 1.05e+00 1.04e+00 1.02e+01 1.01e+01 9.91e+01 9.76e+01 9.60e+00 9.44e+01 Figure 16. Contours of static pressure for CFJ with LT 1.0 Dp, at Mp 0.6. X/D = 0.1 X/D =0.5 X/D =1.0 X/D =1.5 X/D = 2.0 X/D =2.5 X/D =3.0 X/D =4.0 1.2 1.1 1.0 0.9 0.8 0 5 10 15 20 R/Dp Ps/Patm Figure 17. VELOCITY CONTOURS OF COFLOWING JET USING CFD In this section velocity contour plots of coflowing jets with different lip thicknesses and bypass ratios are plotted. These contour plots are plotted for Mach number 0.6, since higher values of Mach number do not cause much variation in the jet flow field behavior. When compared to total pressure contours, velocity contours would have a wider jet spread because of the isentropic relation that is followed in converting total pressure and static pressure in terms of Mach number. The values of velocity shown in the contour plots as shown in left hand side of Fig. 19 are in the unit of m/s. 2.20e+02 2.09e+02 1.98e+02 1.87e+02 1.76e+02 1.65e+02 1.54e+02 1.43e+02 1.32e+02 1.21e+02 1.10e+02 9.91e+01 8.81e+01 7.71e+01 6.61e+01 5.50e+01 4.40e+01 3.30e+01 2.20e+01 1.10e+01 0.00e+00 Figure 19. Velocity contour plot for coflowing jets with Lip thickness 1.5 Dp, bypass ratio 6.0 and Mach number 0.6. Figure 19. Velocity contour plot for coflowing jets with Lip thickness 1.5 Dp, bypass ratio 6.0 and Mach number 0.6. RADIAL STATIC PRESSURE VARIATION Coflowing jet with relatively high Mach number have significant variation in static pressure particularly at sonic Mach numbers, as seen in Fig. 10 (variation up to 11.1%) for lip thickness 1.0 Dp. The reason for static pressure variation along the radial direction can be explained as follows: for large values of lip thickness, huge recirculation zone will be generated just adjacent to the nozzle wall. This is because of intense turbulent mixing of two jets near the nozzle wall, leading to static pressure variation across the region (Matsumoto et al. 1973). Also, when bypass ratio and J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Mach number increase for CFJ with relatively higher values of lip thickness, recirculation zone becomes dominant, and hence static pressure varies profoundly adjacent to the flow field. When bypass ratio and Mach number are found relatively less, only marginal variation in static pressure is observed. Such characteristics can be useful in applications such as fuel air mixing in combustion chamber of aircraft (low sub sonic Mach number), ram jet (Mach number ≤ 0.3) and scramjet (supersonic Mach number), jet mixing in aircraft exhaust, mixing with bluff bodies such as afterburner. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 ACKNOWLEDGMENTS The authors thank Dr. K. Vijayaraja, Head, Department of Aeronautical Engineering, KCG College of Technology, for permitting to conduct experiments. Editors and authors are thankful to Fundação Conrado Wessel for providing the financial support for publishing this article. CONCLUSION Coflowing nozzle subjected to subsonic jet flow is studied to correlate the influence of lip thickness and Mach number over the total pressure and static pressure. It is noted that for very small lip thickness, i.e., 0.2 Dp static pressure was maintained constant along the jet. Because of this, the potential core length and decay characteristics did not vary significantly. For this case, changing the Mach number did not show any noticeable variation in the flow characteristic. Substantial influence of Mach number is noted in the Coflowing jet having finite lip thickness. In other words, the CFJ having considerable lip thickness i.e., 1.0 Dp and 1.5 Dp had noticeable variation in the total pressure and static pressure. Increasing the Mach number results in mixing enhancement along the coflowing jet. However, it shortened the potential core length by 10% for LT 1.0 Dp and 8% for 1.5 Dp, because of its ability to vary the pressure along the jet. It is observed that along the axis of the jet, static pressure varies from 3% for Mach 0.6 to 11% for Mach 1.0. Such incident is novel because the phenomenon does not occur in coflow jet with negligible lip thickness (LT 0.2 Dp), but occurs for coflowing jets with finite lip thicknesses (LT 1.0 Dp and LT 1.5 Dp). Increasing the lip thickness promotes wake dominance between primary and secondary jets. This enables better interaction along the length of the jet stream and enhances mixing. Variation in static pressure is profound with finite lip thickness. Shankar RKN, Raja ND 12 The mixing process of primary and secondary jets is being affected by wake behind the nozzle wall and turbulence intensity increases as the lip thickness increases (Shankar et al. 2016; 2017; 2018). The turbulence level in the near field is mainly due to the wake formation behind the wall which is responsible for the increase in turbulence intensity with lip thickness. The numerical work carried out in the present study is focused to get an insight on the flow field characteristics and the major research focuses on experimental study. AUTHOR’S CONTRIBUTION Conceptualization, Naren Shankar R; Methodology, Naren Shankar R; Validation, Naren Shankar R; Writing – Original Draft, Naren Shankar R and Dilip Raja N; Writing – Review and Editing, Naren Shankar R and Dilip Raja N; Supervision, Naren Shankar R and Dilip Raja N. TURBULENCE CONTOURS OF COFLOWING JET USING CFD Turbulent viscosity ratio is measured by finding the ratio of eddy viscosity (νT) to the molecular viscosity (ν). Figure 20 shows the ratio of variation of turbulent viscosity in the near field of the coflowing jet. It is observed that, for finite value of lip thickness (1.5 Dp in the present study) substantial value of turbulent intensity occurs behind the nozzle wall. 5.09e+03 4.83e+03 4.58e+03 4.32e+03 4.07e+03 3.82e+03 3.56e+03 3.31e+03 3.05e+03 2.80e+03 2.54e+03 2.29e+03 2.03e+03 1.78e+03 1.53e+03 1.27e+03 1.02e+03 7.63e+02 5.09e+02 2.54e+02 6.56e+02 Figure 20. Near field contours of turbulent viscosity ratio for lip thickness 1.5 Dp primary jet exit Mach number 0.6 and bypass ratio 6.0. Figure 20. Near field contours of turbulent viscosity ratio for lip thickness 1.5 Dp primary jet exit Mach number 0.6 and bypass ratio 6.0. J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 REFERENCES Buresti G, Talamelli A, Petagna P (1994) Experimental characterization of the velocity field of a coaxial jet configuration. Exp Therm Fluid Sci. 9(2):135-146. https://doi.org/10.1016/0894-1777(94)90106-6 Matsumoto R, Kimoto K, Tsuchimoto N (1973) A study on double concentric jets: 1st report, experimental results of air-air flow. Bulletin JSME. 16:93:529-540. https://doi.org/10.1299/jsme1958.16.529x Matsumoto R, Kimoto K, Tsuchimoto N (1973) A study on double concentric jets: 1st report, experimental results of air-air flow. Bulletin JSME. 16:93:529-540. https://doi.org/10.1299/jsme1958.16.529x Örlü R, Segalini A, Alfredsson H, Talamelli A (2008) On the passive control of the near-field of coaxial jets by means of vortex shedding. In: J. Aerosp. Technol. Manag., São José dos Campos, v12, e2520, 2020 Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness Novel Characteristics of Subsonic Coflowing Jets With Varying Lip Thickness 13 1st international conference on jets, wakes and separated flows. Berlin: Technical University of Berlin. Papamoschou D (2000) Mixing enhancement using axial flow. In: 38th Aerospace Sciences Meeting and Exhibit. Reno: Meeting Paper. 2000-0093. https://doi.org/10.2514/6.2000-93 Rathakrishnan E (2010) Applied Gas Dynamics. New Jersey: John Wiley. Rathakrishnan E (2010) Applied Gas Dynamics. New Jersey: John Wiley. 2010) Applied Gas Dynamics. New Jersey: John Wiley. Shankar RN, Thanigaiarasu S, Rathakrishnan E (2016) Numerical characterization of lip thickness on subsonic and correctly expanded sonic coflowing jets. T JPN Soc Aeronaut S. 59(3):134-141. https://doi.org/10.2322/tjsass.59.134 Shankar RN (2017) Effect of lip thickness and bypass ratio on characteristics of co flowing jets. (Doctorate Thesis) Chennai: Anna University. http://hdl.handle.net/10603/180944 Shankar RS, Thanigaiarasu S, Elangovan S, Rathakrishnan E (2018) Coflowing jet control using lip thickness variation. Int J Turbo Jet Eng. https://doi.org/10.1515/tjj-2018-0024 Shankar RN, Bennet SK (2019) Characteristics of a coflowing jet with varying lip thickness and constant bypass ratio. Aircr Eng Aerosp Tec. 91(9):1205-1213. https://doi.org/10.1108/AEAT-01-2019-0007 Shupe R, Zaman K, Papamoschou D (2007) Effect of Wedge-Shaped Deflectors on Flow Fields of Dual-Stream Jets. In: 13th AIAA/ CEAS 2007 aeroacoustics conference (28th AIAA aeroacoustics conference). Rome: Meeting Paper 2007-3659. https://doi. org/10.2514/6.2007-3659 Srinivasarao T, Lovaraju P, Rathakrishnan E (2013) Characteristics of coflow jets from orifices. Int J Turbo Jet Eng. 31(2):141-148. https://doi.org/10.1515/tjj-2013-0040 Thanigaiarasu S, Shankar RN, Rathakrishnan E (2018) Influence of bypass ratio on subsonic and correctly expanded sonic coflowing jets with finite lip thickness. P I Mech Eng G-J Aer. 233(7):2536-2548. https://doi.org/10.1177/0954410018782511 Zaman KBMQ, Papamoschou D (2000). Study of mixing enhancement observed with a co-annular nozzle configuration. In: 38th Aerospace Sciences Meeting and Exhibit. Reno: Meeting Paper. 2000-0094. https://doi.org/10.2514/6.2000-94 Zaman KBMQ, Dahl MD (2007) Noise and spreading of subsonic coannular jets-comparison with single equivalent jet. AIAA J. 45(11):2661- 2670. https://doi.org/10.2514/1.29441
https://openalex.org/W3013912386	https://www.nature.com/articles/s41598-020-62157-x.pdf	English	null	Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS)	Scientific reports	2,020	cc-by	8,483	www.nature.com/scientificreports www.nature.com/scientificreports www.nature.com/scientificreports Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) Asgeir Lande1,2, Øystein Fluge3, Elin B. Strand4,5, Siri T. Flåm1, Daysi D. Sosa6, Olav Mella3, T t i E l d2 7 Ol D S t d8 B di t A Li 1 2 7 & M t K Vik 1 7 The specificity and validity of different diagnostic criteria have been questioned, yet there is no agreement on the level of heterogeneity in ME/CFS, and there is no consensus on how to categorize different subgroups4–8. The pathogenesis and etiology of ME/CFS are unknown, with several models having been proposed9. One central hypothesis states that autoimmunity is part of the pathophysiology10,11. ME/CFS has been reported to be partly heritable12,13, consistent with a multifactorial etiology dependent on both genetic and environmental fac- tors This is the prevailing model for a vast number of diseases including established autoimmune diseases (AID) Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a disabling disorder characterized by med- ically unexplained fatigue, post-exertional malaise and a variety of additional symptoms, such as chronic pain, sleep disturbances and cognitive difficulties. ME/CFS is diagnosed on clinical grounds alone, and different sets of criteria specify the mandatory symptoms as well as recommendations for the exclusion of differential diagno- ses1–3. The specificity and validity of different diagnostic criteria have been questioned, yet there is no agreement on the level of heterogeneity in ME/CFS, and there is no consensus on how to categorize different subgroups4–8. The pathogenesis and etiology of ME/CFS are unknown, with several models having been proposed9. One central hypothesis states that autoimmunity is part of the pathophysiology10,11. ME/CFS has been reported to be partly heritable12,13, consistent with a multifactorial etiology dependent on both genetic and environmental fac- tors. This is the prevailing model for a vast number of diseases, including established autoimmune diseases (AID). Several publications report immunological alterations among ME/CFS patients, e.g. changes in natural killer (NK) cell function14,15, cytokine levels16,17, and DNA methylation patterns consistent with immune dysregula- tion18. Some of these findings have failed to reproduce in other studies, which could be due to differences in meth- odology, the complexity and heterogeneity of ME/CFS, and lack of power due to small sample sizes19. Resultingly, the autoimmunity hypothesis warrants further evaluation. A characteristic feature of AID is genetic association with certain human leukocyte antigen (HLA) alleles20. Thus, a thorough investigation of HLA associations in ME/ CFS is relevant, although HLA associations per se cannot be used as evidence regarding disease etiology21. Studies f The pathogenesis and etiology of ME/CFS are unknown, with several models having been proposed9. One central hypothesis states that autoimmunity is part of the pathophysiology10,11. Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) Asgeir Lande1,2, Øystein Fluge3, Elin B. Strand4,5, Siri T. Flåm1, Daysi D. Sosa6, Olav Mella3, T t i E l d2 7 Ol D S t d8 B di t A Li 1 2 7 & M t K Vik 1 7 Asgeir Lande1,2, Øystein Fluge3, Elin B. Strand4,5, Siri T. Flåm1, Daysi D. Sosa6, Olav Mella3, Torstein Egeland2,7, Ola D. Saugstad8, Benedicte A. Lie1,2,7 & Marte K. Viken1,7 The etiology and pathogenesis of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) are unknown, and autoimmunity is one of many proposed underlying mechanisms. Human Leukocyte Antigen (HLA) associations are hallmarks of autoimmune disease, and have not been thoroughly investigated in a large ME/CFS patient cohort. We performed high resolution HLA -A, -B, -C, -DRB1, -DQB1 and -DPB1 genotyping by next generation sequencing in 426 adult, Norwegian ME/CFS patients, diagnosed according to the Canadian Consensus Criteria. HLA associations were assessed by comparing to 4511 healthy and ethnically matched controls. Clinical information was collected through questionnaires completed by patients or relatives. We discovered two independent HLA associations, tagged by the alleles HLA-C07:04 (OR 2.1 [95% CI 1.4–3.1]) and HLA-DQB103:03 (OR 1.5 [95% CI 1.1– 2.0]). These alleles were carried by 7.7% and 12.7% of ME/CFS patients, respectively. The proportion of individuals carrying one or both of these alleles was 19.2% in the patient group and 12.2% in the control group (OR 1.7 [95% CI 1.3–2.2], pnc = 0.00003). ME/CFS is a complex disease, potentially with a substantial heterogeneity. We report novel HLA associations pointing toward the involvement of the immune system in ME/CFS pathogenesis. Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a disabling disorder characterized by med- ically unexplained fatigue, post-exertional malaise and a variety of additional symptoms, such as chronic pain, sleep disturbances and cognitive difficulties. ME/CFS is diagnosed on clinical grounds alone, and different sets of criteria specify the mandatory symptoms as well as recommendations for the exclusion of differential diagno- ses1–3. The specificity and validity of different diagnostic criteria have been questioned, yet there is no agreement on the level of heterogeneity in ME/CFS, and there is no consensus on how to categorize different subgroups4–8.h Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is a disabling disorder characterized by med- ically unexplained fatigue, post-exertional malaise and a variety of additional symptoms, such as chronic pain, sleep disturbances and cognitive difficulties. ME/CFS is diagnosed on clinical grounds alone, and different sets of criteria specify the mandatory symptoms as well as recommendations for the exclusion of differential diagno- ses1–3. Results h Characterization of patient and control groups. We included 426 ME/CFS patients and 4511 healthy, ethnically matched controls. All patients had been diagnosed in Norway according to the 2003 Canadian Consensus Criteria2, except for four patients where the similarly strict 2010 International Consensus Criteria3 were applied. Demographic and clinical characteristics of patients and controls are shown in Table 1. The mean age at diagnosis for ME/CFS patients was 34.7 years, 82.8% were female, and most patients (45.5%) had a disease duration of 5–10 years, from symptom debut to inclusion. 12.5% of the patients had severe or very severe disease (bedridden). An additional 28.6% had moderate to severe disease (strictly housebound). A total of 41.1% of ME/ CFS patients were bed- or housebound, and 86.8% of patients were unable to work full or part time the previous 6 months. HLA alleles associated with ME/CFS. In all patients and controls, we obtained 2nd field resolution geno- types of HLA class I genes HLA -A, -B and -C and class II genes HLA -DRB1, -DQB1 and -DPB1. This resolution distinguishes HLA alleles that encode specific HLA proteins. No significant deviations from Hardy-Weinberg equilibrium were noted at any HLA loci, neither in the patient group nor in the control group (Supplementary Table S1). Allele frequencies for all observed HLA Class I and Class II alleles are presented in Supplementary Table S2. Global association tests for each HLA locus (Supplementary Table S3) were significant for HLA-C (p = 0.04) and HLA-DQB1 (p = 0.04). When comparing individual allele frequencies between patients and con- trols, four HLA risk alleles emerged (Table 2): C07:04 (OR = 2.1 [95% CI 1.4–3.1], pnc = 0.0001, pc = 0.001), B57:01 (OR = 1.6 [95% CI 1.2–2.3], pnc = 0.004, pc < 0.05), DQB103:03 (OR = 1.5 [95% CI 1.1–2.0], pnc = 0.005, pc < 0.05) and B44:02 (OR = 1.3 [95% CI 1.0–1.6], pnc = 0.03, pc = n.s.). In order to evaluate the dependency of these associations, we measured the degree of linkage disequilibrium (LD) between the four alleles within the patient group (Fig. 1). Strong LD was observed between C07:04 and B44:02 (D’ = 0.90) as well as between B57:01 and DQB103:03 (D’ = 0.69), indicating that these alleles may occur on two distinct haplotypes. Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) Asgeir Lande1,2, Øystein Fluge3, Elin B. Strand4,5, Siri T. Flåm1, Daysi D. Sosa6, Olav Mella3, T t i E l d2 7 Ol D S t d8 B di t A Li 1 2 7 & M t K Vik 1 7 ME/CFS has been reported to be partly heritable12,13, consistent with a multifactorial etiology dependent on both genetic and environmental fac- tors. This is the prevailing model for a vast number of diseases, including established autoimmune diseases (AID). Several publications report immunological alterations among ME/CFS patients, e.g. changes in natural killer (NK) cell function14,15, cytokine levels16,17, and DNA methylation patterns consistent with immune dysregula- tion18. Some of these findings have failed to reproduce in other studies, which could be due to differences in meth- odology, the complexity and heterogeneity of ME/CFS, and lack of power due to small sample sizes19. Resultingly, the autoimmunity hypothesis warrants further evaluation. A characteristic feature of AID is genetic association with certain human leukocyte antigen (HLA) alleles20. Thus, a thorough investigation of HLA associations in ME/ CFS is relevant, although HLA associations per se cannot be used as evidence regarding disease etiology21. Studies 1Department of Medical Genetics, University of Oslo and Oslo University Hospital, Oslo, Norway. 2Institute of Clinical Medicine, University of Oslo, Oslo, Norway. 3Department of Oncology and Medical Physics, Haukeland University Hospital and Department of Clinical Science, University of Bergen, Bergen, Norway. 4National Advisory Unit on CFS/ME, Oslo University Hospital, Oslo, Norway. 5Faculty of Health Science, VID Specialized University, Stavanger, Norway. 6CFS/ME Center, Oslo University Hospital, Oslo, Norway. 7Department of Immunology, Oslo University Hospital, Oslo, Norway. 8Department of Pediatric Research, Oslo University Hospital, University of Oslo, Oslo, Norway. email: asgeir.lande@medisin.uio.no Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ Patients Controls Mean age, years (min, max)a,b 39.5 (17, 79) 30.6 (19, 52) Percentage femalesa,c 82.8 59.8 Mean age at diagnosis, years (min, max)d 34.7 (8, 65) — Disease duration, percentage of patientse 1–5 years 18.1 — 5–10 years 45.5 — 10–15 years 26.4 — >15 years 10.1 — Disease severity, measured by activity assessment in DSQ79, percentage of patientsf Cat.1: Bedridden 12.5 — Cat.2: Strictly housebound 28.6 — Cat.3: Light housework 45.7 — Cat.4: Able to work part time 12.2 — Cat.5: Able to work full time 0.8 — Cat.6: Handling some family obligations 0.3 — Cat.7: Handling work and family obligations 0 — Table 1. Demographic and clinical characteristics of ME/CFS patients and healthy controls. aValid number of controls: 4510. Valid number of patients: b426, c424, d345, e288, f385. DePaul Symptom Questionnaire, question no. 7949,50: Cat.1: I am not able to work or do anything, and I am bedridden. Human Leukocyte Antigen alleles associated with Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) Asgeir Lande1,2, Øystein Fluge3, Elin B. Strand4,5, Siri T. Flåm1, Daysi D. Sosa6, Olav Mella3, T t i E l d2 7 Ol D S t d8 B di t A Li 1 2 7 & M t K Vik 1 7 Cat.2: I can walk around the house, but I cannot do light housework. Cat.3: I can do light housework, but I cannot work part-time. Cat.4: I can only work part time at work or on some family responsibilities. Cat.5: I can work full time, but I have no energy left for anything else. Cat.6: I can work full time and finish some family responsibilities but I have no energy left for anything else. Cat.7: I can do all work or family responsibilities without any problems with my energy. Table 1. Demographic and clinical characteristics of ME/CFS patients and healthy controls. aValid number of controls: 4510. Valid number of patients: b426, c424, d345, e288, f385. DePaul Symptom Questionnaire, question no. 7949,50: Cat.1: I am not able to work or do anything, and I am bedridden. Cat.2: I can walk around the house, but I cannot do light housework. Cat.3: I can do light housework, but I cannot work part-time. Cat.4: I can only work part time at work or on some family responsibilities. Cat.5: I can work full time, but I have no energy left for anything else. Cat.6: I can work full time and finish some family responsibilities but I have no energy left for anything else. Cat.7: I can do all work or family responsibilities without any problems with my energy. of HLA associations in CFS have been published, but with great variation in patient inclusion criteria and HLA typing methodology22–30. No reproducible, significant associations are evident across these studies. In the largest study, including 110 patients, the strongest significant association was with HLA-DQ3 with an odds ratio (OR) of 1.8 (95% CI 1.2–2.8)22. Associations with HLA alleles DQA101, DRB113:01 and DQB106:02 have also been reported25,27,29. The great majority of these studies include less than 50 patients, and are underpowered for the detection of moderate to weak associations. Hence, in this study we aimed to conduct a comprehensive investiga- tion of HLA associations in a large ME/CFS cohort, applying modern, high resolution HLA typing. Results h most significant association, we performed Svejgaard analyses between C07:04 and B44:02 as well as between B57:01 and DQB103:03 (Supplementary Data Sheet S4). None of the two alleles on either of the two haplotypes reached significance when testing their independent association, which is not surprising due to the strong LD mentioned above. We report C07:04 and DQB103:03 as tag alleles for the ME/CFS associations, since these alleles occur at the two loci initially showing global association.f We next wanted to make sure that these HLA associations were not due to gender differences between the cases and controls (82.8% vs 59.8% females, respectively). No significant gender differences were observed between the carrier frequencies of these alleles in either cases (Table 3) or controls (C07:04 had a carrier fre- quency of 3.7% in females and 4.0% in males; p = 0.6; DQB103:03 had a carrier frequency of 9.0% in females and 8.3% in males; p = 0.4). Furthermore, after stratifying cases and controls according to gender, heterogeneity was rejected (p > 0.5) between the OR values obtained for females only and males only, indicating no gender differences between the HLA associations observed in ME/CFS. f The two loci HLA-DRB1 and HLA-DQB1 are physically close, and exhibit particularly strong LD. In our data- set, DQB103:03 occurred most frequently with DRB107:01. The haplotype DRB107:01 - DQB103:03 had estimated frequencies of 4.7% and 2.9% in the patient and control group, respectively (OR 1.7 [95% CI 1.2–2.3], pnc = 0.003). Among patients, this haplotype was without exception estimated to carry DQA102:01. Genotype data for the HLA-DQA1 locus was only obtained for patients, and was not available for the controls.h There were two alleles with a negative association with ME/CFS, suggesting a potential protection, namely B08:01 (OR = 0.7 [95% CI 0.6–0.9], pnc = 0.01, pc = n.s.) and DPB102:01 (OR = 0.7 [95% CI 0.6–0.9], pnc = 0.02, pc = n.s.) (Table 2). These alleles were not in LD (D’ = −0.29), indicating that the associations are independent. The most frequent B08:01 haplotype in Norway is the highly conserved so-called autoimmune and ancestral AH8.1 haplotype31 (C07:01-B08:01-DRB103:01-DQB102:01). This haplotype had reduced estimated fre- quency in the patient group compared to the control group (8.2% vs. 10.3%, OR = 0.8, pnc = 0.06), albeit not significantly. HLA risk allele carriers and clinical characteristics. Results h The first haplotype, C07:04 - B44:02, had an estimated frequency of 3.5% in the patient group and 1.7% in the control group, resulting in an OR of 2.1 (95% CI 1.4–3.1, pnc = 0.0002). The second haplotype, B57:01 - DQB103:03, had an estimated frequency of 3.3% in the patient group and 2.0% in the control group, resulting in an OR of 1.7 (95% CI 1.1–2.5 pnc = 0.01). To further evaluate which allele on each of the two haplotypes represents the Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ HLA allele ME/CFS, n (%) Controls, n (%) OR (95% CI) pnc pc* C07:04 33 (3.9) 172 (1.9) 2.1 (1.4–3.1) 0.0001 0.001 B57:01 39 (4.6) 259 (2.9) 1.6 (1.2–2.3) 0.004 <0.05 DQB103:03 56 (6.6) 406 (4.5) 1.5 (1.1–2.0) 0.005 <0.05 B44:02 105 (12.3) 904 (10.0) 1.3 (1.0–1.6) 0.03 0.3 B08:01 83 (9.7) 1146 (12.7) 0.7 (0.6–0.9) 0.01 0.1 DPB102:01 78 (9.2) 1067 (11.8) 0.7 (0.6–1.0) 0.02 0.1 Table 2. HLA alleles showing association (pnc < 0.05) with ME/CFS in 426 patients and 4511 healthy controls. n = number of alleles. Calculated by locus-wise Bonferroni multiple test correction. Table 2. HLA alleles showing association (pnc < 0.05) with ME/CFS in 426 patients and 4511 healthy controls. n = number of alleles. Calculated by locus-wise Bonferroni multiple test correction. Figure 1. Illustration of Linkage disequilibrium (LD) calculated as D’ values in the patient group between the four alleles with a significant association to ME/CFS. D’ measures the strength of LD between two alleles, and ranges from −1 to 0 for negative LD, and from 0 to 1 for positive LD. The number in each square is the D’ value between the two alleles listed diagonally above the square. Blue colors indicate the strength of LD, with darker colors for stronger LD. Figure 1. Illustration of Linkage disequilibrium (LD) calculated as D’ values in the patient group between the four alleles with a significant association to ME/CFS. D’ measures the strength of LD between two alleles, and ranges from −1 to 0 for negative LD, and from 0 to 1 for positive LD. The number in each square is the D’ value between the two alleles listed diagonally above the square. Blue colors indicate the strength of LD, with darker colors for stronger LD. Results h The proportion of individuals carrying the allele C07:04 was 7.7% in the patient group and 3.8% in the control group, while 12.7% of the patients and 8.7% of the controls carried DQB103:03 (Supplementary Data Sheet S4). The proportion of individuals carrying one or both of the two alleles was 19.2% in the patient group and 12.2% in the control group (OR 1.7, pnc = 0.00003, 95% CI[1.3–2.2]). Table 3 shows the distribution of clinical characteristics in the patient group, including stratification for C07:04 and/or DQB103:03. Neither gender, initiating events, comorbidity of depression or fibromyalgia, nor AID or ME/CFS among 1st degree relatives were associated with the risk alleles. However, ME/CFS patients carrying one or both of the risk alleles had a significantly higher proportion of comorbid AID (OR = 2.3 [95% CI 1.2–4.3], pnc = 0.01). Discussion h In this project, we performed high resolution HLA genotyping by next generation sequencing (NGS) in 426 adult, Norwegian ME/CFS patients, diagnosed according to the Canadian Consensus Criteria2. There are no previous publications with comprehensive HLA genotyping by NGS in this patient group. We discovered two independent HLA associations, tagged by the alleles HLA-C07:04 and HLA-DQB103:03. gg y Q To our knowledge, associations with HLA-C alleles have not previously been studied in ME/CFS. In 1994, Keller et al. performed serologic HLA-DR and DQ typing in 110 patients with Chronic fatigue immune dysfunc- tion syndrome (CFIDS)22. The patients were diagnosed with the Holmes Criteria32, and CFIDS was defined as a subgroup with positive findings in viral reactivation patterns and B- and T-cell tests, indicating post-infectious debut and a certain degree of immune dysfunction. The authors found a significant association (OR = 1.8) with the serotype HLA-DQ3. Serologic HLA typing is of low resolution compared to genetic typing33. HLA-DQ3 cor- responds to HLA-DQB103 in genetic nomenclature, where DQB103:03 is one of the three largest subgroups. Higher resolution HLA-DQB1 typing have been performed in two smaller cohorts (<58 patients), and even though statistically not significant, DQB103:03 was observed slightly more frequent among CFS patients, diag- nosed with the Fukuda criteria, than among controls23,25. Hence, the findings in existing literature is compatible with the association between ME/CFS and DQB103:03 in our material.h Q HLA-B08:01 showed reduced frequency in ME/CFS compared to controls in our material. This allele most often occur on the haplotype C07:01-B08:01-DRB103:01-DQB102:01, which was also less prevalent among ME/CFS patients in our material. This ancestral haplotype, AH8.1, is a risk factor for a wide variety of AID, including myasthenia gravis, systemic lupus erythematosus and coeliac disease31, but protective against rheuma- toid arthritis34,35. In the existing literature on HLA and CFS, HLA-DRB1 is the locus most frequently studied. In four out of five studies, the frequency of DR3/DRB103 was lower in the patient group23–25,28, while in the fifth study the frequency was similar in both groups22. Hence, this haplotype seems truly less prevalent among ME/ CFS patients.ht p Some HLA associations previously reported in CFS are not supported by our results27,29. The often cited association with DQA101 reported by Smith et al.25 cannot be evaluated in our material since the HLA-DQA1 locus was not genotyped in the control group. Results h The frequency of comorbid AID was significantly increased also when stratifying for Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ N (valid) Total DQB1 03:03 + DQB1* 03:03 - pnc OR (95% CI) C* 07:04 + C* 07:04 - pnc OR (95% CI) C07:04 + and/or DQB1 03:03 + C07:04 - and DQB1 03:03 - pnc OR (95% CI) Gender: Female 424 351 (82.8%) 44 (81.5%) 307 (83.0%) 0.7 0.9 (0.4–1.8) 25 (75.8%) 326 (83.0%) 0.2 0.6 (0.3–1.3) 65 (79.3%) 286 (83.1%) 0.3 0.7 (0.4–1.3) Symptom start following infectious episode 408 297 (72.8%) 41 (80.4%) 256 (71.7%) 0.2 1.6 (0.8–3.4) 18 (60.0%) 279 (73.8%) 0.1 0.5 (0.2–1.1) 56 (73.7%) 241 (72.6%) 0.8 1.0 (0.6–1.9) Symptom start following vaccination 407 64 (15.7%) 5 (9.8%) 59 (16.6%) 0.2 0.6 (0.2–1.4) 5 (16.7%) 59 (15.6%) 0.9 1.1 (0.4–3.0) 9 (11.8%) 55 (16.6%) 0.3 0.7 (0.3–1.4) Previous or current moderate to severe depressive symptoms 407 39 (9.6%) 5 (9.8%) 34 (9.6%) 1.0 1.0 (0.4–2.8) 2 (6.7%) 37 (9.8%) 0.6 0.7 (0.2–2.9) 7 (9.2%) 32 (9.7%) 0.9 0.9 (0.4–2.2) Comorbidity: Fibromyalgia 408 34 (8.3%) 2 (3.9%) 32 (9.0%) 0.2 0.4 (0.1–1.8) 3 (10.0%) 31 (8.2%) 0.7 1.2 (0.4–4.3) 5 (6.6%) 29 (8.7%) 0.5 0.7 (0.3–2.0) Comorbidity: Autoimmune disease 405 58 (14.3%) 10 (20.0%) 48 (13.5%) 0.2 1.6 (0.8–3.4) 9 (30.0%) 49 (13.1%) 0.01 2.9 (1.2–6.6) 18 (24.0%) 40 (12.1%) 0.01 2.3 (1.2–4.3) 1st-degree relative with autoimmune disease 408 154 (37.7%) 24 (47.1%) 130 (36.4%) 0.1 1.5 (0.9–2.8) 7 (23.3%) 147 (38.9%) 0.1 0.5 (0.2–1.1) 30 (39.5%) 124 (37.3%) 0.7 1.1 (0.7–1.8) 1st-degree relative with CFS/ME 403 59 (14.6%) 11 (22.0%) 48 (13.6%) 0.1 1.8 (0.9–3.7) 3 (10.0%) 56 (15.0%) 0.5 0.6 (0.2–2.1) 13 (17.3%) 46 (14.0%) 0.5 1.3 (0.7–2.5) Table 3. Clinical characteristics of 426 ME/CFS patients: in total and stratified according to the presence of either DQB103:03 or C07:04 or both. Table 3. Clinical characteristics of 426 ME/CFS patients: in total and stratified according to the presence of either DQB103:03 or C07:04 or both. C07:04 alone (OR = 2.9 [95% CI 1.2–6.6], pnc = 0.01), but not when stratifying for DQB103:03 alone (OR = 1.6 [95% CI 0.8–3.4], pnc = n.s.). These patients, carrying HLA risk alleles, had the following AID, ordered by fre- quency: Hashimoto’s thyreoiditis/hypothyreosis, psoriasis, rheumatoid arthritis, alopecia areata and Crohn’s dis- ease or ulcerative colitis. Results h C07:04 alone (OR = 2.9 [95% CI 1.2–6.6], pnc = 0.01), but not when stratifying for DQB103:03 alone (OR = 1.6 [95% CI 0.8–3.4], pnc = n.s.). These patients, carrying HLA risk alleles, had the following AID, ordered by fre- quency: Hashimoto’s thyreoiditis/hypothyreosis, psoriasis, rheumatoid arthritis, alopecia areata and Crohn’s dis- ease or ulcerative colitis. Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x Materials and Methodsh This study is approved by the Norwegian Regional Committees for Medical and Health Research Ethics45. All methods and data handling were performed according to relevant national and institutional regulations and guidelines. All patients gave informed consent. In three cases, written consent was given by a close relative due to the patient being severely ill and unable to sign. A total of 426 adult, Norwegian ME/CFS patients were included. All had been diagnosed in Norway according to the 2003 Canadian Consensus Criteria2, except for four patients where the similarly strict 2010 International Consensus Criteria3 were applied. There were three separate recruit- ment groups for ME/CFS patients: 214 patients were recruited from recent and ongoing trials with Rituximab46,47 and Cyclophosphamide (Rekeland IG et al., submitted, NCT02444091); 116 patients were recruited from the CFS/ME biobank at Oslo University Hospital; 96 patients were recruited via announcements in patient networks, including patient organizations. Patients from the latter two groups were not included in clinical trials. Duplicates within or between the three recruitment groups were excluded. All patients provided the identity of any 1st, 2nd or 3rd degree relatives with ME/CFS, and we excluded close relatives to ensure that only one patient per extended family was included. Norwegian ethnicity was ensured by evaluation of sur- and family names of all patients, country of birth of parents and grandparents as well as self-perceived ethnicity. The control group consisted of 4511 ethnically matched, healthy individuals drawn from the Norwegian Bone Marrow Donor Registry48. Clinical information was collected for the ME/CFS patients through questionnaires completed by patients or close relatives. The categories applied in this study were gender, age at diagnosis, initiating events, disease dura- tion and severity, comorbidities and family history. Most of the questions were based on the DePaul Symptom Questionnaire49,50. Infection or vaccination as initiating event was self-reported, and in many cases, the time from event to symptom debut was not specified. The disease severity was assessed with self-reported activity level dur- ing the previous 6 months, as stated through the DePaul Symptom Questionnaire, question no. 79. HLA genotyping by next generation sequencing. In 426 ME/CFS patients, we performed high resolu- tion, targeted, next generation sequencing (NGS) of HLA class I genes HLA -A, -B and -C and class II genes HLA -DRB1, -DQB1, -DPB1, -DQA1. Discussion h In our patient group, DQA101 occurred on haplotype with the following DQB1 alleles, ordered by frequency: DQB106:02, DQB105:01, DQB106:03, DQB106:04 and DQB105:03, and neither of these (pnc > 0.1, Supplementary Table S2), nor all combined (OR = 1.0, pnc = 0.6) were associated with ME/CFS.h The present HLA study in ME/CFS is to our knowledge the largest performed to date (other studies comprise ≤ 110 CFS patients). Our study had 80% power to discover HLA-associations with OR ≥ 1.5 given an allele fre- quency > 0.05. Interestingly, both C07:04 and DQB103:03 remained significant after Bonferroni multiple test correction. Notably, we performed locus-wise multiple test correction, i.e. correcting for the number of alleles tested at each locus, since alleles at different HLA loci are in strong LD, and therefore do not represent independ- ent tests. The Bonferroni method is considered a strict multiple test correction36, but on the other hand locus-wise correction does not take into account the lack of complete LD between the investigated loci. Taken together, our Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ results need verification in independent cohorts. In general, established HLA associations are reproducible across different populations, but susceptibility loci can also vary between populations20,21,37,38. Therefore, HLA associa- tions in ME/CFS should also be investigated in populations of different ancestry.i results need verification in independent cohorts. In general, established HLA associations are reproducible across different populations, but susceptibility loci can also vary between populations20,21,37,38. Therefore, HLA associa- tions in ME/CFS should also be investigated in populations of different ancestry.i g p pf y Both of the ME/CFS associations observed in our data set were evident at 2nd field resolution (i.e. C07:04 and DQB103:03), which distinguishes alleles encoding amino acid differences. Interestingly, the other C07 and DQB103 alleles were not associated, emphasizing the importance of high resolution HLA genotyping. Q p g p g g yp g We report C07:04 and DQB103:03 as tag alleles for two independent HLA risk associations in ME/CFS, as these alleles are in linkage equilibrium (D’ = 0.06). However, they could still be markers for either one common, or two independently associated, variants outside the loci tested in this study. Alternatively, the associated alleles reported herein could constitute a functional relevance themselves. HLA class I alleles, like C07:04, could influ- ence disease risk through their interactions with CD8 positive cytotoxic T lymphocytes20,39. Discussion h Disturbances in CD8 positive T lymphocytes have been reported in ME/CFS, although the results are somewhat conflicting10. Another important function of HLA class I alleles is to serve as ligands for NK cell receptors. Altered numbers and func- tion of NK cells have been reported by several independent researchers in ME/CFS14,15,40. The other associated allele, DQB103:03, is an HLA class II allele, which is also interesting in regard to the hypothesis of autoimmun- ity in ME/CFS. In certain well studied AID, associated HLA class II alleles have been shown to exhibit unique peptide binding properties, as well as HLA-TCR restriction, directly influencing the acquired immune response, e.g. with the production of specific auto-antibodies41. A dysregulated activity of CD4 positive T lymphocytes, the principal cell type interfering with HLA class II alleles, have been discussed as a central mechanism in ME/CFS42. Several studies report increased levels of specific auto-antibodies in ME/CFS patients43, e.g. to neurotransmitter receptors, although most of these lack verification in additional cohorts. p gi Another interesting question is whether these HLA associations are driven by subgroups of patients, and thereby representing stronger risk alleles. This is a relevant aspect in a complex disease like ME/CFS, where different causal mechanisms may be at play in different subgroups. In our study, ME/CFS patients carrying HLA risk alleles had a significantly higher comorbidity of established AID. We are not aware of any publications reporting associations between C07:04 or DQB103:03 and the AID affecting some of our patients. Therefore, it is unlikely that the associations in our material are driven by HLA associations with already established AID. Familial aggregation is observed for many specific AID, as well as for autoimmunity in general44. 1st degree relatives of ME/CFS patients in our study have a high prevalence of AID (Table 3). These observations could potentially be due to an element of autoimmunity in ME/CFS, or within a subgroup. No other patient characteristics were dominant among the ME/ CFS patients with HLA risk alleles, specifically neither self-reported infectious onset nor current disease severity. It can be argued that the lack of validity of self-reported data precludes the detection of possible subgroup identifiers.i g y p p p g pi In conclusion, we report novel HLA associations in a large cohort of ME/CFS patients fulfilling the Canadian Consensus Criteria, thereby supporting the involvement of the immune system in the ME/CFS pathogenesis. Materials and Methodsh Amplification and library preparation were performed with kits from GenDx (Utrecht, The Netherlands) and Illumina (San Diego, USA), 2 ×150 bp paired-end sequencing was performed by The Norwegian Sequencing Centre with Illumina MiSeq Reagent Kit v2 (300-cycles), and HLA genotypes were obtained by analyzing sequencing reads with NGSengine from GenDx, using the IMGT/HLA Database51. The median sequencing depth was above 150 reads per called base. The 4511 healthy Norwegian controls had pre- viously been HLA typed by NGS52. Both patient and control genotypes were analyzed at 2nd field resolution for HLA -A, -B, -C, -DRB1, -DQB1 and -DPB1. HLA alleles can be genotyped at resolution level from 1st field to 4th field. 2nd field resolution distinguishes alleles that encode amino acid differences, i.e. specific HLA proteins, and is therefore of great biological relevance. The genotyping success exceeded 99% in the patient group and 99.9% Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ in the control group for all loci. In the control group, alleles were originally identified at a G group resolution, and certain alleles from the patient group were therefore converted to avoid typing method bias (Supplementary Table S5). ) Data analyses were performed in Unphased 3.0.10 and Pypop 0.7.053,54. Assessment of Hardy-Weinberg equilibrium was performed with a chi-square test with a significance level of 0.05. Haplotype frequencies were estimated with an expectation-maximization method for unknown gametic phase. Global associations for each locus were calculated with a likelihood ratio test, with a rare allele frequency threshold of 0.01. Genetic associ- ations were investigated on allelic and haplotypic levels, and ORs with 95% confidence intervals (95% CI) were calculated with Woolf’s formula comprising Haldane’s correction. Risk allele ORs were calculated also with gen- der stratification, and homogeneity tests were performed with the logit-based estimator. LD calculations and Svejgaard tests were carried out to examine the degree of independence between the associated alleles55. The LD measure D’ was calculated according to the formula D’ = D/Dmax, where Dmax = min [pA(1-pB), (1-pA)pB] for D > 0, Dmax = min [pApB, (1-pA)(1-pB)] for D < 0, D is the standard mathematical definition of LD between alleles A and B56, and p is the frequency of the stated allele. Because of the comparison of multiple allele frequencies we performed locus-wise Bonferroni correction. Materials and Methodsh For each locus, non-corrected p-values were multiplied by the total number of alleles detected at that locus, excluding alleles with a frequency less than 3% in both the control group and the patient group. The significance level was 0.05. Only haplotypes consisting of associated alleles were inves- tigated, and multiple test correction was therefore not applied on the haplotypic level. Investigation of clinical data. The clinical information was gathered separately for each of the three recruitment groups, controlled in one common database, and exported to SPSS57 for statistical analyses. The patient group was stratified according to the presence of specific HLA alleles, and eight dichotomized clinical variables were assessed with OR calculation by binary logistic regression, and chi-square significance testing. References e e e ces 1. Fukuda, K. et al. The chronic fatigue syndrome: a comprehensive approach to its definition and study. International Chronic Fatigue Syndrome Study Group. Annals of internal medicine 121, 953–959 (1994). y y p f ( ) 2. Carruthers, B. M. et al. Myalgic Encephalomyelitis/Chronic Fatigue Syndrome. Journal of Chronic Fatigue Syndrome 11, 7–115 https://doi.org/10.1300/J092v11n01_02 (2003). p g ( ) 3. Carruthers, B. M. et al. Myalgic encephalomyelitis: International Consensus Criteria. Journal of internal medicine 270, 327–338 https://doi.org/10.1111/j.1365-2796.2011.02428.x (2011). g j 4. de Vega, W. C., Erdman, L., Vernon, S. D., Goldenberg, A. & McGowan, P. O. Integration of DNA methylation & health scores identifies subtypes in myalgic encephalomyelitis/chronic fatigue syndrome. Epigenomics 10, 539–557, https://doi.org/10.2217/epi- 2017-0150 (2018). 5. Williams, T. E., Chalder, T., Sharpe, M. & White, P. D. Heterogeneity in chronic fatigue syndrome - empirically defined subgroups from the PACE trial. Psychological medicine 47, 1454–1465, https://doi.org/10.1017/s0033291716003615 (2017).i y g p g ( ) 6. Sunnquist, M., Jason, L. A., Nehrke, P. & Goudsmit, E. M. A Comparison of Case Definitions for Myalgic Encephalomyelitis and Chronic Fatigue Syndrome. Journal of chronic diseases and management 2 (2017).f y g p g 6. Sunnquist, M., Jason, L. A., Nehrke, P. & Goudsmit, E. M. A Comparison of Case Definitions for Chronic Fatigue Syndrome. Journal of chronic diseases and management 2 (2017).f 6. Sunnquist, M., Jason, L. A., Nehrke, P. & Goudsmit, E. M. A Comparison of Case Definitions for Myalgic Encephalomyelitis and Chronic Fatigue Syndrome. Journal of chronic diseases and management 2 (2017). 7. Maclachlan, L. et al. Are current chronic fatigue syndrome criteria diagnosing different disease phenotypes? PloS one 12, e0186885 https://doi.org/10.1371/journal.pone.0186885 (2017). g y f g ( ) 7. Maclachlan, L. et al. Are current chronic fatigue syndrome criteria diagnosing different disease phenotypes? PloS one 12, e0186885 https://doi.org/10.1371/journal.pone.0186885 (2017). p g j p ( ) 8. Jason, L. A., McManimen, S., Sunnquist, M., Newton, J. L. & Strand, E. B. Clinical Criteria Versus a Possible Research Case Definition in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis. Fatigue: biomedicine, health & behavior 5, 89–102, https://doi. org/10.1080/21641846.2017.1299077 (2017).fi g 9. Griffith, J. P. & Zarrouf, F. A. A systematic review of chronic fatigue syndrome: don’t assume it’s depression. Primary care companion to the Journal of clinical psychiatry 10, 120–128 (2008). 10. Sotzny, F. et al. Myalgic Encephalomyelitis/Chronic Fatigue Syndrome - Evidence for an autoimmune disease. Autoimmunity reviews 17, 601–609, https://doi.org/10.1016/j.autrev.2018.01.009 (2018). 11. Received: 2 September 2019; Accepted: 3 March 2020; Published: xx xx xxxx Published: xx xx xxxx Restrictions on the availability of material. Individual genotypes of patients are not made available due to Norwegian privacy regulations and laws. Restrictions on the availability of material. Individual genotypes of patients are not made available due to Norwegian privacy regulations and laws. Received: 2 September 2019; Accepted: 3 March 2020; Published: xx xx xxxx Received: 2 September 2019; Accepted: 3 March 2020; Published: xx xx xxxx References The complexity and diversity of major histocompatibility complex challenge disease association studies. Hla 93, 3–15, https://doi.org/10.1111/tan.13429 (2019). p g 2. Keller, R. H. et al. Association between HLA class II antigens and the chronic fatigue immune dysfunction syndrome. Clinica infectious diseases: an official publication of the Infectious Diseases Society of America 18(Suppl 1), S154–156 (1994). f ffi p f f y f pp 23. Underhill, J. A., Mahalingam, M., Peakman, M. & Wessely, S. Lack of association between HLA genotype and chronic fatigue syndrome. European journal of immunogenetics: official journal of the British Society for Histocompatibility and Immunogenetics 28, 425–428 (2001). 4. Helbig, K. J. et al. Variation in immune response genes and chronic Q fever. Concepts: preliminary test with post-Q fever fatigue syndrome. Genes and immunity 4, 82–85, https://doi.org/10.1038/sj.gene.6363912 (2003). y y p g j g 5. Smith, J. et al. Association of chronic fatigue syndrome with human leucocyte antigen class II alleles. Journal of clinical pathology 58 860–863, https://doi.org/10.1136/jcp.2004.022681 (2005).t 26. Ledina, D. et al. Chronic fatigue syndrome after Q fever. Medical science monitor: international medical journal of experimental and clinical research 13, Cs88–92 (2007).ii , ( ) 27. Carlo-Stella, N. et al. Molecular study of receptor for advanced glycation endproduct gene promoter and identification of specific HLA haplotypes possibly involved in chronic fatigue syndrome. International journal of immunopathology and pharmacology 22, 745–754, https://doi.org/10.1177/039463200902200320 (2009). ( ) 7. Carlo-Stella, N. et al. Molecular study of receptor for advanced glycation endproduct gene promoter and identification of specific HLA haplotypes possibly involved in chronic fatigue syndrome. International journal of immunopathology and pharmacology 22 h d ( ) p g 8. Ortega-Hernandez, O. D. et al. Autoantibodies, polymorphisms in the serotonin pathway, and human leukocyte antigen class I alleles in chronic fatigue syndrome: are they associated with age at onset and specific symptoms? Annals of the New York Academy o p g 28. Ortega-Hernandez, O. D. et al. Autoantibodies, polymorphisms in the serotonin pathway, and human leukocyte antigen class II alleles in chronic fatigue syndrome: are they associated with age at onset and specific symptoms? Annals of the New York Academy of Sciences 1173, 589–599, https://doi.org/10.1111/j.1749-6632.2009.04802.x (2009). alleles in chronic fatigue syndrome: are they associated with age at onset and specific symptoms? Annals of the New York Academy o Sciences 1173, 589–599, https://doi.org/10.1111/j.1749-6632.2009.04802.x (2009). 29. Spitzer, A. R. & Broadman, M. References J. Biochem. Biotechnol. 6, 120–135 (2010). 0. Strand, E. B. et al. Comparing the DePaul Symptom Questionnaire with physician assessments: A preliminary study. Fatigue Biomedicine, Health & Behavior 4, 52–62 (2016). 51. IPD-IMGT/HLA Database, https://www.ebi.ac.uk/ipd/imgt/hla/ (2018) . IPD-IMGT/HLA Database, https://www.ebi.ac.uk/ipd/imgt/hla/ 51. IPD-IMGT/HLA Database, https://www.ebi.ac.uk/ipd/imgt/ p p g 52. Lande, A., Andersen, I., Egeland, T., Lie, B. A. & Viken, M. K. HLA -A, -C, -B, -DRB1, -DQB1 and -DPB1 allele and haplo frequencies in 4514 healthy Norwegians. Human immunology 79, 527–529, https://doi.org/10.1016/j.humimm.2018.04.012 (2 52. Lande, A., Andersen, I., Egeland, T., Lie, B. A. & Viken, M. K. HLA -A, -C, -B, -DRB1, -DQB1 and -DPB1 allele and haplotype frequencies in 4514 healthy Norwegians. Human immunology 79, 527–529, https://doi.org/10.1016/j.humimm.2018.04.012 (2018). 53. Dudbridge, F. Pedigree disequilibrium tests for multilocus haplotypes. Genetic epidemiology 25, 115–121, https://doi.org/10.1002/ gepi.10252 (2003).ht 3. Dudbridge, F. Pedigree disequilibrium tests for multilocus haplotypes. Genetic epidemiology 25, 115–121, https://doi.org/10.1002 gepi.10252 (2003).ht 53. Dudbridge, F. Pedigree disequilibrium tests for multilocus haplotypes. Genetic epidemiology 25, 115–121, https://doi.org/10.1002/ gepi.10252 (2003).ht g p 54. Lancaster, A. K., Single, R. M., Solberg, O. D., Nelson, M. P. & Thomson, G. PyPop update–a software pipeline for large-scale multilocus population genomics. Tissue antigens 69(Suppl 1), 192–197, https://doi.org/10.1111/j.1399-0039.2006.00769.x (2007). g p 4. Lancaster, A. K., Single, R. M., Solberg, O. D., Nelson, M. P. & Thomson, G. PyPop update–a software pipeline for large-scale multilocus population genomics. Tissue antigens 69(Suppl 1), 192–197, https://doi.org/10.1111/j.1399-0039.2006.00769.x (2007). S d A & d A d d d h T ( ) 55. Svejgaard, A. & Ryder, L. P. HLA and disease associations: detecting the strongest association. Tissue antigens 43, 18–27 (1994). 56. Slatkin, M. Linkage disequilibrium–understanding the evolutionary past and mapping the medical future. Nature reviews. Genetics 9, 477–485, https://doi.org/10.1038/nrg2361 (2008). , , p g g ( ) 57. IBM Corp. Released 2017. IBM SPSS Statistics for Macintosh, Version 25.0. Armonk, NY: IBM Corp. References Journal of translational medicine 10, 88, https://doi.org/10.1186/1479-5876-10-88 (2012). 1. Dendrou, C. A., Petersen, J., Rossjohn, J. & Fugger, L. HLA variation and disease. Nature reviews. Immunology 18, 325–339, https:/ doi.org/10.1038/nri.2017.143 (2018). doi.org/10.1038/nri.2017.143 (2018). 42. Brenu, E. W. et al. Immunological abnormalities as potential biomarkers in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis. J l f t l ti l di i 9 81 htt //d i /10 1186/1479 5876 9 81 (2011) doi.org/10.1038/nri.2017.143 (2018). 42. Brenu, E. W. et al. Immunological abnormalities as potential biomarkers in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis. Journal of translational medicine 9, 81, https://doi.org/10.1186/1479-5876-9-81 (2011). g 2. Brenu, E. W. et al. Immunological abnormalities as potential biomarkers in Chronic Fatigue Syndrome/Myalgic Encephalomyelitis Journal of translational medicine 9, 81, https://doi.org/10.1186/1479-5876-9-81 (2011). J f p g ( ) 43. Ortega-Hernandez, O. D. & Shoenfeld, Y. Infection, vaccination, and autoantibodies in chronic fatigue syndrome, cause or coincidence? Annals of the New York Academy of Sciences 1173, 600–609, https://doi.org/10.1111/j.1749-6632.2009.04799.x (2009). f y f p g j 44. Cardenas-Roldan, J., Rojas-Villarraga, A. & Anaya, J. M. How do autoimmune diseases cluster in families? A systematic review meta-analysis. BMC medicine 11, 73, https://doi.org/10.1186/1741-7015-11-73 (2013). y p g 5. Regional committees for medical and health related ethics, https://helseforskning.etikkom.no/prosjekterirek/prosjektregister?p_ dim=34977&_ikbLanguageCode=us&p_operation=new (Folder no: 2015/1547)i &_ikbLanguageCode=us&p_operation=new (Folder no: 2015/15 g dim=34977&_ikbLanguageCode=us&p_operation=new (Fo g g 46. Fluge, O. et al. Benefit from B-lymphocyte depletion using the anti-CD20 antibody rituximab in chronic fatigue syndr double blind and placebo controlled study PloS one 6 e26358 https://doi org/10 1371/journal pone 0026358 (2011) 46. Fluge, O. et al. Benefit from B-lymphocyte depletion using the anti-CD20 antibody rituximab in chronic fatigue double-blind and placebo-controlled study. PloS one 6, e26358, https://doi.org/10.1371/journal.pone.0026358 (2011) 46. Fluge, O. et al. Benefit from B-lymphocyte depletion using the anti-CD20 antibody rituximab in chronic fatigue syndrome. A double-blind and placebo-controlled study. PloS one 6, e26358, https://doi.org/10.1371/journal.pone.0026358 (2011). gi y p y p g y g y double-blind and placebo-controlled study. PloS one 6, e26358, https://doi.org/10.1371/journal.pone.0026358 (2011). p y p g j p 7. Fluge, O. et al. B-Lymphocyte Depletion in Patients With Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: A Randomized Double-Blind, Placebo-Controlled Trial. Annals of internal medicine, https://doi.org/10.7326/m18-1451 (2019). 48. The Norwegian bone marrow donor registry, https://oslo-universitetssykehus.no/fag-og-forskning/nasjonale-og-regionale- tjenester/det-norske-benmargsgiverregisteret (2018)hi 48. The Norwegian bone marrow donor registry, https://oslo-universitetssykehus.no/fag-og-forskning/nasjonale-og-regionale- tjenester/det-norske-benmargsgiverregisteret (2018)hi tjenester/det-norske-benmargsgiverregisteret (2018)h j g g g 49. Jason, L. A. et al. The development of a revised Canadian myalgic encephalitis chronic fatigue syndrome case definition. Am. References A retrospective review of the sleep characteristics in patients with chronic fatigue syndrome and fibromyalgia. Pain practice: the official journal of World Institute of Pain 10, 294–300, https://doi.org/10.1111/j.1533-2500.2009.00352.x (2010). ( ) 30. Pasi, A. et al. Excess of activating killer cell immunoglobulinlike receptors and lack of HLA-Bw4 ligands: a twoedged weapon in chronic fatigue syndrome. Molecular medicine reports 4, 535–540, https://doi.org/10.3892/mmr.2011.447 (2011). g y p p g 31. Candore, G., Lio, D., Colonna Romano, G. & Caruso, C. Pathogenesis of autoimmune diseases associated haplotype: effect of multiple gene interactions Autoimmunity reviews 1 29 35 (2002) 31. Candore, G., Lio, D., Colonna Romano, G. & Caruso, C. Pathogenesis of autoimmune diseases associated with 8.1 ance 31. Candore, G., Lio, D., Colonna Romano, G. & Caruso, C. Pathogenesis of autoimmune di haplotype: effect of multiple gene interactions. Autoimmunity reviews 1, 29–35 (2002).i haplotype: effect of multiple gene interactions. Autoimmunity reviews 1, 29–35 (2002).i yf g y 32. Holmes, G. P. et al. Chronic fatigue syndrome: a working case definition. Annals of internal medicine 108, 387–389 (1988). f P. et al. Chronic fatigue syndrome: a working case definition. Anna f . Holmes, G. P. et al. Chronic fatigue syndrome: a working case def i 3. Dunckley, H. HLA typing by SSO and SSP methods. Methods in molecular biology (Clifton, N.J.) 882, 9–25, https://doi org/10.1007/978-1-61779-842-9_2 (2012). g 34. Sandoughi, M., Fazaeli, A., Bardestani, G. & Hashemi, M. Frequency of HLA-DRB1 alleles in rheumatoid arthritis patients in Zahedan, southeast Iran. Annals of Saudi medicine 31, 171–173, https://doi.org/10.4103/0256-4947.78205 (2011). 35. Viatte, S. & Barton, A. Genetics of rheumatoid arthritis susceptibility, severity, and treatment response. Seminars in immunopathology 39, 395–408, https://doi.org/10.1007/s00281-017-0630-4 (2017). p g 36. Colquhoun, D. An investigation of the false discovery rate and the misinterpretation of p-values. Royal Society open scien 140216, https://doi.org/10.1098/rsos.140216 (2014). 37. Caillat-Zucman, S. Molecular mechanisms of HLA association with autoimmune diseases. Tissue antigens 73, 1–8, https://doi. org/10.1111/j.1399-0039.2008.01167.x (2009). g j 38. Zhou, F. et al. Deep sequencing of the MHC region in the Chinese population contributes to studies of complex disease. Nature genetics 48, 740–746, https://doi.org/10.1038/ng.3576 (2016).h g g g 9. Gough, S. C. & Simmonds, M. J. The HLA Region and Autoimmune Disease: Associations and Mechanisms of Action. Curren genomics 8, 453–465, https://doi.org/10.2174/138920207783591690 (2007). g p g 0. Brenu, E. W. et al. Longitudinal investigation of natural killer cells and cytokines in chronic fatigue syndrome/myalgic encephalomyelitis. References Blomberg, J., Gottfries, C. G., Elfaitouri, A., Rizwan, M. & Rosen, A. Infection Elicited Autoimmunity and Myalgic Encephalomyelitis/Chronic Fatigue Syndrome: An Explanatory Model. Frontiers in immunology 9, 229, https://doi.org/10.3389/ fimmu.2018.00229 (2018). i ( ) 2. Albright, F., Light, K., Light, A., Bateman, L. & Cannon-Albright, L. A. Evidence for a heritable predisposition to Chronic Fatigue Syndrome. BMC neurology 11, 62, https://doi.org/10.1186/1471-2377-11-62 (2011).fi y gy p g 13. Schur, E., Afari, N., Goldberg, J., Buchwald, D. & Sullivan, P. F. Twin analyses of fatigue. Twin research and human genetics: the official journal of the International Society for Twin Studies 10, 729–733, https://doi.org/10.1375/twin.10.5.729 (2007). 14. Rivas, J. L., Palencia, T., Fernandez, G. & Garcia, M. Association of T and NK Cell Phenotype With the Diagnosis of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS). Frontiers in immunology 9, 1028, https://doi.org/10.3389/ fimmu.2018.01028 (2018).f i ( ) 15. Curriu, M. et al. Screening NK-, B- and T-cell phenotype and function in patients suffering from Chronic Fatigue Syndrome. Journal of translational medicine 11, 68, https://doi.org/10.1186/1479-5876-11-68 (2013). 6. Montoya, J. G. et al. Cytokine signature associated with disease severity in chronic fatigue syndrome patients. Proceedings of the National Academy of Sciences of the United States of America, https://doi.org/10.1073/pnas.1710519114 (2017). y f f f p g p 7. Hornig, M. et al. Distinct plasma immune signatures in ME/CFS are present early in the course of illness. Science advances 1, https:/ doi.org/10.1126/sciadv.1400121 (2015).i g 8. Trivedi, M. S. et al. Identification of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome-associated DNA methylation patterns PloS one 13, e0201066, https://doi.org/10.1371/journal.pone.0201066 (2018). p g j p 19. Mensah, F. K. F., Bansal, A. S., Ford, B. & Cambridge, G. Chronic fatigue syndrome and the immune system: Where are h l l Cl l h l h //d / / l ( ) p g j p 9. Mensah, F. K. F., Bansal, A. S., Ford, B. & Cambridge, G. Chronic fatigue syndrome and the immune system: Where are we now Neurophysiologie clinique = Clinical neurophysiology 47 131–138 https://doi org/10 1016/j neucli 2017 02 002 (2017) p y g q p y gy , , p g j ( ) 20. Matzaraki, V., Kumar, V., Wijmenga, C. & Zhernakova, A. The MHC locus and genetic susceptibility to autoimmune and infectious diseases. Genome biology 18, 76, https://doi.org/10.1186/s13059-017-1207-1 (2017). Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ 1. Lokki, M. L. & Paakkanen, R. Competing interests Competing interests A.L., Ø.F., E.S., S.F., D.S., O.M., T.E., B.L. and M.V. all declare that they have no conflict of interest. O.S. is member of EMECC - European M.E. Clinicians Council; otherwise he has no conflict of interest. Competing interests A.L., Ø.F., E.S., S.F., D.S., O.M., T.E., B.L. and M.V. all declare that they have no conflict of interest. O.S. is member of EMECC - European M.E. Clinicians Council; otherwise he has no conflict of interest. Acknowledgements g We are very grateful to the patients who participated in the study, and we thank the Norwegian Bone Marrow Donor Registry for access to controls. We would also like to express our gratitude to Kari Sørland and Kristin Risa at Department of Oncology and Medical Physics, Haukeland University Hospital, Wenche Kristiansen and Hilde Haukeland at the CFS/ME Biobank, Oslo University Hospital, and employees at the Norwegian Sequencing g We are very grateful to the patients who participated in the study, and we thank the Norwegian Bone Marrow Donor Registry for access to controls. We would also like to express our gratitude to Kari Sørland and Kristin Donor Registry for access to controls. We would also like to express our gratitude to Kari Sørland and Kristin Risa at Department of Oncology and Medical Physics, Haukeland University Hospital, Wenche Kristiansen and Hilde Haukeland at the CFS/ME Biobank, Oslo University Hospital, and employees at the Norwegian Sequencing Risa at Department of Oncology and Medical Physics, Haukeland University Hospital, Wenche Kristiansen and Hilde Haukeland at the CFS/ME Biobank, Oslo University Hospital, and employees at the Norwegian Sequencing Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x www.nature.com/scientificreports/ Centre, University in Oslo and Oslo University Hospital. This study was funded by the Kavli Trust and the Research Council of Norway. We express our deepest gratitude. The funders did not influence the research or the manuscript in any way. Author contributions Ø.F., O.M., T.E., O.S., B.L. and M.V. initiated the project and applied for funding. A.L., Ø.F., E.S., D.S., O.M. and M.V. recruited and included patients. A.L., S.F. and M.V. performed sample handling, laboratory work and HLA genotyping. A.L. performed analyses and wrote the first draft of the manuscript. All authors participated in editing the manuscript and approved the final draft. Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s41598-020-62157-x. Correspondence and requests for materials should be addressed to A.L. Correspondence and requests for materials should be addressed to A.L. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Reprints and permissions information is available at www.nature.com/reprints. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. © The Author(s) 2020 Scientific Reports \| (2020) 10:5267 \| https://doi.org/10.1038/s41598-020-62157-x
https://openalex.org/W4285086360	https://dergipark.org.tr/tr/download/article-file/2539359	Turkish	null	Testing the effect of local macroeconomic indicators and global risk factors on the Turkish participation stock market: Evidence from quantile regression approach	Trends in business and economics	2,022	cc-by	9,594	doi: 10.5152/TBE.2022.1018360 doi: 10.5152/TBE.2022.1018360 Research Article Araştırma Makalesi Geliş Tarihi/Received: 03.11.2021 Kabul Tarihi/Accepted: 25.03.2022 Geliş Tarihi/Received: 03.11.2021 Kabul Tarihi/Accepted: 25.03.2022 Sorumlu Yazar/Corresponding Author: Amal ESSAYEM E-mail: amal.essayem@ogr.sakarya. edu.tr E-mail: amal.essayem@ogr.sakarya. edu.tr Cite this article as: Essayem, A., Görmüş, Ş., & Güven, M. (2022). Testing the effect of local macroeconomic indicators and global risk factors on the Turkish participation stock market: Evidence from quantile regression approach. Trends in Business and Economics, 36(3), 258-267. Anahtar Kelimeler: Küresel risk faktörleri, lokal makroekonomik göstergeler, katılım endeksi, kantil regresyon, Türkiye borsası Amal ESSAYEM1 Şakir GÖRMÜŞ2 Murat GÜVEN3 Lokal makroekonomik göstergelerin ve küresel risk faktörlerinin Türkiye katılım endeksi üzerindeki etkisinin test edilmesi: Kantil regresyon yaklaşımı 1Sakarya University, Social Sciences Institute, Research Center for Islamic Economics and Finance, Sakarya, Turkey 2Department of Public Finance, Sakarya University, Faculty of Political Sciences, Sakarya, Turkey 3Department of Statistics, Sakarya University, Faculty of Arts and Sciences, Sakarya, Turkey ABSTRACT The purpose of this study is to investigate the effect of local macroeconomic indicators and global risk factors on the participation index in the Turkish stock market from May 2011 to April 2021. Using the quantile regression approach, we detect the impact of local macroeconomic indicators and global risk factors across different market conditions: bull, bear, and normal. The empirical results demonstrate that, among local macroeconomic indicators, monetary policy-related indi- cators, Consumer Price Index (CPI)(Slope of Government Bond [SGB]), merely influence partici- pation 30 index (KAT30) return in bearish market (bullish market); however, credit default swap negatively affects KAT30 return across all quantiles. When it comes to global risk factors, results show that KAT30 return is negatively affected by the implied volatility index across all quantiles except Q0.75 and Q0.95. This means that the implied volatility index impact on KAT30 return is stronger during the bearish market. Yet, Oil Volatility Index (OVX) and Morgan Stanley Country Index (MSCI) positively impact the index return across upper quantiles. Keywords: Global risk factors, local macroeconomic indicators, participation index, quantile regression, Turkish stock market Cite this article as: Essayem, A., Görmüş, Ş., & Güven, M. (2022). Testing the effect of local macroeconomic indicators and global risk factors on the Turkish participation stock market: Evidence from quantile regression approach. Trends in Business and Economics, 36(3), 258-267. ÖZ Bu çalışmanın amacı, Mayıs 2011-Nisan 2021 döneminde Türkiye hisse senedi piyasasında lokal makroekonomik göstergelerin ve küresel risk faktörlerinin katılım endeksi üzerindeki etki- sini araştırmaktır. Kantil regresyon yaklaşımı kullanarak, lokal makroekonomik göstergelerin ve küresel risk faktörlerinin düşen(bear), normal, yükselen(bull) piyasa koşullarındaki etkisini tespit etmekteyiz. Ampirik sonuçlar, lokal makroekonomik göstergeler arasında para politikasıyla ilgili göstergelerden CPI(SGB)’nin sadece düşüş piyasasında (yükseliş piyasası) KAT30'u etkilediğini gösterirken, CDS tüm kantillerde KAT30'u negatif etkilemektedir. Küresel risk faktörleri söz konusu olduğunda, sonuçlar KAT30'un Q0.75 ve Q0.95 dışındaki tüm kantillerde VIX'ten negatif etkilendiğini göstermektedir. Bu durum, düşüş piyasasında VIX’in KAT30 üzerindeki etkisinin daha güçlü olduğu göstermektedir. Bununla birlikte, OVX ve MSCI'nin etkisi, üst kantillerde pozitif ve anlamlıdır. Introduction To the extent of our knowledge, there is no research investigat- ing the impact of macroeconomic indicators and global risk fac- tors (GRF) on the participation of the stock index in Turkey. We extend two empirical works, which are related to our paper. Nai- far (2016) examines the co-movement along with the impact of macroeconomic indicators, risk factors, and global financial mar- ket conditions on the global Islamic stock market indices. Bahloul and Ben Amor (2020) explore the effect of local macroeconomic indicators (LMIs) with global factors on the Middle East and North African (MENA) region stock returns (SR hereinafter). Hence, our study probes the effect of global and LMI on the Turkish participa- tion index by employing the quantile regression approach (QR-A) for monthly data from April 2011 to April 2021. The contribution of this study is four-fold. First, we check the effect of LMI and GRF on the participation index in the Turkish stock market. Second, using QR-A, we detect the impact of local and global factors across different market conditions: bull, bear, and normal. Third, looking for the relationship between MSCI Global and the Participation index, we check for potential diversi- fication opportunities. Fourth, we use the lagged oil price volatil- ity index, which has not been used in the existing state of the art on Islamic equity return dynamics. To this extent, we should mention that the existence of shariah- compliant stock market indices gives religiously conscious inves- tors an alternative investment scheme in line with the Shariah principles and conventional investors a diversification invest- ment opportunity (Arshad et al., 2013; Fatima et al., 2019; Jawadi et al., 2018; Usman et al., 2019). The rest of the study is structured as follows: the second section presents the theoretical framework and the existing literature. The third section illustrates the quantile regression methodology and the fourth section illustrates data and preliminary statistics. The fifth and sixth sections provide estimation results and the discussion and the conclusion, respectively. According to The Capital Asset Pricing Model (CAPM), systematic risk factors have an effect on stock market returns. In fact, quan- tifying systemic risk in emerging stock markets is challenging since they are passing through a process of financial integration which makes them exposed to a set of local and global factors. Therefore, we need to consider these factors to correctly esti- mate stock market returns (Bahloul & Ben Amor, 2021). Introduction Content of this journal is licensed under a Creative Commons Attribution 4.0 International License. Content of this journal is licensed under a Creative Commons Attribution 4.0 International License. For the last couple of years, the Islamic capital market has been under the spotlight. It has been gaining momentum in both academia and the financial industry. This rapid growth of the Islamic stock market 259 Naifar, 2016). Yet, these studies intend to study the global partici- pation indices and do not provide country-specific implications. Naifar, 2016). Yet, these studies intend to study the global partici- pation indices and do not provide country-specific implications. accompanied by the increasing concern of religiously conscious investors to invest according to the norms of shariah motivated investors to shift their funds to markets providing participation (Islamic) indices. Over the last decade, the Turkish stock market has been experiencing exponential growth in shariah-compliant investments and continues to grow at a steady pace (Bayram & Abdullah Othman, 2019). The literature related to participation indices has two shortcom- ings. First, global factors have not been incorporated in studies related to the Turkish participation index. Second, previous stud- ies only investigate average dependence without considering extreme market conditions. For instance, in 2011, the Participation 30 Index was launched by the Turkish stock market. This index includes shares of the top 30 companies with the highest public market capitalization traded at Borsa Istanbul, which is at the same time shariah compliant. Participation indices in Turkey, like all Islamic Market indices, have been formed according to certain rules. Generally, Islamic indi- ces construction is based on Shariah screening detailed in stan- dard 21 published by The Accounting and Auditing Organization for Islamic Financial Institutions (AAOIFI). This means that listed companies are strictly banned from investing in projects involv- ing interest-based activities, alcohol, gambling, pork, entertain- ment, etc. Moreover, companies are required to respect some financial ratios to fulfill the shariah screening requirements. Both the ratio of total interest-bearing debt of companies to market capitalization and the ratio of interest-taking deposits and secu- rities to market capitalization must be lower than 30%, while the ratio of income from illicit projects stated above to global income must be less than 5% (AAOIFI, 2015). Introduction Although the Islamic stock market has been known for being decoupled from the conventional market because of the strict compliance rules, literature shows that it is integrated with the global market (Balcilar et al., 2015). Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 Stock Returns and Global Risk Factorsi Since the 2008 financial crisis, the correlation between the SR around the world has increased revealing the existence of conta- gion effects (Haddad et al., 2020). Taking into consideration the changes in stock market linkages studies, recent studies start to rely heavily on GRF to detect dependencies and shed light on international diversification opportunities (Bahloul & Ben Amor, 2021; Balcilar et al., 2015; Haddad et al., 2020; Hammoudeh et al., 2014; Lin & Su, 2020; Naifar 2016; Yarovaya et al., 2021). For instance, the impact of inflation on SR has been documented by Fisher (1930) stating that there is a positive link between infla- tion and stock market. However, Fama (1990) states that infla- tion negatively impacts SR arguing that the inflation harms real economic activity. Naifar (2016) investigates the impact of local and GRF on the Global Dow Jones Islamic index by applying quan- tile regression model for the period between January 2003 and October 2014. Results indicate that inflation impacts Islamic stock return for the intermediate and upper quantiles; yet, it was insignificant for lower quantiles. Bahloul and Ben Amor (2020) conduct similar research for MENA countries for the period from January 2007 to January 2018 using quantile regression. Their study finds a negative relationship between inflation and stock performance for lower quantiles for most MENA countries. Among the commonly used global factors, the implied volatility index (VIX) is a proxy to quantify stock market uncertainty. Robert Whaley, the founder of VIX, argues that the VIX increases during market turmoil. If the expected VIX increases (decreases), inves- tors expect higher (lower) returns (Whaley, 2009). Moreover, VIX index is computed employing the implied volatilities on S&P 500 index options (SPX). The index stands for the market’s expec- tation of the S&P 500 index’s 30-day future volatility. When it comes to its influence on SR, VIX impact can be explained as follows: when option premiums are high, the expected future volatility rises, and investor’s fear increases which will cause a decrease in equity index return (Naifar, 2016). Nevertheless, empirical studies present different results regarding this matter for the Islamic Equity market. Hadad et al. (2020) investigate the time-varying impact of GRF on the return of Islamic indices and volatility spillovers for seven DJIM from April 2003 to November 2018. They found that the seven DJIM stock indices are weakly associated with the movement of global risk factors. Literature Review According to the CAPM, the link between stock’s expected returns and systematic risk factors is predictable. The model sug- gests that market integration implies that the return of assets belonging to the same class is supposed to be similar, disregard- ing the transaction’s location. Thus, it is required to identify the link between the expected returns and the affiliated risks in order to determine potential excess returns that the investor might gain. As mentioned above, studies have been examining the connection between the Islamic stock market and its conventional counter- part. The literature conglomerates a set of researchers studying a wide range of subjects. Several studies have been focusing on the risk–return analysis and comparison to determine the exist- ing link between diversification opportunities and performance of the Islamic financial market focusing on its performance com- pared to the conventional market pre-, throughout, and/or post- financial crisis or checking the decoupling hypothesis (Ahmed & Farooq, 2018; Ali et al., 2018; Al-Khazali et al., 2014; Aloui et al., 2016; Camgöz et al., 2019; Paltrinieri et al., 2019; Safiullah & Shamsuddin, 2021; Trabelsi & Naifar, 2017). A different set of stud- ies highlight the determinants affecting the Islamic stock market (Chau et al., 2014; Dewandaru et al., 2014). The recent state of the art examines the dynamics of Islamic equity indices and provides empirical results for the dependence structure and the co-move- ment between global risk factors, local risk factors, and Islamic stock indices (Haddad et al., 2020; Hammoudeh et al., 2014; Having said that, quantifying risk in emerging stock markets is much more complicated. Bekaert and Harvey (1995) state that the majority of emerging markets are comparatively segmented. In addition, Guesmi and Nguyen (2011) argue that despite the general increase in their integration level overtime, the emerging markets are characterized by a higher segmentation from the global market. This means that financial markets follow a gradual integration process. Thus, investors face both GRF and local risk factors. In this regard, expected returns are captured by analyzing both global and local risk factors. Thus, the level of financial integration of an emerging stock market is explained by its related local and GRF (Bahloul & Ben Amor, 2021). 1 Please check Table 1 for Literature review summary. Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 Literature Review Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 260 predictor of future recession in the real economy (Estrella & Mish- kin, 1998). This is elucidated by the expectation theory of interest rate term structure which states that a long-term rate is in fact a term premium plus the mean of future values of short-term (ST) interest rates. If investors expect a market downturn, the future values of ST rate would decrease as economic agents expect a relaxed monetary policy. This means that the expectation of lower future ST interest rates reduces longer-term rates. Hence, the term spread increases. Bhaduri and Saraogi (2010) examine the link between the yield spread and Indian SR by employing data from May 1996 to May 2008. They conclude that the yield spread can accurately identify entry and exit points for the Indian stock market. Resnick and Shoesmith (2002) also find similar results for the US stock market. Moreover, Naifar (2016) states that while ST interest rates are governed by central banks, LT interest rates depend on market forces, and the slope of the term structure can track embedded term risk premiums. Thus, this spread is con- sidered the tracker of investors’ compensation for bearing inter- est rate risk. Naifar (2016)’s empirical results show a positive link between the slope of the yield curve and DJIMI returns. Yet, the Islamic stock markets are frequently decoupled from the conventional stock market due to the tight constraints on the former. The core of this dichotomy lies in the rules that stipulate Islamic investments code (Balcilar et al., 2015). In fact, Islamic indices are designed to meet shariah requirements, and the funds’ management is heavily regulated to eliminate market manipulations and price gouging (Ho et al., 2014). However, globalization and the recurrent re-structuring of the financial markets usher in the reshaping of the international financial market (Delle Foglie & Panetta, 2020). This might raise the question of whether the Islamic financial markets dynamic is interlinked with the conventional market conditions or indepen- dently moving. Our study aims to answer the following questions: Do global macroeconomic indicators and GRF affect the Turkish participation in SR? Does participation in stock market indices provide an alternative investment scheme, considering the par- ticular impact of emerging market integration on portfolio diver- sification and hedging strategies?1 Stock Returns and Local Macroeconomic Indicators The impact of macroeconomic factors on stock market dynam- ics is a well-documented subject in the literature. Most studies present empirical evidence for the impact of common macro- economic indicators, especially monetary policy indicators on SR (Fama, 1981, 1990; Mun, 2012; Fama, 1990; Murthy et al., 2016). Stock Returns and Global Risk Factorsi (2021) study the effect of gold, oil, bitcoin, VIX, and economic policy uncertainty (EPU) indices on the return of Islamic and conventional stock market during the coronavirus disease-2019 pandemic outbreak using the Vector Autoregressive Moving Average-Baba Engel, Kraft & Kroner-Asymmetric Generalized Conditional Heteroscedastic- ity (VARMA-BEKK-AGARCH) approach on daily data from April 11, 2019, to May 4, 2020. The study shows a negative link between the returns of the Islamic stock market and the VIX. However, Hadad et al. (2020) show that Islamic equity markets are known for lower sensitivity to global risk shocks including VIX shocks. This result supports the claim stating that sharia-compliant assets have a higher resilience to prevalent risk shocks. In addition, Banerjee et al. (2007) and Giot (2005) found a positive link between stock market performance and VIX arguing that an increased VIX index reflects an oversold market. a negative relationship between oil prices and SR (Driesprong et al., 2008; Filis, 2010; Kling, 1985; Sadorsky, 1999), recent papers found a positive relationship (Narayan & Narayan, 2010; Zhu et al., 2016). A different strand of the literature analyzes the linkage between crude oil and Islamic indices (Chang et al., 2020; Ftiti & Hadhri, 2018; Ghorbel et al., 2014; Mezghani & Boujelbène, 2018). Chang et al. (2020) studied the asymmetric effects of oil price on the DJIM and 10 sectorial stock indices by applying quantile-on-quantile approach using daily data over the period January 1996 to April 2019. They found that lower/upper quan- tiles of oil prices negatively impact the upper/lower quantiles of the DJIM. A similar study by Lin and Su (2020) found a negative link between oil market uncertainty and Islamic SR, especially when the Islamic stock market is bearish (low quantiles), using a quantile-on-quantile approach. Their results are in line with the findings of Yarovaya et al. (2021). However, Jawad et al. (2018) analyzed the link between Islamic stock markets and oil prices. They used a copula approach and three derivations of the value at risk measure: Conditional Value at Risk (CoVaR), Delta CoVaR, and VaR. Their results show a positive dependence between oil prices and Islamic SR. This is in line with the study by Naifar (2016) and Hammoudeh et al. (2014). Another common global risk factor is the EPU. Economic pol- icy uncertainty is a US news-based uncertainty index that has become a trending factor used in recent studies. Turkish Stock Returns and Local Macroeconomic Indicators– Global Risk Factors Regarding the Turkish context, few studies have analyzed the effect of local and global factors on the Turkish stock market. Çonkir et al. (2021) analyzed the effect of the VIX index on The Borsa Istanbul Index (BIST) 30 by applying the Granger causality test and VAR model using monthly data from 2015 to 2019. They found one-way causality between VIX and BIST 30. A similar result has been found by Kaya and Coşkun (2015) who applied Granger causality and regression analysis on daily data from 1995 to 2014. They conclude that VIX negatively affects BIST 100. In addition, Kaya (2015) checks for the cointegration between BIST 100 and VIX using daily data for the period from 2009 to 2013. He con- cludes that a long-term relationship exists between the BIST 100 and VIX. This is in line with the results of Öner, İçellioğlu, and Öner (2018) finding a one-way relationship of the long and short hori- zons between VIX and BIST 100. Furthermore, Tursoy and Faisal (2018) investigate the long- and short-term dynamics between Turkish stock prices, gold prices, and crude oil prices by applying Autoregressive Distributed Lag Model (ARDL) model to monthly data from January 1986 to November 2016. Their results show that a positive link exists between crude oil and stock prices. In addition, Erdoğan et al. (2021) use the Nonlinear Autoregressive Distributed Lag Model (NARDL) framework to look at the short- run and long-run asymmetric effects of risk factors on SR in Tur- key using data from 1997 to 2020. Their results show that the Turkish stock market is more reactive to bad news transmitted/ channeled through global factors than those transmitted by local macroeconomic variables. They claim that an increase in EPU One of the prominent GRF in the literature is oil price. The grow- ing financialization of oil price has made it a key factor in portfolio strategies (Lin & Su, 2020). Henriques and Sadorsky (2011) state that financial characteristics of oil price cause extreme fluctua- tions. Thus, oil price volatility is one of the important variables when constructing a portfolio strategy. According to the theory of equity valuation, the stock price is considered as an aggregate of the discounted values of expected future cash flows regard- ing distinct investment perspectives. Stock Returns and Global Risk Factorsi Economic policy uncertainty is perceived as a proxy for the expectations of investors with regard to U.S. government’s future response to economic problems (Liang et al., 2020). The empirical literature shows that EPU has a considerable impact on stock return. From the theoretical perspective, it has been stated that EPU has a negative impact on the SR (Liang et al., 2020). However, empiri- cal results have been divergent, while Christou et al. (2017), Donadelli (2015), and Lean and Nguyen (2014) indicate a positive linkage between EPU and SR. Brogaard and Detzel (2015), Kang et al. (2016), and Bahloul and Ben Amor (2020) found a negative linkage. When it comes to Islamic stock indices, Hammoudeh et al. (2016) analyzed the impact of US federal funds rates and EPU on the indices of the Islamic, US, European, and Asian stock markets using the Structural Vector Error Correction Model (SVEC) and the Structural Vector Autoregressive Model (SVAR) models on daily data from 1999 to 2013. They found a negative impact of EPU on all indices. Their result is in line with Yarovaya et al. (2021). However, Hammoudeh et al. (2014), Nazlioglu et al. (2015), and Ftiti and Hadhri (2018) found that Islamic SR are unaffected by EPU shocks. The MSCI world index is employed as an estimator to check for the integration of the local stock market with the global stock market. Jouini (2013) investigates the effect of MSCI on SR of Gulf Cooperation Council (GCC) countries. He used the VAR and the VEC models using weekly data from 2005 to 2008 and found a positive link between MSCI and Bahrain, Kuwait, and Oman mar- kets. Yet, he found that MSCI negatively affects Saudi Arabia, Qatar, and UAE markets. Bahloul and Ben Amor (2020) found a positive link between SR and MSCI for MENA countries decreas- ing from lower to upper quantiles. Stock Returns and Global Risk Factorsi Another local indicator affecting the SR is the country’s specific credit default swap (CDS). Coronado, Corzo, and Lazcano (2012) state that credit risk information is embedded in the asset’s pay- off. They explain that if countries dedicate a large sum of their income to external debt, governments will dedicate a lower sum for investment that will affect value creation in the economy and sub- sequently, its future growth. In this case, consumption and invest- ment decrease, leading to lower revenue for companies, which will negatively affect their stock prices and thus the returns. Nai- far (2016) finds that the US sovereign 5-year CDS index positively affects Dow Jones Islamic Market Index (DJIMI) returns among quantiles throughout and post the global financial crisis. Han et al. (2017) investigate the relationship between equity market and credit market taking into consideration the information embedded in the slope of credit spread, the latter proxied by the firm’s CDS term structure using monthly data of 776 US-based firms between August 2002 and December 2012. Their results show that the slope of the CDS negatively affects cross-sectional SR. Paltrinieri et al. (2018) investigate the repercussions of oil prices and VIX index on 17 socially responsible investment, conven- tional, and Islamic stock indices using DCC (Dynamic conditional correlation)-GARCH model for the period 2005–2015. They found a negative linkage between the VIX on both conventional and Islamic index. Their findings are similar to those of Naifar (2016), who mentioned that the DJIM index returns and VIX index pres- ent an asymmetric correlated movement (trend) as the degree of dependence increases during a bullish market, yet depen- dence weakens during a bearish market. Furthermore, Ham- moudeh et al. (2014) analyze the relationship between the DJIM and global equity indices of the United States, Asia, and Europe To predict future real economic activity, the term structure is used. In fact, the term spread is considered an important 261 by applying a copula approach to data between January 4, 1999, and July 22, 2013. Authors employ a set of risk factors such as VIX and oil. Their results reveal a negative link between the DJIM and each of the VIX. Moreover, Yarayova et al. Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 Data Description and Preliminary Statistics The data set employed in this study consists of the monthly data for the Turkish participation stock index returns represented by (KAT30) for the period from April 2011 to April 2021. Prior to 2011, some of the data were missing; furthermore, the period is a decade characterized by the European debt crisis (the crisis peaked between 2010 and 2012) and drastic political changes and volatility in the region—for instance the Arab Spring—that caused the decline of the stock markets. This period reflects the impact of these crises on the dynamics of the Turkish stock mar- ket return over time. In this model, i and j denote the LMI and global risk factors. α0 and εt indicate intercept and disturbance. βi represents the sensitivity of the participation index return in the Turkish stock market to the ith LMI and δ j shows the sensitivity of the partici- pation index return in the Turkish stock market to the jth global risk factors. We provided the description of the variables as follows: PIRt: Participation index return in the Turkish stock market at time t; PIRt: Participation index return in the Turkish stock market at time t; For local factors, we employed three macroeconomic indicators which are inflation rate (CPI), sovereign credit risk (CDS), and the term structure slope (SGB). SGB shows the difference between Tur- key’s 10-year Treasury bond rates and 3-year Treasury bond rates. LMIt: The realization of the ith LMI in Turkey at time t; LMIt: The realization of the ith LMI in Turkey at time t; GRFt: The realization of the jth GRF in Turkey at time t. GRFt: The realization of the jth GRF in Turkey at time t. For global factors, we utilized the US VIX, EPU, MSCI world index, and the CBOE oil volatility index OVX.2 To estimate this model, we used the quantile regression model developed by Koenker and Bassett (1978). The QR-A differs from OLS since it considers not only the mean value but also extreme values in a data set. Therefore, QR-A presents a broad relationship between regressed and regressors in a regression model (Koen- ker, 2005). In Table 2, we provide the preliminary statistics of the variables. Table 2 shows that KAT30 presents a mean of 2192,23 points and a standard deviation of 410,25 points. Turkish Stock Returns and Local Macroeconomic Indicators– Global Risk Factors This means that the stock price incorporates a set of macroeconomic events and indica- tors, such as inflation, market participants’ trust, and oil shock (Hedi Arouri & Khuong Nguyen, 2010; Jouini, 2013). Furthermore, volatile oil prices may cause a low-risk premium, which might negatively affect cash flows and thus SR (Chang et al., 2020). To put it in a nutshell, the fluctuations in the oil market may pass to the stock market by affecting investors’ expectations as the oil market can mirror the situation of the global economy (Liu & Chen, 2017; Zhu et al., 2016). Under the conventional stock mar- ket, mixed results have been found. While previous papers found 262 τth conditional quantile of the dependent variable y. Therefore, if the values of ( ) remain the same among the quantiles, the structure of dependence is named to be constant. If these val- ues increase/decrease among the quantiles, the structure of dependence is named to be increasing/decreasing. If these val- ues become similar/dissimilar for high and low quantiles, the structure of dependence is named to be symmetric/asymmetric (Koenker, 2005; Naifar, 2016). It is also noted that there is condi tional/unconditional dependency if there is an exogenous/no exogenous variable in x. depresses real SR and an increase in oil price leads to a decrease in the SR. Lin and Su (2020) analyzed the relationship between the Islamic stock market and oil market uncertainty (OVX) by applying quantile-on-quantile approach to four countries includ- ing Turkey using daily data from 2013 to 2019. They conclude that OVX variations impact asymmetrically the Turkish participation index. They found that the estimated coefficients are negative. As shown above, none of the previous studies has analyzed the impact of local and GRF on the Turkish participation SR across the changing market conditions. For a given τ, the coefficients of independent variable x can be obtained by minimizing the weighted absolute difference between dependent variable y and independent variable x as: Methodology Following Bahloul and Ben Amor (2021) and Naifar (2016), we analyzed the relative effect of LMI and GRF on the participation index return. For this purpose, we used the following model. ˆ( ) argmin ( ) ( ) ( ) 1 1 y n n n x n n Y x PIR LMI GRF t i i N it j j M jt t 0 1 1 (1) (1) Data Description and Preliminary Statistics The world index (MSCI) records the highest average compared to the rest of the vari- ables (43864,73), with the highest standard deviation (8515,01). A positive skewness characterizes our data set. The kurtosis sta- tistic exceeds 3 (except for MSCI), with SGB and OVX scoring the highest kurtosis levels. The Jarque–Bera test shows that the null hypothesis is rejected for most series (except for KAT30), which means that the distribution is not normal. The quantile regression by Koenker and Bassett (1978) is shown as: The quantile regression by Koenker and Bassett (1978) is shown as: Q x x y( ) ( ) (2) (2) In this model, it is assumed that the dependent variable y is lin- early dependent on x. Q x y( ) τ indicates the conditional quantile of y. In this model, it is assumed that the dependent variable y is lin- early dependent on x. Q x y( ) τ indicates the conditional quantile of y. Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 MSCI, VIX, and EPU data are collected from DataStream, yahoo finance, and economicuncertainty.com, respectively. The rest of the Results We applied QR-A to capture the dynamics of KAT30 among all the quantiles. For this purpose, we first controlled the stationarity of the variables using unit root tests: Augmented Dickey–Fuller (ADF) test by Dickey and Fuller (1979) and Philips–Perron (PP) test by Phillips and Perron (1988). Table 3 and Table 4 show the unit root test results. Augmented Dickey–Fuller and PP indicate the rejection of the null hypothesis of the unit root for all variables at level except for KAT 30, CPI, EPU, and CDS which are stationary at the first difference. These results present that all variables are stationary (at different levels). We estimated our model by apply- ing a quantile regression model. Q x a F a x x y y k k k ( ) inf ( / ) ( ) (3) (3) The dependent variable y has a conditional probability distribu- tion function on x denoted by F a x y( / ) . In our study, we use LMI (CPI, SGB, CDS) and GRF (EPU, VIX, MSCI, OVX) to identify the determinants of participation index return (KAT30) in the Turkish stock market. We investigated KAT30 among the seven quantiles (0.05, 0.1, 0.25., 0.5, 0.75, 0.90, 0.95). For the quantile regression model, ( ) quantifies the degree of dependence between x and The dependent variable y has a conditional probability distribu- tion function on x denoted by F a x y( / ) . In our study, we use LMI (CPI, SGB, CDS) and GRF (EPU, VIX, MSCI, OVX) to identify the determinants of participation index return (KAT30) in the Turkish stock market. We investigated KAT30 among the seven quantiles (0.05, 0.1, 0.25., 0.5, 0.75, 0.90, 0.95). For the quantile regression model, ( ) quantifies the degree of dependence between x and Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 263 Table 1. Literature Review Matrix Reference Market Methodology Result Naifar (2016) Global Quantile regression Inflation impacts Islamic stock return of the intermediate and upper quantiles Bahloul and Ben Amor (2020) MENA countries Quantile regression Negative relationship between inflation and stock performance for lower quantiles Naifar (2016) Global Quantile regression The U.S. sovereign 5-year CDS index positively affects Dow Jones Islamic Market Index (DJIMI) Han et al. Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 Results (2017) United States A combination of empirical methods The slope of the CDS negatively affects cross-sectional SR Naifar (2016) Global Quantile regression Positive link between the slope of the yield spread and DJIMI returns Paltrinieri et al. (2018) Asia-Pacific, Europe, the United States, UAE DCC-GGARCH model Negative linkage between the implied volatility index (VIX) on both conventional and Islamic index Naifar (2016) Global Quantile regression Negative link between the VIX and DJIMI returns Hammoudeh et al. (2014) United States, Asia, and Europe Copula approach Negative link between the DJIM and each of the VIX Yarovaya et al. (2021) Global VARMA-BEKK-AGARCH approach Negative link between the returns of Islamic stock market and the VIX Hammoudeh, et al. (2016) United States, Europe, and Asia SVEC and SVAR models Negative impact of EPU on all indices Reference Market Methodology Result Chang et al. (2020) Global Quantile-on-quantile approach Lower/upper quantiles of oil prices negatively impact the upper/ lower quantiles of the DJIM Lin and Su (2020) Canada, Japan, Turkey, and Kuwait Quantile-on-quantile approach Negative link between oil market uncertainty and Islamic SR Yarovaya et al. (2021) Global VARMA-BEKK-AGARCH approach Negative link between oil market uncertainty and Islamic SR Jouini (2013) GCC countries VAR and the VEC models Positive link between MSCI and Bahrain, Kuwait, and Oman markets MSCI negatively affects Saudi Arabia, Qatar, and UAE markets Bahloul and Ben Amor (2020) MENA Quantile regression Positive link between SR and MSCI for MENA countries decreasing from lower to upper quantiles Tursoy and Faisal (2018) Turkey ARDL framework Positive link exists between crude oil and stock prices Erdoğan et al. (2021) Turkey NARDL framework Increasing EPU depresses real SR and an increase in oil price leads to a decrease in the SR Lin and Su (2020) Canada, Japan, Turkey, and Kuwait Quantile-on-quantile approach OVX negatively impacts the Turkish participation index Note: MENA = Middle East and North African; CDS = credit default swap; VIX = implied volatility index; EPU = economic policy uncertainity. Table 5 presents the quantile regression results and illustrates the effect of local and global factors on the bearish market (Q0.05, Q0.1, Q0.25), normal (Q0.5), and the Bullish market (Q0.75, Q 0.9, Q0.95) (see Kangalli Uyar et al., 2021; Naifar, 2016; Zhu et al., 2016). Results Table 5 presents the quantile regression results and illustrates the effect of local and global factors on the bearish market (Q0.05, Q0.1, Q0.25), normal (Q0.5), and the Bullish market (Q0.75, Q 0.9, Q0.95) (see Kangalli Uyar et al., 2021; Naifar, 2016; Zhu et al., 2016). For the local factors, we noticed that CPI does not affect KAT30 except for the last quantile of the bearish market (Q 0.25). Credit default swap negatively affects KAT30 across all quantiles as expected except for Q0.75. This result indicates that CDS nega- tively affects KAT30 across bearish market’s quantiles. When it Table 2. Preliminary Statistics KAT30 CPI EPU MSCI CDS SGB VIX OVX Mean 929.49 10.63 2.02 43864.73 −10.82 236.61 17.96 36.41 Median 798.91 9 1.60 42327.50 −12.09 −0.29 16.01 32.44 Maximum 2379.9 25 9.90 67329.00 3.08 9468.26 53.54 170.55 Minimum 490.05 6 1.70 28064.00 −18.96 −5.01 9.51 15.61 Standard deviation 410.25 4.08 1.87 8515.01 5.27 1484.38 7.17 18.08 Skewness 2.01 1.50 2.98 0.42 1.02 6.08 2.10 3.98 Kurtosis 7.13 4.97 12.40 2.79 3.42 38.02 8.65 27.90 Jarque-Bera 166.98 64.74* 620.35* 3.85* 21.75* 6874.42 248.35* 3418.04* * indicate 1% level of significance. 264 Table 3. Unit Root Tests Results at level ADF PP KAT30 0.85 0.24 CPI −2.97 −2.64 EPU −9.29* −9.36* MSCI −7.61* −8.46* CDS −2.42 −2.30 SGB −4.43* −8.13* VIX −4.48* −4.48* OVX −4.87* −4.81* and * indicate 5% and 1% level of significance, respectively. Note: ADF = Augmented Dickey–Fuller; PP = Philips–Perron; EPU = economic policy uncertainity; CDS = credit default swap; VIX = implied volatility index; KAT30 = participation index 30; CPI = consumer price index; MSCI = world index, Morgan Stanley country index; SGB = Term structure slope; OVX = Oil volatility index. Table 4. Unit Root Tests Results at First Difference ADF PP ΔKAT30 −3.04 −10.60* ΔCPI −8.09* −8.78* ΔEPU −8.79* −62.37* ΔMSCI −1.59* −4.65* ΔCDS −11.42* −13.12* ΔSGB −13.02* −33.33* ΔVIX −9.82* −18.94* ΔOVX −12.24* −34.67* and * indicate 5% and 1% level of significance, respectively. Note: ADF = Augmented Dickey–Fuller; PP = Philips–Perron; EPU = economic policy uncertainity; CDS = credit default swap; VIX = implied volatility index. When it comes to global factors, results show that for most quan- tiles, VIX has a negative impact on KAT30 as expected except for Q 0.75 and Q 0.95. Discussion and Conclusion This study analyzes the effect of LMI and GRF on the participation stock index in Turkey, using monthly data from May 2011 to April 2021. There is a wide set of research that analyzes the effect of LMI and GRF on the Turkish stock market. Yet, there are no stud- ies investigating the effect of LMI and GRF on Turkish participa- tion indices. The motivation of this study arises from this gap in the literature. This study bridges the gap in the empirical litera- ture by analyzing the effect of local macroeconomic and GRF on the Turkish participation index KAT30 returns using a quantile regression method to study, first, dependencies between KAT30 and risk factors, second, to examine the integration of the Turk- ish stock market, and third, to check for potential diversification opportunities. Table 4. Unit Root Tests Results at First Difference ADF PP ΔKAT30 −3.04 −10.60* ΔCPI −8.09* −8.78* ΔEPU −8.79* −62.37* ΔMSCI −1.59* −4.65* ΔCDS −11.42* −13.12* ΔSGB −13.02* −33.33* ΔVIX −9.82* −18.94* ΔOVX −12.24* −34.67* and * indicate 5% and 1% level of significance, respectively. Note: ADF = Augmented Dickey–Fuller; PP = Philips–Perron; EPU = economic policy uncertainity; CDS = credit default swap; VIX = implied volatility index. Our results show that, among local factors, monetary policy- related indicators (CPI and SGB) merely influence KAT30; how- ever, CDS affects the index across all quantiles. This means that KAT30 is comparatively insensitive to interest rate channels rep- resented by inflation and the spread of the term structure. This result is as expected considering that KAT30 firms’ indebtedness ratios and their dependency on interest income are low as men- tioned in the introduction section. Yet, the participation index is affected by the country’s global risk reflected by the CDS; this is expected as the country’s credit risk information is embedded in stock prices, which will in turn affect their returns. In fact, CDS reflects the country’s ability to pay its external debts which means that an increase in CDS causes an increase in external debt risk premium. Therefore, while the cost of external debts increases, the ability of external indebtedness decreases. In this case, lower consumption and investment levels will decrease the revenues of companies, which will negatively affect their stock prices and thus their returns. comes to the bullish market state, at its start (Q 0.75), investors react positively to the good news in the economy. Results Economic policy uncertainty ( has a positive impact on KAT30 (Q0.1 and Q0.9), while it has a negative signifi- cant impact on KAT30 (Q0.95). These results are almost close to the results of Hammoudeh et al. (2014), Nazlioglu et al. (2015), and Ftiti and Hadhri (2018) stating that Islamic SR is unrespon- sive to shocks of EPU. The world index (MSCI) impact on KAT30 is insignificant except for the last upper quantiles (Q0.9 andQ0.95). In addition, OVX positively impacts the upper quantiles (Q0.9 and Q0.95) and positively affects KAT30 during a bullish market. Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 Discussion and Conclusion However, this effect does not hold long and the trend reverses (CDS negatively affecting KAT30). The impact of SGB on KAT30 is insignificant across all quantiles except for the last one (Q 0.95) that shows a positive impact of SGB on KAT30. The latter is in line with Naifar’s (2016) results. That is, even if the SGB increases during a bullish market state, corporate profits increase and so do KAT30 returns. Table 5. Quantile Regression Results Q0.05 Q0.1 Q0.25 Q0.5 Q0.75 Q0.9 Q0.95 Δlog CPI −0.061 −0.030 −0.075* −0.019 0.017 0.013 −0.07 ΔCDS −0.015* −0.011* −0.014* −0.012* 0.012* −0.008* −0.01* SGB −2.07E-06 9.50E-07 5.16E-07 3.48E-06 5.23E-06 6.34E-06 −1.11E-05* Δlog VIX −0.075* −0.073 −0.090* −0.050** −0.042 −0.084* −0.05 EPU 4.6E-17 4.6E-17* 6.6E-18 2.1E-17 1.2E-16 2.1E-16** −1.1E-05* Δlog MSCI −0.292 0.342 0.275 0.163 −0.079 0.359* 0.48 Δlog OVXlag −0.047 −0.020 −0.029 0.049 0.022 0.065 0.08 α −.069* −.051* −.026* .009 .024 .014 .04* Adj R² .27 .23 .18 .14 .15 .22 .3 *, , and * indicate 1%, 5%,and 10% level of significance, respectively. 265 When it comes to GRF, results show that KAT30 is gener- ally affected by VIX across the low quantiles. This means that the VIX impact on KAT30 is stronger during market downturn. This relative asymmetric co-movement is in line with Whaley’s (2009) result, who argues that VIX is more of an indicator that reflects the fear of investors from a recession than it is an indi- cator of investors’ excitement. This shows that religiously con- scious investors behave the same way as conventional investors during bearish markets. Yet, OVX and MSCI are mostly sig- nificant across upper quantiles. MSCI positively affects KAT30 index across upper quantiles. This means that the decoupling hypothesis is affirmed for Turkish participation indices during the normal and bearish phase. The Turkish participation stock market is integrated with the global stock market only during the bullish market. Also, this result implies that during market decline or recession (bearish market), KAT30 might be a safe haven for investors. Thus, Turkish participation indices might be a diversification tool for stock market investors. Lastly, OVX positively affects KAT30 across upper quantiles. The bullish market reflects a good economic condition and thus demand for crude oil increases followed by an increase in its price. This means increasing oil price volatility during the bullish market increases the KAT30 index. Declaration of Interests: The authors declare that they have no compet- ing interest. Çonkir, D., Meriç, E., & Esen, E. (2021). Analysis of the relationship between the fear index (VIX) and emerging markets: A study on investor sentiment (Turkey). Journal of the Human and Social Science Researches, 10(1), 52–84. Funding: The authors declare that this study had received no financial support. Coronado, M., Corzo, M. T., & Lazcano, L. (2012). A case for Europe: The relationship between sovereign CDs and stock indexes. SSRN Electronic Journal, 9(2), 32–63. [CrossRef] 553–571. [CrossRef] Dickey, D. A., & Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with A unit root. Journal of the American Statistical Association, 74(366a), 427–431. [CrossRef] Çıkar Çatışması: Yazarlar, çıkar çatışması olmadığını beyan etmiştir. Çıkar Çatışması: Yazarlar, çıkar çatışması olmadığını beyan etmiştir. Finansal Destek: Yazarlar, bu çalışma için finansal destek almadıklarını beyan etmişlerdir. Donadelli, M. (2015). Asian stock markets, US economic policy uncertainty and US macro-shocks. New Zealand Economic Papers, 49(2), 103–133. [CrossRef] beyan etmişlerdir. [CrossRef] Banerjee, P. S., Doran, J. S., & Peterson, D. R. (2007). Implied volatility and future portfolio returns. Journal of Banking and Finance, 31(10), 3183–3199. [CrossRef] 3183–3199. [CrossRef] Bekaert, G., & Harvey, C. R. (1995). Time‐varying world market integration. Journal of Finance, 50(2), 403–444. [CrossRef] Bekaert, G., Harvey, C. R., Lundblad, C. T., & Siegel, S. (2011). What seg- ments equity markets? Review of Financial Studies, 24(12), 3841–3890. [CrossRef] Bhaduri, S., & Saraogi, R. (2010). The predictive power of the yield spread in timing the stock market. Emerging Markets Review, 11(3), 261–272. [CrossRef] 3–18. [CrossRef] Camgöz, M., Köse, A., & Seval, B. (2019). Risk and return characteristics of Islamic indices: An empirical approach. Istanbul Business Research, 47(2), 124–153. These findings are relevant and valuable as they provide a new perspective on the dynamics of KAT30 returns. Thus, this study provides a tool for fundamental investors targeting a sound long- term investment by identifying entry and exit strategy points considering bullish and bearish market conditions. Chang, B. H., Sharif, A., Aman, A., Suki, N. M., Salman, A., & Khan, S. A. R. (2020). The asymmetric effects of oil price on sectoral Islamic stocks: New evidence from quantile-on-quantile regression approach. Resources Policy, 65. [CrossRef] Chau, F., Deesomsak, R., & Wang, J. (2014). Political uncertainty and stock market volatility in the Middle East and North African (MENA) countries. Journal of International Financial Markets, Institutions and Money, 28(1), 1–19. [CrossRef] Peer-review: Externally peer-reviewed. Peer-review: Externally peer-reviewed. Author Contributions: Concept – A.E., Ş.G., M.G.; Design – A.E.; Supervi- sion – Ş.G.; Data Collection and/or Processing – A.E., M.G.; Analysis and/or Interpretation – A.E., Ş.G., M.G.; Literature Review – A.E., Ş.G., M.G.; Writ- ing – A.E.; Critical Review – A.E., Ş.G., M.G. Christou, C., Cunado, J., Gupta, R., & Hassapis, C. (2017). Economic policy uncertainty and stock returns in PacificRim countries: Evidence based on a Bayesian panel VAR model. Journal of Multinational Financial Management, 40, 92–102. [CrossRef] Declaration of Interests: The authors declare that they have no compet- ing interest. Discussion and Conclusion Furthermore, this result indicates that oil shocks affect the Turkish participation indices with lags; hence, the gradual information diffusion hypothesis is affirmed. This means that some investors in the Turkish stock market do not incorporate new information directly into their strategy, and investors underreact to new information by responding to them at belated points in time. Ali, S., Shahzad, S. J. H., Raza, N., & Al-Yahyaee, K. H. (2018) Stock market efficiency: A comparative analysis of Islamic and conventional stock markets. Physica. Part A, 503, 139–153. [CrossRef] Al-Khazali, O., Lean, H. H., & Samet, A. (2014). Do Islamic stock indexes outperform conventional stock indexes? A stochastic dominance approach. Pacific Basin Finance Journal, 28, 29–46. [CrossRef] Aloui, C., Hkiri, B., Lau, C. K. M., & Yarovaya, L. (2016). Investors’ sentiment and US Islamic and conventional indexes nexus: A time–frequency and US Islamic and conventional indexes nexus: A time fre analysis. Finance Research Letters, 19, 54–59. [CrossRef] analysis. Finance Research Letters, 19, 54–59. [CrossRef] Arshad, S., Aun, S., & Rizvi, R. (2013). Interactions between Islamic stock indices and business cycles: Focusing On Asia Pacific. Australian Journal of Basic and Applied Sciences, 7(9), 1–9. Bahloul, S., & Ben Amor, N. (2021). A quantile regression approach to evaluate the relative impact of global and local factors on the MENA stock markets. International Journal of Emerging Markets. [CrossRef] Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 References Driesprong, G., Jacobsen, B., & Maat, B. (2008). Striking oil: Another puzzle? Journal of Financial Economics, 89(2), 307–327. [CrossRef] AAOIFI. (2015). Shari’ah Standards. Kingdom of Bahrain: Dar Alamaiman, 559–580. Hakem Değerlendirmesi: Dış bağımsız. Yazar Katkıları: Fikir – A.E., Ş.G., M.G.; Tasarım – A.E.; Denetleme – Ş.G.; Veri Toplanması ve/veya İşlemesi – A.E., Ş.G., M.G.; Analiz ve/veya Yorum – A.E., Ş.G., M.G.; Literatür Taraması – A.E., Ş.G., M.G.; Yazıyı Yazan – A.E.; Eleştirel İnceleme – A.E., Ş.G., M.G. Dewandaru, G., Rizvi, S. A. R., Masih, R., Masih, M., & Alhabshi, S. O. (2014). Stock market co-movements: Islamic versus conventional equity indices with multi-timescales analysis. Economic Systems, 38(4), 553–571. [CrossRef] g [CrossRef] Brogaard, J., & Detzel, A. (2015). The asset-pricing implications of govern- ment economic policy uncertainty. Management Science, 61(1), 3–18. [CrossRef] p [CrossRef] Ahmed, N., & Farooq, O. (2018). Does the degree of Shari’ah compliance affect the volatility? Evidence from the MENA region. Research in International Business and Finance, 45, 150–157. [CrossRef] Erdoğan, L., Ceylan, R., & Abdul-Rahman, M. (2021). The impact of domes- tic and GRFon Turkish stock market: Evidence from the NARDL approach. Emerging Markets Finance and Trade, 1–14. 266 Estrella, A., & Mishkin, F. S. (1998). Predicting U.S. recessions: Financial variables as leading indicators. Review of Economics and Statistics, 80(1), 45–61. [CrossRef]l the global financial crisis. Applied Financial Economics, 24(21), 1367–1373. [CrossRef] Liang, C. C., Troy, C., & Rouyer, E. (2020). U.S. uncertainty and Asian stock prices: Evidence from the asymmetric NARDL model. North American Journal of Economics and Finance, 51, 101046. Fama, E. F. (1981). Stock returns, real activity, inflation, and money. Ameri- can Economic Association Stock Returns the American Economic Review, 71, 545–565. Lin, B., & Su, T. (2020). The linkages between oil market uncertainty and Islamic stock markets: Evidence from quantile-on-quantile approach. Energy Economics, 88. [CrossRef] Fama, E. F. (1990). Stock returns, expected returns, and real activity. Jour- nal of Finance, 45(4), 1089–1108. [CrossRef] Fama, E. F. (1990). Stock returns, expected return nal of Finance, 45(4), 1089–1108. [CrossRef] Fama, E. F. (1990). Stock returns, expected return Liu, H. Y., & Chen, X. L. (2017). The imported price, inflation and exchange Liu, H. Y., & Chen, X. L. (2017). The imported price, inflation and exchange rate pass-through in China. Cogent Economics and Finance, 5(1). [CrossRef] Fatima, A., Rashid, A., & Khan, A. (2019). Asymmetric impact of shocks on Islamic stock indices: A cross country analysis. Journal of Islamic Marketing, 10(1), 2–86. [CrossRef] Marketing, 10(1), 2–86. [CrossRef] [CrossRef] Narayan, P. K., & Narayan, S. (2010). Modelling the impact of oil prices on Vietnam’s stock prices. Applied Energy, 87(1), 356–361. [CrossRef] Guesmi, K., & Nguyen, D. K. (2011). How strong is the global integration of emerging market regions? An empirical assessment. Economic Modelling, 28(6), 2517–2527. [CrossRef] Nazlioglu, S., Hammoudeh, S., & Gupta, R. (2015). Volatility transmission between Islamic and conventional equity markets: Evidence from causality-in-variance test. Applied Economics, 47(46), 1–16. [CrossRef] Haddad, H. B. Ben, Mezghani, I., & Al Dohaiman, M. (2020). Common shocks, common transmission mechanisms and time-varying connectedness among Dow Jones Islamic stock market indices and global risk factors. Economic Systems, 44(2). [CrossRef] Öner, H., Şarkaya İçellİoğlu, C., & Öner, S. (2018). Volatilite Endeksi (VIX) ile gelişmekte olan ülke hisse senedi piyasası endeksleri arasındaki engel-Granger eş-bütünleşme ve Granger nedensellik analizi. Finansal Araştırmalar ve Çalışmalar Dergisi, 10(18), 110–124. [CrossRef] Hammoudeh, S., Kim, W. J., & Sarafrazi, S. (2016). Sources of fluctuations in Islamic, U.S., EU, and Asia Equity Markets: The roles of economic uncertainty, interest rates, and stock indexes. Emerging Markets Finance and Trade, 52(5), 1195–1209. Paltrinieri, A., Floreani, J., Kappen, J. A., Mitchell, M. C., & Chawla, K. (2019). Islamic, socially responsible, and conventional market comove- ments: Evidence from stock indices. Thunderbird International Busi- ness Review, 61(5), 719–733. [CrossRef] Hammoudeh, S., Mensi, W., Reboredo, J. C., & Nguyen, D. K. (2014). Dynamic dependence of the global Islamic equity index with global conventional equity market indices and risk factors. Pacific Basin Finance Journal, 30, 189–206. [CrossRef] Resnick, B. G., & Shoesmith, G. L. (2002). Using the yield curve to time the stock market. Financial Analysts Journal, 58(3), 82–90. [CrossRef] Sadorsky, P. (1999). Oil price shocks and stock market activity. Energy Economics, 21(5), 449–469. [CrossRef] Sadorsky, P. (1999). Oil price shocks and stock Economics, 21(5), 449–469. [CrossRef]i Economics, 21(5), 449–469. [CrossRef] Hedi Arouri, M. E., & Khuong Nguyen, D. (2010). Oil prices, stock markets and portfolio investment: Evidence from sector analysis in Europe over the last decade. Energy Policy, 38(8), 4528–4539. [CrossRef] Safiullah, M., & Shamsuddin, A. (2021). Asset pricing factors in Islamic equity returns. International Review of Finance, 21(2), 523–554. [CrossRef] [CrossRef] Naifar, N. (2016). Do global risk factors and macroeconomic conditions affect global Islamic index dynamics? A quantile regression approach. Naifar, N. (2016). Do global risk factors and macroeconomic conditions affect global Islamic index dynamics? A quantile regression approach. Quarterly Review of Economics and Finance, 61, 29–39. [CrossRef] Giot, P. (2005). Relationships between implied volatility indexes and stock index returns. Journal of Portfolio Management, 31(3), 92–100. [CrossRef] Quarterly Review of Economics and Finance, 61, 29–39. [CrossRef] Quarterly Review of Economics and Finance, 61, 29–39. [CrossRef] Narayan, P. K., & Narayan, S. (2010). Modelling the impact of oil prices on Vietnam’s stock prices Applied Energy 87(1) 356–361 [CrossRef] q y [CrossRef] Ho, C. S. F., Abd Rahman, N. A., Yusuf, N. H. M., & Zamzamin, Z. (2014). Performance of global Islamic versus conventional share indices: International evidence. Pacific Basin Finance Journal, 28, 110–121. [CrossRef] Shahzad, S. J. H., Mensi, W., Hammoudeh, S., Rehman, M. U., & Al-Yahyaee, K. H. (2018). Extreme dependence and risk spillovers between oil and Islamic stock markets. Emerging Markets Review, 34, 42–63. [CrossRef] [CrossRef] Jawadi, F., Jawadi, N., & Cheffou, A. I. (2018). A statistical analysis of uncer- tainty for conventional and ethical stock markets a statistical analy- sis of uncertainty for conventional and ethical stock indexes. Quarterly Review of Economics and Finance, 74, 9–17. Trabelsi, N., & Naifar, N. (2017). Are Islamic stock indexes exposed to sys- temic risk? Multivariate GARCH estimation of CoVaR. Research in International Business and Finance, 42, 727–744. [CrossRef] Tursoy, T., & Faisal, F. (2018). The impact of gold and crude oil prices on stock market in Turkey: Empirical evidences from ARDL bounds test and combined cointegration. Resources Policy, 55, 49–54. [CrossRef] Jouini, J. (2013). Return and volatility interaction between oil prices and stock markets in Saudi Arabia. Journal of Policy Modeling, 35(6), 1124–1144. [CrossRef] [CrossRef] Filis, G. (2010). Macro economy, stock market and oil prices: Do meaning- ful relationships exist among their cyclical fluctuations? Energy Eco- nomics, 32(4), 877–886. [CrossRef] Mezghani, T., & Boujelbène, M. (2018). The contagion effect between the oil market, and the Islamic and conventional stock markets of the GCC country Behavioral explanation, 11(2), 157–181. g . [CrossRef] Fisher, I. (1930). The Theory of Interest. The Macmillan Mun, K. C. (2012). The joint response of stock and foreign exchange mar- kets to macroeconomic surprises: Using US and Japanese data. Journal of Banking and Finance, 36(2), 383–394. [CrossRef] Ftiti, Z., & Hadhri, S. (2018). Can economic policy uncertainty, oil prices, and investor sentiment predict Islamic stock returns? A multi-scale perspective. Pacific-Basin Finance Journal, 53, 40–55. Murthy, U., Anthony, P., & Vighnesvaran, R. (2016). Factors affecting Kuala Lumpur Composite Index (KLCI) stock return in Malaysia. Inter- national Journal of Business and Management, 12(1), 122–132. [CrossRef] Ghorbel, A., Abdelhedi, M., Boujelbene, Y., Ghorbel, A., Abdelhedi, M., & Boujelbene, Y. (2014). Assessing the impact of crude oil price and investor sentiment on Islamic indices: Subprime crisis. Journal of African Business, 15(1), 13–24. [CrossRef] national Journal of Business and Management, 12(1), 122–132. [CrossRef] Amaç Bu çalışma, farklı piyasa durumlarında lokal makroekonomik göstergelerin ve küresel risk faktörlerinin Türk hisse senedi piyasasında katılım endeksi üzerindeki etkisini araştırmaktadır. 1124–1144. [CrossRef] Usman, M., Ali, M., Jibran, Q., Amir-ud-din, R., & Akhter, W. (2019). Decou- pling hypothesis of Islamic stocks: Evidence from copula CoVaR approach. Borsa Istanbul Review, 19(1), 56–63. Kang, W., Ratti, R. A., & Vespignani, J. (2016). The impact of oil price shocks on the U.S. stock market: A note on the roles of U.S. and non-U.S. oil production. Economics Letters, 145, 176–181. [CrossRef] Whaley, R. E. (2009). Understanding the VIX. Journal of Portfolio Manage- ment, 35(3), 98–105. [CrossRef] Kangalli Uyar, S. G., Uyar, U., & Balkan, E. (2022). The role of precious metals in extreme market conditions: Evidence from stock markets. Studies in Economics and Finance, 39(1), 63–78. [CrossRef] Yarovaya, L., Elsayed, A. H., & Hammoudeh, S. (2021). Determinants of spillovers between Islamic and conventional financial markets: Exploring the safe haven assets during the COVID-19 pandemic. Finance Research Letters, 43. [CrossRef] Kling, J. L. (1985). Oil price shocks and stock market behavior. Journal of Portfolio Management, 12(1), 34–39. [CrossRef] Koenker, R. (2005). Quantile Regression. Cambridge University Press. Koenker, R. (2005). Quantile Regression. Cambridge University Press. Koenker, R., & Bassett, G., Jr. (1978). Regression quantiles. Econometrica, 46(1), 33–50. [CrossRef] Zhu, H., Guo, Y., You, W., & Xu, Y. (2016). The heterogeneity dependence between crude oil price changes and industry stock returns in China: Evidence from a quantile regression approach. Energy Economics, 55, 30–41. [CrossRef] Koenker, R., & Bassett, G., Jr. (1978). Regression quantiles. Econometrica, 46(1), 33–50. [CrossRef] Koenker, R., & Bassett, G., Jr. (1 46(1), 33–50. [CrossRef] Lean, H. H., & Nguyen, D. K. (2014). Policy uncertainty and performance characteristics of sustainable investments across regions around Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360 267 Araştırmanın Metodu Lokal ve küresel faktörlerin farklı piyasa koşulları üzerindeki etkisini tespit etmek için Mayıs 2011'den Nisan 2021'e kadar olan aylık veriler kullanılarak kantil regresyon yaklaşımı ile analiz edilmektedir. Bulgular ve Tartışma Sonuçlar, lokal makroekonomik göstergeler arasında para politikasıyla ilgili göstergelerden CPI(SGB)’nin sadece düşüş piyasasında (yük- seliş piyasası) KAT30'u etkilediğini gösterirken, CDS tüm kantillerde KAT30'u negatif etkilemektedir.Bu, KAT30'un enflasyon ve vade yapısının yayılması ile temsil edilen faiz oranına nispeten duyarsız olduğunu göstermektedir. Küresel risk faktörleri söz konusu oldu- ğunda, sonuçlar KAT30'un Q0.75 ve Q0.95 dışındaki tüm kantillerde VIX'ten negatif etkilendiğini göstermektedir. Bu durum, düşüş piya- sasında VIX’in KAT30 üzerindeki etkisinin daha güçlü olduğu göstermektedir. Bununla birlikte, OVX ve MSCI'nin etkisi, üst kantillerde pozitif ve anlamlıdır. Bu, KAT30 üzerindeki VIX etkisinin pazarın gerilemesi sırasında daha güçlü olduğu anlamına gelmektedir. Bu göreli asimetrik ortak hareket, VIX'in yatırımcıların heyecanının bir göstergesi olmaktan çok, yatırımcıların resesyon korkusunu yansıtan bir gösterge olduğunu savunan Whaley (2009)’un sonucu ile uyumludur. Bu, dini hassasiyeti olan ve olmayan yatırımcıların düşüş piyasa- larında aynı şekilde davrandığını göstermektedir. Ancak, OVX ve MSCI'nin etkisi çoğunlukla üst kantillerde anlamlılık göstermektedir. MSCI, üst kantillerde KAT30 endeksini olumlu yönde etkilemektedir. Bu, normal ve düşüş döneminde Türkiye katılım endeksleri için ayrışma hipotezini desteklemektedir. Trends in Business and Economics l 2022 36(3): 258-267 l doi: 10.5152/TBE.2022.1018360
https://openalex.org/W4200389663	https://ria.ua.pt/bitstream/10773/32773/1/%5b499%5dRosa-Torres-fractalfract-cholera.pdf	English	null	Fractional-Order Modelling and Optimal Control of Cholera Transmission	Fractal and fractional	2,021	cc-by	10,247	Keywords: cholera; compartmental mathematical models; fractional-order optimal control Citation: Rosa, S.; Torres, D.F.M. Fractional-Order Modelling and Optimal Control of Cholera Transmission. Fractal Fract. 2021, 5, 261. https://doi.org/10.3390/ fractalfract5040261 Citation: Rosa, S.; Torres, D.F.M. Fractional-Order Modelling and Optimal Control of Cholera Transmission. Fractal Fract. 2021, 5, 261. https://doi.org/10.3390/ fractalfract5040261 MSC: 34A08, 49M05, 92C60 Silvério Rosa 1,,† and Delﬁm F. M. Torres 2,† 1 Instituto de Telecomunicações (IT) and Department of Mathematics, Universidade da Beira Interior, 6201-001 Covilhã, Portugal † These authors contributed equally to this work. Abstract: A Caputo-type fractional-order mathematical model for “metapopulation cholera transmis- sion” was recently proposed in [Chaos Solitons Fractals 117 (2018), 37–49]. A sensitivity analysis of that model is done here to show the accuracy relevance of parameter estimation. Then, a fractional optimal control (FOC) problem is formulated and numerically solved. A cost-effectiveness analysis is performed to assess the relevance of studied control measures. Moreover, such analysis allows us to assess the cost and effectiveness of the control measures during intervention. We conclude that the FOC system is more effective only in part of the time interval. For this reason, we propose a system where the derivative order varies along the time interval, being fractional or classical when more advantageous. Such variable-order fractional model, that we call a FractInt system, shows to be the most effective in the control of the disease. Citation: Rosa, S.; Torres, D.F.M. Fractional-Order Modelling and Optimal Control of Cholera Transmission. Fractal Fract. 2021, 5, 261. https://doi.org/10.3390/ fractalfract5040261 Academic Editors: Vassili Kolokoltsov and Omar Abu Arqub Received: 20 October 2021 Accepted: 2 December 2021 Published: 7 December 2021 Citation: Rosa, S.; Torres, D.F.M. Fractional-Order Modelling and Optimal Control of Cholera Transmission. Fractal Fract. 2021, 5, 261. https://doi.org/10.3390/ fractalfract5040261 Academic Editors: Vassili Kolokoltsov and Omar Abu Arqub Received: 20 October 2021 Accepted: 2 December 2021 Published: 7 December 2021 Article Fractional-Order Modelling and Optimal Control of Cholera Transmission Silvério Rosa 1,,† and Delﬁm F. M. Torres 2,† fractal and fractional fractal and fractional fractal and fractional f fractal and fractional Citation: Rosa, S.; Torres, D.F.M. Fractional-Order Modelling and Optimal Control of Cholera Transmission. Fractal Fract. 2021, 5, 261. https://doi.org/10.3390/ fractalfract5040261 1. Introduction Fractional calculus is an old subject that raised as a consequence of a pertinent question that L’Hôpital asked Leibniz in a letter about the possible meaning of a derivative of order 1/2. Recently, many researchers have focused their attention in modelling real-world phenomena using fractional-order derivatives. The dynamics of those problems have been modelled and studied by using the concept of fractional-order derivatives. Such problems appear in, for example, biology, physics, ecology, engineering, and various other ﬁelds of applied sciences, see, e.g., [1–3]. Academic Editors: Vassili Kolokoltsov and Omar Abu Arqub Received: 20 October 2021 Accepted: 2 December 2021 Published: 7 December 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional afﬁl- iations. Cholera is a gastroenteritis infection, contracted after consuming an infectious dose or inoculum size of the pathogenic vibrio cholerae [4]. The mode of transmission consists of two pathways: the primary route, where individuals consumes the pathogen from vibrio contaminated water and seafood; the secondary route being characterised by individuals consuming unhygienic or soiled food that is infested with pathogenic vibrios from an infected person. This secondary route of transmission is also commonly referred to as person-to- person contact [4]. Cholera infection has affected many parts of the world. However, its devastating force has been more pronounced in impoverished communities [5,6]. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). Cholera is one of the most studied infections in recent years. Mathematical models are used to study and understand the dynamics of infection, as well as offer suggestions toward its control, see, e.g., [4,7–12]. Most models used in the study of Cholera have been based on systems of integer-order ordinary differential equations. However, those models do not fully account for memory, as well as non-local properties exhibited by the epidemic system. Non-local behaviour asserts that the subsequent state of the model depends on https://www.mdpi.com/journal/fractalfract Fractal Fract. 2021, 5, 261. https://doi.org/10.3390/fractalfract5040261 Fractal Fract. 2021, 5, 261 2 of 16 both the current and historical states. 1. Introduction Fractional order differential systems have been proposed as more suitable to describe epidemic dynamics of diseases [13–15]. Recently, a fractional order differential system was used in the study of cholera transmission in adjacent communities [10]. Individuals in adjacent communities often frequent their home range, an aspect which constitutes memory that is present in such a model. Here, we start to do a sensitivity analysis to the fractional-order model of [10] in order to determine which model parameters are most inﬂuential on the disease dynamics. After that, fractional optimal control (FOC) is applied, as a generalisation of the classical optimal control system of [11]. In contrast with other works, for example [16], we could not ﬁnd a derivative order for which the FOC system is more effective. However, we noticed that the FOC system is more effective in part of the time interval. Hence, we propose a system where the derivative order varies along the interval, being fractional or classical when more advantageous. Such system, a variable-order fractional one, is named here a FractInt system, shown to be useful in the control of the disease. y The paper is organised as follows. In Section 2, the fractional order model formulation is presented. Our main results are then given in Section 3: sensitivity analysis of the parameters of the model, taking into account the derivative order (Section 3.1); fractional optimal control of the model (Section 3.2); numerical simulations and cost-effectiveness of the fractional model (Section 3.3); and the new variable-order FractInt system (Section 3.3.2). We end with Section 4 of conclusions. 2. Fractional-Order Cholera Model A metapopulation model for cholera transmission is considered, dividing the popula- tion into mutually exclusive distinct groups and using deterministic continuous transitions between those groups, also known as states. The model describes the dynamics of a pop- ulation exposed to infection by the pathogen vibrio cholerae. The human population is divided into three compartments: susceptible individuals (S); infectious individuals (I); and recovered individuals (R). The Caputo fractional-order system of differential equations is as follows [10]:                            C 0Dα t S1 =π1 + a2S2 + ωR1 −(1 −u) β1B1S1 K + B1 −(1 −m)ϱ1I1S1 −(a1 + µ1 + v)S1, C 0Dα t I1 =(1 −u) β1B1S1 K + B1 + (1 −m)ϱ1I1S1 + b2I2 −Q1I1, C 0Dα t R1 =vS1 + γ1I1 −(µ1 + ω + c1)R1 + c2R2, C 0Dα t B1 =σ1I1 −Q2B1, (1) (1) for the ﬁrst sub-population, and for the ﬁrst sub-population, and                            C 0Dα t S2 =π2 + a1S1 + ωR2 −(1 −u) β2B2S2 K + B2 −(1 −m)ϱ2I2S2 −(a2 + µ2 + v)S2, C 0Dα t I2 =(1 −u) β2B2S2 K + B2 + (1 −m)ϱ2I2S2 + b1I1 −Q3I2, C 0Dα t R2 =vS2 + γ2I2 −(µ2 + ω + c2)R2 + c1R1, C 0Dα t B2 =σ2I2 −Q4B2, (2) (2) for the second sub-population, where C 0Dα t denotes the left Caputo fractional order deriva- tive of order α [1], 0 < α ⩽1, Q1 = µ1 + δ1 + γ1 + b1, Q2 = µp −g1, Q3 = µ2 + δ2 + γ2 + b2, and Q4 = µp −g2. for the second sub-population, where C 0Dα t denotes the left Caputo fractional order deriva- tive of order α [1], 0 < α ⩽1, Q1 = µ1 + δ1 + γ1 + b1, Q2 = µp −g1, Q3 = µ2 + δ2 + γ2 + b2, and Q4 = µp −g2. Fractal Fract. 2021, 5, 261 3 of 16 3 of 16 We note that the equations of model (1)–(2) do not have appropriate time dimensions. 2. Fractional-Order Cholera Model Indeed, on the left-hand side the dimension is (time)−α while on the right-hand side the dimension is (time)−1 (see, e.g., [17,18] for more details). Therefore, we claim that the accurate way of writing system (1) is                              C 0Dα t S1 =πα 1 + aα 2S2 + ωαR1 −(1 −u) βα 1B1S1 K + B1 −(1 −m)ϱα 1 I1S1 −(aα 1 + µα 1 + v)S1, C 0Dα t I1 =(1 −u) βα 1B1S1 K + B1 + (1 −m)ϱα 1 I1S1 + bα 2 I2 −Q1I1, C 0Dα t R1 =vS1 + γα 1 I1 −(µα 1 + ωα + cα 1)R1 + cα 2R2, C 0Dα t B1 =σα 1 I1 −Q2B1, (3) (3) for the ﬁrst sub-population, and for the ﬁrst sub-population, and for the ﬁrst sub-population, and                            C 0Dα t S2 =πα 2 + aα 1S1 + ωαR2 −(1 −u) βα 2B2S2 K + B2 −(1 −m)ϱα 2 I2S2 −(aα 2 + µα 2 + v)S2, C 0Dα t I2 =(1 −u) βα 2B2S2 K + B2 + (1 −m)ϱα 2 I2S2 + bα 1 I1 −Q3I2, C 0Dα t R2 =vS2 + γα 2 I2 −(µα 2 + ωα + cα 2)R2 + cα 1R1, C 0Dα t B2 = σα 2 I2 −Q4B2, (4) (4) the accurate way of writing system (2), with Q1 = µα 1 + δα 1 + γα 1 + bα 1, Q2 = µα p −gα 1, Q3 = µα 2 + δα 2 + γα 2 + bα 2, and Q4 = µα p −gα 2. 3. Main Results We begin by doing a sensitivity analysis to the parameters of the model, in order to identify those for which a small perturbation leads to relevant quantitative changes in the dynamics. 3.1. Sensitivity Analysis Two distinct ways to compute the basic reproduction numbers, R01 and R02, of the two sub-populations of the model, are available in [10,11]. To know which one is proper, we determine them by using the next-generation matrix method [19]. We obtain the community speciﬁc reproduction numbers as R01 = (πα 1(µα 2 + aα 2 + v) + aα 2πα 2)((1 −u)βα 1σα 1 + (1 −m)Q2ϱα 1K) Q1Q2(µα 1 + aα 1 + v)(µα 2 + aα 2 + v)(1 −Φ1)K , (5) (5) for the ﬁrst community, and R02 = (πα 2(µα 1 + aα 1 + v) + aα 1πα 1)((1 −u)βα 2σα 2 + (1 −m)Q4ϱα 2K) Q3Q4(µα 1 + aα 1 + v)(µα 2 + aα 2 + v)(1 −Φ1)K , (6) (6) for the second community, where Φ1 = aα 1aα 2 (µα 1 + aα 1 + v)(µα 2 + aα 2 + v). 4 of 16 Fractal Fract. 2021, 5, 261 The values of used parameters are presented in Table 1, which were taken from [10], with exception of π1, π2 and ϱ2. The ﬁrst two parameters are equal and are deﬁned as π1 = π2 = 1.08 × 10−4, π1 = π2 = 1.08 × 10−4, which is bigger than the values proposed in [10], in order to ensure an endemic scenario (R01 > 1 and R02 > 1). Indeed, it is the presence of an endemic situation that motivates us, in Section 3.2, to apply optimal control theory to tackle the cholera problem. The latter parameter, ϱ2, is also changed with respect to [10] and is deﬁned as ϱ2 = 0.1875. This particular value makes speciﬁc reproduction numbers of the two communities, R01 and R02, clearly different from each other as we can see in Figure 1. This means that the two communities are distinct and that they could not be considered as one unique community. Table 1. Values of model’s parameters. Table 1. Values of model’s parameters. Table 1. Values of model’s parameters. Table 1. Values of model’s parameters. 3.1. Sensitivity Analysis Sensitivity of R01 (top) and R02 (bottom), evaluated for the parameter values given in Table 1 with α = 1. Parameter ΥR01 · Parameter ΥR01 · Parameter ΥR01 · π1, π2 0.500 µ2, a1 −0.454 β1, σ1 2 × 10−6 ϱ1 0.999 K −2 × 10−6 µ1 −0.547 a2 0.454 b1 −0.343 δ1 −0.143 γ1 −0.514 µp −6 × 10−6 g1 4 × 10−6 Parameter ΥR02 · Parameter ΥR02 · Parameter ΥR02 · π1, π2 0.500 µ2 −0.456 β2, σ2 1 × 10−5 ϱ2 0.999 K −1 × 10−5 µ1, a2 −0.545 a1 0.545 b2 −0.259 δ2 −0.416 γ2 −0.324 µp −3 × 10−5 g2 2 × 10−5 For the fractional model, the sensitivity index depends on the derivative order α. We can see this in Figure 2, where: (a) the impact of variation of α in the sensitivity index of β1 is displayed for the ﬁrst community; (b) the impact of variation of α in the sensitivity index of ϱ2 is exhibited for the second community. The graphics of the other parameters are not shown because they exhibited similar behaviours. The parameters whose index value for α = 1 is close to zero, as β1, do not vary much if we consider lower values of α, as we see in Figure 2a. On the other hand, parameters whose index value in Table 2 are not as close to zero as the previous one, as ϱ2, vary signiﬁcantly if we consider lower values of α, as we see in Figure 2b, and their sensitivity decreases with the decrease in α. Table 2. Sensitivity of R01 (top) and R02 (bottom), evaluated for the parameter values given in Table 1 with α = 1. Parameter ΥR01 · Parameter ΥR01 · Parameter ΥR01 · π1, π2 0.500 µ2, a1 −0.454 β1, σ1 2 × 10−6 ϱ1 0.999 K −2 × 10−6 µ1 −0.547 a2 0.454 b1 −0.343 δ1 −0.143 γ1 −0.514 µp −6 × 10−6 g1 4 × 10−6 Parameter ΥR02 · Parameter ΥR02 · Parameter ΥR02 · π1, π2 0.500 µ2 −0.456 β2, σ2 1 × 10−5 ϱ2 0.999 K −1 × 10−5 µ1, a2 −0.545 a1 0.545 b2 −0.259 δ2 −0.416 γ2 −0.324 µp −3 × 10−5 g2 2 × 10−5 Table 2. Sensitivity of R01 (top) and R02 (bottom), evaluated for the parameter values given in Table 1 with α = 1. 3.1. Sensitivity Analysis Parameter Value Source β1 0.00125 [10] β2 0.0125 [10] K 106 [8] µ1, µ2 8.4×10−5 [20,21] δ1 0.0125 [12,22] δ2 0.045 [12,22] γ1 0.045 [12,23] γ2 0.035 [12,23] µp 1.06 [8,24–26] g1, g2 0.73 [10] ϱ1 0.102 – σ1, σ2 50 [10] We start by considering that rates u, v, and m are all zero in the sensitivity analysis, unless speciﬁed otherwise. We start by considering that rates u, v, and m are all zero in the sensitivity analysis, unless speciﬁed otherwise. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 Figure 1. Impact of the variation of derivative order, α, in the basic reproduction numbers of both communities (3) and (4), respectively, R01 and R02. Deﬁnition 1 (See [27,28]). The normalised forward sensitivity index of R0i, which is differentiable with respect to a given parameter p, is deﬁned by ΥR0i p = ∂R0i ∂p p R0i , i=1,2. (7) (7) Table 2 presents the values of the sensitivity index of the parameters of the model, obtained by the normalised sensitivity index (7), for the classical case (α = 1) of connected Fractal Fract. 2021, 5, 261 5 of 16 communities. These values have a meaning. For instance, ΥR01 ϱ1 = +0.999 means that increasing (decreasing) ϱ1 by a given percentage increases (decreases) always R01 by nearly that same percentage. Sensitive parameters should be carefully evaluated, once a small perturbation in such parameter leads to signiﬁcant quantitative changes. On the other hand, the estimation of a parameter with a small value for the sensitivity index does not require as much attention to evaluate, because a small perturbation in that parameter leads to small adjustments [29]. According with Table 2, we should pay special attention to the estimation of sensitive parameters ϱ1 and ϱ2. In contrast, the estimation of K, µp, β1, β2, σ1, σ2, g1, and g2 do not require as much attention because of its low sensitivity. The missing parameters are those whose index value is zero. Table 2. 3.1. Sensitivity Analysis We see that the evolution of the sensitivity index is analogous for the two reproduction numbers. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.5 1 1.5 2 2.5 R01 R02 Figure 3. Evolution of the sensitivity index for the basic reproduction numbers of both communities, R01 and R02, with the variation of the derivative order, α. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.5 1 1.5 2 2.5 R01 R02 Figure 3. Evolution of the sensitivity index for the basic reproduction numbers of both communities, R01 and R02, with the variation of the derivative order, α. 3.1. Sensitivity Analysis For the fractional model, the sensitivity index depends on the derivative order α. We can see this in Figure 2, where: (a) the impact of variation of α in the sensitivity index of β1 is displayed for the ﬁrst community; (b) the impact of variation of α in the sensitivity index of ϱ2 is exhibited for the second community. The graphics of the other parameters are not shown because they exhibited similar behaviours. The parameters whose index value for α = 1 is close to zero, as β1, do not vary much if we consider lower values of α, as we see in Figure 2a. On the other hand, parameters whose index value in Table 2 are not as close to zero as the previous one, as ϱ2, vary signiﬁcantly if we consider lower values of α, as we see in Figure 2b, and their sensitivity decreases with the decrease in α. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 2 4 6 8 10 12 10-6 (a) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 (b) Figure 2. (a) Evolution of the sensitivity index of parameter β1, evaluated for α, with respect to R01 (ﬁrst community); (b) Evolution of the sensitivity index of parameter ϱ2, evaluated for α, with respect to R02 (second community). 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 2 4 6 8 10 12 10-6 (a) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 (b) (b) (a) Figure 2. (a) Evolution of the sensitivity index of parameter β1, evaluated for α, with respect to R01 (ﬁrst community); (b) Evolution of the sensitivity index of parameter ϱ2, evaluated for α, with respect to R02 (second community). 6 of 16 Fractal Fract. 2021, 5, 261 Figure 3 presents the evolution of the sensitivity index for the specific reproduction Figure 3 presents the evolution of the sensitivity index for the specific reproduction numbers of the two communities, R01 and R02, with the variation of the derivative order. We see that the evolution of the sensitivity index is analogous for the two reproduction numbers. g p y p p numbers of the two communities, R01 and R02, with the variation of the derivative order. 3.2. Fractional Optimal Control of the Model Modelling dynamic control systems optimally is a very important issue in applied sciences and engineering [30]. In this section, our aim is to minimise the number of cholera infected individuals and, simultaneously, to reduce the associated cost. This is achieved trough: (i) the use of vaccination into communities, as an effective time-dependent measure- control v(t); (ii) the use of clean treated water, a preventive measure control u(t); (iii) and the implementation of proper hygiene, another preventive measure control m(t), in order to control person-to-person contact. Thus, we consider the following fractional optimal control problem: min J (I1, I2, u, v,m) = Z t f 0 k1I1 + k2I2 + k3u2 + k4v2 + k5m2 dt (8) (8) subject to system (3)–(4) with given initial conditions subject to system (3)–(4) with given initial conditions S1(0) = S10 ⩾0, I1(0) = I10 ⩾0, . . . , R2(0) = R20 ⩾0, B2(0) = B20 ⩾0. (9) (9) Note that we are using a quadratic cost functional on the controls, as an approx- imation to the real non-linear functional. Indeed, as in [31], our optimal control prob- lem depends on the assumption that the cost takes a non-linear form. The parameters 0 < k1, k2, k3, k4, k5 < +∞are positive weights and t f is the duration of the control pro- gram. In addition, k3, k4, and k5 represent the costs of applying controls efforts u, v, and m, respectively. The set of admissible control functions is (10) U = n (u(·), v(·), m(·)) ∈L∞(0, t f ) : 0 ⩽u ⩽umax, 0 ⩽v ⩽vmax, 0 ⩽m ⩽mmax o . (10) The Pontryagin maximum principle (PMP) for fractional optimal control is used to solve the problem [32,33]. The Hamiltonian of the resulting optimal control problem is deﬁned as 7 of 16 Fractal Fract. 3.2. Fractional Optimal Control of the Model , 8, verify C 0Dα t ξ1(t′) = ϱα 1 I1(m −1)(ξ1 −ξ2) + aα 1(ξ5 −ξ1) + B1βα 1(ξ1 −ξ2)(u −1) −B1µα 1ξ1 −Kµα 1ξ1 B1 + K + (ξ3 −ξ1)v, ξ1(t′) = ϱα 1 I1(m −1)(ξ1 −ξ2) + aα 1(ξ5 −ξ1) + B1βα 1(ξ1 −ξ2)(u −1) −B1µα 1ξ1 −Kµα 1ξ1 + C 0Dα t ξ3(t′) = ωα(ξ1 −ξ3) −(µα 1 + cα 1)ξ3 + cα 1ξ7, C 0Dα t ξ4(t′) = K(−Kξ4Q2 + βα 1(ξ1 −ξ2)S1(u −1)) −B2 1ξ4Q2 −2B1Kξ4Q2 (B1 + K)2 , C 0Dα t ξ5(t′) = aα 2(ξ1 −ξ5) + ϱα 2 I2(m −1)(ξ5 −ξ6) + B2βα 2(ξ5 −ξ6)(u −1) −B2µα 2ξ5 −Kµα 2ξ5 B2 + K + (ξ7 −ξ5)v, C 0Dα t ξ6(t′) = k2 + bα 2ξ2 + γα 2ξ7 + ϱα 2(m −1)ξ5S2 −ξ6(Q3 + ϱα 2(m −1)S2) + ξ8σα 2 , C 0Dα t ξ7(t′) = ωαξ5 + cα 2ξ3 −(µα 2 + ωα + cα 2)ξ7, C 0Dα t ξ8(t′) = K(−Kξ8Q4 + βα 2(ξ5 −ξ6)S2(u −1)) −B2 2ξ8Q4 −2B2Kξ8Q4 (B2 + K)2 , (12 0 t ξ ( ) (ξ ξ ) (µ1 1)ξ 1ξ C 0Dα t ξ4(t′) = K(−Kξ4Q2 + βα 1(ξ1 −ξ2)S1(u −1)) −B2 1ξ4Q2 −2B1Kξ4Q2 (B1 + K)2 , C 0Dα t ξ5(t′) = aα 2(ξ1 −ξ5) + ϱα 2 I2(m −1)(ξ5 −ξ6) + B2βα 2(ξ5 −ξ6)(u −1) −B2µα 2ξ5 −Kµα 2ξ5 B2 + K + (ξ7 −ξ5)v, CDαξ6(t′) = k2 + bαξ2 + γαξ7 + ϱα(m −1)ξ5S2 −ξ6(Q3 + ϱα(m −1)S2) (12) (12) C 0Dα t ξ7(t′) = ωαξ5 + cα 2ξ3 −(µα 2 + ωα + cα 2)ξ7, C 0Dα t ξ8(t′) = K(−Kξ8Q4 + βα 2(ξ5 −ξ6)S2(u −1)) −B2 2ξ8Q4 −2B2Kξ8Q4 (B2 + K)2 , with t′ = t f −t. In turn, the optimality conditions of PMP establish that the optimal controls u∗, v∗, and m∗are deﬁned by with t′ = t f −t. In turn, the optimality conditions of PMP establish that the optimal controls u∗, v∗, and m∗are deﬁned by u∗= min max 0, B1βα 1(ξ2 −ξ1)S1 2k3(B1 + K) + B2βα 2(ξ6 −ξ5)S2 2k3(B2 + K) , umax , v∗= min max 0, S1(ξ1 −ξ3) + S2(ξ5 −ξ7) 2k4 , vmax , m∗= min max 0, ϱα 1 I1(ξ2 −ξ1)S1 + ϱα 2 I2(ξ6 −ξ5)S2 2k5 , mmax . 3.2. Fractional Optimal Control of the Model 2021, 5, 261 H = k1I1 + k2I2 + k3u2 + k4v2 + k5m2 + ξ1 πα 1 + aα 2S2 + ωαR1 −(1 −u) βα 1B1S1 K + B1 −(1 −m)ϱα 1 I1S1 −(aα 1 + µα 1 + v)S1 + ξ2 (1 −u) βα 1B1S1 K + B1 + (1 −m)ϱα 1 I1S1 + bα 2 I2 −Q1I1 H = k1I1 + k2I2 + k3u2 + k4v2 + k5m2 + ξ1 πα 1 + aα 2S2 + ωαR1 −(1 −u) βα 1B1S1 K + B1 −(1 −m)ϱα 1 I1S1 −(aα 1 + µα 1 + v)S1 + ξ2 (1 −u) βα 1B1S1 K + B1 + (1 −m)ϱα 1 I1S1 + bα 2 I2 −Q1I1 + ξ3(vS1 + γα 1 I1 −(µα 1 + ωα + cα 1)R1 + cα 2R2) + ξ4(σα 1 I1 −Q2B1) + ξ5 πα 2 + aα 1S1 + ωαR2 −(1 −u) βα 2B2S2 K + B2 −(1 −m)ϱα 2 I2S2 −(aα 2 + µα 2 + v)S2 + ξ6 (1 −u) βα 2B2S2 K + B2 + (1 −m)ϱα 2 I2S2 + bα 1 I1 −Q3I2 + ξ7(vS2 + γα 2 I2 −(µα 2 + ωα + cα 2)R2 + cα 1R1) + ξ8(σα 2 I2 −Q4B2) (11 (11) + ξ6 (1 −u) βα 2B2S2 K + B2 + (1 −m)ϱα 2 I2S2 + bα 1 I1 −Q3I2 + ξ7(vS2 + γα 2 I2 −(µα 2 + ωα + cα 2)R2 + cα 1R1) + ξ8(σα 2 I2 −Q4B2) and the adjoint system asserts that the co-state variables ξi(t), i = 1, . . . 3.3. Numerical Results and Cost-Effectiveness Analysis 3.3. Numerical Results and Cost-Effectiveness Analysis We start by calculating the relevance of the three measures used in the control of the disease. We do it by using the sensitivity index, presented in Deﬁnition 1, as proposed in [34]. In this case, the sensitivity indices are presented as functions of the control param- eters in Figure 4, using the parametric values from Table 1 and considering the classical model (that is, α = 1). The resulting graphics show that: (a) the curve of vaccination is the one that most rapidly moves away from zero, meaning that the vaccination program has a big impact for small rates of application; (b) proper hygiene measures are also important in the control of cholera, having a bigger impact for greater rates of application of it, being the only control that has precisely the same impact in both communities; (c) the domestic water treatment is useless in the control of cholera transmission, when used simultaneously with the two previous controls. So, in what follows, the third control, variable u, is ignored. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 10-5 R01 R02 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 R01 R02 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 -12 -10 -8 -6 -4 -2 0 R01 R02 Figure 4. Sensitivity index of the basic reproduction numbers with respect to the control variables u (top left), v (top right), and m (bottom). 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 R01 R02 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 -1.2 -1 -0.8 -0.6 -0.4 -0.2 0 10-5 R01 R02 Figure 4. Sensitivity index of the basic reproduction numbers with respect to the control variables u (top left), v (top right), and m (bottom). One can deal with many different problems arising in different ﬁelds of sciences and engineering by applying some appropriate discretisation [35]. Here, the Pontryagin Maximum Principle is used to numerically solve the optimal control problem, as discussed in Section 3.2, both in classical and fractional cases, using the predict-evaluate-correct- evaluate (PECE) method of Adams–Basforth–Moulton [36], coded by us in MATLAB. 3.2. Fractional Optimal Control of the Model (13) (13) In addition, the following transversality conditions hold: In addition, the following transversality conditions hold: In addition, the following transversality conditions hold: ξi(t f ) = 0, i = 1, . . . , 8. (14) (14) Fractal Fract. 2021, 5, 261 8 of 16 3.3. Numerical Results and Cost-Effectiveness Analysis Firstly, we solve system (3)–(4) by the PECE procedure, with the initial values for the state variables as given in Table 3 and a guess for the controls over the time interval [0, t f ], that way obtaining the values of the state variables. Similarly to [16], a change of variable is applied to the adjoint system and to the transversality conditions, obtaining the fractional initial value problem (12)–(14). Such IVP is also solved with the PECE algorithm, and the values of the co-state variables ξi, i = 1, . . . , 8, are determined. The controls are then updated by a convex combination of the controls of previous iteration and the current values, computed according to (13). This procedure is repeated iteratively until the values of all the variables and the values of the controls are almost identical to the ones of the previous iteration. The solutions of the classical model were successfully conﬁrmed by a classical forward–backward scheme, also implemented by us in MATLAB. Fractal Fract. 2021, 5, 261 9 of 16 In the numerical experiments, we consider weights k1 = 4, k2 = 2.4, k3 = 1.6, and k4 = k5 = 1. These values have the same relations between them than the homonyms weights in [11], with the numerical advantage of being closer to the value one. We also use vmax = mmax = 1, while the other parameters are ﬁxed according to Table 1. p g According with Figure 1, R01 > 1 and R02 > 1 for α ⩾0.68 (endemic scenario). In order to be able to compare the FOCP results for several derivative orders, we consider the initial conditions, given by Table 3, which correspond to the non-trivial endemic equilibrium for system (3) and (4) for the classical cholera model (α = 1). Table 3. Initial conditions for the fractional optimal control problem of Section 3.2 with parameters given by Table 1, corresponding to the endemic equilibrium of cholera model (3)–(4) with classical derivative order. S1(0) I1(0) R1(0) B1(0) S2(0) I2(0) R2(0) B2(0) 0.53144 0.001997 0.01028 0.30254 0.44222 0.002380 0.01082 0.36065 Without loss of generality, we consider the fractional order derivatives α = 1.0, 0.9 and 0.8. In Figures 5–7, we ﬁnd the solutions of the fractional optimal control problem for those values of α. 3.3. Numerical Results and Cost-Effectiveness Analysis We can see that a change in the value of α corresponds to signiﬁcant variations of the state and control variables. Beyond those values of α, others values were also tested, but the results do not changed qualitatively. g q y The efﬁcacy function [37], exhibited in Figure 8, is deﬁned as F(t) = i(0) −i∗(t) i(0) = 1 −i∗(t) i(0) , (15) (15) where i∗(t) = I∗ 1 (t) + I∗ 2 (t) is the optimal solution associated with the fractional optimal control problem and i(0) = I1(0) + I2(0) is the correspondent initial condition. This function measures the proportional variation in the number of infected individuals, of both communities, after the application of the control measures, {v∗, m∗}, by comparing the number of infectious individuals at time t with its initial value i(0). We observe that the graphic of F(t) exhibits the inverse tendency of infected individuals curves, growing and reaching the maximum at the end of the time interval. To assess the cost and the effectiveness of the proposed fractional control measure during the intervention period, some summary measures are presented. The total cases averted by the intervention during the time period t f is deﬁned in [37] by AV = t f i(0) − Z t f 0 i∗(t) dt, (16) (16) where i∗(t) is the optimal solution associated with the fractional optimal controls and i(0) is the correspondent initial condition. Note that this initial condition is obtained as the equilibrium proportion i of systems (3) and (4), which is independent of time, so that Z t f t f i(0) = Z t f 0 i dt represents the total infectious cases over a given period of t f days. Z 0 Effectiveness is deﬁned as the proportion of cases averted on the total possible cases under no intervention [37]: F = AV i(0)t f = 1 − Z t f 0 i∗(t) dt i(0)t f . (17) (17) The total cost associated with the intervention is deﬁned in [37] by The total cost associated with the intervention is deﬁned in [37] by TC = Z t f 0 (C1 v∗(t)s∗(t) + C2 m∗(t)i∗(t))dt, (18) (18) Fractal Fract. 3.3. Numerical Results and Cost-Effectiveness Analysis 2021, 5, 261 10 of 16 10 of 16 where s∗(t) = S∗ 1(t) + S∗ 2(t) and Ci corresponds to the per person unit cost of the two possible interventions: (i) vaccination at any time t of susceptible individuals (C1); and (ii) infected individuals practising proper hygiene (C2). Following [37,38], the average cost-effectiveness ratio is given by ACER = TC AV . (19) (19) The cost-effectiveness measures are summarised in Table 4. The results show the effectiveness of controls to reduce cholera infectious individuals and the leadership in doing so by the classical model (α = 1). Table 4. Summary of cost-effectiveness measures for classical and fractional (0 < α < 1) cholera disease optimal control problems. Parameters according to Tables 1 and 3 with C1 = C2 = 1. α AV TC ACER F 1.0 0.316716 0.90049 2.84322 0.723582 0.9 0.297175 1.17595 3.95708 0.678938 0.8 0.280311 1.35978 4.85099 0.640408 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 0.002 I1(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0005 0.001 0.0015 0.002 0.0025 I2(t) =1 =0.9 =0.8 (b) Figure 5. Variables I1 and I2 of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Evolution of infected individuals of 1st community; (b) Evolution of infected individuals of 2nd community. Table 4. Summary of cost-effectiveness measures for classical and fractional (0 < α < 1) cholera disease optimal control problems. Parameters according to Tables 1 and 3 with C1 = C2 = 1. Table 4. Summary of cost-effectiveness measures for classical and fractional (0 < α < 1) cholera disease optimal control problems. Parameters according to Tables 1 and 3 with C1 = C2 = 1. Table 4. Summary of cost-effectiveness measures for classical and fractional (0 < α < 1) cholera disease optimal control problems. Parameters according to Tables 1 and 3 with C1 = C2 = 1. 3.3. Numerical Results and Cost-Effectiveness Analysis α AV TC ACER F 1.0 0.316716 0.90049 2.84322 0.723582 0.9 0.297175 1.17595 3.95708 0.678938 0.8 0.280311 1.35978 4.85099 0.640408 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 0.002 I1(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0005 0.001 0.0015 0.002 0.0025 I2(t) =1 =0.9 =0.8 (b) Figure 5. Variables I1 and I2 of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Evolution of infected individuals of 1st community; (b) Evolution of infected individuals of 2nd community. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 0.002 I1(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0005 0.001 0.0015 0.002 0.0025 I2(t) =1 =0.9 =0.8 (b) (b) (a) Figure 5. Variables I1 and I2 of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Evolution of infected individuals of 1st community; (b) Evolution of infected individuals of 2nd community. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 B1(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 B1(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 B2(t) =1 =0.9 =0.8 (b) Figure 6. Variables B1 and B2 of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Variation of vibrio population in 1st community; (b) Variation of vibrio population in 2nd community. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 B2(t) =1 =0.9 =0.8 (b) (b) (a) Figure 6. Variables B1 and B2 of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Variation of vibrio population in 1st community; (b) Variation of vibrio population in 2nd community. 11 of 16 Fractal Fract. 3.3. Numerical Results and Cost-Effectiveness Analysis 2021, 5, 261 0 10 20 30 40 50 60 70 80 90 100 time 0 0.02 0.04 0.06 0.08 0.1 0.12 v(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 m(t) =1 =0.9 =0.8 (b) Figure 7. Control variables v and m of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Contour of vaccination, v; (b) Evolution of control of hygiene, m. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.02 0.04 0.06 0.08 0.1 0.12 v(t) =1 =0.9 =0.8 (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 m(t) =1 =0.9 =0.8 (b) (b) (a) Figure 7. Control variables v and m of the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. (a) Contour of vaccination, v; (b) Evolution of control of hygiene, m. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 F(t) =1 =0.9 =0.8 0 10 20 30 40 50 60 70 80 90 100 time 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 F(t) =1 =0.9 =0.8 Figure 8. Evolution of the efﬁcacy function (15) for the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. Figure 8. Evolution of the efﬁcacy function (15) for the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. Figure 8. Evolution of the efﬁcacy function (15) for the FOCP (8)–(13) with values from Table 1 and fractional order derivatives α = 1.0, 0.9 and 0.8. 3.3.1. Optimal Control Strategies and Cost-Effectiveness Analysis 3.3.1. Optimal Control Strategies and Cost-Effectiveness Analysis Now, we analyse the cost-effectiveness of alternative combinations of two possible control measures: • strategy A—implementing the vaccination control, v; • strategy B—implementing proper hygiene control, m; • strategy C—implementing both controls, v and m. • strategy C—implementing both controls, v and m. To analyse the cost-effectiveness of the three alternative strategies A, B, and C, we use the Incremental Cost-Effectiveness Ratio (ICER) [38]. This ratio is used to compare the dif- ferences between costs and health outcomes of two alternative intervention strategies that compete for the same resources, being often described as the additional cost per additional health outcome. We start by ranking the strategies in order of increasing effectiveness, assessed by the total averted cases AV, deﬁned in (16). The ICER for the classical model (α = 1), is calculated as follows: ICER(B) =ACER(B) = 0.216276, ICER(A) =0.900865 −0.0084106 0.316411 −0.038888 = 3.2157877, ICER(C) =1.900494 −0.900865 0.316716 −0.316411 = −1.216393. Results are shown in Table 5. Strategy A is the most costly one, so we exclude this strat- egy from the set of alternatives. We align the remaining strategies by increasing effectiveness Fractal Fract. 2021, 5, 261 12 of 16 12 of 16 (AV) and recalculate the ICER: ICER(B) = ACER(B) = 0.216276 and ICER(C) = 3.210922. Hence, we conclude that strategy B (implementing only the control of hygiene measure, m) is the most cost-effective strategy. (AV) and recalculate the ICER: ICER(B) = ACER(B) = 0.216276 and ICER(C) = 3.210922. Hence, we conclude that strategy B (implementing only the control of hygiene measure, m) is the most cost-effective strategy. Table 5. Incremental cost-effectiveness ratio for strategies A, B, and C. Parameters according to Tables 1 and 3 with C1 = C2 = 1 and α = 1. Strategies AV TC ACER ICER B 0.038888 0.0084106 0.216276 0.216276 A 0.316411 0.900865 2.84713 3.2157877 C 0.316716 0.900494 2.84322 −1.216393 Table 5. Incremental cost-effectiveness ratio for strategies A, B, and C. Parameters according to Tables 1 and 3 with C1 = C2 = 1 and α = 1. The ICER was also computed for the fractional model, considering the same strategies, with the three derivative orders used previously: α = 0.9, α = 0.8 and α = 0.68. Even in those cases, we obtain the same conclusion: strategy B is the most cost-effective. 3.3.1. Optimal Control Strategies and Cost-Effectiveness Analysis gy Comparing the cost-effectiveness of strategy B for above derivative orders, we start by getting the values of ICER presented in Table 6. Once the values of Total Cost (TC) diminish with the decrease in the derivative order, proceeding as above, when the values of ICER were computed for the classical α = 1 model, we eliminate successively the scheme with the highest TC. Consequently, the “strategies” α = 1.0, α = 0.9, and α = 0.8 are excluded, by this order. y Our conclusion is: the most cost-effective scheme is strategy B with α = 0.68. Table 6. Incremental cost-effectiveness ratio for strategy B and several derivative orders. Same conditions as Table 5. α AV TC ICER 1.0 0.038888 0.0084106 0.216276 0.9 0.147496 0.003549 −22.3419 0.8 0.203782 0.001705 −30.5191 0.68 0.237211 0.000845 −38.859 Table 6. Incremental cost-effectiveness ratio for strategy B and several derivative orders. Same conditions as Table 5. 3.3.2. The Variable-Order FractInt System Table 7. Cost-effectiveness measures for the FractInt system. Parameters according to Tables 1 and 3 with C1 = C2 = 1. AV TC ACER F 0.332078 1.10967 3.3416 0.758679 Table 7. Cost-effectiveness measures for the FractInt system. Parameters according to Tables 1 and 3 with C1 = C2 = 1. Our results show that it is effective to control cholera infection through optimal control and that the FracInt model is more efﬁcient than the classical one (cf. Table 4 and Figure 11), being the most effective model. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 0.002 I1(t) =1 =0.68 (t) (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0005 0.001 0.0015 0.002 0.0025 I2(t) =1 =0.68 (t) (b) Figure 9. Comparison of variables I1 and I2 of the FractInt system with the ones of the FOCP (8)–(13) with α = 1 and α = 0.68, considering the parameter values from Table 1. (a) Infected individuals of 1st community; (b) Infected individuals of 2nd community. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0005 0.001 0.0015 0.002 0.0025 I2(t) =1 =0.68 (t) (b) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 0.002 I1(t) =1 =0.68 (t) (a) (a) (b) Figure 9. Comparison of variables I1 and I2 of the FractInt system with the ones of the FOCP (8)–(13) with α = 1 and α = 0.68, considering the parameter values from Table 1. (a) Infected individuals of 1st community; (b) Infected individuals of 2nd community. Figure 9. Comparison of variables I1 and I2 of the FractInt system with the ones of the FOCP (8)–(13) with α = 1 and α = 0.68, considering the parameter values from Table 1. (a) Infected individuals of 1st community; (b) Infected individuals of 2nd community. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.02 0.04 0.06 0.08 0.1 0.12 v(t) =1 =0.68 (t) (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 m(t) =1 =0.68 (t) (b) Figure 10. 3.3.2. The Variable-Order FractInt System 3.3.2. The Variable-Order FractInt System The solution of the Fractional Optimal Control Problem, exhibited in Section 3.3 for some derivative order values, evidences that the fractional-order model can be more effective in part of the time interval—as shown in Figure 8—but classical model is more effective if we consider all the interval, as shown in Table 4. In this section, we consider that the derivative order varies along the interval, being fractional or classical when more advantageous. According with Figure 8, α should be fractional at beginning and become classical/one after a certain time. Such class of Variable-Order Fractional (VOF) systems [39] are baptised here as FractInt. The derivative order of the FractInt system varies according with α(t) = ( α0 if 0 ⩽t ⩽t′, 1 if t′ < t ⩽100, (20) (20) where 0 < α0 < 1. In practice, we noticed that the resulting system is as more effective as smaller the value of α0. In view of an endemic scenario, we consider in our simulations α0 = 0.68, which the lowest value that can guarantee it. With respect to the switching time, the value considered is t′ = 7, which is the one to which it corresponds the maximum value of efﬁciency. The FractInt system is numerically solved with the procedure described above, at the beginning of Section 3.3. In each iteration of the procedure, the predict-evaluate-correct- evaluate method is applied successively to each one of the two initial value problems— associated to the state system and to the adjoint system of the FOCP (8)–(13)—to perform Fractal Fract. 2021, 5, 261 13 of 16 the integration of them in two steps. These steps correspond to the two branches of the derivative order function, deﬁned by (20). In such process, the ﬁnal solution of ﬁrst step (ﬁrst branch) is the initial solution of the second step. Solutions of the FractInt system are reproduced in Figures 9 and 10 along with solutions of the classical and fractional models (α = 1 and α = 0.68, respectively). We can see that the solutions of the FracInt system start to follow the solutions of the fractional model and end by following the solutions of the classical model. This behaviour can be observed in Figure 11, where the efﬁcacy function for these three models is displayed. g y p y The cost-effectiveness measures for the FractInt system are summarised in Table 7. 4. Conclusions Nowadays, Cholera infection is still an healthcare problem in many parts of the world, namely in impoverished regions where it has devastating effects. Its control is the goal of many studies in last years. In this work, a fractional-order mathematical model for Cholera with two connected communities, proposed in [10], is studied and generalised. The two communities in appreciation are distinct and therefore they could not be molten in one unique community. A sensitivity analysis of the model is done to show the importance of estimation of parameters. A fractional optimal control (FOC) problem is then formulated and numeri- cally solved. The relevance of studied controls is assessed using a cost-effectiveness analysis. Such analysis allow us to neglect the control u (domestic water treatment) since it proved to be useless when used in combination with remaining controls. The numerical results show that the FOC system is more effective only in part of the time interval. Therefore, we propose a system where the derivative order varies along the time interval, being fractional or integer when more advantageous in the control of infection. Such variable-order fractional model, baptised here as FractInt, shows to be the most effective in the control of the disease. Author Contributions: Conceptualisation, S.R. and D.F.M.T.; methodology, S.R. and D.F.M.T.; soft- ware, S.R.; validation, S.R. and D.F.M.T.; formal analysis, S.R. and D.F.M.T.; investigation, S.R. and D.F.M.T.; writing—original draft preparation, S.R. and D.F.M.T.; writing—review and editing, S.R. and D.F.M.T.; visualisation, S.R. and D.F.M.T. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by Fundação para a Ciência e a Tecnologia (FCT, the Portuguese Foundation for Science and Technology) through IT, Grant Number UIDB/50008/2020 (S.R.), and CIDMA, Grant Number UIDB/04106/2020 (D.F.M.T.). Institutional Review Board Statement: Not applicable. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Data Availability Statement: Not applicable. Acknowledgments: The authors would like to thank John B. H. Njagarah, from Botswana Interna- tional University of Science and Technology, for having provided most of the values of parameters used in this work, which are presented in Table 1. Conﬂicts of Interest: The authors declare no conﬂict of interest. 3.3.2. The Variable-Order FractInt System Comparison of variables v and m of the FractInt system with the ones of the FOCP (8)–(13) with α = 1 and α = 0.68, considering the parameter values from Table 1. (a) Contour of vaccination, v; (b) Evolution of the hygiene control, m. 0 10 20 30 40 50 60 70 80 90 100 time 0 0.02 0.04 0.06 0.08 0.1 0.12 v(t) =1 =0.68 (t) (a) 0 10 20 30 40 50 60 70 80 90 100 time 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 m(t) =1 =0.68 (t) (b) (a) (b) Figure 10. Comparison of variables v and m of the FractInt system with the ones of the FOCP (8)–(13) with α = 1 and α = 0.68, considering the parameter values from Table 1. (a) Contour of vaccination, v; (b) Evolution of the hygiene control, m. 14 of 16 Fractal Fract. 2021, 5, 261 0 10 20 30 40 50 60 70 80 90 100 time 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 F(t) =1 =0.68 (t) Figure 11. Evolution of the efﬁcacy function for the FractInt system and for the FOCP with α = 1 and α = 0.68, considering the parameter values from Table 1. Figure 11. Evolution of the efﬁcacy function for the FractInt system and for the FOCP with α = 1 and α = 0.68, considering the parameter values from Table 1. 4. Conclusions The funders had no role in the design of the study; in the collection, analyses, or interpretation of the data; in the writing of the manuscript; nor in the decision to publish the results. 15 of 16 Fractal Fract. 2021, 5, 261 References Boukhouima, A.; Lotﬁ, E.M.; Mahrouf, M.; Rosa, S.; Torres, D.F.M.; Yousﬁ, N. Stability analysis and optimal control of a fractional HIV-AIDS epidemic model with memory and general incidence rate. Eur. Phys. J. Plus 2021, 136, 1–20. [CrossRef] 13. Boukhouima, A.; Lotﬁ, E.M.; Mahrouf, M.; Rosa, S.; Torres, D.F.M.; Yousﬁ, N. Stability analysis and optimal control of a fractional HIV-AIDS epidemic model with memory and general incidence rate. Eur. Phys. J. Plus 2021, 136, 1–20. [CrossRef] 14. Ndaïrou, F.; Torres, D.F.M. Mathematical analysis of a fractional COVID-19 model applied to Wuhan, Spain and Portugal. Axioms p y g y J , , 14. Ndaïrou, F.; Torres, D.F.M. Mathematical analysis of a fractional COVID-19 model applied to Wuhan, Sp 2021, 10, 135. [CrossRef] 15. Ammi, M.R.S.; Tahiri, M.; Torres, D.F.M. Global stability of a Caputo fractional SIRS model with general incidence rate. Math. Comput. Sci. 2021, 15, 91–105. [CrossRef] p 16. Rosa, S.; Torres, D.F.M. Optimal control of a fractional order epidemic model with application to human respiratory syncytial virus infection. Chaos Solitons Fractals 2018, 117, 142–149. [CrossRef] 17. Almeida, R. Analysis of a fractional SEIR model with treatment. Appl. Math. Lett. 2018, 84, 56–62. [Cro alysis of a fractional SEIR model with treatment. Appl. Math. Lett. 2018, 84, 56–62. [CrossRef] 17. Almeida, R. Analysis of a fractional SEIR model with treatment. Appl 17. Almeida, R. Analysis of a fractional SEIR model with treatment. Appl. Math. Lett. 2018, 84, 56–62. [CrossRef] 17. Almeida, R. Analysis of a fractional SEIR model with treatment. Appl. Math. Lett. 2018, 84, 56–62. [CrossRef] 18 C lh A R M Pi t C M A I i HIV id i f di ti t t i i t d f t t d CT 18. Carvalho, A.R.M.; Pinto, C.M.A. Immune response in HIV epidemics for distinct transmission rate response. Math. Model. Nat. Phenom. 2019, 14, 307. [CrossRef] 19. van den Driessche, P.; Watmough, J. Reproduction numbers and sub-threshold endemic equilibria for compartmental models of disease transmission. Math. Biosci. 2002, 180, 29–48. [CrossRef] 20. Statistics South Africa. Community Survey 2016 Provincial Proﬁle: Kwazulu Natal. 2016. Available online: http://cs2016.statssa. gov.za/wp-content/uploads/2018/07/KZN.pdf (accessed on 23 July 2019). g p p p y 21. Jamison, D.T.; Feachem, R.G.; Makgoba, M.W.; Bos, E.R.; Baingana, F.K.; Hofman, K.J.; Rogo, K.O. (Eds.) Disease and Mortality in Sub-Saharan Africa; World Bank Publications: Washington, DC, USA, 2006. 22. Kumate, J.; Sepúlveda, J.; Gutiérrez, G. References 1. Podlubny, I. Fractional Differential Equations; Mathematics in Science and Engineering; Academic Press, Inc.: San Diego, CA, USA, 1999; Volume 198. 1. Podlubny, I. Fractional Differential Equations; Mathematics in Science and Engineering; Academic Press, Inc.: San Diego, CA, USA, 1999; Volume 198. 2. Arqub, O.A. Computational algorithm for solving singular Fredholm time-fractional partial integrodifferential equations with error estimates. J. Appl. Math. Comput. 2019, 59, 227–243. [CrossRef] pp p gfeh, N.; Abu Arqub, O. A fractional Tikhonov regularization method for an inverse backward and source e-space fractional diffusion equations. Chaos Solitons Fractals 2021, 150, 111127. [CrossRef] pp p 3. Djennadi, S.; Shawagfeh, N.; Abu Arqub, O. A fractional Tikhonov regularization method for an inve problems in the time-space fractional diffusion equations. Chaos Solitons Fractals 2021, 150, 111127. [Cr 4. Mukandavire, Z.; Liao, S.; Wang, J.; Gaff, H.; Smith, D.L.; Morris, J.G. Estimating the reproductive numbers for the 2008–2009 Cholera outbreaks in Zimbabwe. Proc. Natl. Acad. Sci. USA 2011, 108, 8767–8772. [CrossRef] [ ] 5. Lemos-Paião, A.P.; Silva, C.J.; Torres, D.F.M. An epidemic model for cholera with optimal control treatment. J. Comput. Appl. Math. 2017, 318, 168–180. [CrossRef] 6. Lemos-Paião, A.P.; Silva, C.J.; Torres, D.F.M. A cholera mathematical model with vaccination and the biggest outbreak of world’s history. AIMS Math. 2018, 3, 448–463. [CrossRef] 7. Lemos-Paião, A.P.; Silva, C.J.; Torres, D.F.M.; Venturino, E. Optimal control of aquatic diseases: A case study of Yemen’s cholera outbreak. J. Optim. Theory Appl. 2020, 185, 1008–1030. [CrossRef] 8. Codeço, C.T. Endemic and epidemic dynamics of Cholera: The role of the aquatic reservoir. BMC Infect. Diseases 2001, 1, 14. [CrossRef] [PubMed] 8. Codeço, C.T. Endemic and epidemic dynamics of Cholera: The role of the aquatic reservoir. BMC Infect. Diseases 2001, 1, 14. [CrossRef] [PubMed] 9. Capasso, V.; Paveri-Fontana, S. Mathematical model for the 1973 Cholera epidemic in the european mediterranean region. Revue 9. Capasso, V.; Paveri-Fontana, S. Mathematical model for the 1973 Cholera epidemic in the european med d’Epidemiologie et de Sante Publique 1979, 27, 121–132. p g q 10. Njagarah, J.B.H.; Tabi, C.B. Spatial synchrony in fractional order metapopulation Cholera transmission. Chaos Solitons Fractals 2018, 117, 37–49. [CrossRef] 11. Njagarah, J.B.H.; Nyabadza, F. Modelling optimal control of Cholera in communities linked by migration. Comput. Math. Methods Med. 2015, 2015, 898264. [CrossRef] [PubMed] 12. RNeilan, L.M.; Schaefer, E.; Gaff, H.; Fister, K.R.; Lenhart, S. Modeling optimal intervention strategies for Cholera. Bull. Math. Biol. 2010, 72, 2004–2018. 13. 31. Abraha, T.; al Basir, F.; Obsu, L.L.; Torres, D.F.M. Farming awareness based optimum interventions for crop pest control. Math. Biosci. Eng. 2021, 18, 5364–5391. [CrossRef] [PubMed] , , 33. Bergounioux, M.; Bourdin, L. Pontryagin maximum principle for general Caputo fractional optimal con cost and terminal constraints. ESAIM Control Optim. Calc. Var. 2020, 26, 35. [CrossRef] References Cholera epidemiology in latin america and perspectives for eradication. Bull. l’Institut Pasteur 1998, 96, 217–226. [CrossRef] , E.L.; Pascual, M.; Bouma, M.J. Inapparent infections and Cholera dynamics. Nature 2008, 454, 877–880 d] 23. King, A.A.; Ionides, E.L.; Pascual, M.; Bouma, M.J. Inapparent infections and Cholera dynamics. N [CrossRef] [PubMed] 24. Hartley, D.M.; Morris, J.G., Jr.; Smith, D.L. Hyperinfectivity: A critical element in the ability of v. cholerae to cause epidemics? PLoS Med. 2005, 3, 7. [CrossRef] [PubMed] [ ] [ ] 25. Mukandavire, Z.; Tripathi, A.; Chiyaka, C.; Musuka, G.; Nyabadza, F.; Mwambi, H. Modelling and analysis of the intrinsic dynamics of Cholera. Differ. Equ. Dyn. Syst. 2011, 19, 253–265. [CrossRef] ff q y y lwell, R.R. Fate of vibrio Cholerae o1 in seawater microcosms. Water Res. 1996, 30, 47–50. [CrossRef] 26. Munro, P.M.; Colwell, R.R. Fate of vibrio Cholerae o1 in seawater microcosms. Water Res. 1996, 30, 27. Chitnis, N.; Hyman, J.M.; Cushing, J.M. Determining important parameters in the spread of malaria through the sensitivity analysis of a mathematical model. Bull. Math. Biol. 2008, 70, 1272–1296. [CrossRef] [PubMed] y [ ] [ ] 28. Rodrigues, H.S.; Monteiro, M.T.T.; Torres, D.F.M. Sensitivity analysis in a dengue epidemiological model. Conf. Pap. Math. 2013, 2013, 721406. [CrossRef] [ ] 29. Mikucki, M.A. Sensitivity Analysis of the Basic Reproduction Number and Other Quantities for Infectious Disease Models. Ph.D. Thesis, Colorado State University, Fort Collins, CO, USA, 2012. y 30. Arqub, O.A.; Shawagfeh, N. Solving optimal control problems of Fredholm constraint optimality via the reproducing kernel Hilbert space method with error estimates and convergence analysis. Math. Methods Appl. Sci. 2021, 44, 7915–7932. [CrossRef] Fractal Fract. 2021, 5, 261 16 of 16 31. Abraha, T.; al Basir, F.; Obsu, L.L.; Torres, D.F.M. Farming awareness based optimum interventions for crop pest control. Math. Biosci. Eng. 2021, 18, 5364–5391. [CrossRef] [PubMed] 31. Abraha, T.; al Basir, F.; Obsu, L.L.; Torres, D.F.M. Farming awareness based optimum interventions for crop pest control. Math. Biosci. Eng. 2021, 18, 5364–5391. [CrossRef] [PubMed] g [ ] [ ] 32. Almeida, R.; Pooseh, S.; Torres, D.F.M. Computational Methods in the Fractional Calculus of Variations; Imperial College Press: London, UK, 2015. g 32. Almeida, R.; Pooseh, S.; Torres, D.F.M. Computational Methods in the Fractional Calculus of Variations; Imperial College Press: London, UK, 2015. 33. Bergounioux, M.; Bourdin, L. Pontryagin maximum principle for general Caputo fractional optimal control problems with Bolza cost and terminal constraints. g 32. Almeida, R.; Pooseh, S.; Torres, D.F.M. Computational Methods in the Fractional Calculus of Variations; Imperial College Press: London, UK, 2015. References ESAIM Control Optim. Calc. Var. 2020, 26, 35. [CrossRef] 33. Bergounioux, M.; Bourdin, L. Pontryagin maximum principle for general Caputo fractional optimal control problems with Bolza cost and terminal constraints. ESAIM Control Optim. Calc. Var. 2020, 26, 35. [CrossRef] j p y p p y 35. Arqub, O.A.; Rashaideh, H. The RKHS method for numerical treatment for integrodifferential algebraic systems of temporal two-point BVPs. Neural Comput. Appl. 2018, 30, 2595–2606. [CrossRef] 35. Arqub, O.A.; Rashaideh, H. The RKHS method for numerical treatment for integrodifferential algebraic systems of temporal two-point BVPs. Neural Comput. Appl. 2018, 30, 2595–2606. [CrossRef] p p pp [ ] 36. Diethelm, K.; Ford, N.J.; Freed, A.D.; Luchko, Y. Algorithms for the fractional calculus: A selection of numerical methods. Comput. Methods Appl. Mech. Eng. 2005, 194, 743–773. [CrossRef] p p pp 36. Diethelm, K.; Ford, N.J.; Freed, A.D.; Luchko, Y. Algorithms for the fractional calculus: A selection of numerical methods. Comput. Methods Appl. Mech. Eng. 2005, 194, 743–773. [CrossRef] 36. Diethelm, K.; Ford, N.J.; Freed, A.D.; Luchko, Y. Algorithm Methods Appl. Mech. Eng. 2005, 194, 743–773. [CrossRef] pp g 37. Rodrigues, P.; Silva, C.J.; Torres, D.F.M. Cost-effectiveness analysis of optimal control measures for tuberculosis. Bull. Math. Biol. 2014, 76, 2627–2645. [CrossRef] [PubMed] pp g 37. Rodrigues, P.; Silva, C.J.; Torres, D.F.M. Cost-effectiveness analysis of optimal control measures for tuberculosis. Bull. Math. Biol. 2014, 76, 2627–2645. [CrossRef] [PubMed] 38. Okosun, K.O.; Rachid, O.; Marcus, N. Optimal control strategies and cost-effectiveness analysis of a malaria model. BioSystems 2013, 111, 83–101. [CrossRef] [PubMed] 38. Okosun, K.O.; Rachid, O.; Marcus, N. Op 2013, 111, 83–101. [CrossRef] [PubMed] 38. Okosun, K.O.; Rachid, O.; Marcus, N. Optimal control strategies and cost-effectiveness analysis of a malaria model. BioSystems 2013, 111, 83–101. [CrossRef] [PubMed] 39. Almeida, R.; Tavares, D.; Torres, D.F.M. T
https://openalex.org/W4289144016	https://zenodo.org/records/1928684/files/%D0%9B%D0%A4%D0%94.pdf	Russian	null	Измерение фактора избыточного шума лавинного фотодиода Hamamatsu APD S8664-1010	Zenodo (CERN European Organization for Nuclear Research)	2,018	cc-by	4,154	МИНИСТЕРСТВО ОБРАЗОВАНИЯ И НАУКИ РОССИЙСКОЙ ФЕДЕРАЦИИ ФЕДЕРАЛЬНОЕ ГОСУДАРСТВЕННОЕ АВТОНОМНОЕ ОБРАЗОВАТЕЛЬНОЕ УЧРЕЖДЕНИЕ ВЫСШЕГО ОБРАЗОВАНИЯ «НОВОСИБИРСКИЙ НАЦИОНАЛЬНЫЙ ИССЛЕДОВАТЕЛЬСКИЙ ГОСУДАРСТВЕННЫЙ УНИВЕРСИТЕТ». Усков А. А. Научный руководитель: с.н.с. ИЯФ СО РАН, к. ф.-м. н. Д. А. Епифано «Измерение фактора избыточного шума лавинного фотодиода Hamamatsu APD S8664-1010» Аннотация Одной из главных подсистем детектора Belle II является электромагнитный калориметр на основе сцинтилляционных кристаллов CsI(Tl). В торцевых частях этого калориметра планируется заменить кристаллы CsI(Tl) на кристаллы чистого CsI. В качестве фоточувствительных элементов в новых счётчиках калориметра планируется использовать кремниевые лавинные фотодиоды (ЛФД) Hamamatsu APD S8664-1010. В данной работе были проведены измерения основных характеристик лавинных фотодиодов Hamamatsu APD S8664-1010: темнового тока, коэффициента усиления и фактора избыточного шума для двух длин волн детектируемого света, 400 нм и 630 нм. Проведенные измерения нужны для определения рабочей точки фотодиода. Также, по результатам измерений фактора избыточного шума планируется выработать требования к внутренней структуре оптимизированных ЛФД. Введение и цели работы В 2018 году в Японии (лаборатория КЕК, г. Цукуба) начнутся эксперименты на Супер В-фабрике (детектор Belle II на асимметричном e+ e- коллайдере SuperKEKB) [1]. Проектная светимость фабрики, 8 · 1035 1 / (см2 · с), позволит в десятки раз увеличить объём экспериментальной информации, набранной в области Upsilon(4S)-мезонного резонанса (энергия в системе центра масс 2E = 10.58 ГэВ) на предыдущих В-фабриках Belle и BABAR [2]. Широкая программа исследований на Супер -фабрике включает всестороннее изучение физики тяжёлых b- и c-кварков, а также тау-лептона. На новом уровне точности будет изучено нарушение СР симметрии в распадах В- и D-мезонов, а также тау-лептонов. Одной из основных задач Супер -фабрики является поиск эффектов, которые не объясняются в рамках Стандартной Модели, поиск т.н. Новой Физики. Работы на Супер - фабрике будут не только конкурировать, но и дополнять исследования, которые проводятся в экспериментах на Большом Адронном Коллайдере в ЦЕРН. Высокая светимость Супер В-фабрики, и, как следствие, высокий уровень пучкового фона предъявляют очень жёсткие требования к детектору Belle II. Одной из главных подсистем Belle II является электромагнитный калориметр на основе сцинтилляционных кристаллов CsI(Tl), в создании и эксплуатации которого принимают активное участие физики из ИЯФ им. Г. И. Будкера СО РАН. Для уменьшения шумов наложения (т.н. pile-up шумов) было предложено заменить кристаллы CsI(Tl) (с временем высвечивания около 1 мкс) на кристаллы чистого CsI (с временем высвечивания около 30 нс) в торцевых частях калориметра. Световыход кристаллов чистого CsI на порядок меньше чем у CsI(Tl), поэтому для преобразования световой вспышки в электрический сигнал 1 Световыход кристаллов чистого CsI на порядок меньше чем у CsI(Tl), поэтому для преобразования световой вспышки в электрический сигнал приходится использовать фоточувствительные элементы с внутренним 1 усилением. Рассматривались два варианта фоточувствительных элементов: вакуумные фотопентоды и кремниевые лавинные фотодиоды (ЛФД). В работе [3] исследовались характеристики счётчика на основе кристалла чистого CsI (с размерами 6 x 6 x 30 см3), пластины со спектросмещающим люминофором NOL-9 [4] и четырёх ЛФД Hamamatsu APD S8664-55 [5]. Были измерены электронные шумы счётчика (в энергетических единицах) на уровне 400 кэВ с 4-канальным зарядочувствительным предусилителем CAEN A1422B045F3 и усилителем-формирователем (CR-4RC) с временем формирования 30 нс, что удовлетворяет проектным требованиям к калориметру. вакуумные фотопентоды и кремниевые лавинные фотодиоды (ЛФД). В работе [3] исследовались характеристики счётчика на основе кристалла чистого CsI (с размерами 6 x 6 x 30 см3), пластины со спектросмещающим люминофором NOL-9 [4] и четырёх ЛФД Hamamatsu APD S8664-55 [5]. Введение и цели работы Были измерены электронные шумы счётчика (в энергетических единицах) на уровне 400 кэВ с 4-канальным зарядочувствительным предусилителем CAEN A1422B045F3 и усилителем-формирователем (CR-4RC) с временем формирования 30 нс, что удовлетворяет проектным требованиям к калориметру. Люминофор NOL-9 эффективно переизлучает сцинтилляционный свет кристалла чистого CsI (с длиной волны 320 нм) в область видимого света с длиной волны 588 нм, где квантовая эффективность ЛФД S8664-55 максимальна. Исторически характеристики ЛФД S8664-1010 приводятся компанией Hamamatsu для света с длиной волны 420 нм (длина волны сцинтилляционного света кристалла вольфрамата свинца (PbWO4)). Целью данной работы было измерить фактор избыточного шума (F) ЛФД Hamamatsu S8664-1010 в зависимости от напряжения смещения для длин волн детектируемого света 400 нм и 630 нм. Фактор избыточного шума кремниевого ЛФД может сильно зависеть от длины волны детектируемого света и вносить заметный вклад в энергетическое разрешение калориметра на основе кристаллов чистого CsI, спектросместителя с NOL-9 и ЛФД Hamamatsu S8664-1010 в области малых энергий гамма-квантов [6]. По результатам измерений F планируется выработать требования к внутренней структуре оптимизированных ЛФД (а именно расстоянию от входного окна ЛФД до плоскости p-n-перехода с высоким электрическим полем) с целью минимизации F. 2 2 Описание экспериментальной установки В измерениях использовались: лавинный фотодиод Hamamatsu S8664-1010, произведённые в ИЯФ - высоковольтный источник напряжения (ВВИ), аттенюатор (АТТ), АЦП, дискриминатор (Д), линия задержки, формирователь импульсов; а также вольтметр B7-22A, вольтметр B7-40/4, источник питания постоянного тока GW Instek GPD-74303S, 2 светодиода: ARL-3214URC-6cd (красный, 620 – 635 nm) и L-7104UVC (ультрафиолето- вый, 385-415 nm), черная металлическая коробка, большая металлическая коробка, два резистора по 10 МОм, один резистор 100 МОм, генератор калибровочного импульса GW Instek AFG-72225, осциллограф Tektronix TDS 220, усилитель-формирователь (УФ) Ortec 570, зарядово-чувствительный предусилитель ORTEC 142IH (ЗЧПУ) (Рис. 1, 2, 3, 4). В измерениях использовались: лавинный фотодиод Hamamatsu S8664-1010, произведённые в ИЯФ - высоковольтный источник напряжения (ВВИ), Рис. 1. Блок-схема установки Рис. 1. Блок-схема установки Рис. 1. Блок-схема установки 3 Рис. 2. Экспериментальная установка Осциллограф Tektronix TDS 220 Вольтметр B7-40/4 Вольтметр B7-22A Источник питания постоянного тока GW Instek GPD-74303S Генератор калибровочного импульса GW Instek AFG-72225 Источник питания постоянного тока GW Instek GPD-74303S Рис. 2. Экспериментальная установка Рис. 3. Модули, входящие в состав спектрометрического тракта Осциллограф Tektronix TDS 220 Вольтметр B7-40/4 Вольтметр B7-22A тока GW Instek GPD-74303S Генератор либровочного мпульса GW tek AFG-72225 Осциллограф Tektronix TDS 220 Генератор калибровочного импульса GW Instek AFG-72225 Вольтметр B7-40/4 Рис. 2. Экспериментальная установка Рис. 3. Модули, входящие в состав спектрометрического тракта Рис. 3. Модули, входящие в состав спектрометрического тракта Рис. 3. Модули, входящие в состав спектрометрического тракта 4 Рис. 4. Лавинный фотодиод, подключенный в цепь для измер светодиод Лавинный фотодиод светодиод Лавинный фотодиод 5 Рис. 4. Лавинный фотодиод, подключенный в цепь для измерений светодиод Лавинный фотодиод светодиод Лавинный фотодиод Рис. 4. Лавинный фотодиод, подключенный в цепь для измерений 5 Спектрометрический тракт с лавинным фотодиодом, электронный шум Фотодиод – детектор оптического излучения, преобразующий свет в электрический сигнал. Особенностью ЛФД является эффект внутреннего усиления, появляющийся за счёт лавинного умножения заряда. При подаче достаточно большого напряжения обратного смещения на p-n переход, на длине свободного пробега электроны (дырки) успевают набрать энергию достаточную для выбивания новых электронов из валентной зоны в зону проводимости. Таким образом нарастает лавинный эффект. Одними из важнейших характеристик фотодиода являются: темновой ток, коэффициент усиления (𝑔) и напряжение пробоя. Темновой ток - ток, протекающий через фотодиод в отсутствие поглощенных фотонов, возникающий из-за тепловой генерации электронов или дырок в полупроводнике. Коэффициент усиления 𝑔 при данном напряжении на ЛФД, показывает во сколько раз фототок (ток через фотодиод, возникающий из-за попадания света) больше фототока при малом напряжении, когда лавинного усиления нет. Известно, что при малом 𝑈ЛФД = 10 В усиление действительно отсутствует. 𝑔(𝑈ЛФД) = 𝐼фото(𝑈ЛФД) 𝐼фото(10 В) = 𝐼свет(𝑈ЛФД) −𝐼темн(𝑈ЛФД) 𝐼свет(10 В) −𝐼темн(10 В) Напряжение пробоя ЛФД – это такая величина напряжения, при приближении к которой ток, протекающий через фотодиод резко возрастает. Таким образом p-n переход может быть уничтожен, при превышении напряжения пробоя. Для измерения электронных шумов в схеме с ЛФД был собран спектрометрический тракт, показанный на Рис.1. 6 Принцип измерения шумов такой: калибровочный генератор создаёт прямоугольные импульсы (Рис. 5), которые через емкость C𝑐𝑎𝑙= (1,50 ± 0,02) пФ подаются на калибровочный вход предусилителя. Рис. 5. Схематичное изображение калибровочного импульса Принцип измерения шумов такой: калибровочный генератор создаёт прямоугольные импульсы (Рис. 5), которые через емкость C𝑐𝑎𝑙= (1,50 ± 0,02) пФ подаются на калибровочный вход предусилителя. Принцип измерения шумов такой: калибровочный генератор создаёт прямоугольные импульсы (Рис. 5), которые через емкость прямоугольные импульсы (Рис. 5), которые через емкость Рис. 5. Схематичное изображение калибровочного импульса Рис. 5. Схематичное изображение калибровочного импульса Рис. 5. Схематичное изображение калибровочного импульса Таким образом в предусилитель инжектируется в течение нескольких наносекунд известный заряд 𝑞= C𝑐𝑎𝑙𝑈𝑐𝑎𝑙. Предусилитель интегрирует входной токовый импульс (осуществляет преобразование импульса тока в импульс напряжения) и импульс напряжения на выходе имеет вид 𝑈= 𝐻(𝑡) 𝑞 C𝑓𝑏𝑒 − 𝑡 C𝑓𝑏R𝑓𝑏 , где 𝐻(𝑡) – функция Хэвисайда, C𝑓𝑏, R𝑓𝑏 емкость и сопротивление обратной связи предусилителя соответственно. После этого сигнал поступает на УФ (усилитель формирователь), схема которого показана на Рис.6. 7 Рис. 6. Схема простейшего CR-RC усилителя-формирователя U ВЫХ U ВХ U 7 Рис. 6. Схема простейшего CR-RC усилителя-формирователя Видно, что он состоит из дифференцирующей и интегрирующей цепочек. УФ осуществляет фильтрацию и усиление входного сигнала, сигнал на выходе УФ имеет вид 𝑈вых = 𝐴𝑡𝑒(−𝑡 𝜏), 𝜏= 3 мкс На самом деле, настоящий УФ Ortec-570 устроен сложнее (дополнительные интегрирующие звенья и т. д.), но в любом случае график сигнала на его выходе представляет собой колоколообразную кривую и амплитуда в максимуме пропорциональна втёкшему заряду 𝑞. Поэтому измерив этот максимум с помощью АЦП, можно получить величину заряда. После УФ сигнал разветвляется, одна линия поступает через аттенюатор (АТТ) на вход аналого-цифрового преобразователя (АЦП), другая линия через дискриминатор идёт на линию задержки и потом через формирователь импульсов на управляющий вход АЦП. Во время эксперимента с осциллографа ведётся контроль за тем, чтобы запускать АЦП от сигнала, а не от шумов, для чего настраивают порог дискриминатора и величину задержки. Электронный шум в цепи состоит из двух компонент: тепловой шум и дробовой шум. После УФ сигнал разветвляется, одна линия поступает через аттенюатор (АТТ) на вход аналого-цифрового преобразователя (АЦП), другая линия через дискриминатор идёт на линию задержки и потом через формирователь импульсов на управляющий вход АЦП. Во время эксперимента с осциллографа ведётся контроль за тем, чтобы запускать АЦП от сигнала, а не от шумов, для чего настраивают порог дискриминатора и величину задержки. 𝑄2 ̅̅̅̅ = 𝑄2 ̅̅̅̅ тепл + 2𝑒𝜏𝐼𝑔𝑓𝑠ℎ𝐹, где 𝑄2 ̅̅̅̅ – дисперсия шумового заряда (приведённого ко входу предусилителя), 𝑄2 ̅̅̅̅ тепл – дисперсия шумового заряда, связанного с тепловыми флуктуациями зарядов в цепи (приведённого ко входу предусилителя). В данном случае интересует именно вторая компонента 2𝑒𝜏𝐼𝑔𝑓𝑠ℎ𝐹 – дробовой шум, где 𝐼 – ток протекающий через фотодиод (является суммой темнового и фото тока), 𝜏= 3 мкс – время формирования УФ и 𝑓𝑠ℎ= 1.55 ± 0.06 , зависящий от реальных особенностей УФ (эффективное время интегрирования). Флуктуации лавины (как её электронной, так и дырочной компоненты) описываются фактором избыточного шума, 𝐹. Рис. 5. Схематичное изображение калибровочного импульса Если при одном напряжении измерить 𝑄2 ̅̅̅̅ свет и 𝑄2 ̅̅̅̅ темн шумы с засветкой и без, то их разница 𝑄2 ̅̅̅̅ свет − 8 𝑄2 ̅̅̅̅ т н = 2𝑒𝜏𝐼𝑔𝑓𝑠ℎ𝐹 ведь остальные члены одинаковы и сократятся. Отсюда темн формула (1) для расчётов 𝐹(𝑈ЛФД) = 𝑄2 ̅̅̅̅свет− 𝑄2 ̅̅̅̅темн 2𝑒𝜏𝐼фото(𝑈ЛФД)𝑔(𝑈ЛФД)𝑓𝑠ℎ 𝐼фото = 𝐼свет −𝐼темн Методика проведения и результаты эксперимента Измерение темнового тока Измерение темнового тока Чтобы обеспечить хорошую светоизоляцию и не было «засветки», сам диод должен находиться в непроницаемой защитной коробке, дополнительно обернутой темной тканью. Эксперимент начинается с измерения темнового тока, для этого устанавливается высокое напряжение с источника, ЛФД подключен к своим зажимам (в соответствие с “цоколёвкой”, таким образом, что на p-n переход подано обратное напряжение). Соответственно, по падению напряжения на вольтметре B7-40/4 и по известным сопротивлениям можно установить величину тока в цепи. По закону Ома Iтемн = 𝑈𝑉 Rэф , где 𝑈𝑉 падение напряжения на вольтметре. Эффективное сопротивление параллельно соединённых сопротивления и вольтметра Rэф = 5.05 мОм . Поскольку во всех измерениях интерес представляют именно зависимости от напряжения на фотодиодных зажимах (далее 𝑈ЛФД ), можно, отняв из 𝑈ВВИ (напряжения на ВВИ) падение на эквивалентном сопротивлении цепочки резисторов ( R∑= 116.5мОм) получить требуемое 𝑈ЛФД = 𝑈ВВИ −R∑ Iтемн или численно 𝑈ЛФД = 𝑈ВВИ −23.1 ∙𝑈𝑉. 9 9 Рис. 7. Схема для измерения темнового тока U ЛФД Рис. 7. Схема для измерения темнового тока Перед измерением темнового тока ЛФД был измерен ток утечек (Таблица 1), который связан с большим, но не бесконечным сопротивлением изоляции между гнёздами входного разъёма. U ВВИ, В Uv вольтметр, мВ U ЛФД, В I утечки, нA 102,4 1,2 102,4 0,24 203 2,1 202,9 0,42 305 3 304,9 0,59 405 3,9 404,9 0,77 505 5 504,9 0,99 Таблица 1. Ток утечки на разъёме ЛФД в зависимости от напряжения ВВИ. 10 Ток утечек в зависимости от напряжения, подаваемого на разъём, показан на Рис. 8. Из рис. видно, что сопротивление изоляции между гнёздами разъёма около 500 ГОм, что является достаточным для проводимых измерений. 11 около 500 ГОм, что является достаточным для проводимых измерений Рис. 8. Зависимость тока утечки от напряжения на разъёме ЛФД U ВВИ, В Uv вольтметр, мВ U ЛФД, В I темновой, нA 10,1 1,5 10,1 0,3 51,2 5,5 51,1 1,1 102,1 9,4 101,9 1,9 151,3 14 151,0 2,8 203 26 202,4 5,2 253 40,8 252,1 8,1 305 61,8 303,6 12,2 356 95,7 353,8 19,0 386 139,6 382,8 27,6 405 191,6 400,6 37,9 415 236,7 409,5 46,9 425 303,1 418,0 60,0 0 100 200 300 400 500 600 0 0,2 0,4 0,6 0,8 1 1,2 U ЛФД, В I утечки, нA 11 Рис. 8. Методика проведения и результаты эксперимента Зависимость тока утечки от напряжения на разъёме ЛФД U ВВИ, В Uv вольтметр, мВ U ЛФД, В I темновой, нA 10,1 1,5 10,1 0,3 51,2 5,5 51,1 1,1 102,1 9,4 101,9 1,9 151,3 14 151,0 2,8 203 26 202,4 5,2 253 40,8 252,1 8,1 305 61,8 303,6 12,2 356 95,7 353,8 19,0 386 139,6 382,8 27,6 405 191,6 400,6 37,9 415 236,7 409,5 46,9 425 303,1 418,0 60,0 0 100 200 300 400 500 600 0 0,2 0,4 0,6 0,8 1 1,2 U ЛФД, В I утечки, нA 0 100 200 300 400 500 600 0 0,2 0,4 0,6 0,8 1 1,2 U ЛФД, В I утечки, нA Рис. 8. Зависимость тока утечки от напряжения на разъёме ЛФД Рис. 8. Зависимость тока утечки от напряжения на разъёме ЛФД 11 Рис. 8. Зависимость тока утечки от напряжения на разъёме ЛФД U ВВИ, В Uv вольтметр, мВ U ЛФД, В I темновой, нA 10,1 1,5 10,1 0,3 51,2 5,5 51,1 1,1 102,1 9,4 101,9 1,9 151,3 14 151,0 2,8 203 26 202,4 5,2 253 40,8 252,1 8,1 305 61,8 303,6 12,2 356 95,7 353,8 19,0 386 139,6 382,8 27,6 405 191,6 400,6 37,9 415 236,7 409,5 46,9 425 303,1 418,0 60,0 11 435 402,1 425,7 79,6 445 553,6 432,2 109,6 455 773,6 437,1 153,2 465 1057,1 440,6 209,3 Таблица. 2. Темновой ток ЛФД (I темн) в зависимости от напряжения смещения ЛФД (U ЛФД). Данные, представленные в Таблице 2, использованы для построения графика на рис. 9. Рис. 9. Темновой ток в зависимости от напряжения смещения ЛФД 0,1 1 10 100 1000 0 50 100 150 200 250 300 350 400 450 500 I, нA U ЛФД, В Р 9 Т й ЛФД 0,1 1 10 100 1000 0 50 100 150 200 250 300 350 400 450 500 I, нA U ЛФД, В Рис. 9. Темновой ток в зависимости от напряжения смещения ЛФД Из рис.9 видно, что напряжение пробоя ЛФД около 450 В. Из рис.9 видно, что напряжение пробоя ЛФД около 450 В. Имеет значение температура при которой проводятся измерения, ведь вместе с ней меняются характеристики ЛФД. В процессе измерений фиксировались 12 также показания термометра. При измерениях температура менялась не более чем на 0,5 °C, каждую серию измерений проводили за один раз, специально чтобы не менять условия, поэтому связанные с температурой погрешности незначительны. Фототок Проведя такие же по принципу измерения, но с засветкой фотодиодом можно построить зависимость 𝐼свет(𝑈ЛФД). В работе используются два типа диодов: первый ARL-3214URC-6cd (красный с длинной волны 620 – 635 нм) и L- 7104UVC (ультрафиолетовый в диапазоне 385-415 нм). При измерении коэффициента усиления постоянная засветка при 𝑈ЛФД = 10В устанавливалась такой, чтобы фототок через ЛФД был равен около 5нА. Коэффициент усиления Лавинный фотодиод обеспечивает внутреннее усиление за счёт эффекта лавинного умножения. Известно, что при 𝑈ЛФД = 10В усиления нет. Коэффициент усиления 𝑔(𝑈ЛФД) = 𝐼фото(𝑈ЛФД) 𝐼фото(10 В) = 𝐼свет(𝑈ЛФД)−𝐼темн(𝑈ЛФД) 𝐼свет(10 В)−𝐼темн(10 В). Планки погрешностей маленькие, и оказались меньше маркеров-точек, что видно из рис. 10 и рис. 11. Планки погрешностей маленькие, и оказались меньше маркеров-точек, что видно из рис. 10 и рис. 11. 13 Рис. 10. Зависимость коэффициента усиления g от напряжения смещения ЛФД для красного светодиода. 0 50 100 150 200 250 300 350 400 450 500 0,1 1 10 100 1000 U ЛФД, В g 0 50 100 150 200 250 300 350 400 450 500 0,1 1 10 100 1000 U ЛФД, В g Рис. 10. Зависимость коэффициента усиления g от напряжения смещения ЛФД для красного светодиода. Рис. 11. Зависимость коэффициента усиления g от напряжения смещения ЛФД для ультрафиолетового светодиода. 0 50 100 150 200 250 300 350 400 450 500 0,1 1 10 100 1000 U ЛФД, В g Рис. 11. Зависимость коэффициента усиления g от напряжения смещения ЛФД для ультрафиолетового светодиода. 14 Измерение фактора избыточного шума Измерение фактора избыточного шума Запустив большое количество калибровочных импульсов, с помощью АЦП набирают статистику из соответствующего количества амплитуд сигнала, поступившего с УФ, затем строят гистограмму и фитируют её гауссовским распределением. Таким образом, получают среднее значение амплитуды за множество измерений и среднеквадратическое отклонение 𝜎= √𝑄2 ̅̅̅̅. Фактор избыточного шума рассчитывался по формуле (1) он показан на ножество измерений и среднеквадратическое отклонение 𝜎= √𝑄2 ̅̅̅̅. Фактор избыточного шума рассчитывался по формуле (1), он показан на Рис. 12 и 13. Рис. 12. Зависимость фактора избыточного шума от напряжения смещения ЛФД для ультрафиолетового света 0 50 100 150 200 250 300 350 400 450 500 0 0,5 1 1,5 2 2,5 U ЛФД, В F 0 50 100 150 200 250 300 350 400 450 500 0 0,5 1 1,5 2 2,5 U ЛФД, В F Рис. 12. Зависимость фактора избыточного шума от напряжения смещения ЛФД для ультрафиолетового света 15 Рис. 13. Зависимость фактора избыточного шума от напряжения на ЛФД для красного света 0 50 100 150 200 250 300 350 400 450 500 0 1 2 3 4 5 6 7 8 U ЛФД, В F 0 50 100 150 200 250 300 350 400 450 500 0 1 2 3 4 5 6 7 8 U ЛФД, В F Рис. 13. Зависимость фактора избыточного шума от напряжения на ЛФД для красного света Для напряжений смещения ЛФД меньше 350 В погрешность измерения F велика. Показаны величины F для напряжений смещения Для напряжений смещения ЛФД меньше 350 В погрешность измерения F велика. Показаны величины F для напряжений смещения ЛФД 200, 250, 300 В; при меньших напряжениях погрешность F становится равной около 100%. Таким образом, в диапазоне напряжений смещения ЛФД Для напряжений смещения ЛФД меньше 350 В погрешность измерения F велика. Показаны величины F для напряжений смещения ЛФД 200, 250, 300 В; при меньших напряжениях погрешность F становится равной около 100%. Таким образом, в диапазоне напряжений смещения ЛФД до 300 В фактор избыточного шума согласуется с теоретически ожидаемым F=1 в пределах больших погрешностей измерения. Также представляет интерес зависимость фактора избыточного шума от коэффициента уcиления, показанная на рис. 14 и рис 15. 16 Рис. 14. Зависимость фактора избыточного шума от коэффициента усиления для ультрафиолетового светодиода Рис. 15. Зависимость фактора избыточного шума от коэффициента усиления для красного светодиода Для того, чтобы эти результаты для разных длин волн были более наглядными, построен график (см. рис. 16). Измерение фактора избыточного шума 0 20 40 60 80 100 120 140 160 180 200 0 0,5 1 1,5 2 2,5 g F 0 20 40 60 80 100 120 140 160 180 200 0 1 2 3 4 5 6 7 8 g F Рис. 14. Зависимость фактора избыточного шума от коэффициента усиления для ультрафиолетового светодиода 0 20 40 60 80 100 120 140 160 180 200 0 0,5 1 1,5 2 2,5 g F 0 20 40 60 80 100 120 140 160 180 200 0 0,5 1 1,5 2 2,5 g F g Рис. 14. Зависимость фактора избыточного шума от коэффициента усиления для ультрафиолетового светодиода Рис. 15. Зависимость фактора избыточного шума от коэффициента усиления для красного светодиода Для того, чтобы эти результаты для разных длин волн были более наглядными, построен график (см. рис. 16). 0 20 40 60 80 100 120 140 160 180 200 0 1 2 3 4 5 6 7 8 g F Рис. 14. Зависимость фактора избыточного шума от коэффициента усиления для ультрафиолетового светодиода 0 20 40 60 80 100 120 140 160 180 200 0 1 2 3 4 5 6 7 8 g F g Рис. 15. Зависимость фактора избыточного шума от коэффициента усиления для красного светодиода Для того, чтобы эти результаты для разных длин волн были более Рис. 15. Зависимость фактора избыточного шума от коэффициента усиления для красного светодиода Для того, чтобы эти результаты для разных длин волн были более наглядными, построен график (см. рис. 16). 17 17 Рис. 16. Фактор избыточного шума в зависимости от напряжения на ЛФД, комбинированный график для ультрафиолетового и красного светодиодов U ЛФД, В U ЛФД, В Рис. 16. Фактор избыточного шума в зависимости от напряжения на ЛФД, комбинированный график для ультрафиолетового и красного светодиодов Выводы и заключение В курсовой работе мной были проведены измерения основных характеристик лавинного фотодиода Hamamatsu APD S8664-1010: темнового тока, коэффициента усиления и фактора избыточного шума для двух длин волн детектируемого света, 400 нм и 630 нм. Из результатов можно сделать следующие выводы: - темновой ток стремительно увеличивается с приближением напряжения на ЛФД к напряжению пробоя. - коэффициент усиления ЛФД (g) при увеличении напряжения возрастает по кривой, напоминающей экспоненту, при этом его значение немного меньше для красной длины волны, как видно из рис.17. - коэффициент усиления ЛФД (g) при увеличении напряжения возрастает по кривой, напоминающей экспоненту, при этом его значение немного меньше для красной длины волны, как видно из рис.17. 18 Рис. 17. Коэффициент усиления (g) в зависимости от напряжения ЛФД, комбинированный график для ультрафиолетового и красного светодиодов 0 50 100 150 200 250 300 350 400 450 500 0,1 1 10 100 1000 U ЛФД, В g Ряд2 U_APD,V Ультрафиолетовый светодиод Красный светодиод Рис. 17. Коэффициент усиления (g) в зависимости от напряжения ЛФД, комбинированный график для ультрафиолетового и красного светодиодов 0 50 100 150 200 250 300 350 400 450 500 0,1 1 10 100 1000 U ЛФД, В g Ряд2 U_APD,V Ультрафиолетовый светодиод Красный светодиод Рис. 17. Коэффициент усиления (g) в зависимости от напряжения ЛФД, комбинированный график для ультрафиолетового и красного светодиодов - зависимость фактора избыточного шума от коэффициента усиления, различна для ультрафиолетового и красного светодиодов, причём с красным светодиодом (т. е. для большей длины волны) показатель F при одинаковых g выше, и разница немного увеличивается с ростом показателя g (рис. 18). - зависимость фактора избыточного шума от коэффициента усиления, различна для ультрафиолетового и красного светодиодов, причём с красным светодиодом (т. е. для большей длины волны) показатель F при одинаковых g выше, и разница немного увеличивается с ростом показателя g (рис. 18). 19 Рис. 18. Фактор избыточного шума в зависимости от коэффициента усиления, комбинированный график для ультрафиолетового и красного светодиодов Рис. 18. Фактор избыточного шума в зависимости от коэффициента усиления, комбинированный график для ультрафиолетового и красного светодиодов Заметное увеличение фактора избыточного шума ЛФД при засветке ЛФД красным светом (с длиной волны 630 нм) по сравнению со случаем засветки ультрафиолетовым светом (с длиной волны 400 нм) связана с большим (более чем в 30 раз) увеличением длины поглощения красного света в кремнии. ультрафиолетового 1,36 ± 0,03. В счётчиках на основе кристаллов чистого CsI и ЛФД Hamamatsu S8664-1010 использование дополнительного спектросместителя с NOL-9 приведёт к увеличению вклада статистики фотоэлектронов в энергетическое разрешение калориметра не более чем на (√ 1.78 1.36 −1) ∙100% = 14%, что удовлетворяет проектным требованиям к калориметру. Список использованной литературы 1. T. Aushev et al., arXiv:1002.5012 [hep-ex], T. Abe et al. [Belle-II Collaboration], arXiv:1011.0352 [physics.ins-det] 2. A. J. Bevan et al. [BaBar and Belle Collaborations], Eur. Phys. J C 74 (2014) 3026 3. H. Aihara et al., PoS (PhotoDet2015) 052, http://cmd.inp.nsk.su/~epifanov/PhotoDet2015_052.pdf 4. http://www.luminnotech.com/ 5. http://www.hamamatsu.com/jp/en/S8664-55.html 6. Д. А. Епифанов, доклад на 22-ом общем собрании коллаборации Belle II 19-23 Октября 2015 г. В счётчиках на основе кристаллов чистого CsI и В счётчиках на основе кристаллов чистого CsI и ЛФД Hamamatsu S8664-1010 использование дополнительного спектросместителя с NOL-9 приведёт к увеличению вклада статистики фотоэлектронов в энергетическое разрешение калориметра не более чем на (√ 1.78 1.36 −1) ∙100% = 14%, что удовлетворяет проектным требованиям к на (√ 1.78 1.36 −1) ∙100% = 14%, что удовлетворяет проектным требованиям к калориметру. калориметру. калориметру. Выводы и заключение При этом фактор избыточного шума для красного света начинает определяться не только флуктуациями развития лавины, но и чисто геометрическим фактором флуктуации коэффициента усиления ЛФД (индуцированным большим разбросом точки поглощения красного света в ЛФД). Обычно с ЛФД работают при таком напряжении, которое обеспечивает коэффициент усиления g = 50. При этом значении g, как видно из рис.18, 20 20 фактор избыточного шума для красного света равен 1,78 ± 0,05, а для ультрафиолетового - 1,36 ± 0,03. Список использованной литературы 1. T. Aushev et al., arXiv:1002.5012 [hep-ex], T. Abe et al. [Belle-II Collaboration], arXiv:1011.0352 [physics.ins-det] 2. A. J. Bevan et al. [BaBar and Belle Collaborations], Eur. Phys. J C 74 (2014) 3026 3. H. Aihara et al., PoS (PhotoDet2015) 052, http://cmd.inp.nsk.su/~epifanov/PhotoDet2015_052.pdf 4. http://www.luminnotech.com/ 5. http://www.hamamatsu.com/jp/en/S8664-55.html 6. Д. А. Епифанов, доклад на 22-ом общем собрании коллаборации Belle II 19-23 Октября 2015 г. 1. T. Aushev et al., arXiv:1002.5012 [hep-ex], T. Abe et al. [Belle-II Collaboration], arXiv:1011.0352 [physics.ins-det] 1. T. Aushev et al., arXiv:1002.5012 [hep-ex], T. Abe et al. [Belle-II Collaboration], arXiv:1011.0352 [physics.ins-det] 2. A. J. Bevan et al. [BaBar and Belle Collaborations], Eur. Phys. J C 74 (2014) 3026 3. H. Aihara et al., PoS (PhotoDet2015) 052, http://cmd.inp.nsk.su/~epifanov/PhotoDet2015_052.pdf 4. http://www.luminnotech.com/ 5. http://www.hamamatsu.com/jp/en/S8664-55.html 6. Д. А. Епифанов, доклад на 22-ом общем собрании коллаборации Belle II 19-23 Октября 2015 г. 21
W2997297795.txt	https://www.mdpi.com/1996-1944/13/1/180/pdf?version=1577877202	en		Ultrahigh Energy Storage Properties in (Sr0.7Bi0.2)TiO3-Bi(Mg0.5Zr0.5)O3 Lead-Free Ceramics and Potential for High-Temperature Capacitors	Materials	2,020	cc-by	7,393	materials Article Ultrahigh Energy Storage Properties in (Sr0.7Bi0.2)TiO3-Bi(Mg0.5Zr0.5)O3 Lead-Free Ceramics and Potential for High-Temperature Capacitors Xi Kong , Letao Yang, Zhenxiang Cheng and Shujun Zhang * ISEM, Australian Institute of Innovative Materials, University of Wollongong, Wollongong NSW 2500, Australia; xk759@uowmail.edu.au (X.K.); ly742@uowmail.edu.au (L.Y.); cheng@uow.edu.au (Z.C.) * Correspondence: shujun@uow.edu.au Received: 24 November 2019; Accepted: 28 December 2019; Published: 1 January 2020 Abstract: Due to the enhanced demand for numerous electrical energy storage applications, including applications at elevated temperatures, dielectric capacitors with optimized energy storage properties have attracted extensive attention. In this study, a series of lead-free strontium bismuth titanate based relaxor ferroelectric ceramics have been successfully synthesized by high temperature solid-state reaction. The ultrahigh recoverable energy storage density of 4.2 J/cm3 under 380 kV/cm, with the high efficiency of 88%, was obtained in the sample with x = 0.06. Of particular importance is that this ceramic composition exhibits excellent energy storage performance over a wide work temperature up to 150 ◦ C, with strong fatigue endurance and fast discharge speed. All these merits demonstrate the studied ceramic system is a potential candidate for high-temperature capacitors as energy storage devices. Keywords: dielectric capacitors; lead-free; high temperature; energy storage 1. Introduction Dielectric capacitors have been widely used as energy-storage components in various electronic and electrical applications [1,2], with more focus in high-power and pulse-power applications, because of their higher power density, faster charge–discharge rate, and robustness when compared with fuel cells and batteries [3]. However, the energy density of dielectric capacitors is usually lower than other energy storage devices, which limits their applications. In addition, modern dielectric capacitors are challenged by operating under high-temperature environments, for example, the power inverters used for hybrid vehicles are operated at temperatures over 140 ◦ C [4] and the devices for exploring underground oil or gas are required to withstand temperature over 200 ◦ C [5]. The low glass transition point of present commercial polymer capacitors limit their applications at elevated temperature [6]. In contrast, ceramic capacitors benefit from their inorganic structure, and so are more capable for hard environments. However, the energy efficiency of ceramic capacitors usually reduces as the operating temperature increases due to the leakage current and/or space charge effect [2]. The reduced energy efficiency at elevated temperature corresponds to severe energy dissipation in the form of heat, raising the concerns of device stability and reliability. Therefore, a demand exists for exploration of high temperature dielectric ceramics with good energy storage properties. The energy density and efficiency of capacitors can be calculated by the following the Equations: Pmax Z W = EdP, (1) 0 Materials 2020, 13, 180; doi:10.3390/ma13010180 www.mdpi.com/journal/materials Materials 2020, 13, 180 2 of 13 Z Pmax Wrec = EdP, (2) Pr Wrec × 100%, (3) W where W and W rec represent the stored energy density and recoverable energy density. The energy efficiency (η) is defined by the ratio between W rec and W. It can be seen that high maximum polarization (Pmax ), low remnant polarization (Pr ), and high breakdown strength (BDS) relating to the various electric field will greatly benefit the recoverable energy density and energy efficiency. However, the Pmax or Pr for many dielectric materials are strongly temperature-dependent [7], leading to inferior stability of the dielectric capacitors. In addition, at elevated temperature, the trapped charges existing in the ceramics gain enough energy and contribute to the conductivity, resulting in an increased leakage current and decreased energy efficiency, which will accelerate thermal breakdown of the ceramics by generating more heat [8,9]. The opposite way, the capacitors with high energy efficiency could operate at elevated temperature without additional cooling devices, which reduces the volume and cost of the whole system. Many lead-containing materials attracted wide attention as energy storage capacitors in last few decades due to their excellent properties, such as PbZrO3 and (Pb,La)(Zr,Ti)O3 (PLZT)-based systems [10,11]. However, considering the environmental and health concerns of the lead component, lead-free dielectrics have been the mainstay in research over the last 20 years. Strontium titanate (SrTiO3 , ST) has been extensively studied for energy storage applications in the past few years due to its high permittivity (~300), low dielectric loss (<1%) and relatively high BDS (~200 kV/cm) [12]. Many efforts have been attempted for improving the energy storage properties of ST, such as modification with the Bi3+ cation, Ca2+ cation or rare earth cations for increasing permittivity [13–15], meanwhile enhancing the BDS [12]. Of particular importance is that the bismuth-containing relaxor ferroelectrics have attracted increasing attention as promising alternatives for lead-free relaxors, due to the fact that the Bi3+ ions possess the lone-pair electronic configuration 6s2 being analogous to Pb2+ ions, which can be hybridized with O 2p orbitals, resulting the high saturation polarization. It was reported that bismuth-modified strontium titanate ((Sr0.7 Bi0.2 )TiO3 abbreviated to SBT) dielectric material exhibited strong relaxor-like behavior, showing smeared phase transition temperature and high polarization [14], while the energy storage density of 1.63 J/cm3 at 217.6 kV/cm was reported for relaxor ceramic (Sr0.85 Bi0.1 )TiO3 [16]. In addition, SBT has been selected as an endmember to form solid solutions with other compounds to improve the dielectric and energy storage properties. It was reported that a 0.55(Na0.5 Bi0.5 )TiO3 –0.45SBT multilayer ceramic capacitor showed a high energy density of 9.5 J/cm3 and energy efficiency of 92% at 720 kV/cm [17]. Although the maximum energy density of 48.5 J/cm3 was obtained in 0.9SBT–0.1BiFeO3 thin-film capacitor, the energy efficiency was relatively low (48%) [18]. Based on the unique relaxor characteristic of Bi-based compound, in order to improve the temperature stability and energy storage density, numerous Bi(Me0 Me”)O3 (Me0 = Mg2+ , Zn2+ ; Me” = Ti4+ , Nb5+ , et al.) endmembers have been attempted for dielectric solid solutions. For example, BaTiO3 –Bi(Me0 Me”)O3 ceramics transfer from typical ferroelectric to relaxor with increasing Bi(Me0 Me”)O3 concentration, showing improved temperature stability, low dielectric loss, as well as improved energy efficiency [9,19–23]. Among these solid solutions, 0.7BaTiO3 –0.3Bi(Mg0.5 Zr0.5 )O3 was found to exhibit high stability over a wide temperature range from −20 ◦ C to 430 ◦ C with low permittivity variation (< ±15%) and relatively low loss (~2%) [24]. Therefore, in this work, Bi(Mg0.5 Zr0.5 )O3 -modified SBT relaxor ferroelectric solid solution (1-x)(Sr0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ((1-x)SBT-xBMZ) was fabricated, and its energy storage properties were studied in detail. A high energy storage density of 4.2 J/cm3 at applied electric field of 380 kV/cm was obtained for the x = 0.06 sample. The energy efficiency at 200 kV/cm was maintained above 90% over the temperature range of 25 ◦ C to 125 ◦ C, exhibiting an excellent temperature stability. η = Materials 2020, 13, 180 3 of 13 2. Materials and Methods (1-x)SBT-xBMZ ceramics with x = 4%, 6%, 8%, 10%, 15% (abbreviated as 4BMZ, 6BMZ, 8BMZ, 10BMZ, and 15BMZ, respectively) were fabricated by a conventional high-temperature solid state reaction method using reagent powders SrCO3 (99.9%, Alfa Aesar, Haverhill, MA, USA), TiO2 (99.0%, Strem Chemicals, Newburyport, MA, USA), Bi2 O3 (99.9%, Alfa Aesar), MgO (99.0%, Alfa Aesar), and ZrO2 (99.7%, Alfa Aesar). These powders were baked at 150 ◦ C for 12 h to remove the absorbed moisture and then stoichiometrically weighed with the addition of 1–2% excess MgO [25]. The mixed powders were ball milled for 12 h with ethanol as medium. Thereafter, the powders were re-milled again after calcination at 900 ◦ C for 2 h. The median particle size of re-milled powers were around 0.5–0.7 µm. Then 0.2 wt % Rhoplex binder was added to the slurry of each composition and the mixtures were ball-milled for 2 h. The slurries with binder were dried, granulated, and sieved. Then the dried powders were pressed into 15-mm-diameter pellets. In order to increase the green density, these pellets were pressed under an isostatic pressure of 200 MPa for 10 min after burning out the binder at 600 ◦ C. The pellets were sintered at 1200–1250 ◦ C for 2 h with self-source powder in a closed crucible. For dielectric and electrical measurements, the silver electrodes and gold electrodes were prepared on polished parallel sample surfaces, respectively. In addition, the pellets were polished to ~100 µm for P-E and breakdown strength measurements. The X-ray diffraction patterns were recorded on a diffractometer (GBC MMA XRD) using Cu Kα radiation with the ground pellets. The microstructure information of polished and thermally etched surfaces were collected using scanning electron microscopy (SEM) (JSM-7500FA, JEOL, Akishima Tokyo). Grain size distributions were calculated using Nano Measurer by measuring the selected area of SEM images. In order to measure the dielectric properties of the studied samples, silver electrodes were fired on both sides of polished samples. The dielectric data was collected on a LCR meter (4980AL, Keysight, Santa Rosa, CA, USA) from 100 Hz to 1 MHz over temperature range of −100 ◦ C to 400 ◦ C. For the electrical measurements, various electric fields were applied to the sample at 10 Hz and room temperature using a ferroelectric tester (TF2000, aixACCT, Aachen, Germany) to obtain P-E hysteresis loops. Thermal stability and fatigue endurance were measured at 200 kV/cm from 25–150 ◦ C and 1 to 105 cycles, respectively. The charge–discharge rate was measured with charge–discharge system (PK-CRP1701, PolyK, State College, PA, USA) based on the resistance–capacitance circuit with 2 kΩ load resistance. 3. Results and Discussion 3.1. Phase Structure and Microstructure Characterization The XRD patterns of (1−x)SBT-xBMZ ceramics with x = 4–15% are shown in Figure 1a. It can be seen that all samples exhibit pseudocubic perovskite structure, as evidenced by the single (111) and (200) peaks. Figure 1b shows the enlarged (200) peaks shifting to a lower angle with increasing BMZ concentration, corresponding to the expansion of unit cell volume. It can be attributed to the larger effective ionic radius (r) of (Mg0.5 Zr0.5 )3+ (r (Mg0.5 Zr0.5 )3+ = 0.72 Å calculated from the average ionic radii of Mg2+ and Zr4+ ) than that of Ti4+ (r(Ti4+ ) = 0.605 Å) on the B-site. The lattice parameters are obtained from XRD patterns, the linearly increasing trend with BMZ concentration from x = 4% to 15% can be seen in Figure 1c. An additional peak around 30◦ corresponding to the secondary phase Bi2 Ti2 O7 was observed for BMZ in the range of 6–8%, above which, the secondary phase was changed to Sr2 Bi4 Ti5 O18 [25]. The densities of SBT-BMZ samples are above 95% theoretical, being 5.46 g/cm3 for x = 0.04, slightly increasing to 5.68 g/cm3 for x = 0.15. Materials Materials2020, 2020,13, 13,180 x FOR PEER REVIEW 44 ofof1313 (a) (b) (c) Figure1.1.(a) (a)XRD XRDpatterns patterns (1−x)(Sr )TiO 3-xBi(Mg0.5Zr 0.5)O )O33 ceramics; (b) enlarged enlargeddiffraction diffraction Figure of of (1−x)(Sr BiBi )TiO ceramics; (b) 0.70.7 0.20.2 3 -xBi(Mg 0.5 Zr0.5 ◦ toto ◦ of of 0.7 Bi 0.2 )TiO 3-xBi(Mg0.5Zr Zr 0.5)O3 ceramics; the lattice parameters peaksfrom from (1−x)(Sr peaks 4646° 4747° (1−x)(Sr Bi )TiO -xBi(Mg ceramics; (c) the lattice parameters asasa a 0.7 0.2 3 0.5 0.5 3 function of BMZ concentration. function of BMZ concentration. The TheSEM SEMimages imagesofofstudied studiedceramics ceramicswith withthermally thermallyetched etchedsurface surfaceare areshown shownininFigure Figure2.2.All All (1−x)SBT-xBMZ not only show minimal porosity but also fine grains with average grain (1−x)SBT-xBMZceramics ceramics not only show minimal porosity but have also have fine grains with average size smaller than 1 µm, Figurein2i,j. The dense microstructure with fine grains are grains expected grain size smaller thanas1 shown μm, asin shown Figure 2i–j. The dense microstructure with fine are toexpected benefit the enhancement of BDS [26].of BDS [26]. to benefit the enhancement Materials 2020, 13, 180 Materials 2020, 13, x FOR PEER REVIEW 5 of 13 5 of 13 (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) Figure 0.7Bi0.2 )TiO3-xBi(Mg0.5Zr0.5)O3 ceramics: (a) x = 4%; (b) x = 6%; (c) x = 8%; (d) x = Figure 2. 2. SEM SEM micrographs micrographsof ofthe thepolished polishedand andthermally thermallyetched etchedsurfaces surfacesfor for(1−x)(Sr (1−x)(Sr 0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ceramics: (a) x = 4%; (b) x = 6%; (c) x = 8%; 0.7 Bi 0.2 )TiO 3 -xBi(Mg 0.5 Zr 0.5 )O 3 ceramics: (f) x = 4%; = 6%; 10%; (e) x = 15%; The grain size of (1−x)(Sr (d) x = 10%; (e) x = 15%; The grain size of (1−x)(Sr0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ceramics: (g) (f) xx = 4%; (h) (g) xx == 8%; 6%; (i) (h)xx==10%; 8%; (j) (i) xx ==15%. 10%; (j) x = 15%. Materials 2020, 13, x FOR PEER REVIEW 6 of 13 Materials 2020, 13, 180 6 of 13 3.2. Dielectric and Energy-Storage Characterizations for (1−x)SBT-xBMZ Ceramics The dielectric properties ofCharacterizations BT-Bi(Me’Me”)Ofor 3-based ceramics have been widely investigated to 3.2. Dielectric and Energy-Storage (1−x)SBT-xBMZ Ceramics acquire deeper understanding, including the relationship of relative dielectric permittivity (εr) and The dielectric properties of BT-Bi(Me0 Me”)O3 -based ceramics have been widely investigated to the component Bi(Me’Me”)O 3 endmember. Several previous studies show that the phase transition acquire deeper understanding, including the relationship of relative dielectric permittivity (ε ) and r peaks became broader and smeared with increasing Bi(Me’Me”)O3 substitutions in BT-Bi(Me’Me”)O 3 0 Me”)O endmember. Several previous studies show that the phase transition the component Bi(Me 3 solid solutions, leading to the enhancement of thermal stability [23,24,27,28]. Figure 3a–e shows the 0 Me”)O substitutions in BT-Bi(Me0 Me”)O peaks becamedependence broader and of smeared with increasing Bi(Me 3 3 temperature dielectric permittivity (εr) and loss (tanδ) for (1−x)SBT-xBMZ ceramics. solid solutions, leading to the enhancement of thermal stability [23,24,27,28]. Figure 3a–e shows the The smeared phase transition and frequency dispersion behavior of permittivity and dielectric loss temperature of dielectric permittivity (εr ) and loss (tanδ) for a(1−x)SBT-xBMZ are observeddependence near the dielectric maximum temperature (Tm ), exhibiting typical relaxor ceramics. behavior. The smeared phase transition and frequency dispersion behavior of permittivity and dielectric loss− Generally, the diffuseness of phase transition can be evaluated by modified Curie–Weiss law 1/ε are observed near the dielectric maximum temperature (T ), exhibiting a typical relaxor behavior. m 1/εm = (T − Tm)γ/C, where εm and Tm represent maximum permittivity and the corresponding Generally, the diffuseness of phase transition can be evaluated by modified Curie–Weiss lawwhile 1/ε − 1/ε temperature, respectively. The value γ = 2 represents a complete diffuse phase transition γ =m1 =represents (T − Tm )γa/C, where ε and T represent maximum permittivity and the corresponding temperature, m m transition [29]. In order to reduce the effects of space charge, the γ classical ferroelectric respectively. The value γ = 2 represents a completeusing diffuse phase transition γ =Figure 1 represents a values for (1-x)SBT-xBMZ ceramics are calculated permittivity data atwhile 1 MHz. 3f gives classical ferroelectric transition [29]. In order to reduce the effects of space charge, the γ values for the γ values of (1-x)SBT-xBMZ ceramics, being above 1.5, indicating relaxor characteristics. With (1-x)SBT-xBMZ ceramics are calculated using (ε permittivity data at 1 MHz. Figure 3f gives the γ values increasing BMZ, the maximum permittivity m) at 1 kHz for ceramics with x up to 10% are similar, of (1-x)SBT-xBMZ ceramics, being above indicating relaxor than characteristics. With increasing BMZ, but the Tm increases gradually from −681.5, °C to −27 °C, higher the Tm of (Sr 0.7Bi0.2)TiO3 ceramics the maximum permittivity (ε ) at 1 kHz for ceramics with x up to 10% are similar, but the Tm increases m which is around −100 °C [30]. The lower permittivity and higher dielectric loss for samples with x = ◦ C to −27 ◦ C, higher than the T of (Sr Bi )TiO ceramics which is around gradually from −68 m 0.7 0.2 3 15% may be related to the secondary phase. Of particular interest is that the dielectric loss is less than ◦ [30]. The lower permittivity and higher dielectric loss for samples with x = 15% may be related −100 3% forCall the studied ceramics over a wide temperature range, as shown in Table 1, exhibiting the to the secondary phase. Of particular that the dielectric loss is lessInthan 3% for all the studied potential of high energy efficiencyinterest in theissame temperature range. addition, dielectric loss ceramics over a wide temperature range, as shown in Table 1, exhibiting the potential of high energy corresponds the reliability of capacitors, lower generated heat leading to the associated low efficiency same breakdown. temperatureItrange. loss corresponds thearound reliability possibilityinofthe thermal shouldIn beaddition, noted thatdielectric an additional dielectric peak 290 of °C capacitors, lower generated heat leading to the associated low possibility of thermal breakdown. It was observed for 10BMZ and 15BMZ, as shown in Figure 3d,e. This observation can be explained by ◦ should be notedphases that an(Sr additional dielectric peak around 290 C was observed for 10BMZ and 15BMZ, the secondary 2Bi4Ti4O15) which was reported to possess a Curie temperature around 285 as shown in Figure 3d,e. This observation can be explained by the secondary phases (Sr2 Bi4 Ti4 O15 ) °C [31]. which was reported to possess a Curie temperature around 285 ◦ C [31]. (a) (b) (c) (d) Figure 3. Cont. Materials 2020, 13, 180 7 of 13 Materials 2020, 13, x FOR PEER REVIEW Materials 2020, 13, x FOR PEER REVIEW 7 of 13 7 of 13 (e) (f) (e) (f) Figure 3. Temperature dependence r and tanδ of (1−x)(Sr0.7Bi0.2)TiO3-xBi(Mg0.5Zr0.5)O3 ceramics: (a) Figure 3. Temperature dependence ofofεrεand tanδ of (1−x)(Sr0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ceramics: Figure Temperature dependence Bi0.2)TiO 3-xBi(Mg0.5Zr0.5)O3 ceramics: (a) m) as a function of ln (T − Tm) for x = 3. 4%; (b) x = 6%; (c) x = 8%; (d)ofxε=r and 10%;tanδ (e) xof= (1−x)(Sr (f)0.7ln − 1/ε (a) x = 4%; (b) x = 6%; (c) x = 8%; (d) x = 10%; (e) x15%; = 15%; (f)(1/ε ln (1/ε − 1/εm ) as a function of ln (T − Tm ) m) as a function of ln (T − Tm) for x = 4%; (b) x = 6%; (c) x = 8%; (d) x = 10%; (e) x = 15%; (f) ln (1/ε − 1/ε (1−x)(Sr0.7Bi0.2)TiO3-xBi(Mg0.5Zr0.5)O3 ceramics at 1 MHz. for(1−x)(Sr (1−x)(Sr Bi )TiO -xBi(Mg Zr )O ceramics at 1 MHz. 0.7 0.2 3 0.5 0.5 3 0.7Bi0.2)TiO3-xBi(Mg0.5Zr0.5)O3 ceramics at 1 MHz. Table 1. The summary of dielectric performance at 1 kHz and energy storage properties at 200 kV/cm Table Thesummary summary of performance at 1at kHz and energy storage properties at 200 kV/cm The of3dielectric dielectric performance 1 kHz and energy storage properties at 200 kV/cm Table 1.1.(1−x)(Sr for 0.7Bi0.2)TiO -xBi(Mg 0.5Zr 0.5)O3 ceramics. for (1−x)(Sr 0.7Bi0.2)TiO3-xBi(Mg0.5Zr0.5)O3 ceramics. for (1−x)(Sr0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ceramics. η @ RT Wrec @ RT Sample Tm (°C) εm ε @ RT tanδ @ RT T-range (°C) tanδ < 3% Wrec @ RT 3 η @ RT ) (%) (J/cm Sample T m Tm(◦(°C) RT tanδ @ @ RT RT T-range T-range tanδ 3% W rec @ RT3 (J/cm3 ) Sample C) εmεm εε@@RT tanδ (◦ (°C) C) tanδ < <3% η @ RT (%) ) (%) 92 (J/cm 1.50 4BMZ −68 1020 865 0.06% −80–400 4BMZ 0.06% −80–400 1.501.71 4BMZ −68 −64 1020 10201020 865 865 880 0.06% −80–400 1.50 92 92 92 6BMZ −68 0.06% −75–380 6BMZ −64 1020 880 0.06% −75–380 1.71 6BMZ −64 1020 880 0.06% −75–380 1.71 92 75 92 8BMZ −56 1110 990 0.08% −70–390 1.59 8BMZ −56 1110 990 0.08% −70–390 1.59 8BMZ 0.08% −70–390 1.591.34 75 81 75 10BMZ −56 −54 1110940 990 870 0.04% −70–350 10BMZ −54 940 870 0.04% −70–350 1.34 81 10BMZ −54 940 870 0.04% −70–350 1.34 81 77 15BMZ −27 −27 890 890 865 865 −40–340 15BMZ 0.2%0.2% −40–340 1.221.22 77 15BMZ −27 890 865 0.2% −40–340 1.22 77 The P–E loops at 200 kV/cm for (1−x)SBT-xBMZ ceramics with different BMZ content are TheP–E P–E loops at for ceramics with different BMZ content The loops at 200200 kV/cm for4a. (1−x)SBT-xBMZ with different BMZ content areare compared compared and shown in kV/cm Figure It(1−x)SBT-xBMZ is clear to seeceramics that the 4BMZ and 6BMZ ceramics exhibit slim compared and shown in Figure 4a. It is clear to see that the 4BMZ and 6BMZ ceramics exhibit slim and shown in Figure It is clear to see high that the 4BMZ and 6BMZ ceramics exhibit slim P–EBMZ loops at P–E loops at 2004a. kV/cm, indicating energy efficiencies above 92%. With increasing P–E loops indicating at 200 PkV/cm, indicating highdecreases, energy efficiencies above 92%. With increasing 200 kV/cm, energy above 92%. to With BMZ concentration, concentration, max high increases firstefficiencies then relevant the increasing dielectric permittivity, as BMZ listed inPmax concentration, Pmax increases first found then decreases, relevant to the as listedenergy in Tablefirst 1. The highest Pmax was for but its large Pr dielectric would reduce thein recoverable increases then decreases, relevant to 8BMZ, the dielectric permittivity, aspermittivity, listed Table 1. The highest Table 1. The highest P max was found for 8BMZ, but its large Pr would reduce the recoverable energy density and decreases energy efficiency. The hysteresis increases obviously in 8BMZ, 10BMZ and Pmax was found for 8BMZ, but its large Pr would reduce the recoverable energy density and decreases density andsamples, decreasesresulting energy efficiency. The hysteresis increases obviously in 8BMZ, 10BMZ and in 15BMZ in increases the decreased Wrec with increasing BMZ and concentration, as shown energy efficiency. The hysteresis obviously in 8BMZ, 10BMZ 15BMZ samples, resulting 15BMZ samples, resulting in the decreased W rec with increasing BMZ concentration, as shown in Figure 4b. The increased hysteresis may be related to the increased secondary phases. Based on the in Figure the decreased W rec with increasing be BMZ concentration, as secondary shown inphases. FigureBased 4b. The increased 4b. The increased hysteresis related to the increased on the preliminary results, we chose themay 6BMZ sample for further evaluation of the energy storage property. hysteresis may be related to the the6BMZ increased secondary Based the preliminary results, we preliminary results, we chose sample for furtherphases. evaluation of theon energy storage property. chose the 6BMZ sample for further evaluation of the energy storage property. (a) (a) (b) (b) Figure 4. (a) The P–E loops at 200 kV/cm for (1−x)(Sr0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ceramics; (b) the stored energy density (W), recoverable energy density (W r ), and energy efficiency (η) for (1−x)(Sr0.7 Bi0.2 )TiO3 -xBi(Mg0.5 Zr0.5 )O3 ceramics as functions of different BMZ concentration at 200 kV/cm and ambient temperature. Figure 4. (a) The P–E loops at 200 kV/cm for (1−x)(Sr0.7Bi0.2)TiO3-xBi(Mg0.5Zr0.5)O3 ceramics; (b) the stored energy density (W), recoverable energy density (Wr), and energy efficiency (η) for (1−x)(Sr0.7Bi0.2)TiO3-xBi(Mg0.5Zr0.5)O3 ceramics as functions of different BMZ concentration at 200 kV/cm and ambient temperature. Materials 2020, 13, 180 8 of 13 3.3. Energy Storage Properties of 0.94SBT-0.06BMZ Ceramic 3.3.The Energy Storage Properties 0.94SBT-0.06BMZ unipolar P–E loops ofof6BMZ ceramic are Ceramic measured at 10 Hz with different applied electric fields as shown in Figure 5a. It is that all P–Eare loops are slimatwith negligible hysteresis which are The unipolar P–E loops of clear 6BMZ ceramic measured 10 Hz with different applied electric associated with the small value of P r, contributing to the high energy efficiency. Pr and Pmax as a fields as shown in Figure 5a. It is clear that all P–E loops are slim with negligible hysteresis which function of electric field 6BMZ ceramic are shown into Figure 5b. When electric field increased are associated with thefor small value of Pr , contributing the high energythe efficiency. Pr and Pmax as a from 50 to 380 kV/cm, the P max value enhances monotonically from 6.7 μC/cm2 to 30.3 μC/cm2, while function of electric field for 6BMZ ceramic are shown in Figure 5b. When the electric field increased the Pr remains almostthe thePsamevalue values. Thus, the ΔP (ΔP = Pmax − P6.7 r) keeps increasing as function 2 , while from 50 to 380atkV/cm, enhances monotonically from µC/cm2 to 30.3 µC/cm max ofthe applied field, at which is beneficial for enlarging Wrec∆P and results given in Figure The Pr remains almost the same values. Thus, the (∆Pη;=the Pmax − Pr )are keeps increasing as5c. function maximum W rec of 4.2 J/cm3 was obtained at 380 kV/cm. On the other hand, the energy efficiency of applied field, which is beneficial for enlarging W rec and η; the results are given in Figure 5c. The exhibits remarkable applied up to 320 with values than 90%, maximum W rec of stability 4.2 J/cm3over was the obtained at field 380 kV/cm. OnkV/cm the other hand, thelarger energy efficiency above which the η slightly drops 88%. exhibits remarkable stability overtothe applied field up to 320 kV/cm with values larger than 90%, above which the η slightly drops to 88%. (a) (b) (c) (d) Figure 5.5.(a)(a) The unipolar , Pr, and Figure The unipolarP–E P–Eloops loopsfor for6BMZ 6BMZceramic; ceramic;(b) (b)the theelectric electricfiled fileddependent dependentPPmax max , Pr , and rec, and η versus applied electric field of 6BMZ ceramic; ΔP∆P forfor 6BMZ ceramic; (c)(c) the calculated W,W,WW 6BMZ ceramic; the calculated , and η versus applied electric field of 6BMZ ceramic; rec (d) Weibull distribution of the breakdown strength data. (d) Weibull distribution of the breakdown strength data. High BDS is advantageous in dielectric materials for obtaining high W as shown in Equation (1). The Weibull distribution has been used as a capable analysis method to explore the breakdown strength of ceramics, as given in the following equations: Xi = ln(Ei ), (4) Materials 2020, 13, 180 9 of 13 Yi = ln ln Pi = !! 1 , 1 − Pi (5) i n+1 (6) where i is the serial number of sample. Ei , Pi and n are the breakdown strength of each test sample, the probability of breakdown and the summation samples, respectively. The ascending sort of breakdown strength for samples is shown as E1 ≤ E2 ≤ . . . Ei . . . ≤ En . The Weibull modulus (m) can be obtained by the slope of the fitted linear line of Xi and Yi , which is one of the important parameters that represents the reliability of experimental data. The Weibull characteristic breakdown strength (Eb ) can be calculated by the intercept between the fitted linear line and Yi = 0. As shown in Figure 5d, the m value for 6BMZ ceramic is 17, leading to a high level of confidence. The calculated Eb of 6BMZ is 460 kV/cm which is much higher than other lead-free energy storage ceramics [32–36], as shown in Table 2. The energy storage properties of BaTiO3 -Bi(Me0 Me”)O3 ceramics have been widely studied, and the obtained energy densities have been found to be around 1–3 J/cm3 in previous research as shown in Table 2. The dielectric maxima temperatures of most of the studied BaTiO3 -Bi(Me0 Me”)O3 ceramics are around or above room temperature, leading to a relatively high dielectric loss at room temperature. However, in this work, the Tm of 6BMZ is far below room temperature, resulting in a low dielectric loss at room temperature and above. The depressed tanδ is believed to be associated with high breakdown strength, and hence the high energy density as well as the high energy efficiency. The ultrahigh breakdown strength of 6BMZ ceramic leads to improved W rec of 4.2 J/cm3 , which is favorable for energy storage applications. Table 2. Comparison of energy-storage properties between the 0.94SBT-0.06BMZ ceramic and other lead-free ceramics. Composition W rec (J/cm3 ) η (%) Eapp (kV/cm) Eb (kV/cm) Ref. 0.85BaTiO3 -0.15Bi(Zn2/3 Nb1/3 )O3 0.9BaTiO3 -0.1Bi(Mg2/3 Nb1/3 )O3 0.61BiFeO3 -0.33BaTiO3 -0.06Ba(Mg1/3 Nb2/3 )O3 0.88BaTiO3 -0.12Bi(Mg1/2 Ti1/2 )O3 0.88BaTiO3 -0.12Bi(Li0.5 Nb0.5 )O3 0.85BaTiO3 -0.15Bi(Zn0.5 Sn0.5 )O3 0.9BaTiO3 -0.1Bi(Zn0.5 Zr0.5 )O3 (Na0.25 Bi0.25 Sr0.5 )(Ti0.8 Sn0.2 )O3 0.94(Sr0.7 Bi0.2 )TiO3 -0.06Bi(Mg0.5 Zr0.5 )O3 0.79 1.13 1.56 1.81 2.032 2.21 2.46 3.4 4.2 93.5 95.8 75 ~88 88 91.6 ~73 90 88 131 143.5 125 224 270 230 264 310 380 262 ~270 535.5 340 280 266.5 330 460 [22] [32] [33] [20] [9] [34] [35] [36] This work For high-temperature energy-storage applications, thermal stability plays a key role when working at elevated temperatures. For example, the operating temperature can be higher than 140 ◦ C in hybrid electric vehicles, thus requiring some electronic components to be protected by cooling systems [37]. Dielectric materials that work and maintain stability at high temperature could not only reduce the cost but also the volume of the whole energy storage system. The temperature-dependent unipolar P–E loops for 6BMZ ceramic at various temperatures from ambient to 150 ◦ C at 200 kV/cm are shown in Figure 6a. It can be seen that all the P–E loops are slim, indicating the high temperature stability. The Pmax value was found to slightly decrease, which is related to the decreased permittivity at elevated temperature, while the Pr remains at a low value over a wide temperature range, indicating the 6BMZ ceramic capable of operating at high temperature. Figure 6b shows W rec and η as functions of temperature, the decrease of W rec is on the order of 14% due to reduced permittivity at elevated temperature. Of particular significance is that the variation of η is minimal up to 130 ◦ C, associated with the low dielectric loss and temperature-insensitive Pr . The excellent thermal stability of η for 6BMZ ceramic make it very promising for energy storage applications working at elevated temperature. Materials 2020, 13, x FOR PEER REVIEW 10 of 13 associated with the low dielectric loss and temperature-insensitive Pr. The excellent thermal stability of η for 6BMZ ceramic make it very promising for energy storagePrapplications working elevated associated with the low dielectric loss and temperature-insensitive . The excellent thermalatstability temperature. of η for 6BMZ ceramic make it very promising for energy storage applications working at elevated temperature. Materials 2020, 13, 180 10 of 13 (a) (b) (a) P–E loops for 6BMZ ceramic at different temperature (b) at 200 kV/cm; (b) Wr Figure 6. (a) The unipolar and η of 6BMZ ceramic as functions of temperature. Figure 6. 6. (a)(a) The unipolar P–E loops forfor 6BMZ ceramic at at different temperature at at 200 kV/cm; (b)(b) WW r Figure The unipolar P–E loops 6BMZ ceramic different temperature 200 kV/cm; r and η of 6BMZ ceramic asas functions ofof temperature. and η of 6BMZ ceramic functions temperature. 3.4. Fatigue Endurance and Charge–Discharge Rate for 0.94SBT-0.06BMZ Ceramic 3.4. Fatigue Endurance and Charge–Discharge Rate 0.94SBT-0.06BMZ Ceramic 3.4. Fatigue Endurance and Charge–Discharge Rate forfor 0.94SBT-0.06BMZ Ceramic To further investigate the energy storage properties for 6BMZ ceramic, the fatigue cycling endurance and a charge-discharge rate experiment were performed. Strong fatigue endurance is To further investigate the energy storage properties for 6BMZ ceramic, thefatigue fatigue cycling To further investigate the energy storage properties for 6BMZ ceramic, the cycling essential for energy storage capacitors in practical applications. Figure 7a displays the unipolar P–E endurance and a charge-discharge rate experiment were performed. Strong fatigue endurance endurance and a charge-discharge rate experiment were performed. Strong fatigue endurance is is loops for 6BMZ ceramic at 200 kV/cm underapplications. different cycle numbers. The energy storage essential energy storage capacitors practical applications. Figure displays the unipolar P–E essential forfor energy storage capacitors inin practical Figure 7a7a displays the unipolar P–E 5 cycles, where the unipolar P–E loops performances of the 6BMZ ceramic are reproducible after 10 loops for 6BMZ ceramic at 200 kV/cm under different cycle numbers. The energy storage performances loops for 6BMZ ceramic at 200 kV/cm under different cycle numbers. The energy storage maintain the of same the cycling, small be seen in the inset figure of loops Figure of the 6BMZ ceramic are over reproducible after 105 difference cycles, the unipolar P–E loops maintain the performances the shape 6BMZ ceramic are reproducible afterwhere 105can cycles, where the unipolar P–E 7a. The variation of W rec and η for fatigue cycling numbers are given in Figure 7b; here the W rec value same shape over the cycling, small difference can be seen in the inset figure of Figure 7a. The variation maintain the same shape over the cycling, small difference can be seen in the inset figure of Figure remains almost whilecycling the slightly drops from 95% torec 92% 10rec5 value cycles, ofThe W rec and η for fatigue cycling numbers areηgiven in Figure 7b; here the W value remains almost 7a. variation ofthe Wrecsame and ηvalue for fatigue numbers are given in Figure 7b; hereafter the W 5 demonstrating 6BMZ sample possesses reliability. the same valuethat while the ηvalue slightly drops from 95% tocycling 92% after 10 cycles, 6BMZ 5 cycles, remains almost the same while the ηexcellent slightly drops from 95% todemonstrating 92% after 10that sample possesses excellent cycling reliability. demonstrating that 6BMZ sample possesses excellent cycling reliability. (a) (b) (a) P–E (b) Figure cycling Figure7.7.(a) (a)Unipolar Unipolar P–Eloops loopsfor for6BMZ 6BMZceramic ceramicunder underdifferent differentfatigue fatigue cyclingnumber; number; (b) (b) W Wrrand and 5 5 for 6BMZ ceramic. η as functions of charging–discharging cycles from 1 to 10 η as functions of charging–discharging cycles from 1 to 10 for 6BMZ ceramic. Figure 7. (a) Unipolar P–E loops for 6BMZ ceramic under different fatigue cycling number; (b) Wr and η as functions of charging–discharging cycles from 1 to 105 for 6BMZ ceramic. Thefast fast discharge discharge rate rate is isRcritical critical for for pulse-power pulse-power applications. applications. The Thedischarged dischargedenergy energydensity density The 2(𝑡)𝑑𝑡/𝑉𝑜𝑙, ( ) can be calculated by W = R i t dt/Vol, where R , i(t) and Vol represent the total load resistor where Rapplications. canThe be calculated by 𝑊rate 𝑅 L𝑖 for pulse-power LL, i(t) and Vol represent the total load resistor dis = fast discharge is critical The discharged energy density (2 kΩ), the discharge current through R , and the sample volume, respectively. The calculated W dis at L can be calculated by 𝑊 = 𝑅 𝑖 (𝑡)𝑑𝑡/𝑉𝑜𝑙, where RL, i(t) and Vol represent the total load resistor an applied electric field of 200 kV/cm as a function of time for 6BMZ ceramic is given in Figure 8. The discharge time τ0.9 is the time required to release 90% of total W dis value. The τ0.9 for 6BMZ at room temperature is 0.62 µs, indicating the great potential of 6BMZ for pulse power capacitors. (2 kΩ), the discharge current through RL, and the sample volume, respectively. The calculated Wdis at an applied electric field of 200 kV/cm as a function of time for 6BMZ ceramic is given in Figure 8. The discharge time τ0.9 is the time required to release 90% of total Wdis value. The τ0.9 for 6BMZ at room temperature is 0.62 μs, indicating the great potential of 6BMZ for pulse power capacitors. Materials 2020, 13, 180 11 of 13 Figure 8. Discharge energy density as a function of time for 6BMZ ceramic at 200 kV/cm and ambient temperature. Figure 8. Discharge energy density as a function of time for 6BMZ ceramic at 200 kV/cm and ambient 4. Conclusions temperature. In summary, lead-free (1−x)SBT-xBMZ ceramics were fabricated by conventional solid state 4. Conclusions processing. The highest recoverable energy density of 4.2 J/cm3 and high energy efficiency of 88% were lead-free ceramics were fabricated by conventional solid state obtainedInatsummary, applied electric field(1−x)SBT-xBMZ of 380 kV/cm for 0.94(Sr0.7 Bi0.2 )TiO 3 -0.06Bi(Mg 0.5 Zr0.5 )O3 ceramic. 3 and high energy efficiency of 88% The highest energy stability density of 4.2 aJ/cm Theprocessing. 6BMZ ceramic shows arecoverable good temperature over wide temperature range from room were obtained electric field ceramic of 380 also kV/cm for 0.94(Sr 0.7Bi 0.2)TiO3stability -0.06Bi(Mg 0.5Zr temperature to 150 ◦at C. applied In addition, the 6BMZ exhibits excellent cycling over 100.55 )O3 ceramic. The 6BMZcycles. ceramicTogether shows awith goodits temperature stability from charging–discharging fast discharge rate, over the τa0.9wide is ontemperature the order ofrange 0.62 µs, room temperature to 150 °C. In addition, the 6BMZ ceramic also exhibits excellent cycling stability demonstrating that the 0.94(Sr0.7 Bi0.2 )TiO3 -0.06Bi(Mg0.5 Zr0.5 )O3 ceramic shows promising potential as over 105dielectric charging–discharging cycles. Together withenergy its faststorage discharge rate, the τ0.9 is on the order of a lead-free for high-temperature, high-power applications. 0.62 μs, demonstrating that the 0.94(Sr0.7Bi0.2)TiO3-0.06Bi(Mg0.5Zr0.5)O3 ceramic shows promising Author Contributions: Conceptualization, and S.Z.; validation, X.K. andenergy S.Z.; formal analysis, X.K. and potential as a lead-free dielectric for X.K. high-temperature, high-power storage applications. L.Y.; investigation, X.K. and L.Y.; resources, X.K.; data curation, X.K.; writing—original draft preparation, X.K.; writing—review and editing, X.K., L.Y., Z.C. and S.Z.; visualization, X.K.; supervision, S.Z. and Z.C. All authors Author Contributions: X.K. S.Z.; validation, X.K. and S.Z.; formal analysis, X.K. and have read and agreed to the Conceptualization, published version of the and manuscript. L.Y.; investigation, X.K. and L.Y.; resources, X.K.; data curation, X.K.;grant writing—original draft preparation, X.K.; Funding: This research was funded by the Australian Research Council, number FT140100698. writing—review and editing, X.K., L.Y., Z.C. and S.Z.; visualization, X.K.; supervision, S.Z. and Z.C. Acknowledgments: The authors acknowledge the use of facilities within the UOW Electron Microscopy Centre. TheFunding: authors also likewas to thank Dr.by Andrew Nattestad for preparing thegrant electrodes. Thiswould research funded the Australian Research Council, number FT140100698. Conflicts of Interest: The authors declare no conflict interest. Acknowledgments: The authors acknowledge theofuse of facilities within the UOW Electron Microscopy Centre. The authors also would like to thank Dr. Andrew Nattestad for preparing the electrodes. References 1. 2. 3. 4. 5. Conflicts of Interest: The authors declare no conflict of interest. Hao, X. A review on the dielectric materials for high energy-storage application. J. Adv. Dielectr. 2013, 3, 1330001. [CrossRef] References Yang, L.; Kong, X.; Li, F.; Hao, H.; Cheng, Z.; Liu, H.; Li, J.-F.; Zhang, S. Perovskite lead-free dielectrics for 1. energy Hao,storage X. A review on the Prog. dielectric materials for102, high energy-storage applications. Mater. Sci. 2019, 72–108. [CrossRef]application. J. Adv. Dielectr. 2013, 3, 1330001. Zou, K.; Dan, Y.; Xu, H.; Zhang, Q.; Lu, Y.; Huang, H.; He, Y. Recent advances in lead-free dielectric materials 2. for energy Yang, L.; Kong, Mater. X.; Li, F.; Cheng, Liu, H.;[CrossRef] Li, J.-F.; Zhang, S. Perovskite lead-free dielectrics for storage. Res.Hao, Bull.H.; 2019, 113,Z.; 190–201. energyR.W.; storage applications. Prog. Mater. Sci. 2019, 102, 72–108. Johnson, Evans, J.L.; Jacobsen, P.; Thompson, J.R.; Christopher, M. The changing automotive environment: High-temperature electronics. IEEE Trans. Electron. Packag. Manuf. 2004, 27, 164–176. [CrossRef] Li, Q.; Yao, F.-Z.; Liu, Y.; Zhang, G.; Wang, H.; Wang, Q. High-Temperature Dielectric Materials for Electrical Energy Storage. Annu. Rev. Mater. Res. 2018, 48, 219–243. [CrossRef] Materials 2020, 13, 180 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 12 of 13 Li, Q.; Chen, L.; Gadinski, M.R.; Zhang, S.; Zhang, G.; Li, H.U.; Iagodkine, E.; Haque, A.; Chen, L.-Q.; Jackson, T.N.; et al. Flexible high-temperature dielectric materials from polymer nanocomposites. Nature 2015, 523, 576. [CrossRef] Ogihara, H.; Randall, C.A.; Trolier-McKinstry, S. High-energy density capacitors utilizing 0.7BaTiO3 –0.3BiScO3 ceramics. J. Am. Ceram. Soc. 2009, 92, 1719–1724. [CrossRef] Shao, T.; Du, H.; Ma, H.; Qu, S.; Wang, J.; Wang, J.; Wei, X.; Xu, Z. Potassium–sodium niobate based lead-free ceramics: Novel electrical energy storage materials. J. Mater. Chem. A 2017, 5, 554–563. [CrossRef] Li, W.-B.; Zhou, D.; Pang, L.-X.; Xu, R.; Guo, H.-H. Novel barium titanate based capacitors with high energy density and fast discharge performance. J. Mater. Chem. A 2017, 5, 19607–19612. [CrossRef] Ge, J.; Remiens, D.; Dong, X.; Chen, Y.; Costecalde, J.; Gao, F.; Cao, F.; Wang, G. Enhancement of energy storage in epitaxial PbZrO3 antiferroelectric films using strain engineering. Appl. Phys. Lett. 2014, 105, 112908. [CrossRef] Wang, Y.; Hao, X.; Yang, J.; Xu, J.; Zhao, D. Fabrication and energy-storage performance of (Pb, La)(Zr, Ti)O3 antiferroelectric thick films derived from polyvinylpyrrolidone-modified chemical solution. J. Appl. Phys. 2012, 112, 034105. [CrossRef] Wang, Z.; Cao, M.; Yao, Z.; Song, Z.; Li, G.; Hu, W.; Hao, H.; Liu, H. Dielectric relaxation behavior and energy storage properties in SrTiO3 ceramics with trace amounts of ZrO2 additives. Ceram. Int. 2014, 40, 14127–14132. [CrossRef] Zhang, G.-F.; Liu, H.; Yao, Z.; Cao, M.; Hao, H. Effects of Ca doping on the energy storage properties of (Sr, Ca)TiO3 paraelectric ceramics. J. Mater. Sci. Mater. Electron. 2015, 26, 2726–2732. [CrossRef] Chen, A.; Zhi, Y. Dielectric relaxor and ferroelectric relaxor: Bi-doped paraelectric SrTiO3 . J. Appl. Phys. 2002, 91, 1487–1494. Shen, Z.Y.; Li, Y.M.; Hu, Q.G.; Luo, W.Q.; Wang, Z.M. Dielectric properties of B–site charge balanced Dy–doped SrTiO3 ceramics for energy storage. J. Electroceramics 2015, 34, 236–240. [CrossRef] Zhang, G.-F.; Cao, M.; Hao, H.; Liu, H. Energy storage characteristics in Sr(1-1.5x) Bix TiO3 ceramics. Ferroelectrics 2013, 447, 86–94. [CrossRef] Li, J.; Li, F.; Xu, Z.; Zhang, S. Multilayer Lead-Free Ceramic Capacitors with Ultrahigh Energy Density and Efficiency. Adv. Mater. 2018, 30, 1802155. [CrossRef] Song, B.; Wu, S.; Li, F.; Chen, P.; Shen, B.; Zhai, J. Excellent energy storage density and charge–discharge performance of a novel Bi0.2 Sr0.7 TiO3 –BiFeO3 thin film. J. Mater. Chem. C 2019, 7, 10891–10900. [CrossRef] Beuerlein, M.A.; Kumar, N.; Usher, T.M.; Brown-Shaklee, H.J.; Raengthon, N.; Reaney, I.M.; Cann, D.P.; Jones, J.L.; Brennecka, G.L. Current understanding of structure–processing–property relationships in BaTiO3 –Bi(M)O3 dielectrics. J. Am. Ceram. Soc. 2016, 99, 2849–2870. [CrossRef] Hu, Q.; Jin, L.; Wang, T.; Li, C.; Xing, Z.; Wei, X. Dielectric and temperature stable energy storage properties of 0.88BaTiO3 –0.12Bi(Mg1/2 Ti1/2 )O3 bulk ceramics. J. Alloy. Compd. 2015, 640, 416–420. [CrossRef] Zhang, Q.; Li, Z. Weakly coupled relaxor behavior of BaTiO3 -Bi(Mg1/2 Ti1/2 )O3 lead-free ceramics. J. Adv. Dielectr. 2013, 3, 1320001. [CrossRef] Wu, L.; Wang, X.; Li, L. Lead-free BaTiO3 –Bi (Zn2/3 Nb1/3 )O3 weakly coupled relaxor ferroelectric materials for energy storage. RSC Adv. 2016, 6, 14273–14282. [CrossRef] Jiang, X.; Hao, H.; Zhang, S.; Lv, J.; Cao, M.; Yao, Z.; Liu, H. Enhanced energy storage and fast discharge properties of BaTiO3 based ceramics modified by Bi(Mg1/2 Zr1/2 )O3 . J. Eur. Ceram. Soc. 2019, 39, 1103–1109. [CrossRef] Zeb, A.; Milne, S.J. Temperature-stable dielectric properties from −20 ◦ C to 430 ◦ C in the system BaTiO3 –Bi(Mg0. 5 Zr0. 5 )O3 . J. Eur. Ceram. Soc. 2014, 34, 3159–3166. [CrossRef] Kong, X.; Yang, L.; Cheng, Z.; Zhang, S. Bi-modified SrTiO3 -based ceramics for high-temperature energy storage applications. J. Am. Ceram. Soc. 2019, 1–10. [CrossRef] Gerson, R.; Marshall, T.C. Dielectric breakdown of porous ceramics. J. Appl. Phys. 1959, 30, 1650–1653. [CrossRef] Xiong, B.; Hao, H.; Zhang, S.; Liu, H.; Cao, M. Structure, Dielectric Properties and Temperature Stability of BaTiO3 –Bi(Mg1/2 Ti1/2 )O3 Perovskite Solid Solutions. J. Am. Ceram. Soc. 2011, 94, 3412–3417. [CrossRef] Choi, D.H.; Baker, A.; Lanagan, M.; Trolier-McKinstry, S.; Randall, C. Structural and Dielectric Properties in (1−x)BaTiO3 –xBi(Mg1 /2Ti1/2 )O3 Ceramics (0.1 ≤ x ≤ 0.5) and Potential for High-Voltage Multilayer Capacitors. J. Am. Ceram. Soc. 2013, 96, 2197–2202. [CrossRef] Materials 2020, 13, 180 29. 30. 31. 32. 33. 34. 35. 36. 37. 13 of 13 Shvartsman, V.V.; Lupascu, D.C. Lead-Free Relaxor Ferroelectrics. J. Am. Ceram. Soc. 2012, 95, 1–26. [CrossRef] Sakurai, M.; Kanehara, K.; Takeda, H.; Tsurumi, T.; Hoshina, T. Wideband dielectric spectroscopy of (Sr0.7 Bi0.2 )TiO3 ceramics and its microscopic mechanism of polarization. J. Ceram. Soc. Jpn. 2016, 124, 664–667. [CrossRef] Subbarao, E. A family of ferroelectric bismuth compounds. J. Phys. Chem. Solids 1962, 23, 665–676. [CrossRef] Wang, T.; Jin, L.; Li, C.; Hu, Q.; Wei, X. Relaxor ferroelectric BaTiO3 –Bi (Mg2/3 Nb1/3 )O3 ceramics for energy storage application. J. Am. Ceram. Soc. 2015, 98, 559–566. [CrossRef] Zheng, D.; Zuo, R.; Zhang, D.; Li, Y. Novel BiFeO3 –BaTiO3 –Ba(Mg1/3 Nb2/3 )O3 Lead-Free Relaxor Ferroelectric Ceramics for Energy-Storage Capacitors. J. Am. Ceram. Soc. 2015, 98, 2692–2695. [CrossRef] Zhou, M.; Liang, R.; Zhou, Z.; Dong, X. Novel BaTiO3 -based lead-free ceramic capacitors featuring high energy storage density, high power density, and excellent stability. J. Mater. Chem. C 2018, 6, 8528–8537. [CrossRef] Yuan, Q.; Yao, F.; Wang, Y.; Ma, R.; Wang, H. Relaxor ferroelectric 0.9BaTiO3 –0.1Bi(Zn0.5 Zr0.5 )O3 ceramic capacitors with high energy density and temperature stable energy storage properties. J. Mater. Chem. C 2017, 5, 9552–9558. [CrossRef] Yang, L.; Kong, X.; Cheng, Z.; Zhang, S. Ultra-high energy storage performance with mitigated polarization saturation in lead-free relaxors. J. Mater. Chem. A 2019, 7, 8573–8580. [CrossRef] Rabuffi, M.; Picci, G. Status quo and future prospects for metallized polypropylene energy storage capacitors. IEEE Trans. Plasma Sci. 2002, 30, 1939–1942. [CrossRef] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
https://openalex.org/W2597002197	https://europepmc.org/articles/pmc5354111?pdf=render	English	null	Contrasting determinants for the introduction and establishment success of exotic birds in Taiwan using decision trees models	PeerJ	2,017	cc-by	6,980	How to cite this article Liang et al. (2017), Contrasting determinants for the introduction and establishment success of exotic birds in Taiwan using decision trees models. PeerJ 5:e3092; DOI 10.7717/peerj.3092 Contrasting determinants for the introduction and establishment success of exotic birds in Taiwan using decision trees models Shih-Hsiung Liang1, Bruno Andreas Walther2 and Bao-Sen Shieh3,4 1 Department of Biotechnology, National Kaohsiung Normal University, Kaohsiung, Taiwan 2 Master Program in Global Health and Development, College of Public Health, Taipei Medical University, Taipei, Taiwan 3 Department of Biomedical Science and Environmental Biology, Kaohsiung Medical University, Kaohsiung Taiwan 4 Department of Medical Research, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan ABSTRACT Therefore, we suggest that future success for Distributed under Creative Commons CC-BY 4.0 OPEN ACCESS introduction and establishment of exotic birds may be gauged by simply looking at previous success in invading other countries. Finally, we found that species traits related to reproduction were more important in establishment models than in introduction models; importantly, these determinants were not averaged but either minimum or maximum values of species traits. Therefore, we suggest that in addition to averaged values, reproductive potential represented by minimum and maximum values of species traits should be considered in invasion studies. Subjects Biodiversity, Biogeography, Ecology, Mathematical Biology Keywords Biological invasion, Alien birds, Model comparison, Gradient boosting, Random forest j y, g g p y, gy, gy Keywords Biological invasion, Alien birds, Model comparison, Gradient boosting, Random forest ABSTRACT Background. Biological invasions have become a major threat to biodiversity, and identifying determinants underlying success at different stages of the invasion process is essential for both prevention management and testing ecological theories. To investigate variables associated with different stages of the invasion process in a local region such as Taiwan, potential problems using traditional parametric analyses include too many variables of different data types (nominal, ordinal, and interval) and a relatively small data set with too many missing values. Methods. We therefore used five decision tree models instead and compared their performance. Our dataset contains 283 exotic bird species which were transported to Taiwan; of these 283 species, 95 species escaped to the field successfully (introduction success); of these 95 introduced species, 36 species reproduced in the field of Taiwan successfully (establishment success). For each species, we collected 22 variables associated with human selectivity and species traits which may determine success during the introduction stage and establishment stage. For each decision tree model, we performed three variable treatments: (I) including all 22 variables, (II) excluding nominal variables, and (III) excluding nominal variables and replacing ordinal values with binary ones. Five performance measures were used to compare models, namely, area under the receiver operating characteristic curve (AUROC), specificity, precision, recall, and accuracy. Submitted 5 December 2016 Accepted 14 February 2017 Published 14 March 2017 Corresponding author Bao-Sen Shieh, bsshieh@kmu.edu.tw Academic editor Michael Wink Additional Information and Declarations can be found on page 12 DOI 10.7717/peerj.3092 Copyright 2017 Liang et al. Distributed under Creative Commons CC-BY 4.0 Results. The gradient boosting models performed best overall among the five decision tree models for both introduction and establishment success and across variable treat- ments. The most important variables for predicting introduction success were the bird family, the number of invaded countries, and variables associated with environmental adaptation, whereas the most important variables for predicting establishment success were the number of invaded countries and variables associated with reproduction. Discussion. Our final optimal models achieved relatively high performance values, and we discuss differences in performance with regard to sample size and variable treatments. Our results showed that, for both the establishment model and introduction model, the number of invaded countries was the most important or second most important determinant, respectively. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 INTRODUCTION Biological invasions have become a major threat to biodiversity (Pimentel, Zuniga & Morrison, 2005). Hence, some studies of biological invasion have focused on how to prevent the invasion or how to eradicate the invasive species (Dana, Jeschke & García- de-Lomas, 2014). As more and more invasive species have spread into the wild, invasive species have also become important subjects in testing ecological theories in relation to niche and competition (e.g., Broennimann et al., 2007; Allen et al., 2015). Both prevention management and testing ecological theories require the identification of the key factors underlying success at different stages in the invasion process (Duncan, Blackburn & Sol, 2003); moreover, factors that are important to explain the invasion success have been suggested to be different at each stage of the invasion process (Kolar & Lodge, 2002; Williamson, 2006; Dawson, Burslem & Hulme, 2009). Compared with other vertebrate taxa, birds have a higher number of invasive species and invasion success rates in a study focusing on Europe and North America (Jeschke & Strayer, 2006). Previous studies on exotic birds have identified two major categories of factors associated with their success at the introduction and establishment stages: human selectivity factors and species traits. Human selectivity factors consist of factors such as taxa and geography selected non-randomly by humans during the transport or introduction stages of exotic birds (Duncan, Blackburn & Sol, 2003). Species traits, on the other hand, then play an important role during the introduction and establishment stages (Blackburn, Cassey & Lockwood, 2009). In Taiwan, at least 290 exotic species of pet birds have been imported, and a 9.7% rate of invasion success was estimated (Shieh et al., 2006). For the transport stage, non-random selectivity of exotic birds imported to Taiwan was associated with bird family, native geographic range, body size, and song production of species (Su, Cassey & Blackburn, 2014); as to the later stages of invasion, pet trade factors such as song attractiveness were significantly associated with introduction success but not establishment success (Su, Cassey & Blackburn, 2016). For the exotic birds of Taiwan, species traits that help to avoid stochastic extinction or to constrain establishment (cf. Sol, 2008) have not been investigated with regard to their influences on different stages of the invasion process. To investigate the effects of these Liang et al. INTRODUCTION (2017), PeerJ, DOI 10.7717/peerj.3092 2/15 factors which are associated with both human selectivity and species traits onto different stages of the invasion process in a local region such as Taiwan, two potential problems using traditional parametric analyses have been identified as (1) a relatively small data set with too many missing values and (2) too many variables of different types (nominal, ordinal, and interval). Machine learning is a new, advanced analytical method which overcomes many of the restrictions of traditional parametric analyses. We chose the decision tree method, a machine learning algorithm, because its advantages include no need to input data for missing values and no assumptions about the distribution of the data; therefore, this method is ideal for dealing with mixed data types, such as nominal, ordinal and interval variables (Olinsky, Kennedy & Brayton Kennedy, 2014). In studies of biological invasion, the decision tree method was first applied to investigating a data set of 45 fish species for risk assessment in the Great Lakes (Kolar & Lodge, 2002). In another recent study, Chen, Peng & Yang (2015) found that decision tree methods not only work best with nominal variables but also have higher performance values than traditional parametric methods in predicting alien herb invasion. In a comparative study of trait-based risk assessment for invasive species which included a bird data set, Keller, Kocev & Džeroski (2011) found that random forests (an ensemble method that creates multiple decision tree sub-models) was one of the two best performing methods. Vall-llosera & Sol (2009) investigated only one of the four stages of the invasion process, namely establishment success, in a global risk assessment for invasive birds. They found that their tree model had an overall predictive accuracy as high as the conventional statistical models (generalized linear mixed models). Besides these two studies, which only focused on the establishment stage for exotic birds using decision tree models, there are, to our knowledge, no other studies which used decision tree methods to analyze the determinants for both the introduction and establishment stages of exotic birds. Consequently, we decided to use decision tree methods to assess factors associated with human selectivity and species traits which determine the success during the introduction and establishment stages of exotic birds in Taiwan. We used five decision tree models which differed in regard to resampling the data set and compared their performance. INTRODUCTION An optimal prediction model was chosen based on five performance measures, and the relative importance of factors in the optimal model for introduction success and for establishment success was examined and compared. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Species of the data set Species of the data set The four stages of the invasion process were defined in Duncan, Blackburn & Sol (2003) as transport, introduction, establishment, and spread. In this study, we focused on the introduction and establishment stages. For a species to reach the introduction stage, it must have passed the transport stage. Therefore, we selected all the exotic species which had been transported to Taiwan’s main island (not including surrounding islands, such as Lanyu and Kinmen Island) as documented in Shieh et al. (2006) which included the results of Chi (1995), Severinghaus (1999) and Lin (2004). Whether a transported species Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 3/15 has passed the subsequent stages of the invasion process was based (1) on escaping records in the field (introduction success) and (2) breeding record in the field (establishment success). We followed the detailed methods of how to define introduction success and establishment success which were given in Su, Cassey & Blackburn (2016). However, Su, Cassey & Blackburn (2016) based their decision of establishment success on the respective species having been recorded to be breeding at least twice; instead, we based it on at least one record of fledglings actually having left the nest successfully. In order to record all the escaping and breeding records of bird species up to 2015, we continuously (1) checked information from the Chinese Wild Bird Federation (http://www.bird.org.tw/) database which is the main collector of wild bird data in Taiwan, as well as other Taiwanese websites dedicated to natural history observations of birds, (2) remained in contact with local ornithologists, birdwatchers and bird societies, and (3) included any relevant publications (e.g., Walther, 2011; Walther, 2014 for red- whiskered bulbul, Pycnonotus jocosus; Fan et al., 2009 for white-rumped shama, Copsychus malabaricus, or Shieh, Lin & Liang, 2016 for Asian glossy starling, Aplonis panayensis). Most of this updated information was published recently in a project report for the Taiwan Forestry Bureau (Liang & Shieh, 2016). Despite following the methods as described in Su, Cassey & Blackburn (2016), we independently collected all the data used in this analysis beginning in 2004 and ending in 2015. Our dataset thus contains 283 full species (although we entered subspecies in our dataset, for this analysis, we only used full species), which were transported to Taiwan (see above). Of these 283 species, 95 species escaped to the field successfully (introduction success). Species of the data set Of these 95 species, 36 species reproduced in the field of Taiwan successfully (establishment success) (see Table S1 for species list). Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Decision trees models and variable treatments To investigate the possible effects of nominal variables (family and order) and ordinal variables on the performance of the decision tree models, we conducted three variable treatments for modeling: (I) including all variables, (II) excluding nominal variables, and (III) excluding nominal variables and replacing ordinal values with binary ones; e.g., changing habitat values of 0–4 to 0 (natural habitats) and habitat values of 5–6 to 1 (artificial habitats). For each variable treatment, we used five decision tree models (DT_no bagging, DT_bagging 90%, DT_bagging 100%, gradient boosting, and HP forest) to predict the outcomes of introduction and establishment, respectively. Modeling processes and comparisons of model performance were implemented using SAS Enterprise Miner 13.1 (for diagrams of process flow, see supplementary files Figs. S1 & S2). Because of the small data set, no data partition was implemented; that is, all data were used as training data. Instead, other methods, such as bagging and cross validation, which have been suggested for the use with small data sets (SAS Institute Inc., 2013), were used in the present study. DT_no bagging is the traditional classification tree method by constructing a layered tree model with the following settings: splitting rule = Gini, cross validation with 10 subsets and 100 repeats. The DT_bagging 90% and DT_bagging 100% methods used the same setting of splitting rule and cross validation as the DT_no bagging method but with bagging 90% or 100% of the data set for 50 times, respectively. Gradient boosting is a boosting method that resamples the data set to produce a series of decision trees which together form a single predictive model which has been found to be less prone to overfitting the data than a single decision tree (Georges, 2008). HP Forest is the random forest method which builds many parallel trees forming a forest; a tree in the forest is a sample without replacement from all the available observations, and the input variables that are considered for splitting a node are randomly selected from all the available inputs (Hall et al., 2014). Variables We collected 22 variables for each species, including two nominal ones (order and family), six ordinal ones (latitude overlap with Taiwan: 0–2, migration pattern: 0–3, nesting location: 0–3, feeding: 1–3, diet: 1–6, and habitat: 0–6), three binary ones (hole nest, Taiwan genus_resident, dichromatism), and 11 interval ones (clutch size: clutch, maximum clutch size: Mclutch, incubation days: incubation, minimum incubation days: Minincub, body length: length, maximum body length: Mlength, body mass: Mass, maximum body mass: Mmass, the number of invaded countries: Invcountry_Max, distribution range (km2): Range, the number of subspecies: subspecies) (see supplementary file Table S2 for code descriptions of variables). The variable Taiwan genus_resident was based on the information in Hsiao & Li (2014). For the other variables, we gathered species information from the books of Del Hoyo et al. (1992–2011), Dunning Jr (1993), and internet databases of IUCN (http://www.iucn.org) and BirdLife International Datazone (http://datazone.birdlife.org) (see Table S1 for associated information of each species and Table S2 for code descriptions of variables). When we collected the values for reproduction and body size for each species, we usually found a given range instead of fixed values in the references. In order to account for the maximum adaptation and reproduction potential in the invasion process, we used maximum values such as maximum body mass or minimum Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 4/15 values such as minimum incubation days in addition to averaged values. To determine the number of invaded countries (Invcountry_Max), we counted the total (or maximum) number of countries in which occurrences of introduced populations of each respective species were reported. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Decision trees models and variable treatments We calculated five performance measures to compare models, namely, the area under the receiver operating characteristic curve (AUROC), the specificity which measures the fraction of negative events that were correctly labeled, the precision which measures the fraction of positively labeled outcomes that were correctly labeled, the recall which measures the fraction of positive events that were correctly labeled, and the accuracy which measures the fraction of all events that were correctly labeled (accuracy = 1 −misclassification rate) (Söhngen, Chang & Schomburg, 2011). These five performance measures have the same range (0–1), and we gave each measure equal weight in evaluating the model performance in accordance with Chen, Peng & Yang (2015). The higher the values of these five performance measures are, the better the model performs; therefore, we summed up the five values (from hereupon called the ‘‘total score’’) and chose the model with the Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 5/15 Figure 1 The visual output of the introduction model based on the classification tree method for exotic birds of Taiwan generated from the dataset of 283 transported species, of which 95 species successfully escaped in the field (see Table S1 for associated information of each species and Table S2 for code descriptions of variables). Figure 1 The visual output of the introduction model based on the classification tree method for exotic birds of Taiwan generated from the dataset of 283 transported species, of which 95 species successfully escaped in the field (see Table S1 for associated information of each species and Table S2 for code descriptions of variables). highest sum as our final optimal model. We then compared the relative importance of each of the variables in the optimal introduction model and establishment model. For illustrative purposes, we chose the visual output of the resulting trees of DT_no bagging of variable treatment I for our figures (Figs. 1 and 2). Such visual outputs are not possible for the other four methods (namely, DT_bagging 90%, DT_ bagging 100%, gradient boosting, and HP forest). 6/15 Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Liang et al. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Decision trees models and variable treatments (2017), PeerJ, DOI 10.7717/peerj.3092 Figure 2 The visual output of the establishment model based on the classification tree method for ex- otic birds of Taiwan generated from the dataset of 95 introduced species, of which 36 species success- fully reproduced in the field (see Table S1 for associated information of each species and Table S2 for code descriptions of variables). Figure 2 The visual output of the establishment model based on the classification tree method for ex- otic birds of Taiwan generated from the dataset of 95 introduced species, of which 36 species success- fully reproduced in the field (see Table S1 for associated information of each species and Table S2 for code descriptions of variables). Figure 2 The visual output of the establishment model based on the classification tree method for ex- otic birds of Taiwan generated from the dataset of 95 introduced species, of which 36 species success- fully reproduced in the field (see Table S1 for associated information of each species and Table S2 for code descriptions of variables). We used the decision tree models described above to build various versions of two kinds of models: (1) introduction success prediction models and (2) establishment success prediction models. However, for brevity’s sake, from hereupon we will call them introduction models and establishment models, respectively. RESULTS Across the three variable treatments and for both the introduction models (Table 1) and establishment models (Table 2), the gradient boosting models always achieved the highest score among the five decision tree models (i.e., it performed best overall). However, this overall best performance did not mean that gradient boosting always performed best when comparing values of the five performance measures. For instance, Table 1 (see also supplementary file Fig. S3 for receiver operating characteristic curves, and supplementary file Fig. S4 for classification charts) shows that gradient boosting only performed best for accuracy in variable treatment I and II; otherwise, other models always performed better using the other four performance measures. Nevertheless, across all three treatments, the total score is always highest for gradient boosting for the introduction models (Table 1). Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 7/15 Table 1 Comparison of five performance measures among five introduction models of exotic birds in Taiwan, separately for three variable treatments (see ‘Methods’ for details). Model AUROC Specificity Precision Recall Accuracy Total Variable treatment I DT_no bagging 0.894 0.830 0.722 0.874 0.845 4.164 DT_bagging 90% 0.970 0.936 0.782 0.453 0.774 3.914 DT_bagging 100% 0.976 0.910 0.742 0.516 0.777 3.921 Gradient boosting 0.936 0.941 0.869 0.768 0.883 4.398 HP Forest 0.903 0.963 0.873 0.505 0.809 4.053 Variable treatment II DT_no bagging 0.904 0.872 0.765 0.821 0.855 4.217 DT_bagging 90% 0.949 0.899 0.683 0.432 0.742 3.705 DT_bagging 100% 0.955 0.910 0.742 0.516 0.777 3.900 Gradient Boosting 0.924 0.915 0.816 0.747 0.859 4.261 HP Forest 0.894 0.963 0.848 0.411 0.777 3.893 Variable treatment III DT_no bagging 0.910 0.888 0.781 0.789 0.855 4.224 DT_bagging 90% 0.946 0.910 0.691 0.400 0.739 3.685 DT_bagging 100% 0.953 0.888 0.700 0.516 0.763 3.820 Gradient Boosting 0.919 0.926 0.827 0.705 0.852 4.229 HP Forest 0.888 0.957 0.840 0.442 0.784 3.912 Table 2 Comparison of five performance measures among five establishment models of exotic birds in Taiwan, separately for three variable treatments (see ‘Methods’ for details). Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 RESULTS Model AUROC Specificity Precision Recall Accuracy Total Variable treatment I DT_no bagging 0.839 0.898 0.806 0.694 0.821 4.059 DT_bagging 90% 0.945 0.932 0.800 0.444 0.747 3.869 DT_bagging 100% 0.963 0.949 0.842 0.444 0.758 3.957 Gradient Boosting 0.985 1.000 1.000 0.861 0.947 4.793 HP Forest 0.901 0.983 0.875 0.194 0.684 3.638 Variable treatment II DT_no bagging 0.839 0.898 0.806 0.694 0.821 4.059 DT_bagging 90% 0.942 0.932 0.800 0.444 0.747 3.866 DT_bagging 100% 0.963 0.949 0.842 0.444 0.758 3.957 Gradient boosting 0.976 0.983 0.969 0.861 0.937 4.726 HP Forest 0.914 1.000 1.000 0.167 0.684 3.765 Variable treatment III DT_no bagging 0.839 0.898 0.806 0.694 0.821 4.059 DT_bagging 90% 0.936 0.932 0.800 0.444 0.747 3.860 DT_bagging 100% 0.940 0.949 0.842 0.444 0.758 3.934 Gradient boosting 0.971 1.000 1.000 0.778 0.916 4.665 HP Forest 0.912 1.000 1.000 0.139 0.674 3.725 Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Relative Importance Variable Figure 3 Relative importance of variables in the prediction models using the gradient boosting ap- proach (grey bars for introduction models and black bars for establishment models). For descriptions of codes for variables, see Table S2. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Relative Importance Variable Figure 3 Relative importance of variables in the prediction models using the gradient boosting ap- proach (grey bars for introduction models and black bars for establishment models). For descriptions of codes for variables, see Table S2. Figure 3 Relative importance of variables in the prediction models using the gradient boosting ap- proach (grey bars for introduction models and black bars for establishment models). For descriptions of codes for variables, see Table S2. Figure 3 Relative importance of variables in the prediction models using the gradient boosting ap- proach (grey bars for introduction models and black bars for establishment models). For descriptions of codes for variables, see Table S2. Figure 3 Relative importance of variables in the prediction models using the gradient boosting ap- proach (grey bars for introduction models and black bars for establishment models). For descriptions of codes for variables, see Table S2. For the establishment models (Table 2, see also supplementary file Fig. S5 for receiver operating characteristic curves, and supplementary file Fig. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 RESULTS S6 for classification charts), however, gradient boosting has the highest total score for all the three treatments and also for most of the five performance measures (the only exceptions being specificity and precision in variable treatment II). Therefore, we considered gradient boosting the optimal model for both the introduction models and establishment models and only considered its results from hereupon. Looking across the three different variable treatment methods I–III, gradient boosting performed best with variable treatment I for the introduction models (Table 1) as well as the establishment models (Table 2). For variable treatments II and III, the total score decreased by only 0.169 (4%) and 0.128 (3%), respectively. We also note that this decreasing trend across variable treatments is maintained for most of the five performance measures. Furthermore, the values of the performance measures are all >0.7 and 60% are >0.9, which means that the performance was consistently high or very high. In the optimal introduction model, family and the number of invaded countries (Invcountry_Max) were the most important variables, and their relative importance values were 1 and 0.888, respectively (Fig. 3). The top six variables with an importance value >0.3 also included maximum body mass (0.394), order (0.384), latitude overlap with Taiwan (0.354), and distribution range (0.345). For the introduction model based on the classification tree method (Fig. 1), the number of invaded countries was the most important determinant, as it appeared at the top of the tree, which means that the 84 species with Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 9/15 any record of invading other countries had a 66.7% chance of successful introduction to Taiwan. Among these 84 species, the 72 species which had a migration pattern categorized as sedentary (0), local movement (1) or partial migration (2) had a 73.6% chance of successful introduction, while the 12 species categorized as migrants (3) had only a 25.0% chance of successful introduction. Among the 199 species which had no record of invading other countries, family was chosen as an important determinant of successful introduction. In the optimal establishment model, the number of invaded countries and distribution range were the most important variables, and their relative importance values were 1 and 0.826, respectively (Fig. 3). The top six variables with an importance value >0.6 also included minimum incubation days (Minincub, 0.647), migration pattern (Migration, 0.633), clutch size (Clutch, 0.62), and habitat type (Habitat, 0.616). RESULTS The relative importance of the variable family decreased to 0.569 which is therefore much lower than in the optimal introduction model (see above). For the establishment model based on the classification tree method (Fig. 2), the number of invaded countries was again the most important determinant, as it appeared at the top of the tree. In this case it means that the 39 species with a record of invading at least two countries had a 59.0% chance of successful establishment in Taiwan, while the 56 species with a record of invading fewer than two countries had only a 23.2% chance of successful establishment. Among the 39 species noted above, the 21 species with a maximum clutch size (Mclutch) <5.5 had an 81.0% chance of successful establishment, while the other 18 with a maximum clutch size of ≥5.5 had only a 33.3% chance of successful establishment. Finally, among the 56 species noted above, the eight species with a body length (Length) ≥36.5 cm had a 62.5% chance of successful establishment. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Model comparisons and variable treatment comparisons Model comparisons and variable treatment comparisons Our results showed that for the complete data set of 283 transported species or for the data set of 95 introduced species, the gradient boosting method performed better than the other four decision tree methods. While we calculated five performance measures, the only other study which used the decision tree method on a bird data set was Keller, Kocev & Džeroski (2011) who calculated only AUROC and accuracy values. Considering AUROC values first, the AUROC values of gradient boosting of our study were over 0.919 in the introduction models and over 0.971 in the establishment models; thus, they were all higher than our values for the random forests method. This is in contrast to the results of Keller, Kocev & Džeroski (2011) who found that, based on the AUROC values, random forests performed better than gradient boosting for both their New Zealand and Australia bird data sets. Specifically, AUROC values for gradient boosting for their New Zealand (79 species with 11 traits) and Australia (52 species with 11 traits) data sets were 0.682 and 0.681, respectively, whereas AUROC values for random forests were 0.731 and 0.745, respectively. Pearce & Ferrier (2000) suggested that AUROC values between 0.7 and 0.9 indicate a reasonable discrimination ability of models, and values higher than 0.9 indicate a very good discrimination ability of models. The higher AUROC values of our study might Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 have resulted from the inclusion of more variables (up to 22 variables) rather than larger samples used for analysis. In our study, both the introduction model and establishment model used 22 variables, and we found higher AUROC values (0.971–0.985) in the smaller data set (namely, the establishment model with 95 species) than in the larger data set (namely, the introduction model with 283 species) (AUROC values 0.919–0.936). We therefore suggest that even a small data set (less than 100 species) with up to 22 variables can achieve a prediction model of good performance using the gradient boosting method. have resulted from the inclusion of more variables (up to 22 variables) rather than larger samples used for analysis. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Model comparisons and variable treatment comparisons In our study, both the introduction model and establishment model used 22 variables, and we found higher AUROC values (0.971–0.985) in the smaller data set (namely, the establishment model with 95 species) than in the larger data set (namely, the introduction model with 283 species) (AUROC values 0.919–0.936). We therefore suggest that even a small data set (less than 100 species) with up to 22 variables can achieve a prediction model of good performance using the gradient boosting method. Comparing the performances of variable treatment I with variable treatments II and III, we found little difference on model performance. Treatment II excluded nominal variables, and treatment III changed ordinal variables of species traits into binary variables, but neither one of these changes really had much discernable influence on overall performance. Our results therefore provide evidence to support the use of ordinal variables of species traits, and that there is no need to convert ordinal variables of species traits to binary ones for their use in decision tree models. Comparing the performances of variable treatment I with variable treatments II and III, we found little difference on model performance. Treatment II excluded nominal variables, and treatment III changed ordinal variables of species traits into binary variables, but neither one of these changes really had much discernable influence on overall performance. Our results therefore provide evidence to support the use of ordinal variables of species traits, and that there is no need to convert ordinal variables of species traits to binary ones for their use in decision tree models. Predictors of introduction and establishment success in exotic birds Predictors of introduction and establishment success in exotic birds Perhaps the most interesting and novel result of our study is that, for both the establishment model and introduction model, the number of previously invaded countries was the most important or second most important determinant in all the models. Therefore, our study suggests that future success for introduction and establishment of birds can be gauged by simply looking at previous success in invading other countries or regions. Future studies should include this variable to confirm our supposition because it might be a very simple and straightforward way to predict the potential invasion success of a species: if it has been successful before, it will probably be successful again. While this variable could not have been established a few decades ago, we now have a global track record of successful species invasions, and we might therefore be able to use it to better predict future local or regional invasions. Furthermore, global studies could investigate what species traits and other relevant factors, e.g., local ecological factors, are related to the number of successfully invaded countries; or, given the differential size of countries, the actual area invaded. Another important determinant was family. While family was the most important variable in the optimal introduction model, it dropped to being only the seventh most important variable in the optimal establishment model. In other words, family was an important determinant of introduction but not establishment in Taiwan. Our results thus differ from those of a global study which found that bird family was also a good predictor for establishment success (Lockwood, 1999). The discrepancy between this study and our study could result from the fact that exotic birds in Taiwan are primarily introduced for aesthetic reasons but not for hunting (Shieh et al., 2006; Su, Cassey & Blackburn, 2016), while the global data set included many hunted species. Several species traits were also chosen as important determinants for the introduction and establishment models. For the optimal introduction model, the top three selected species traits were maximum body mass (Mmass), latitude overlap with Taiwan (Overlap), and distribution range (Range). Among these three variables, maximum body mass was ranked the most important, and it also had a relative importance greater than that of two Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 11/15 other closely related measures, specifically, the averaged body mass (Mass) and body length (Length). Predictors of introduction and establishment success in exotic birds One possible explanation is that birds are usually heavier in captivity under well fed condition. Our data set contained primarily pet species (and not game species, which are prevalent in many other studies), and the body mass of pet birds might be higher than the average body mass of their wild congeners and therefore closer to the maximum attainable body mass. In order to consider the representability and the maximum adaptation potential in the invasion process, we therefore suggest that including maximum body mass may be important in order not to miss a potentially important determinant for the invasion success of exotic pet birds in particular. For example, Su, Cassey & Blackburn (2016) did not find that body mass had any influence on introduction success. However, they only used averaged body mass, and perhaps their result would have been different if they had also included maximum body mass. Furthermore, Cassey’s (2001) global study found that averaged body mass was significantly correlated with introduction success which further supports the role of some measure of body mass being an important determinant of introduction success. Finally, several species traits related to reproduction were also important, such as minimum incubation days (Minincub), clutch size (Clutch), dichromatism, and nesting location (Nesting); however, these determinants were more important in establishment success than in introduction success. Furthermore, given that some top ranking variables were associated with maximum or minimum values of species traits, we suggest that in addition to averagedvalues,reproductive potentialrepresentedbyminimumandmaximum values of species traits should be considered in prediction models of invasion studies. We conclude that decision tree models are efficient for the analysis of small data sets with mixed types of variables, including nominal, ordinal and interval variables, in predicting the invasion success of exotic birds. Our results further demonstrate that the most important determinants in predicting introduction success of exotic birds in Taiwan were the bird family, the number of invaded countries, and variables associated with environmental adaptation, whereas the most important determinants in predicting establishment success were the number of invaded countries and variables associated with reproduction. ACKNOWLEDGEMENTS We thank Liviu Parau and two anonymous reviewers for their valuable comments and suggestions on previous drafts of the manuscript. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Competing Interests Competing Interests The authors declare there are no competing interests. Supplemental Information Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.3092#supplemental-information. Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.3092#supplemental-information. Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.3092#supplemental-information. Author Contributions • Shih-Hsiung Liang and Bao-Sen Shieh conceived and designed the experiments, performed the experiments, analyzed the data, contributed reagents/materials/analysis tools, wrote the paper, prepared figures and/or tables, reviewed drafts of the paper. • Bruno Andreas Walther conceived and designed the experiments, contributed reagents/materials/analysis tools, wrote the paper, prepared figures and/or tables, reviewed drafts of the paper. Data Availability The following information was supplied regarding data availability: The following information was supplied regarding data availability: The raw data has been supplied as a Supplementary File. The raw data has been supplied as a Supplementary File. Supplemental Information Funding This work was supported by the Ministry of Science and Technology, Taiwan, R.O.C. (grant no. MOST 105 - 2311 - B - 037 – 002) and the Forestry Bureau of the Taiwanese Government. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Grant Disclosures The following grant information was disclosed by the authors: The following grant information was disclosed by the authors: The following grant information was disclosed by the authors: Ministry of Science and Technology, Taiwan, R.O.C.: MOST 105 - 2311 - B - 037 – 002. Forestry Bureau of the Taiwanese Government. Ministry of Science and Technology, Taiwan, R.O.C.: MOST 105 - 2311 - B - 037 – 002. Ministry of Science and Technology, Taiwan, R.O.C.: MOST 105 - 2311 - B - 037 – 002 Forestry Bureau of the Taiwanese Government. Forestry Bureau of the Taiwanese Government. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 REFERENCES Allen CR, Angeler DG, Moulton MP, Holling CS. 2015. The importance of scaling for detecting community patterns: success and failure in assemblages of introduced species. Diversity 7:229–241 DOI 10.3390/d7030229. Allen CR, Angeler DG, Moulton MP, Holling CS. 2015. The importance of scaling for detecting community patterns: success and failure in assemblages of introduced species. Diversity 7:229–241 DOI 10.3390/d7030229. Blackburn TM, Cassey P, Lockwood JL. 2009. The role of species traits in the establishment success of exotic birds. Global Change Biology 15:2852–2860 DOI 10.1111/j.1365-2486.2008.01841.x. DOI 10.1111/j.1365-2486.2008.01841.x. Broennimann O, Treier UA, Müller-schärer H, Thuiller W, Peterson AT, Guisan A. 2007. Evidence of climatic niche shift during biological invasion. Ecology Letters 10:701–709 DOI 10.1111/j.1461-0248.2007.01060.x. Cassey P. 2001. Are there body size implications for the success of globally introduced land birds? Ecography 24:413–420 DOI 10.1034/j.1600-0587.2001.d01-197.x. Chen L, Peng S, Yang B. 2015. Predicting alien herb invasion with machine learning models: biogeographical and life-history traits both matter. Biological Invasions 17:2187–2198 DOI 10.1007/s10530-015-0870-y. Chi W-L. 1995. An investigation report on pet bird trade in Taiwan. Taipei: Green Consumer’s Foundation (In Chinese). Dana ED, Jeschke JM, García-de-Lomas J. 2014. Decision tools for managing biological invasions: existing biases and future needs. Oryx 48:56–63 DOI 10.1017/S0030605312001263. DOI 10.1017/S0030605312001263. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 Dawson W, Burslem DF, Hulme PE. 2009. Factors explaining alien plant invasion success in a tropical ecosystem differ at each stage of invasion. Journal of Ecology 97:657–665 DOI 10.1111/j.1365-2745.2009.01519.x. Del Hoyo J, Elliott A, Sargatal J, Christie DA. 1992–2011. Handbook of the birds of the world. 1–16. Barcelona: Lynx Edicions. Duncan RP, Blackburn TM, Sol D. 2003. The ecology of bird introductions. Annual Re- view of Ecology, Evolution, and Systematics 34:71–98 Dunning Jr JB. 1993. CRC handbook of avian body masses. Boca Raton: CRC Press. Fan M-W, Lin R-S, Fang W, Lin Y-H. 2009. The distribution and abundance of the alien invasive white-rumped shama (Copsychus malabaricus) in Taiwan. Taiwania 54:248–254. Georges J. 2008. Applied analytics using SAS R⃝Enterprise MinerTM 5.3. Cary: SAS Publishing. Hall P, Dean J, Kabul IK, Silva J. 2014. An overview of machine learning with SAS R⃝enter- prise minerTM. Cary: SAS Institute Inc. Hsiao M-C, Li C-L. 2014. Field guide of wild bird in Taiwan. Taipei: Wild Bird Society of Taiwan. Jeschke JM, Strayer DL. 2006. Determinants of vertebrate invasion success in Europe and North America. Global Change Biology 12:1608–1619 DOI 10.1111/j.1365-2486.2006.01213.x. Keller RP, Kocev D, Džeroski S. 2011. Traitbased risk assessment for invasive species: high performance across diverse taxonomic groups, geographic ranges and machine learning/statistical tools. Diversity and Distributions 17:451–461 DOI 10.1111/j.1472-4642.2011.00748.x. Kolar CS, Lodge DM. 2002. Ecological predictions and risk assessment for alien fishes in North America. Science 298(5596):1233–1236 DOI 10.1126/science.1075753. Liang S-H, Shieh B-S. 2016. Application and analysis of the life history database to assess invasion impact of exotic animals (in Chinese with English abstract). Available at http://conservation.forest.gov.tw/File.aspx?fno=61675. Lin R-S. 2004. The black terrors of Hitchcock—current status of invasive birds in Taiwan. Nature Conservation Quarterly 48:38–43 (In Chinese). Lockwood JL. 1999. Using taxonomy to predict success among introduced avifauna: rela- tive importance of transport and establishment. Conservation Biology 13(3):560–567 DOI 10.1046/j.1523-1739.1999.98155.x. Olinsky A, Kennedy K, Brayton Kennedy B. 2014. Assessing gradient boosting in the reduction of misclassification error in the prediction of success for actuarial majors. Case Studies in Business, Industry and Government Statistics 5:12–16. Pearce J, Ferrier S. 2000. Evaluating the predictive performance of habitat mod- els developed using logistic regression. Ecological Modelling 133:225–245 DOI 10.1016/S0304-3800(00)00322-7. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 14/15 Pimentel D, Zuniga R, Morrison D. 2005. Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics 52:273–288 DOI 10.1016/j.ecolecon.2004.10.002. SAS Institute Inc. 2013. SAS R⃝enterprise minerTM 13.1 extension nodes: developer’s guide. Cary: SAS Institute Inc. Severinghaus L-L. 1999. Exotic birds in Taiwan. Wild Birds 7:45–58. Shieh B-S, Lin Y-H, Lee T-W, Chang C-C, Cheng K-T. 2006. Pet trade as sources of introduced bird species in Taiwan. Taiwania 51:81–86. Shieh B-S, Lin C-J, Liang S-H. 2016. Breeding biology of the invasive Asian Glossy Starling (Aplonis panayensis) in urban parks of Kaohsiung City, southern Taiwan. Taiwan Journal of Forest Science 31:63–70. Söhngen C, Chang A, Schomburg D. 2011. Development of a classification scheme for disease-related enzyme information. BMC Bioinformatics 12:329 DOI 10.1186/1471-2105-12-329. Söhngen C, Chang A, Schomburg D. 2011. Development of a classification scheme for disease-related enzyme information. BMC Bioinformatics 12:329 Sol D. 2008. Do successful invaders exist? Pre-adaptations to novel environments in terrestrial vertebrates. In: Biological invasions. Berlin Heidelberg: Springer, 127–141. Su S, Cassey P, Blackburn TM. 2014. Patterns of non-randomness in the composition and characteristics of the Taiwanese bird trade. Biological Invasions 16:2563–2575 DOI 10.1007/s10530-014-0686-1. Su S, Cassey P, Blackburn TM. 2016. Liang et al. (2017), PeerJ, DOI 10.7717/peerj.3092 DOI 10.1111/j.1365-2486.2006.01213.x. The wildlife pet trade as a driver of introduction and establishment in alien birds in Taiwan. Biological Invasions 18:215–229 DOI 10.1007/s10530-015-1003-3. Vall-llosera M, Sol D. 2009. A global risk assessment for the success of bird introduc- tions. Journal of Applied Ecology 46:787–795 DOI 10.1111/j.1365-2664.2009.01674.x. Walther BA. 2011. First documented nesting of the red-whiskered bulbul Pycnonotus jocosus in Taiwan. Taiwan Journal of Biodiversity 13:121–133. Walther BA. 2014. Second documented nesting of the red-whiskered bulbul Pycnonotus jocosus in Taiwan. Feather 266:20–25. Williamson M. 2006. Explaining and predicting the success of invading species at different stages of invasion. Biological Invasions 8:1561–1568 DOI 10.1007/s10530-005-5849-7. 15/15
https://openalex.org/W4313729354	https://shilap.org/revista/article/download/295/242	English	null	Dichagyris (Albocosta) batanga Gyulai & Saldaitis, sp. n. from China (Lepidoptera: Noctuidae)	SHILAP. Sociedad Hispano-Luso-Americana de Lepidopterología/SHILAP	2,021	cc-by	1,452	Abstract The diagnosis and description of a new species Dichagyris (Albocosta) batanga Gyulai & Saldaitis, sp. n. is given, with four colour figures and four genitalia figures. KEY WORDS: Lepidoptera, Noctuidae, Dichagyris, new species, China. eISSN: 2340-4078 ISSN: 0300-5267 eISSN: 2340-4078 ISSN: 0300-5267 SHILAP Revista de lepidopterología, 49 (194) junio 2021: 203-206 https://doi.org/10.57065/shilap.295 Resumen Se da la diagnosis y descripción de una especie nueva Dichagyris (Albocosta) batanga Gyulai & Saldaitis, sp. n., con cuatro figuras en color y cuatro figuras de genitalia. PALABRAS CLAVE: Lepidoptera, Noctuidae, Dichagyris, nueva especie, China. Introducción For the most recent revision of the subgenus Albocosta Fibiger & Lafontaine 1997 of the genus Dichagyris Lederer, 1857, with checklist of taxa, distribution, taxonomic interpretation, and new descriptions see GYULAI (2021). The further dissections of the late autumnal (from the second half of September to the end of November) Albocosta specimens from China, led to the recognition of a new species, was collected by the second author near Batang, Sichuan. The new one is the sister species of the Dichagyris (Albocosta) stentzi (Lederer, 1853), which is a widely distributed species in Asia (Russia (Altai, Tuva, Sayan and Baikal area, Transbaikalia, Amur region, Primorye territory, Sakhalin, Kuriles; pers. comm. V. Kononenko), Kazakhstan, Kyrgyzstan, Mongolia, Korea, China, Tibet, Nepal, India, Pakistan). All of the further Albocosta species dissected from late autumnal materials from China proved to Dichagyris (Albocosta) triangularis sinangularis Gyulai, 2021. Dichagyris (Albocosta) batanga Gyulai & Saldaitis, sp. n. from China (Lepidoptera: Noctuidae) P. Gyulai & A. Saldaitis PT = paratype f = female PT = paratype f = female PT = paratype f = female Abbreviations AFM = Alessandro Floriani (Milan, Italy) HNHM = Hungarian Natural History Museum (Budapest, Hungary) PGYM = collection of Péter Gyulai (Miskolc, Hungary) GYP = genitalia slide of Péter Gyulai HT = holotype 203 P. GYULAI & A. SALDAITIS Description of new taxa Dichagyris (Albocosta) batanga Gyulai & Saldaitis, sp. n. (Figs 1-2, 5-6) Holotype 1 0 (Fig. 1): CHINA, W Sichuan, 25 km N from Batang, 3100 m, dry valley, 30º12,049’N, 99º14,078’E, 19-20-IX-2007, leg. A. Saldaitis, slide no. GYP 5350 (coll. PGYM, later to be deposited in the HNHM). Paratype: 1 0, with same data, slide no. GYP 5367 (coll. AFM). Diagnosis: The females of the new species (Figs 1-2) are slightly larger than most of the females of the closest relative D. (A.) stentzi (Figs 3-4); forewing length is 18-19 mm wingspans 37-38 mm, versus 15-19 mm and 30-38 mm. The new species differs in its broader triangular basal black dash, darker costal area, less elongate orbicular stigma costally and slightly less sinuous subterminal line in the forewings and darker, fuscous, unicolourous hindwings. Figures 1-4.– Dichagyris (Albocosta) spp., adults. 1. D. (A.) batanga Gyulai & Saldaitis, sp. n., HT, 0, China, Sichuan (coll. PGYM); 2. D. (A.) batanga Gyulai & Saldaitis sp. n., PT, 0, China, Sichuan (coll. AFM); 3. D. (A.) stentzi (Led.), 0, Russia, Far East, Vladivostok distr., Nachodka (coll. PGYM); 4. D. (A.) stentzi (Led.), 0, China, Sichuan, Chuan Zhusi/Huang Long, Juraka pass, (coll. PGYM). 1 2 3 4 1 2 1 2 3 4 3 4 Figures 1-4.– Dichagyris (Albocosta) spp., adults. 1. D. (A.) batanga Gyulai & Saldaitis, sp. n., HT, 0, China, Sichuan (coll. PGYM); 2. D. (A.) batanga Gyulai & Saldaitis sp. n., PT, 0, China, Sichuan (coll. AFM); 3. D. (A.) stentzi (Led.), 0, Russia, Far East, Vladivostok distr., Nachodka (coll. PGYM); 4. D. (A.) stentzi (Led.), 0, China, Sichuan, Chuan Zhusi/Huang Long, Juraka pass, (coll. PGYM). In the female genitalia (Figs 5-6), the most conspicuous difference is the large lateral diverticulum in the inner side of the appendix bursae, which is very unique and well differs from that section of the D. (A.) stentzi (Figs 7-8) and those of the further relative species (see: GYULAI, 2021), as well. The In the female genitalia (Figs 5-6), the most conspicuous difference is the large lateral diverticulum in the inner side of the appendix bursae, which is very unique and well differs from that section of the D. (A.) stentzi (Figs 7-8) and those of the further relative species (see: GYULAI, 2021), as well. The 204 SHILAP Revta. lepid., 49 (194) junio 2021 DICHAGYRIS (ALBOCOSTA) BATANGA GYULAI & SALDAITIS, SP. N. Description of new taxa FROM CHINA appendix bursae of the holotype is affected in shape by the presence of spermatophores, however the large lateral diverticulum is well defined, too. Description (Figs 1-2): Forewing length 18-19 mm wingspans 37-38 mm. Palpi covered with dark red-brown scales, third segment tiny. Vertex light brown, collar blackish brown with broad light brownish-ochreous base; vesture of thorax and legs dark reddish-brown, that of abdomen brown. Forewings triangular, apex pointed. Ground colour of forewings dark reddish-brown, slightly lighter in the marginal area. Costa broadly pale brownish-ochreous until the tip of the reniform stigma, conjoining with the same-coloured orbicular macula and the light brownish reniform stigma; claviform stigma small, blackish. Basal dash black, broadly triangular. Transverse lines faint, antemedian line wavy, postmedian line evenly arcuate, subterminal line finely lacy with a small black dash in the costa; fringe brown. Hindwings evenly fuscous brown, discal spot a fine arch; fringe pale brown. Female genitalia (Figs 5-6): The main characters are the followings: setose, quadrangular papillae anales, short apophyses anteriores and much longer apophyses posteriores; sclerotized, broadly V-shaped antrum with more sclerotized, pincer-like, bilateral, symmetrical lobi; tubular, membranous, posteriorly broaden, longitudinally wrinkled ductus bursae; saccate appendix bursae and corpus bursae with numerous fine longitudinal wrinkles; the former one bears a large lateral diverticulum in the inner side. Bionomics and distribution: Two females were collected at ultraviolet light during two nights on 19-20-IX-2007 in remote part of west China Sichuan province near the Batang. Dichagyris (Albocosta) batanga was collected at altitude ranging 3100 meters in mountain river dry valley rarely covered by mixed forests dominated by various deciduous trees and bushes. Etymology: The specific epithet refers to the type locality. Acknowledgements The authors are grateful to V. S. Kononenko (Laboratory of Entomology, Vladivostok, Russia) for information on the distribution of Albocosta in Russia; to Adrienne Gyulai-Garai (Miskolc, Hungary) for much help in computer works and Alessandro Floriani (Milan, Italy) for access to his rich private collection provided. BIBLIOGRAFÍA FIBIGER, M. & LAFONTAINE, D., 1997.– Noctuidae III.– In M. FIBIGER. Noctuidae Europaeae, 3: 418 pp. Entomological Press, Sorø. FIBIGER, M. & LAFONTAINE, D., 1997.– Noctuidae III.– In M. FIBIGER. Noctuidae Europaeae, 3: 418 pp. Entomological Press, Sorø. GYULAI, P., 2021.– Notes on the subgenus Albocosta Fibiger & Lafontaine 1997, with descriptions of five new taxa from Asia (Lepidoptera, Noctuidae.– Ecologica Montenegrina, 40: 26-45. Available from http://dx.doi.org/10.37828/em.2021.40.2. GYULAI, P., 2021.– Notes on the subgenus Albocosta Fibiger & Lafontaine 1997, with descriptions of five new taxa from Asia (Lepidoptera, Noctuidae.– Ecologica Montenegrina, 40: 26-45. Available from http://dx.doi.org/10.37828/em.2021.40.2. p g LEDERER, J., 1857.– Die Noctuinen Europa’s mit Zuziehung einiger bisher meist dazu gezählter Arten Noctuinen Europa’s: 251 pp. Friedrich Manz, Wien. LEDERER, J., 1857.– Die Noctuinen Europa’s mit Zuziehung einiger bisher meist dazu gezählter Arten Noctuinen Europa’s: 251 pp. Friedrich Manz, Wien. *P. G. Mélyvölgy, 13/A H-3530 Miskolc HUNGRÍA / HUNGARY E-mail: adriennegyulai@gmail.com https://orcid.org/0000-0003-3878-2880 (Recibido para publicación / Received for publication 1-IV-2021) (Revisado y aceptado / Revised and accepted 20-IV-2021) (Publicado / Published 30-VI-2021) (Recibido para publicación / Received for publication 1-IV-2021) (Revisado y aceptado / Revised and accepted 20-IV-2021) (Publicado / Published 30-VI-2021) SHILAP Revta. lepid., 49 (194) junio 2021 205 P. GYULAI & A. SALDAITIS Figures 5-8.– Dichagyris (Albocosta) spp., female genitalia; 5. D. (A.) batanga Gyulai & Saldaitis, sp. n., HT, China, Sichuan, GYP 5350 (coll. PGYM); 6. D. (A.) batanga Gyulai & Saldaitis, sp. n., PT, China, Sichuan GYP 5367 (coll. AFM); 7. D. (A.) stentzi (Led.), Russia, Far East, GYP 5432 (coll. PGYM); 8. D. (A.) stentzi (Led.), China, Sichuan GYP 3269 (coll. PGYM). 5 6 7 8 Figures 5-8.– Dichagyris (Albocosta) spp., female genitalia; 5. D. (A.) batanga Gyulai & Saldaitis, sp. n., HT, China, Sichuan, GYP 5350 (coll. PGYM); 6. D. (A.) batanga Gyulai & Saldaitis, sp. n., PT, China, Sichuan GYP 5367 (coll. AFM); 7. D. (A.) stentzi (Led.), Russia, Far East, GYP 5432 (coll. PGYM); 8. D. (A.) stentzi (Led.), China, Sichuan GYP 3269 (coll. PGYM). 5 6 7 8 5 5 6 5 6 7 8 7 7 8 Figures 5-8.– Dichagyris (Albocosta) spp., female genitalia; 5. D. (A.) batanga Gyulai & Saldaitis, sp. n., HT, China, Sichuan, GYP 5350 (coll. PGYM); 6. D. (A.) batanga Gyulai & Saldaitis, sp. n., PT, China, Sichuan GYP 5367 (coll. AFM); 7. D. (A.) stentzi (Led.), Russia, Far East, GYP 5432 (coll. PGYM); 8. D. BIBLIOGRAFÍA (A.) stentzi (Led.), China, Sichuan GYP 3269 (coll. PGYM). 206 SHILAP Revta. lepid., 49 (194) junio 2021
https://openalex.org/W2979287996	https://biologicalproceduresonline.biomedcentral.com/track/pdf/10.1186/s12575-019-0107-z	English	null	DNA Methylation Validation Methods: a Coherent Review with Practical Comparison	Biological procedures online	2,019	cc-by	9,973	© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. DNA Methylation Validation Methods: a Coherent Review with Practical Comparison Šárka Šestáková1,2, Cyril Šálek1,2 and Hana Remešová2* Abstract Here, we present a practical overview of four commonly used validation methods for DNA methylation assessment: methylation specific restriction endonucleases (MSRE) analysis, pyrosequencing, methylation specific high-resolution DNA melting (MS-HRM) and quantitative methylation specific polymerase chain reaction (qMSP). Using these methods, we measured DNA methylation levels of three loci in human genome among which one was highly methylated, one intermediately methylated and one unmethylated. We compared the methods in terms of primer design demands, methods’ feasibility, accuracy, time and money consumption, and usability for clinical diagnostics. Pyrosequencing and MS-HRM proved to be the most convenient methods. Using pyrosequencing, it is possible to analyze every CpG in a chosen region. The price of the instrument may represent the main limitation of this methodology. MS- HRM is a simple PCR-based method. The measurement was quick, cheap and very accurate. MSRE analysis is based on a methylation specific digestion of DNA. It does not require a bisulfite conversion of DNA as the other methods. MSRE analysis was very easy to perform, however, it was not suitable for intermediately methylated regions and it was also quite expensive. qMSP is a qPCR-based method that uses primers designed specifically for methylated and unmethylated alleles of a chosen region. This was the least accurate method and also the primer design and optimization of PCR conditions were highly demanding. Keywords: DNA methylation, Validation methods, MSRE, Pyrosequencing, MS-HRM, qMSP Nowadays, there is a rapid expansion of high- throughput methods for DNA methylation assessment with single-base resolution. Array techniques can examine as much as 850,000 CpGs at once [5] and all CpG sites, over 28 million in human genome [6], can be analyzed using whole-genome bisulfite sequencing. These methods provide not only an overview of the methylation status of a certain genome, but also a methylation level of each stud- ied CpG. Despite the advantages of these genome-wide ap- proaches, it is still essential to have a proper technique for validation of DNA methylation results for chosen loci. The ideal validation method should be sensitive, quick, cost ef- fective and suitable for screening of large sets of clinical samples to acquire statistically significant data. (2019) 21:19 (2019) 21:19 Šestáková et al. Biological Procedures Online https://doi.org/10.1186/s12575-019-0107-z Background DNA methylation plays a fundamental role in many crucial biological processes such as embryonic development, gene imprinting, and gene expression regulation. In mammals, the DNA methylation occurs almost exclusively in CpG di- nucleotides where a methyl group is attached to the fifth carbon of cytosine base, creating 5-methylcytosine. The biological effect of DNA methylation depends not only on its presence or absence but mainly on its exact location in the genome [1]. Aberrant DNA methylation has been proved as an inducing mechanism in many cancers and is connected to other complex disorders (e.g. diabetes and cardiovascular diseases, neurodegenerative and psychiatric disorders) [2]. Therefore, DNA methylation profiles are ex- amined as biomarkers for diagnosis, prognosis, treatment response and disease monitoring [3, 4]. In this review, we assessed the methylation status of cer- tain CpGs using four most common methods for DNA methylation validation. These methods were: quantitative PCR with prior digestion by methylation specific restric- tion endonucleases (MSRE), pyrosequencing, methylation specific high-resolution DNA melting (MS-HRM) analysis, and quantitative methylation specific polymerase chain * Correspondence: hana.remesova@uhkt.cz 2Institute of Hematology and Blood Transfusion, Prague, Czech Republic Full list of author information is available at the end of the article * Correspondence: hana.remesova@uhkt.cz 2Institute of Hematology and Blood Transfusion, Prague, Czech Republic Full list of author information is available at the end of the article Šestáková et al. Biological Procedures Online (2019) 21:19 Page 2 of 11 Šestáková et al. Biological Procedures Online that after BS conversion the DNA strands are no longer complementary. This must be kept in mind when choos- ing a method for quantification of BS converted DNA. The reliability of methylation analysis is dependent on a complete BS conversion. Unconverted cytosines, if present, would be mistaken for methylated loci and the analysis will produce biased results. Formerly, the conversion method required a high DNA input and exposure to high bisulfite salt concentration under high temperatures and low pH. These harsh conditions resulted in significant DNA frag- mentation and loss [15]. Nowadays, there is a wide variety of commercial kits available that are able to convert as low as 100 pg of DNA in less than 2 h. These kits, nice compari- son is available here [16], use convenient column system and guarantee more than 99% conversion efficiency. reaction (qMSP). For proper evaluation of each method, we selected 3 distinct CpG sites within the human genome that were > 99% methylated (methylated “M” locus), around 50% methylated (intermediately methylated “IM” locus), and < 1% methylated (unmethylated “U” locus). that after BS conversion the DNA strands are no longer complementary. This must be kept in mind when choos- ing a method for quantification of BS converted DNA. The reliability of methylation analysis is dependent on a complete BS conversion. Unconverted cytosines, if present, would be mistaken for methylated loci and the analysis will produce biased results. Formerly, the conversion method required a high DNA input and exposure to high bisulfite salt concentration under high temperatures and low pH. These harsh conditions resulted in significant DNA frag- mentation and loss [15]. Nowadays, there is a wide variety of commercial kits available that are able to convert as low as 100 pg of DNA in less than 2 h. These kits, nice compari- son is available here [16], use convenient column system and guarantee more than 99% conversion efficiency. Overview of Evaluated Validation Methods Methods Based on DNA Digestion by MSRE g y Restriction methods for quantification of DNA methyla- tion are simple, rapid and do not require bisulfite conver- sion of DNA. Selective digestion of DNA by methylation specific restriction enzymes (HpaII, AatII, ClaI, etc.) was historically the first method used for assessing DNA methylation levels [7]. High specificity is characteristic for this method, however, only the specific restriction sites can be analyzed which is an important limitation. y The analysis is based on a selective DNA cleavage by MSRE which will not cut its restriction site when a methylated cytosine is present. The most frequently used enzyme is HpaII with recognition sequence CCGG. It is also possible to use a pair of isoschisomeric enzymes, where one is methylation sensitive and the other is not. Most common pair is HpaII/MspI where MspI also cleaves CCGG sequence but regardless of its methylation status. In older protocols, resulting fragments were ana- lyzed on a gel or by a southern blot and the location of methylated sites was estimated from the fragments’ sizes [8, 9]. Newer approaches employ quantitative PCR (qPCR) [10]. In order to determine methylation of a spe- cific region, DNA is digested by MSRE and subsequently analyzed with qPCR using primers surrounding the se- quence of interest. Methylation percentage is counted from threshold cycles (Ct) measured for digested and undigested control DNA. For this approach, it is possible to buy easy-to-use commercial kits with mixes of MSREs to target more sites, e.g. OneStep qMethyl kit from Zymo Research [11]. Pyrosequencing P i Pyrosequencing is a sequencing method used for quanti- tative methylation analysis of bisulfite converted DNA. For its relative simplicity, speed and comparable results, pyrosequencing can be preferred to cloning [17], a method used as a gold standard for identification of al- lele specific methylation patterns [18]. Another advan- tage of pyrosequencing is that it is suitable for both CpG poor and CpG rich regions. Main drawback of this method is that only shorter regions (maximum 350 bp) can be analyzed. However, this disadvantage can be overcome by using more sequencing primers on one amplicon or by a serial pyrosequencing [19, 20]. Pyrosequencing process can be divided into three steps: (i) PCR amplification and tagging using a biotinylated pri- mer, (ii) isolation of the PCR product with streptavidin beads and hybridization with a sequencing primer, and (iii) sequencing. During the sequencing step, nucleotides are added in a predefined order depending on the sequence of interest. The technology is based on a release of pyro- phosphate (PPi) during nucleotide incorporation when complementary to the template DNA strand (the purified PCR product). An ATP sulfurylase then uses PPi and ad- enosine phosphosulfate to produce ATP. ATP is utilized by luciferase which converts luciferin to oxyluciferin. The in- tensity of produced light is detected and translated as a peak on a pyrogram [21]. Methylation percentage is then calculated from the ratio of heights of a cytosine peak (methylated signal) and the sum of cytosine and thymine peaks (methylated and unmethylated signal) for each cyto- sine in a CpG dinucleotide. Primers can be easily designed with free online software such as Primer3Plus [12] (http://www.bioinformatics.nl/cgi- bin/primer3plus/primer3plus.cgi) or Primer-BLAST [13] (https://www.ncbi.nlm.nih.gov/tools/primer-blast/). How- ever, it is required that at least two restriction sites are in- side the amplicon to reliably measure the DNA methylation. Thus, it is not possible to investigate the methylation level of only one particular CpG site. Methods Requiring Bisulfite Converted DNA There should be at least four non-CpG cyto- sines in each primer to assure that only a properly BS con- verted DNA will be amplified. Presence of a CpG and therefore a use of a degenerated primer in not recom- mended because it may lead to a preferential amplification of a specific subset of molecules [21]. However, in some of our previous experiments we used degenerated primers without any difficulties. One of the PCR primers must be labeled on its 5’end by biotin and this primer should be purified by HPLC or an equivalent procedure to assure zero contamination by free biotin molecules. The orienta- tion of a sequencing assay depends on which primer is tagged. It is also essential to incorporate all biotinylated primers into amplicons during the PCR step. Otherwise, these primers might compete with the amplicons during the streptavidin binding. It is recommended to use 0.1 μM biotinylated primer and 0.2 μM unlabeled primer concen- trations and 45–50 PCR cycles. It is also possible to use a universal biotinylated primer and a tailed reverse primer in 5:(0.01–1) ratio [20, 25]. The use of a universal biotinylated primer significantly reduces the costs when having more pyrosequencing assays for various regions. On the other hand, it sometimes requires deeper PCR optimization to gain a sufficient amount of the PCR product. The amount and size of the amplified PCR product as well as a negative PCR control should be always checked by an agarose gel electrophoresis to prevent further complications. Sequencing primer should be 15–20 bp long with a melting temperat re bet een 45 and 55 °C The most Having a strong amplicon with no side product, and therefore a high-quality primer design, is crucial for this assay. One way is to order the primers from commercial companies. For example, Qiagen offers a full assay design for desired regions and it is also possible to buy a prede- signed primer set. The other option is to use a free soft- ware for bisulfite primer design such as MethPrimer [22] (http://www.ucsf.edu/urogene/methprimer/index1.html), A set of assay validation reactions, listed in appendix B of PyroMark Q24 User Manual 2016, should be always performed before using the assay to analyze samples. Methods Requiring Bisulfite Converted DNA Bisulfite (BS) conversion of DNA is a crucial step in most DNA methylation analyses. Already in 1970, it was discovered that sodium bisulfite mediates the deamin- ation of cytosine into uracil while the methylated cyto- sine is left intact [14]. After PCR amplification, the converted residues are read as thymines and methylated cytosines will remain cytosines. It is important to note As mentioned above, this method is suitable for regions 80–200 bp long. One reason is that longer amplicons could form secondary structures and loops that would im- pede the sequencing reaction. The second issue arises dur- ing the sequencing procedure where nucleotides are added in each sequencing cycle. The volume in reaction wells increases which causes dilution of all reagents and Page 3 of 11 Šestáková et al. Biological Procedures Online (2019) 21:19 Šestáková et al. Biological Procedures Online on the 3’end which should be unique in the amplicon. Also, it is not recommended to use the non-biotinylated PCR primer as a sequencing primer. The sequencing pri- mer should differ from the PCR primer in at least one additional nucleotide on the 3’end [21]. Nevertheless, we tried to use the non-biotinylated primer as a sequencing primer in some of our previous experiments and the py- rosequencing was successful. thus a decrease of the signal. At the same time, the back- ground signal rises during the sequencing due to an in- complete degradation of previously added nucleotides [21]. Because of that, a signal measured after 90–100 cy- cles has low quality and the results are not credible [20]. cles has low quality and the results are not credible [20]. Having a strong amplicon with no side product, and therefore a high-quality primer design, is crucial for this assay. One way is to order the primers from commercial companies. For example, Qiagen offers a full assay design for desired regions and it is also possible to buy a prede- signed primer set. The other option is to use a free soft- ware for bisulfite primer design such as MethPrimer [22] (http://www.ucsf.edu/urogene/methprimer/index1.html), Bisearch [23] (http://bisearch.enzim.hu/) or MethylPrimer Express by Applied biosystems (http://www.appliedbiosys tems.com/methylprimerexpress). Moreover, it is important to check for potential primer dimers formation or self-complementarity of the primers, e.g. with Multiple Pri- mer Analyzer (https://www.thermofisher.com/cz/en/home/ brands/thermo-scientific/molecular-biology/molecular-biol- ogy-learning-center/molecular-biology-resource-library/ thermo-scientific-web-tools/multiple-primer-analyzer.html). Primers should be 15–30 bp long (20 bp is optimal) with a melting temperature between 50 and 69 °C (optimally 60 °C) [24]. Methods Requiring Bisulfite Converted DNA These controls are (i) PCR reaction without template DNA, (ii) PCR product without sequencing primer, (iii) sequencing primer without PCR product, (iv) biotinyl- ated primer without PCR product, and (v) sequencing primer and biotinylated primer together without PCR product. Moreover, in each assay, controls of BS conver- sion should be included in the dispensation order [21]. The BS conversion ratio can be evaluated when a dis- pensation of cytosine nucleotides is incorporated before or after thymines which are supposedly converted cyto- sines in the sequence. In case of an unsuccessful BS con- version, peaks would appear in the pyrogram for these additional dispensations. p f g p Bisearch [23] (http://bisearch.enzim.hu/) or MethylPrimer Express by Applied biosystems (http://www.appliedbiosys tems.com/methylprimerexpress). Moreover, it is important to check for potential primer dimers formation or self-complementarity of the primers, e.g. with Multiple Pri- mer Analyzer (https://www.thermofisher.com/cz/en/home/ brands/thermo-scientific/molecular-biology/molecular-biol- ogy-learning-center/molecular-biology-resource-library/ thermo-scientific-web-tools/multiple-primer-analyzer.html). Methylation Specific HRM Analysis MS-HRM is a method based on different melting tem- peratures (Tm) of methylated and unmethylated DNA. Tm is defined as a temperature at which the two DNA strands dissociate and is characterized by a sudden drop of fluorescence signal due to a release of an intercalating dye, e.g. SYBR Green, EvaGreen or SYTO9. Tm depends on the DNA base composition because CG base pairs are connected by three hydrogen bonds and AT pairs only by two. This also enables to differentiate between methylated and unmethylated DNA after a BS conver- sion where the unmethylated cytosines are converted to uraciles and after PCR changed to thymines. MS-HRM comprises of PCR for amplification of a chosen region followed by HRM analysis with ramping by only 0.1 °C [26]. It is recommended to use quantita- tive PCR for the amplification as an additional quality control [27]. For the DNA methylation assessment, DNA standards are analyzed together with the samples. Standards are prepared by diluting fully methylated BS converted DNA by fully unmethylated BS converted DNA and are usually 100, 75, 50, 25, 10 and 0% methyl- ated. By comparing the HRM curves of standards and samples it is possible to determine an approximate methylation level [26]. Several more quantitative ap- proaches for establishing the DNA methylation were de- veloped. It is possible to construct a linear curve by plotting the temperature at which 50% of DNA is disso- ciated (T50) against the methylation percentage of the Sequencing primer should be 15–20 bp long with a melting temperature between 45 and 55 °C. The most relevant part of the primer are the last four or five bases Page 4 of 11 Page 4 of 11 Šestáková et al. Biological Procedures Online (2019) 21:19 Šestáková et al. Biological Procedures Online this method is more suitable for CpG rich regions, like CpG islands. There should be again at least five thymines, BS converted unmethylated cytosines, included in the se- quence to assure that only a properly BS converted DNA will be amplified [16]. Primers length should be at least 23 bp with melting temperature between 55 and 65 °C. It is recommended that Met and Unm primers have similar melting temperatures. To achieve that, it is often needed to prolong the Unm primer on the 5’end because the BS conversion decreases the Tm of unmethylated DNA [35]. Methylation Specific HRM Analysis The above-mentioned software for finding bisulfite primers for pyrosequencing or MS-HRM can be also used for designing MSP primer sets. standards [28]. Another method estimates the methyla- tion level by using two sets of primers, methylated and unmethylated, for amplification. A differential melting profile is then calculated by normalizing the methylated HRM profile against the unmethylated. The differential fluorescence peak heights are then plotted against the dilution factor which generates a linear calibration curve [29]. Another approach uses peak heights and area under the curve (AUC) of normalized and temperature shifted first derivatives of HRM curves. There is a linear dependency between these values and the methylation percentage [27]. p g [ ] For MS-HRM analysis, the only requirement are primers surrounding the region with CpGs of interest. It is again crucial to obtain a pure PCR product. Primers should be between 15 and 30 bp long with similar melting tempera- tures around 65 °C. This will allow to run the PCR at ap- proximately 60 °C which is important for the method’s specificity [30]. Generally, for this type of primers that sur- round the region of interest, it is not advised to have a CpG inside the primer. However, Wojdacz et al. have shown that inclusion of a CpG in the primer sequence can compensate the PCR bias towards unmethylated alleles and thus signifi- cantly increase the method’s sensitivity [30, 31]. They also claim that MS-HRM is taken as a method to assess methy- lation levels and therefore a slight bias towards the methyl- ated sequences further increases the method’s sensitivity [32]. According to their recommendations, the primers should contain one or two CpG dinucleotides as close to their 5’end as possible [33]. It is necessary to have several thymines, corresponding to unmethylated cytosines, in- cluded in the primer sequence to amplify only properly converted DNA. It is also advisable to check for primer di- mers and loops formation. The amplicon should be kept small, around 100 bp, to reduce the complexity of its melt- ing profile [32]. Nevertheless, it should be considered that a shorter PCR product gives higher sensitivity but limited resolution between methylation levels because of only small difference between methylated and unmethylated DNA. Longer amplicons have more distinguishable methylation profiles [26]. During the PCR, the number of amplification cycles should not exceed 35 because after that a false methyla- tion signal could appear. Methylation Specific PCR In this methodology, DNA methylation is examined by two sets of primers where one is specific for a methyl- ated state (Met) and the other pair for an unmethylated state (Unm) of a certain genomic locus. A set of two PCR reactions is performed and the products are ana- lyzed via a gel electrophoresis [34]. Methylation Specific HRM Analysis It is also crucial to use such an- nealing temperature (Tann) that the primers are specific for the DNA methylation status they were designed for [35]. Thus, it is essential to perform proper control reac- tions in each new MSP experiment such as PCR with both methylated and unmethylated standards, no template con- trol and PCR with non-converted DNA [16]. For a long time, this method was only qualitative. As a result of the electrophoresis, it was possible to recognize that only methylated or unmethylated locus or both loci were amplified. Previous studies did not find any clear correlation between the size of the band on the gel and the amount of DNA examined. Nevertheless, this does not compromise the great sensitivity of this method [16]. Newer protocols employ quantitative PCR and make this method semiquantitative or quantitative. For example, for establishing unmethylated DNA, qPCR is performed with Unm primers together with bisulfite specific primers (BSP) that amplify chosen locus independently of its methylation status. The ratio of unmethylated alleles to total number of amplified molecules is then calculated by either classical ΔΔCt approach with or without correction for primers efficiency [36] or by a demethylation index, as proposed in another study [37]. Apart from using simple qPCR with an intercalating dye, a quantitative method called MethylLight was developed. It uses a pair of methy- lation specific TaqMan probes where each probe, designed for either methylated or unmethylated DNA, is tagged with different fluorophore [38]. When using quantitative MSP approaches, it is advisable to perform a melt analysis after the PCR to check for any side products [35]. Results and Discussion MSRE Analysis With the MSRE approach, we were able to accurately measure methylation levels of the M and U loci. However, for the IM locus, we acquired lower methylation levels than expected. Therefore, we performed two additional MSRE experiments with shortened digestion time to see Despite the relative simplicity of this method, finding convenient methylated and unmethylated primers is sometimes challenging. The primers are designed to span the analyzed region and should contain from one to three CpGs ideally located at the 3’end of the primer. Therefore, Šestáková et al. Biological Procedures Online (2019) 21:19 Page 5 of 11 Šestáková et al. Biological Procedures Online Fig. 1 The influence of digestion time on measured methylation levels in MSRE analysis. Error bars represent the standard deviation (n = 4). M - methylated locus, IM – intermediately methylated locus, U – unmethylated locus agarose gel electrophoresis, the peak height of a single base was always around 50–200 units. In the resulting pyrograms, it was obvious that the sig- nal starts dropping significantly after 45th dispensation cycle which roughly corresponds to a 100 bp region. This is in accordance with the recommendations for this method to keep the studied region short [20]. In our experiments, we measured four CpGs in the M locus and all were highly methylated (95.4 ± 3.1%). In the U locus, we measured three CpGs and all were unmethylated (7.4 ± 3.1%). There were only two CpGs in the IM locus. The IM CpG chosen from the Infinium MethylationEPIC BeadChip data was indeed intermedi- ately methylated 58.5 ± 7.3%. However, the next CpG in- cluded in the sequenced amplicon was rather unmethylated 18.4 ± 3.8%. The final average methylation of the studied region was thus around 37%. This demon- strates the main advantage of the pyrosequencing method which is the base resolution. The other quite beneficial aspect of pyrosequencing is the bisulfite con- version control which allows us to see whether the BS conversion was done properly [21]. We always included at least three of these BS control dispensations in our pyrosequencing assays. Fig. 1 The influence of digestion time on measured methylation levels in MSRE analysis. Error bars represent the standard deviation (n = 4). M - methylated locus, IM – intermediately methylated locus, U – unmethylated locus whether it will increase the methods accuracy for the IM locus. Results and Discussion MSRE Analysis The recommended digestion time was 2 h, so we additionally tried 1.5-h and 1-h digestions. There was no statistically significant difference in DNA methylation levels measured after various digestion times. We achieved only a slight improvement for the IM locus, where the cal- culated methylation percentage rose from 12 to 17% when the digestion time was shortened to 1 h. The methylation levels for M and U loci remained the same. The results for all three loci are shown in Fig. 1. Based on these results, we propose that a shorter digestion time may be used to make the MSRE method faster while retaining the original results. MS-HRM Wojdacz et al. have shown that inclusion of a CpG to the primer sequence can compensate the PCR bias of unmethylated alleles by favoring amplification of methyl- ated alleles [32]. Thus, we designed two sets of HRM primers for the M and U loci. One primer set did not in- clude any CpGs in its sequence. The other set was de- signed according to Wojdacz et al. [30] and each primer had one or two CpGs on its 5’end. It was not possible to design Wojdacz HRM primers for the IM region because of its CpG shortage. Sequences of the primers are listed in Table 1. Pyrosequencing I h In the pyrosequencing procedure, the most important step for a successful analysis is gaining a strong amplicon. However, even when we detected a strong band on our agarose gel after the PCR, we did not achieve the desirable signal during pyrosequencing. Thus, we tried to enhance the binding of the amplified, biotin-labeled PCR product by adding more streptavidin beads into the process. We compared the results after adding 1, 2 and 3 μl of strepta- vidin beads per sample. The 2 μl proved to be ideal for gaining the strongest signal on the pyrogram. Additionally, we prolonged the agitation step to 20 min to increase the number of bound molecules. Also, we noticed that after the agitation, it is essential to proceed immediately to the next step to ensure that the beads are resuspended in the tube and will be taken up efficiently by the probes in the subsequent procedure. According to the manufacturer, for accurate results the peak height of a single base should be at least 40 units in the pyrogram. From our experience, ex- periments where single based peaks are at least 25 units high give reliable results. Nevertheless, when a strong amplicon was acquired, judging by the results of the From acquired MS-HRM data, we constructed calibra- tion curves as proposed by Tse et al. [27] for every pri- mer set. The correlation coefficients (R2) for peak height-based calibration curves together with calculated methylation levels for each locus are summarized in Table 2. The AUC-based calibration curves had slightly lower correlation coefficients. This was probably caused by less exact AUC calculations performed in Excel which we used to keep the data analysis as simple as possible. Still, the AUC-based DNA methylation assessment gave similar results as the peak height approach (see Additional file 1). Interestingly, the results in Table 2 show that the Wojdacz’s improvement of primers’ sequence was quite beneficial for the U region. However, it caused a deviation in measurements of the M region result- ing in worse R2 of the calibration curve and lower Šestáková et al. Pyrosequencing I h Biological Procedures Online (2019) 21:19 Page 6 of 11 Table 1 Primer sequences and characteristics Primer type Forward/ sequencing primer Reverse primer Tann [°C] Product length [bp] M pyrosequencing GGTAGGAGGATGGTTTGAATT/ GGTGGAAATGAAGTAGGTGTGTTTG GTGCCGAGGCTCAGGCAACACTACTCTTACCAAAACAACC 60 373/ 227 IM pyrosequencing GTTAAGGGGGTGTATTTTAGAGA/ GGTAGAGAGAAGTTTTTTTTGTAGG GTGCCGAGGCTCAGGCCTTAACTACTTTCCCAAACTACCT 58 399/ 339 U pyrosequencing GGGGGGGTGTTAGTATTTG/ TTAGTATTTGYGTTGTGGAGTG GTGCCGAGGCTCAGGCCCAAACTAACCTAATAAAACC 58 300/ 290 Universal biotinylated primer 5’biotin-TCTGTGCCGAGGCTCAGGC M MSRE TTTTCTGTGACCTCCTTTGG CAGTGTGACTGCTGGTGAAG 60 243 N MSRE GCAATAGGCGTTAATGTCGT AGGAGTGGCAAAAGAGGACT 60 199 U MSRE CGCTTAGCAATCATCGACTT GAAACAGGCCGCATCCTC 60 265 M MSP Met GTATATTCGGAATTATTTCGTTTTC AATTAACAACCGACAACCG 56 72 M MSP Unm GATGTATATTTGGAATTATTTTGTTTTT AATTAACAACCAACAACCA 56 75 IM MSP Met CGGTTTTTATAGTTTTGAATTAGATC TTATTTATTATCACATCAACTACTTCCG 58 166 IM MSP Unm ATTGGTTTTTATAGTTTTGAATTAGATT TTATTTATTATCACATCAACTACTTCCA 58 168 U MSP Met CGTTGTGGAGTGAAGTGAATC ACCGAACGAACAATAAACGAA 54 210 U MSP Unm TGTGTTGTGGAGTGAAGTGAATT ACCAAACAAACAATAAACAAAAAA 54 212 M HRM TTGGGTGGAAATGAAGTAGGTGTG CCAAACCATTAACCATAACAATA 54–58* 94 IM HRM TTTGGGGAAAAAATATATGGAGTT CTACTAATAAAACCCTTTACTCCCA 54–58* 90 U HRM TTAGTATTTGYGTTGTGGAGTG CCRACACTTACTCTTATTAACRATC 54–58* 93 M HRM Wojdacz CGGGGGGGTGTTAGTATTTG CCCGACACTTACTCTTATTAACRATC 55 110 U HRM Wojdacz TCGTGTTTTTTTTTGGGTGGAAATG GCGACCAAACCATTAACCATAACA 55 104 For MS-HRM experiments Tann was 55 °C, in qMSP experiments Tann of MSP primers was used M methylated locus, IM intermediately methylated locus, U unmethylated locus, MSP methylation specific PCR, Met primers for methylated DNA sequence, Unm primers for unmethylated DNA sequence, HRM high resolution melting analysis, Tann annealing temperature For MS-HRM experiments Tann was 55 °C, in qMSP experiments Tann of MSP primers was used M methylated locus, IM intermediately methylated locus, U unmethylated locus, MSP methylation specific PCR, Met primers for methylated DNA sequence, Unm primers for unmethylated DNA sequence, HRM high resolution melting analysis, Tann annealing temperature For MS-HRM experiments Tann was 55 °C, in qMSP experiments Tann of MSP primers was used M methylated locus, IM intermediately methylated locus, U unmethylated locus, MSP methylation specific PCR, Met primers for methylated DNA sequence, Unm primers for unmethylated DNA sequence, HRM high resolution melting analysis, Tann annealing temperature qMSP calculated methylation levels. It is thus not so straightforwardly beneficial to introduce the CpGs into primers’ sequence. Gaining the optimal results apparently require additional thorough Ta optimization to achieve equal amplification of meth- ylated and unmethylated alleles with Wojdacz’s primers [26]. Nevertheless, with all primer sets, the methylation levels of all three loci were measured accurately enough. We designed both Met and Unm primer sets for each locus to perform the qMSP experiments. Reassuringly, in all investigated samples, the M locus was amplified only by Met primers, the U locus only by Unm primers and the IM locus was amplified by both Met and Unmet primer sets. Regarding the DNA methylated and unmethylated standards, all three loci were always amplified only by the corresponding Met or Unm primer set. The HRM primers were used as BSP, amplifying the region independently of its methylation status. Summary of Ct values and mea- sured efficiencies for all primer sets is shown in Table 3. We were not able to measure the efficiency for IM Unm primer set properly. The deviation between duplicates was higher probably because the efficiency of the primer set was low and the amplification of the first dilution began always after 34th cycle. The measurement was thus less reliable despite the fact that the resulting calibration curve had R2 > 0.99. Therefore, we tried to count the efficiency of MSP primers based on Ct values measured for stan- dards and an assumption that BSP primers have 100% effi- ciency. We used the following equation Ec ¼ 100∙CBSP t CMSP t . Table 2 Correlation coefficients for peak height-based MS-HRM calibration curves and counted methylation levels Locus name Primer set R2 Methylation [%] (n = 10) ± SD M M HRM 0.952 93.61 5.28 M M HRM Wojdacz 0.798 85.49 5.13 IM IM HRM 0.973 29.20 4.71 U U HRM 0.868 2.69 1.04 U U HRM Wojdacz 0.938 0.57 0.81 M methylated locus, IM intermediately methylated locus, U unmethylated locus, R2 square of the correlation coefficient, SD standard deviation. Table 2 Correlation coefficients for peak height-based MS-HRM calibration curves and counted methylation levels Šestáková et al. Overall Methods Comparison This counted efficiency corresponded well with the mea- sured efficiency (Table 3) and we used it in the subsequent analysis for the IM Unm primer set. This counted efficiency corresponded well with the mea- sured efficiency (Table 3) and we used it in the subsequent analysis for the IM Unm primer set. The final results of DNA methylation levels measured by all four methods are shown in Fig. 2. All methods were comparable and correlated with each other with R2 > 0.92 and p-value < 1.2·10−17, except the qMSP method results of which were spoiled by extreme standard deviations. We also provide a review of all costs and mea- surements for each method in Table 5 and a final evalu- ation of few other parameters in Table 6. We analyzed our data using all three approaches reviewed by Housseiny et al. [36]. The relative expres- sion ratio method, developed originally by Pfaffl [39], gave very variable results with extremely high standard deviation and thus was not reliable (see Additional file 2). The other approaches, demethylation index and ΔΔCt, gave quite similar results, reviewed in Table 4. The M locus was highly amplified by the MSP Met primers. The seemingly double amplification of MSP primers compared to BSP primers indicated by ΔΔCt ≅2 is prob- ably caused by the method’s inaccuracy because when we repeated the experiment with five samples, the ΔΔCt results were 1.5 ± 0.3. The ΔΔCt results for U locus were close to 1, meaning that the number of molecules ampli- fied by MSP Unm and BSP primer set was comparable. Regarding the IM locus, the MSP Unm primers ampli- fied around half of the molecules in comparison with the BSP primers, which corresponds with the expected intermediate methylation level of this locus. However, the results of MSP Met primers were spoiled by the dis- proportionately high Ct measured for the methylated DNA standard resulting in a very low ratio of molecules amplified by MSP Met primers in the samples. This could be caused by a higher affinity of Met primers to the methylated DNA standard, compared to samples’ DNA that was rather unmethylated in the IM region. Indisputably, pyrosequencing has most advantages in terms of the DNA methylation assessment of a specific locus. Primer design and interpretation of the results is straightforward with available software. qMSP Biological Procedures Online (2019) 21:19 Page 7 of 11 Table 3 qMSP Ct values and primers’ efficiencies Locus and used primer set Average Ct of samples (n = 10) Ct of standards Efficiency MSP ± SD BSP ± SD MSP BSP MSP measured MSP counted BSP measured M Met 23.18 0.52 22.88 0.27 24.01 22.70 96.47 94.57 83.00 IM Met 34.54 1.23 24.33 0.35 29.23 24.51 81.13 83.85 94.78 IM Unm 32.21 0.50 24.33 0.35 31.99 25.15 125.80 78.63 94.78 U Unm 37.36 0.80 23.04 0.20 37.95 23.40 65.93 61.66 90.52 M methylated locus, IM intermediately methylated locus, U unmethylated locus, Met primers for methylated DNA sequence, Unm primers for unmethylated DNA sequence, MSP methylation specific primers, BSP bisulfite specific primers (methylation independent), SD standard deviation. Table 3 qMSP Ct values and primers’ efficiencies M methylated locus, IM intermediately methylated locus, U unmethylated locus, Met primers for methylated DNA sequence, Unm primers for unmethylated DNA sequence, MSP methylation specific primers, BSP bisulfite specific primers (methylation independent), SD standard deviation. Overall Methods Comparison Only the PCR step may require some optimization for gaining a sufficient amplicon but this is not always necessary. A disadvantage of this method may be the relatively high price of the in- strument. Also, the method is more time consuming be- cause it comprises three steps: PCR, gel electrophoresis and sequencing itself. This also corresponds with the higher price per one measurement. When the pyrosequencing instrumentation is unavail- able, we recommend using MS-HRM. The primer de- sign is feasible for most regions. In our experiment, we had a CpG poor region (IM) as well as CpG dense re- gions (M and U) and were able to design reliable sets of primers for both. As we discussed in the chapter about the results from MS-HRM, deep optimization of primers’ sequence and Tann is not needed for method resolution of 5–10%. This method is very simple as well as cost and time effective. The approximate results can be derived immediately from the melting curves. The exact quantification is not so straightforward when a specific software is not provided. Nevertheless, the cal- culations can be done using free software and Excel, as we have shown. Table 4 Summary of qMSP methylation results calculated using demethylation index and ΔΔCt approach Locus and used primer set Demethylation index ΔΔCt Average (n = 10) ± SD Average (n = 10) ± SD M Met 2.02 0.60 2.09 0.66 IM Met 0.05 0.03 0.03 0.02 IM Unm 0.51 0.10 0.50 0.12 U Unm 1.03 0.30 1.30 0.66 M methylated locus, IM intermediately methylated locus, U unmethylated locus, Met primers for methylated DNA sequence, Unm primers for unmethylated DNA sequence, SD standard deviation Table 4 Summary of qMSP methylation results calculated using demethylation index and ΔΔCt approach The MSRE analysis proved to be a quick and simple method. The main advantage of this approach is that it does not require the BS conversion of DNA. Thus, less DNA is needed to perform the analysis and it also makes the primer design easier. We were able to accurately measured the DNA methylation in M and U regions. Obviously though, the method is not suitable for inter- mediately methylated loci. Even by shortening the Šestáková et al. Biological Procedures Online (2019) 21:19 Page 8 of 11 Fig. 2 Summary of DNA methylation levels measured by the investigated methods. Overall Methods Comparison The average methylation of 10 samples is shown, the error bars represent the SD. Displayed MSRE data were measured after 2-h digestion. Displayed MS-HRM data were acquired using HRM M, HRM IM and HRM U Wojdacz primers. qMSP data shown were calculated using ΔΔCt approach and multiplied by 100 to gain the percentage. For the M locus in qMSP, all values were higher than 100% so we set the mean to 100% to make the figure more comprehensible, the SD was calculated from the original values multiplied by 100. For the IM and U loci in qMSP, we calculated the methylation percentage as 1-(Unm ΔΔCt), the SD was also calculated from the original ΔΔCt values multiplied by 100. M – methylated locus, IM – intermediately methylated locus, U – unmethylated locus Fig. 2 Summary of DNA methylation levels measured by the investigated methods. The average methylation of 10 samples is shown, the error bars represent the SD. Displayed MSRE data were measured after 2-h digestion. Displayed MS-HRM data were acquired using HRM M, HRM IM and HRM U Wojdacz primers. qMSP data shown were calculated using ΔΔCt approach and multiplied by 100 to gain the percentage. For the M locus in qMSP, all values were higher than 100% so we set the mean to 100% to make the figure more comprehensible, the SD was calculated from the original values multiplied by 100. For the IM and U loci in qMSP, we calculated the methylation percentage as 1-(Unm ΔΔCt), the SD was also calculated from the original ΔΔCt values multiplied by 100. M – methylated locus, IM – intermediately methylated locus, U – unmethylated locus digestion time to half of that recommended, the mea- sured DNA methylation of IM locus remained signifi- cantly lower than expected. Also, the method is very costly when compared to the other three. which both Met and Unm primer sets were specific only for the methylated or unmethylated allele respectively but still functional so that it amplified the corresponding DNA standard. When we finally achieved this, the primers had very low efficiency, except for the M Met primer set. Moreover, the exact quantification of measured data was difficult and the results of the M and U loci had extremely high standard deviations within the samples. This method, The last method evaluated was qMSP and this caused the most difficulties. Overall Methods Comparison The primer design was quite challen- ging and nearly impossible for the IM locus because of its lack of CpGs. Overall Methods Comparison Biological Procedures Online (2019) 21:19 Šestáková et al. Biological Procedures Online (2019) 21:19 despite of its simplicity, is also quite expensive because it requires amplification of a chosen region by at least one MSP primer set and BSP primers. despite of its simplicity, is also quite expensive because it requires amplification of a chosen region by at least one MSP primer set and BSP primers. Samples and DNA Standards This study was approved by the Institutional Ethics Committee and all blood donors provided their full con- sent. Mononuclear cells of ten healthy blood donors were harvested from buffy coats by Ficoll gradient cen- trifugation (Histopaque, Sigma-Aldrich, St.Louis, MO, USA). DNA was extracted using MagCore system (RBCBioscience, New Taipei City, Tchaj-wan). Human Methylated & Non-methylated DNA Set (Zymo Re- search, Irvine, CA, USA) was used as methylated and unmethylated standards. Conclusion Even in the era of arrays and next-generation sequencing, it is essential to have a method for validation of acquired DNA methylation data. A quick, cost-effective, and reli- able method that would enable to confirm or reject a po- tential clinical significance of certain methylation changes and could be used in common laboratory practice is still needed. We tested four standard methods that are used for DNA methylation validation: MSRE analysis, pyrosequencing, MS-HRM and qMSP. In terms of overall feasibility, ob- tained DNA methylation information and consistency across various methylation levels, we consider pyrosequenc- ing and MS-HRM approaches as the most suitable. Pyrose- quencing enables base resolution and thus acquisition of a methylation level for each CpG in the region, an indisput- able benefit. MS-HRM can be also designed to investigate a single CpG locus when needed. Otherwise, it provides an overall DNA methylation status of all CpGs inside the stud- ied region. Apparently, MSRE and qMSP are not very applicable for the detection of intermediate levels of DNA methylation. Nonetheless, MSRE does not require BS con- version of DNA and as we showed here, the digestion time can be shortened to one half. This makes the MSRE ana- lysis the simplest and fastest out of the four methods com- pared. The qMSP approach proved to be quite imprecise and demanding so it may be more convenient to keep this method only as a qualitative tool. Primer Design Primer Design For MSRE analysis, online software Primer3Plus (http:// www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus. cgi) was used. For methods that require BS converted DNA, we used Methyl Primer Express Software v1.0 (Thermo Fisher Scientific). For primers’ sequences and characteristics see Table 1. Positions of all primer pairs within the studied regions are shown in Fig. 3. MSRE Analysis OneStep qMethyl Kit (Zymo Research) was used for MSRE analysis. For each sample, DNA (20 ng) was proc- essed through the Test and Reference reactions according to the manufacturer’s protocol. In the PCR step, Tann was set to 60 °C and annealing time was shortened to 45 s. Rotor-Gene Q 2plex HRM Platform (Qiagen, Hilden, Germany) was used to perform the measurements. Materials and Methods Characterization of Analyzed CpGs Characterization of Analyzed CpGs The three analyzed CpGs with different levels of methyla- tion were selected based on healthy donors’ data from Infinium MethylationEPIC BeadChip (Illumina, San Diego, CA, USA) acquired in our previous work [40]. Characteristic of chosen loci is summarized in Table 7. These CpG dinucleotides were also chosen so that they are within CCGG sequence to enable their cutting by MSRE. Bisulfite Conversion DNA (500 ng) was treated with bisulfite using EZ DNA Methylation-Lightning Kit (Zymo Research). For MS- HRM and qMSP experiments, the concentration of BS converted DNA was measured by NanoDrop™One/OneC Microvolume UV-Vis Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and then adjusted to 10 ng·μl−1. Overall Methods Comparison Another issue was to find a suitable Tann at Table 5 Costs summary of each method Method Total cost of analysis [$] Number of samples measured Number of standards measured Total number of measurements Cost per measurement [$] MSRE analysis 576 10 for each locus, Test and Reference reaction, duplicates 2 for each locus, duplicates 144 4 pyrosequencing 162‡ 10 for each locus 2 for each locus* 36 4.5 MS-HRM 85 10 for each locus, duplicates 6 for each locus, duplicates 96 0.9 qMSP 196 10 for each primer set†, duplicates 2 for each primer set†, duplicates 216 0.9 Number of loci = 3 †Number of MSP/HRM primer sets for each locus = 3 ‡price of the pyrosequencing instrument ca 45,000 $ Table 6 Overall evaluation of tested methods Method Base resolution Consistency across methylation levels Analysis of acquired data Method optimization Time consumption Price MSRE analysis – * * * * Pyrosequencing + * * * * MS-HRM – * /(if needed) * qMSP – * * ** * - simple/low, - intermediate, * - demanding/high Table 5 Costs summary of each method Method Total cost of analysis [$] Number of samples measured Number of standards measured Total number of measurements Cost per measurement [$] MSRE analysis 576 10 for each locus, Test and Reference reaction, duplicates 2 for each locus, duplicates 144 4 pyrosequencing 162‡ 10 for each locus* 2 for each locus* 36 4.5 MS-HRM 85 10 for each locus, duplicates 6 for each locus, duplicates 96 0.9 qMSP 196 10 for each primer set†, duplicates 2 for each primer set†, duplicates 216 0.9 Number of loci = 3 †Number of MSP/HRM primer sets for each locus = 3 ‡ $ Table 5 Costs summary of each method †Number of MSP/HRM primer sets for each locus = 3 ‡price of the pyrosequencing instrument ca 45,000 $ Table 6 Overall evaluation of tested methods Method Base resolution Consistency across methylation levels Analysis of acquired data Method optimization Time consumption Price MSRE analysis – * * * * Pyrosequencing + * * * * MS-HRM – * /(if needed) * qMSP – * * ** * - simple/low, - intermediate, * - demanding/high Table 6 Overall evaluation of tested methods Page 9 of 11 Page 9 of 11 Šestáková et al. Pyrosequencing y q g BS converted DNA (10–20 ng) was first amplified using HotStar HiFidelity Polymerase Kit (Qiagen) with final 2.5 mM MgCl2 concentration. To increase primers’ spe- cificity and for easy gel loading, we added CoralLoad Concentrate from PyroMark PCR kit (Qiagen) to the final concentration of 1x. The final concentration of forward and universal biotinylated primer was 0.2 μM. The final concen- tration of reverse tailed primer was 0.04 μM. We used rec- ommended PCR reaction conditions for PyroMark PCR with 48 PCR cycles and Tann according to Table 1. Ampli- con quality (1 μl of PCR reaction) was checked using 2% agarose gel electrophoresis. Pyrosequencing was performed on PyroMark Q24 instrument (Qiagen). Pyrosequencing protocol (User Manual 01/2016) was optimized by adding Table 7 Specifications of selected CpG sites Locus name BeadChip probe ID Cytosine location (hg 19) Beta value for all samples measured with BeadChip Chromosome Position M cg24337108 10 11,797,422 > 0.99 IM cg25722983 1 36,840,028 from 0.45 to 0.55 U cg09655782 4 57,333,859 < 0.1 M methylated locus, IM intermediately methylated locus, U unmethylated locus, BeadChip Infinium MethylationEPIC BeadChip (Illumina), Beta value corresponds to methylation percentage Šestáková et al. Biological Procedures Online (2019) 21:19 Page 10 of 11 Fig. 3 Positions of primer pairs, CpGs and restriction sites within studied regions. CpGs are shown as red and yellow bars on a line representing the DNA sequence. The red CpG is the one originally chosen from Infinium MethylationEPIC BeadChip. The scissors indicate sites that are cut by MSREs. The lighter blue primers were used for initial pyrosequencing PCR and the darker blue primers represent the sequencing primers. The patterned light green HRM primers were designed with a CpG on its 5’end (M/U Wojdacz primers) Fig. 3 Positions of primer pairs, CpGs and restriction sites within studied regions. CpGs are shown as red and yellow bars on a line representing the DNA sequence. The red CpG is the one originally chosen from Infinium MethylationEPIC BeadChip. The scissors indicate sites that are cut by MSREs. The lighter blue primers were used for initial pyrosequencing PCR and the darker blue primers represent the sequencing primers. The patterned light green HRM primers were designed with a CpG on its 5’end (M/U Wojdacz primers) Green PCR Kit (Qiagen). The final concentration of primers was 0.5 μM. We kept recommended cycling con- ditions with 40 cycles and Tann according to Table 1. Additional files Additional file 1: Correlation coefficients for AUC-based MS-HRM calibration curves and counted methylation levels. (DOCX 14 kb) Additional file 2: qMSP methylation results calculated using the relative expression ration. (DOCX 12 kb) Additional file 1: Correlation coefficients for AUC-based MS-HRM calibration curves and counted methylation levels. (DOCX 14 kb) Pyrosequencing In one run, all samples together with methylated and unmethylated DNA standards were amplified with meth- ylated MSP, unmethylated MSP and HRM primers. For each primer set, the amplification efficiency was calculated according to Dorak et al. [42]. We performed qPCR with four dilutions of BS converted DNA of one sample and plotted decadic logarithm of the dilutions against acquired Cts. The efficiency was then calculated from the slope of the calibration curve as follows: E ¼ ½10ð−1 slopeÞ−1 ∙100. All measurements were done using StepOnePlus Real-Time PCR System (Thermo Fisher Scientific). Methylation levels were calculated using all three approaches reviewed in Husseiny et al. [36]. 2 μl of sepharose-coated Streptavidin beads (step 5.3.3.2) and by prolonging the agitation step to 20 min (step 5.3.3.6). MS-HRM Analysis We prepared 100, 75, 50, 25, 10 and 0% methylated stan- dards by mixing BS converted DNA methylated and unmethylated standards. 15 ng of BS converted samples and standards were processed using EpiTect HRM PCR Kit (Qiagen). Reaction conditions were set according to manufacturer’s protocol. The amount of reagents was ad- justed to 20 μl final volume. The final concentration of primers was 0.375 μM. The experiment was performed on Rotor-Gene Q 2plex HRM Platform (Qiagen). For the HRM analysis, the ramping was set from 67.1 to 82.2 °C, rising by 0.1 °C/2 s. Raw data were processed using web- based tool uAnalyze [41]. In the software, we performed baseline normalization and calculated the difference curves for all standards and samples using the 0% methyl- ated standard as a reference curve. Calibration curves were then plotted in Microsoft Excel according to Tse et al. [27] using peak heights and AUC of the standards’ processed data. From the calibration curves, the methyla- tion percentage of analyzed samples was calculated. References In: Dorak MT, editor. Real- time PCR. 1st ed. London: Taylor & Francis; 2007. p. 68–71. 15. Sant KE, Nahar MS, Dolinoy DC. DNA methylation screening and analysis. Methods Mol Biol (Clifton, NJ). 2012;886:385–406. Funding h k 17. Frommer M, Mcdonald LE, Millar DS, Collis CM, Watt F, Grigg GW, et al. A genomic sequencing protocol that yields a positive display of 5- methylcytosine residues in individual DNA strands. 1992,. Funding This work was supported by the Project for Conceptual Development of Research Organization (00023736, Institute of Hematology and Blood Transfusion) provided by Ministry of Health of the Czech Republic. 18. Reed K, Poulin ML, Yan L, Parissenti AM. Comparison of bisulfite sequencing PCR with pyrosequencing for measuring differences in DNA methylation. Anal Biochem. 2010;397(1):96–106. Availability of Data and Materials 19. Tost J, El abdalaoui H, Glynne Gut I. Serial pyrosequencing for quantitative DNA methylation analysis. BioTechniques. 2006;40(6):721–6. The datasets supporting the conclusions of this article are included within the article (and its additional files). 20. Delaney C, Garg SK, Yung R. Analysis of DNA methylation by 20. Delaney C, Garg SK, Yung R. Analysis of DNA methylation by pyrosequencing. Methods Mol Biol. 2015;1343:249. pyrosequencing. Methods Mol Biol. 2015;1343:249. Author details 1 27. Tse MY, Ashbury JE, Zwingerman N, King WD, Taylor SA, Pang SC. A refined, rapid and reproducible high resolution melt (HRM)-based method suitable for quantification of global LINE-1 repetitive element methylation. BMC Res Notes. 2011;4(1):565. Author details 1Institute of Clinical and Experimental Hematology, First Faculty of Medicine, Charles University and Institute of Hematology and Blood Transfusion, Prague, Czech Republic. 2Institute of Hematology and Blood Transfusion, Prague, Czech Republic. 28. Smith E, Jones ME, Drew PA. Quantitation of DNA methylation by melt curve analysis. BMC Cancer. 2009;9(1):123. curve analysis. BMC Cancer. 2009;9(1):123. 29. Malentacchi F, Forni G, Vinci S, Orlando C. Quantitative evaluation of DNA methylation by optimization of a differential-high resolution melt analysis protocol. Nucleic Acids Res. 2009 Jul;37(12):e86. Received: 6 June 2019 Accepted: 9 August 2019 Received: 6 June 2019 Accepted: 9 August 2019 30. Wojdacz TK, Borgbo T, Hansen LL. Primer design versus PCR bias in methylation independent PCR amplifications. Epigenetics. 2009;4(4):231–4. References Direct detection of methylated cytosine in DNA by use of the restriction enzyme MspI. Nucleic Acids Res. 1979;6(6):2125–32. 7. Cedar H, Solage A, Glaser G, Razin A. Direct detection of methylated cytosine in DNA by use of the restriction enzyme MspI. Nucleic Acids Res. 1979;6(6):2125–32. 8. Fraga MF, Esteller M. DNA Methylation: A Profile of Methods and Applications BioTechniques 2002;33(3):632–49 8. Fraga MF, Esteller M. DNA Methylation: A Profile Applications. BioTechniques. 2002;33(3):632–49. 8. Fraga MF, Esteller M. DNA Methylation: A Profile Applications. BioTechniques. 2002;33(3):632–49. 37. Akirav EM, Lebastchi J, Galvan EM, Henegariu O, Akirav M, Ablamunits V, et al. Detection of β cell death in diabetes using differentially methylated circulating DNA. Proc Natl Acad Sci U S A. 2011;108(47):19018–23. Applications. BioTechniques. 2002;33(3):632–49. 9. Singer-Sam J, LeBon JM, Tanguay RL, Riggs AD. A quantitative Hpall-PCR assay to measure methylation of DNA from a small number of cells. Nucleic Acids Res. 1990;18(3):687. 38. Eads CA, Danenberg KD, Kawakami K, Saltz LB, Blake C, Shibata D, et al. MethyLight: a high-throughput assay to measure DNA methylation. Nucleic Acids Res. 2000;28(8):e32. 10. Itoi E, Kokubun S, Hashimoto K, Roach HI. Improved Quantification of DNA Methylation Using Methylation-Sensitive Restriction Enzymes and Real-Time PCR. Epigenetics. 2007;2(2):86–91. 39. Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res. 2001;29(9):45. 11. Kurdyukov S, Bullock M. DNA Methylation Analysis: Choosing the Right Method. Biology. 2016;5(1):3. 40. Šestáková Š, Krejčík Z, Folta A, Cerovská E, Šálek C, Merkerová MD, et al. DNA methylation and hydroxymethylation patterns in acute myeloid leukemia patients with mutations in DNMT3A and IDH1/2 and their combinations. Cancer Biomark: Section A of Disease Markers. 2019;25(1):43–51. 12. Untergasser A, Nijveen H, Rao X, Bisseling T, Geurts R, Leunissen JAM. Primer3Plus, an enhanced web interface to Primer3. Nucleic Acids Res. 2007;35(suppl2):W74. 13. Ye J, Coulouris G, Zaretskaya I, Cutcutache I, Rozen S, Madden TL. Primer- BLAST: A tool to design target-specific primers for polymerase chain reaction. BMC bioinformatics. 2012;13(1):134. 41. Dwight Z, Palais R, Wittwer C. uAnalyze: web-based high-resolution DNA melting analysis with comparison to thermodynamic predictions. IEEE/ACM Trans Comput Biol Bioinform. 2012;9(6):1805–11. 14. Hayatsu H, Wataya Y, Kazushige K. Addition of sodium bisulfite to uracil and to cytosine. J Am Chem Soc. 1970;92(3):724–6. 42. Pfaffl MW. Real-time qPCR amplification efficiency. In: Dorak MT, editor. Real- time PCR. 1st ed. London: Taylor & Francis; 2007. p. 68–71. 42. Pfaffl MW. Real-time qPCR amplification efficiency. Authors’ Contributions f d h Quantitative PCR with subsequent melting curve analysis was performed with 10–15 ng of BS converted DNA. Re- action mix (20 μl) was prepared using QuantiTect SYBR® SS performed the measurements, data analysis and wrote the manuscript. CS revised the manuscript. HR designed the study and wrote the manuscript. All authors read and approved the final version. Page 11 of 11 Page 11 of 11 Šestáková et al. Biological Procedures Online (2019) 21:19 Consent for Publication 24. King CR, Scott Horton T. Pyrosequencing. In: Marsh S, editor. Pyrosequencing protocols Totowa. NJ: Humana Press; 2007. p. 39–55. Pyrosequencing protocols Totowa. NJ: Humana Press; 2007. p 25. Guo D, Milewicz DM. Universal primer applications for pyrosequencing. In: Marsh S, editor. Pyrosequencing protocols Totowa. NJ: Humana Press; 2007. p. 57–62. editor. Pyrosequencing protocols Totowa. NJ: Humana Press; 2007. p. 57–62 Competing Interests 26. Wojdacz TK, Dobrovic A. Methylation-sensitive high resolution melting (MS- HRM): a new approach for sensitive and high-throughput assessment of methylation. Nucleic Acids Res. 2007 Mar;35(6):e41. p g The authors declare that they have no competing interests. Author details 1Institute of Clinical and Experimental Hematology, First Faculty of Medicine Charles University and Institute of Hematology and Blood Transfusion, Prague, Czech Republic. 2Institute of Hematology and Blood Transfusion, Prague, Czech Republic. Ethics Approval and Consent to Participate 21. Tost J, Gut IG. DNA methylation analysis by pyrosequencing. Nat Protoc. 2007;2(9):2265–75. 21. Tost J, Gut IG. DNA methylation analysis by pyrosequencing. Nat Protoc. 2007;2(9):2265–75. This work was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Institute of Hematology and Blood Transfusion, Prague, Czech Republic. All blood donors provided written informed consent. 22. Li L, Dahiya R. MethPrimer: designing primers for methylation PCRs. Bioinformatics. 2002;18(11):1427–31. 22. Li L, Dahiya R. MethPrimer: designing primers for methylation PCRs. Bioinformatics. 2002;18(11):1427–31. 23. Arányi T, Váradi A, Simon I, Tusnády GE. The BiSearch web server. BMC bioinformatics. 2006;7(1):431. 23. Arányi T, Váradi A, Simon I, Tusnády GE. The BiSearch web server. BMC bioinformatics. 2006;7(1):431. 24. King CR, Scott-Horton T. Pyrosequencing. In: Marsh S, editor. 24. King CR, Scott-Horton T. Pyrosequencing. In: Marsh S, editor. Pyrosequencing protocols Totowa. NJ: Humana Press; 2007. p. 39–55. References 1. Schubeler D. Function and information content of DNA methylation. Nature. 2015;517(7534):321. 1. Schubeler D. Function and information content of DNA methylation. Nature. 2015;517(7534):321. 31. Wojdacz TK, Lotte HL. Reversal of PCR bias for improved sensitivity of the DNA h l i l i Bi T h i 2006 41(3) 274 8 31. Wojdacz TK, Lotte HL. Reversal of PCR bias for improved sensitivity of the DNA methylation melting curve assay. BioTechniques. 2006;41(3):274–8. DNA methylation melting curve assay. BioTechniques. 2006;41(3):274– 2. Jin Z, Liu Y. DNA methylation in human diseases. Genes & Diseases. 2018; 5(1):1–8. 2. Jin Z, Liu Y. DNA methylation in human diseases. Genes & Diseases. 2018; 5(1):1–8. 32. Hansen LL, Wojdacz TK, Dobrovic A. Methylation-sensitive high-resolution melting. Nat Protoc. 2008;3(12):1903–8. 3. Levenson VV. DNA methylation as a universal biomarker. Expert Rev Mol Diagn. 2010;10(4):481–8. 33. Wojdacz TK, Hansen LL, Dobrovic A. A new approach to primer design for the control of PCR bias in methylation studies. BMC Res Notes. 2008;1(1):54. g 4. Heyn H, Esteller M. DNA methylation profiling in the clinic: applications and challenges. Nat Rev Genet. 2012;13(10):679–92. 4. Heyn H, Esteller M. DNA methylation profiling in the clinic: applications and challenges. Nat Rev Genet. 2012;13(10):679–92. 34. Herman JG, Graff JR, Myöhänen S, Nelkin BD, Baylin SB. Methylation-Specific PCR: A Novel PCR Assay for Methylation Status of CpG Islands. Proc Natl Acad Sci U S A. 1996;93(18):9821–6. 5. Pidsley R, Zotenko E, Peters TJ, Lawrence MG, Risbridger GP, Molloy P, et al. Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling. Genome Biol. 2016;17(1):208. 5. Pidsley R, Zotenko E, Peters TJ, Lawrence MG, Risbridger GP, Molloy P, et al. Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling. Genome Biol. 2016;17(1):208. 35. Derks S, Lentjes MH, Hellebrekers DM, de Bruïne AP, Herman JG, van Engeland M. Methylation-specific PCR unraveled. Cellular oncology : the official journal of the International Society for Cellular Oncology. 2004;26(5–6):291–9. 6. Lövkvist C, Dodd IB, Sneppen K, Haerter JO. DNA methylation in human epigenomes depends on local topology of CpG sites. Nucleic Acids Res. 2016;44(11):5123–32. 36. Husseiny MI, Kuroda A, Kaye AN, Nair I, Kandeel F, Ferreri K. Development of a Quantitative Methylation-Specific Polymerase Chain Reaction Method for Monitoring Beta Cell Death in Type 1 Diabetes. PLoS One. 2012 Oct 1;7(10):e47942. 7. Cedar H, Solage A, Glaser G, Razin A. Publisher’s Note 16. Hernández HG, Tse MY, Pang SC, Arboleda H, Forero DA. Optimizing methodologies for PCR-based DNA methylation analysis. BioTechniques. 2013;55(4):181–97. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
https://openalex.org/W2083927736	https://europepmc.org/articles/pmc4414347?pdf=render	English	null	Red Squirrel Middens Influence Abundance but Not Diversity of Other Vertebrates	PloS one	2,015	cc-by	8,548	RESEARCH ARTICLE Red Squirrel Middens Influence Abundance but Not Diversity of Other Vertebrates Erin E. Posthumus, John L. Koprowski, Robert J. Steidl Erin E. Posthumus, John L. Koprowski, Robert J. Steidl Wildlife and Fisheries Science, School of Natural Resources and the Environment, University of Arizona, Tucson, Arizona, United States of America Erin E. Posthumus, John L. Koprowski, Robert J. Steidl Wildlife and Fisheries Science, School of Natural Resources and the Environment, University of Arizona, Tucson, Arizona, United States of America Wildlife and Fisheries Science, School of Natural Resources and the Environment, University of Arizona, Tucson, Arizona, United States of America Wildlife and Fisheries Science, School of Natural Resources and the Environment, University of Arizona, Tucson, Arizona, United States of America eposthum@email.arizona.edu OPEN ACCESS Citation: Posthumus EE, Koprowski JL, Steidl RJ (2015) Red Squirrel Middens Influence Abundance but Not Diversity of Other Vertebrates. PLoS ONE 10(4): e0123633. doi:10.1371/journal.pone.0123633 Academic Editor: Tim A. Mousseau, University of South Carolina, UNITED STATES Received: December 18, 2014 Accepted: February 27, 2015 Published: April 29, 2015 Copyright: © 2015 Posthumus et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Academic Editor: Tim A. Mousseau, University of South Carolina, UNITED STATES Received: December 18, 2014 Accepted: February 27, 2015 Published: April 29, 2015 Copyright: © 2015 Posthumus et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Academic Editor: Tim A. Mousseau, University of South Carolina, UNITED STATES Academic Editor: Tim A. Mousseau, University of South Carolina, UNITED STATES Received: December 18, 2014 Accepted: February 27, 2015 Published: April 29, 2015 Received: December 18, 2014 Accepted: February 27, 2015 Published: April 29, 2015 Copyright: © 2015 Posthumus et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Locality data for the federally endangered Mt Graham red squirrel cannot be made publicly available. The Mount Graham red squirrel (Tamiasciurus hudsonicus grahamensis) was listed in 1982 as a federally endangered species under the United States Endangered Species Act of 1973. We are unable to disclose locality data that will pinpoint the nesting locations of this species. Our Supporting Information Dataset includes unique identifiers for each midden location, but we do not disclose latitudinal and longitudinal coordinates. These location data may not Abstract Some animals modify the environment in ways that can influence the resources available to other species. Because red squirrels (Tamiasciurus hudsonicus) create large piles of conifer-cone debris (middens) in which they store cones, squirrels concentrate resources that might affect biodiversity locally. To determine whether other animals are attracted to midden sites beyond their affinity for the same resources that attract red squirrels, we as- sessed associations between middens, mammals, and birds at population and community levels. We surveyed 75 middens where residency rates of red squirrels varied during the previous five years; sampling along this residency gradient permitted us to evaluate the in- fluence of resources at middens beyond the influence of a resident squirrel. At each loca- tion, we quantified vegetation, landscape structure, abundance of conifer cones, and midden structure, and used capture–recapture, distance sampling, and remote cameras to quantify presence, abundance, and species richness of mammals and birds. Red squirrels and the resources they concentrated at middens influenced mammals and birds at the pop- ulation scale and to a lesser extent at the community scale. At middens with higher residen- cy rates of red squirrels, richness of medium and large mammals increased markedly and species richness of birds increased slightly. After accounting for local forest characteristics, however, only species richness of medium-to-large mammals was associated with a red squirrel being resident during surveys. In areas where red squirrels were resident during surveys or in areas with greater amounts of resources concentrated by red squirrels, abun- dances of two of four small mammal species and two of four bird species increased. We conclude that the presence of this ecosystem modifier and the resources it concentrates in- fluence abundance of some mammals and birds, which may have implications for maintain- ing biodiversity across the wide geographic range inhabited by red squirrels and other larderhoarding animals. Examples include beaver, which alter hydrology and productivity by building dams and through their feeding activities [7], prairie dogs, which alter soil structure and vege- tation composition through burrowing and feeding [8], badgers (Taxidea taxus), which create mounds that maintain diversity of prairie flora [13], and woodrats (Neotoma spp.), which alter rates of nitrogen mineralization in soils and that build nests with a unique microclimate used by other animals [14]. Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. Some species also alter the distribution of resources, concentrating resources for themselves and in ways that could increase the availability of resources for other species [15]. Handling and storage of food by animals for later use, termed larderhoarding, enables animals to satisfy energy requirements when resource abundance fluctuates [15]. Red squirrels (Tamiasciurus hudsonicus), a territorial tree squirrel distributed throughout most coniferous forests of the United States and Canada [16], often larderhoard food supplies that last for years [15, 17]. Red squirrels fulfill a number of important ecosystem functions, including dispersing seeds, serving as prey for predators, and creating structure via conspicuous cone-scale piles, known as mid- dens, which are a product of feeding in a single location (Fig 1) [16]. Middens are central to a single red squirrel’s well-defended territory [16], which can vary from 1 to >10 ha [18]. In the southwestern United States, middens are typically located in forest patches with locally dense canopies, high stem densities, thick foliage, and on cooler north-facing slopes [19–20], and have a cool, moist microclimate optimal to larderhoard conifer cones and fungi [21–22]. Mid- dens often are used over multiple generations [17] and can reach 13 m in diameter and 50 cm in depth [23]. The structure of middens facilitates tunneling, nesting, and access to thousands of stored cones, plus seeds dropped during feeding are distributed throughout a red squirrel’s territory [24]. These resources may attract small mammals and birds, which may in turn attract other predatory mammals and birds [25–26]. Positive associations with middens have been reported for marten (Martes americana) that use subnivean tunnels more often when they are near middens [27], and grizzly bears (Ursus arctos horribilis) that excavate middens to obtain seeds of whitebark pine (Pinus albicaulis) [28]. Influence of Red Squirrel Middens on Abundance of Vertebrates function adversely [4], earning them the moniker of keystone species [5]. Although the utility of the keystone-species concept has been debated, evaluating the interaction strength of particular species can increase our understanding of factors that govern ecological processes [1, 6]. A suite of mammals that vary widely both taxonomically and functionally have been hypothesized to function in these ecologically important ways, including beaver (Castor cana- densis) [7], prairie dogs (Cynomys spp.) [8], kangaroo rats (Dipodomys spp.) [9], sea otters (Enhydra lutris) [10], and bison (Bison bison) [11]. Species that redistribute resources or create structures on a large scale have been further classified as keystone modifiers [6] or ecosystem engineers [12]. Examples include beaver, which alter hydrology and productivity by building dams and through their feeding activities [7], prairie dogs, which alter soil structure and vege- tation composition through burrowing and feeding [8], badgers (Taxidea taxus), which create mounds that maintain diversity of prairie flora [13], and woodrats (Neotoma spp.), which alter rates of nitrogen mineralization in soils and that build nests with a unique microclimate used by other animals [14]. function adversely [4], earning them the moniker of keystone species [5]. Although the utility of the keystone-species concept has been debated, evaluating the interaction strength of particular species can increase our understanding of factors that govern ecological processes [1, 6]. be requested. John Koprowski, squirrel@ag.arizona. edu, may be contacted with any questions. be requested. John Koprowski, squirrel@ag.arizona. edu, may be contacted with any questions. Funding: This work was supported by the United States Forest Service Rocky Mountain Research Station, #10-JV-188 (University of Arizona; www.fs. fed.us/rmrs); T&E Incorporated’s Grants for Conservation Research (www.tandeinc.com); and the American Society of Mammalogists’ Grants-in-Aid Program (http://www.mammalsociety.org/committees/ grants-aid). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. A suite of mammals that vary widely both taxonomically and functionally have been hypothesized to function in these ecologically important ways, including beaver (Castor cana- densis) [7], prairie dogs (Cynomys spp.) [8], kangaroo rats (Dipodomys spp.) [9], sea otters (Enhydra lutris) [10], and bison (Bison bison) [11]. Species that redistribute resources or create structures on a large scale have been further classified as keystone modifiers [6] or ecosystem engineers [12]. Introduction Certain species influence ecosystems disproportionally [1–2] or perform unique ecological func- tions [3]. If extirpated, the loss of these species could influence species diversity and ecosystem 1 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Richness and abundance of mammals and birds has been observed to be higher at mid- dens than in surrounding forest [25–26]. Although species richness of vertebrates may be higher at middens, this might be an artifact of species selecting forest features that are coincident with those used by red squirrels [26] rath- er than species leveraging resources concentrated by red squirrels. Red squirrels select sites with high canopy cover, large cone-bearing trees [19–20], and coarse woody debris [26], fea- tures likely important to many other species. For example, deer mice (Peromyscus spp.), chip- munks (Tamias spp.), voles (Microtus spp.), and woodrats consume conifer seeds [29] and den in coarse woody debris [30–31]. Insectivores, such as shrews (Soricidae), may be attracted to the abundant insects supported by conifer seeds and in moist conditions; flying squirrels (Glaucomys spp.) and voles may target woody debris that is conducive to growth of fungi [32]. To determine whether animals are attracted to middens beyond their affinity for resources similar to those selected by red squirrels, we assessed associations between middens, mammals, and birds at population and community scales across a gradient of residency rates by squirrels PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 2 / 14 Influence of Red Squirrel Middens on Abundance of Vertebrates Fig 1. Red squirrel (Tamiasciurus hudsonicus) midden. Photographs of (a) red squirrel (Tamiasciurus hudsonicus) midden cone-scale pile, (b) cached cones inside pit excavated by red squirrel, and (c) stored cones, which may number in the thousands at a single midden. Photo credit, E. E. Posthumus. Mt. Graham, Graham Co. Arizona, 2011–2012. Fig 1. Red squirrel (Tamiasciurus hudsonicus) midden. Photographs of (a) red squirrel (Tamiasciurus hudsonicus) midden cone-scale pile, (b) cached cones inside pit excavated by red squirrel, and (c) stored cones, which may number in the thousands at a single midden. Photo credit, E. E. Posthumus. Mt. Graham, Graham Co. Arizona, 2011–2012. doi:10.1371/journal.pone.0123633.g001 during the previous five years. Our objective was to determine if red squirrels or the resources they concentrate influence populations and communities of vertebrates. If the presence of mid- dens influences other species, we anticipate that richness and abundance of mammals and birds will be higher at locations with higher residency rates and will be associated positively with the presence of red squirrels at middens and abundance of the resources they concentrate. during the previous five years. Our objective was to determine if red squirrels or the resources they concentrate influence populations and communities of vertebrates. If the presence of mid- dens influences other species, we anticipate that richness and abundance of mammals and birds will be higher at locations with higher residency rates and will be associated positively with the presence of red squirrels at middens and abundance of the resources they concentrate. Resources Concentrated by Red Squirrels We recorded whether a red squirrel was resident at each location between May and September of each year (red squirrel residency during surveys), and measured area and depth of cone- scale piles at each midden. We ranked the number of cones cached in each scale pile from 1–4 to reflect 1–25, 26–50, 51–75, and >75 cones, respectively, and used the highest value observed in either September or December of the previous year to approximate cone availability. Habitat Features We surveyed vegetation along 2-m wide and 30-m long transects that radiated from the center of each location in four cardinal directions (total area surveyed = 0.02 ha/location). We used a spherical densiometer to estimate canopy closure [34] and averaged readings taken 0, 2, 10, 20, and 30 m from center of each location in each direction. We counted the number of live and dead trees in three size classes (<20 cm, 20–40 cm, >40 cm dbh), measured diameter at breast height (dbh), and calculated basal area/ha for all live and dead trees. We recorded the number of live trees of each species and calculated diversity at each location with the Shannon-Weiner index. We measured volume of hard downed logs >20 cm in diameter and slope and aspect in absolute degrees from the center of each location. Materials and Methods Field efforts were conducted under permits from the United States Department of Agriculture Forest Service, Arizona Game and Fish Department, and United States Fish and Wildlife 3 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Influence of Red Squirrel Middens on Abundance of Vertebrates Service, and approved by the University of Arizona Institutional Animal Care and Use Com- mittee (IACUC; Protocol #11–248). We studied a 100-ha area of mixed-conifer forest at an ele- vation of 2,870 to 3,050 m in the Pinaleño Mountains, 25 km southwest of Safford, Graham Co., Arizona. The forest is dominated by Douglas-fir (Pseudotsuga menziesii), white fir (Abies concolor), and southwestern white pine (Pinus strobiformis), interspersed with cork-bark fir (Abies lasiocarpa var. arizonica), Engelmann spruce (Picea engelmanii), aspen (Populus tremu- loides) and ponderosa pine (Pinus ponderosa) [33]. Middens in this area have been surveyed for red squirrels every three months since 1996 [33] because endemic Mt. Graham red squirrels (red squirrel; T. h. grahamensis) are listed as endangered in the United States. In 2011, we classified 176 locations based on residency rates of red squirrels during the previ- ous five years, which we defined as the proportion of quarterly surveys where a red squirrel was observed at each midden. Surveys involved visiting middens to determine if each was inhabited based on signs of red squirrel activity (tracks, feeding sign, nesting material) [33]. Because red squirrels are highly territorial and solitary, when middens are inhabited, they are usually de- fended aggressively by the resident adult [33]. From this set of midden locations, we selected 50 in both 2011 and 2012, 10 at random from each of five residency rate classes to capture the en- tire gradient: 0% residency since 1996, 0% residency since 2011, 1–49% and without a resident red squirrel for at least one year prior to the study, 50–75% residency, and > 75% residency. Fif- teen locations were included in both 2011 and 2012 samples, which we treated as independent across years. Distance between locations averaged 75 m (min = 30 m, max = 150 m). Data Analysis We used logistic regression for binomial counts (number of quarterly surveys where a red squirrel was present relative to the total number of quarterly surveys done at each site) to ex- plore relationships between residency rates and habitat features. We used linear regression to explore relationships between residency rates and species richness of small mammals, medium and large mammals, and birds, and added quadratic and cubic components successively until they were no longer significant (P > 0.10) when relationships were curvilinear [40]. To reduce dimensionality among the seven habitat features we measured, we used principal components analysis based on the correlation matrix after first transforming basal area, number of large trees > 40 cm dbh, number of large snags > 40 cm dbh, volume of downed logs, and slope with the natural log and by squaring canopy cover to normalize their distributions. To test whether species richness was associated with a red squirrel being resident during surveys or the resources concentrated by red squirrels after accounting for forest characteristics, we first used Poisson regression to model richness of small mammals, medium and large mammals com- bined, and birds separately as a function of habitat features as described by the first three prin- cipal components. We retained influential components (P < 0.10), then fit a second model to describe the influence of red squirrels that included red squirrel residency during surveys, vol- ume of the cone-scale pile, and index of cached cones. We then used a drop-in-deviance test to contrast the amount of variation in the response explained by the two models and to determine the influence of red squirrels on each response after accounting for habitat features [41]. For all mammal and bird species observed, we used logistic regression to model presence of each species at each location and as a function of residency rates of red squirrels. For the two mammal species captured at > 90% of sites, deer mice (Peromyscus maniculatus) and cliff chip- munks (Tamias dorsalis), we created hierarchical models of abundance based on a capture- recapture framework with the R package ‘unmarked’ [42]. Influence of Red Squirrel Middens on Abundance of Vertebrates individuals captured to species, and weighed, measured [36], and marked each individual with ear tags (Monel #1005–1, National Band and Tag Company, Newport, KY), before releasing animals at the capture location. Handling methods followed American Society of Mammalo- gists [37] and University of Arizona IACUC guidelines. We surveyed medium and large mammals at all locations with remote cameras (Bushnell Trophy Cam 119436c, TrailCamPro, Springfield, MO) [25]. We set three cameras 5–10 m from and facing the center of each survey location and located 1–2 m above the ground for a six-day period in each year established at random between May and September. We set cam- eras at maximum sensitivity to record one photo upon detection, with a 3-sec delay between photographs. We identified animals in photographs to species and counted the number of spe- cies detected over the six-day period for each location. We surveyed birds from each location with 10-min point counts on four consecutive morn- ings (0500–0900 h) concurrent with small mammal trapping [38]. We detected birds by sight and sound and classified distances from the observer as 0–10 m, 11–20 m, 21–30 m, 31–50 m, and 51–100 m, truncating detections at 100 m to include only those birds proximate to mid- dens. We considered only species that were ground foragers to target species that search for seeds or insects on the ground [39]. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Composition of Mammal and Bird Communities We surveyed small mammals at all locations using a trapping-web design centered on each lo- cation [35]. Webs were comprised of eight, 30-m lines radiating at 45° angles from the center of the location. Along each line, we set one folding, galvanized Sherman live trap (7.5 by 9 by 23 cm: Model LFG, H.B. Sherman Trap Co., Tallahassee, FL) at 10, 20, and 30 m from the cen- ter of each location, four traps at 2 m in each cardinal direction, and one trap at the center for a total of 29 traps per location. For one four-night period at each location per year between May and September, we baited traps at sunset with a mix of peanut butter, rolled oats, and alfalfa pellets, checked traps at sunrise, and reset and checked again mid-morning. We identified PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 4 / 14 Influence of Red Squirrel Middens on Abundance of Vertebrates during surveys or the resources concentrated by red squirrels after accounting for habitat fea- tures, we fit two models using a process similar to the one we described for species richness. For all ground-foraging bird species, including American robin (Turdus migratorius), her- mit thrush (Catharus guttatus), Stellar’s jay (Cyanocitta stelleri), and yellow-eyed junco (Junco phaeonotus), we created hierarchical models for abundance based on a distance-sampling framework with ‘unmarked’ [42]. We compared half normal, hazard rate, and uniform detec- tion functions, explored a series of models for detection probability including models that var- ied with behavior and mean low temperature and precipitation, and retained influential covariates (P < 0.10). To determine whether abundance of each species was associated with a red squirrel being resident during surveys or the resources concentrated by red squirrels after accounting for vegetation features, we used a process similar to the one we described for species richness. Habitat Features Surrounding Middens Residency rates of red squirrels increased with canopy cover (7.16 ± 0.57% [SE], χ2 = 157.02, P < 0.001), basal area (15.57 ± 1.36 m²/ha, χ2 = 79.63, P < 0.001), number of large trees > 40 cm dbh (1.78 ± 1.11 trees/ha, χ2 = 33.02, P < 0.001), number of large snags > 40 cm dbh (1.31 ± 1.07 snags/ha, χ2 = 16.62, P < 0.001), steeper slopes (2.65 ± 1.20°, χ2 = 28.50, P < 0.001), more north facing slopes (-0.0016 ± 0.0010°, χ2 = 2.88, P = 0.090), and decreased with the index of tree diversity (-0.53 ± 0.16, χ2 = 10.70, P < 0.001) and volume of downed logs (-2.08 ± 1.19 m3, χ2 = 17.34, P < 0.001). Volume of cone-scale piles (4.75 ± 0.55 m3, t99 = 8.64, P < 0.001) and the index of number of cached cones (2.55 ± 0.30, t99 = 8.47, P < 0.001) in- creased as residency rates increased, indicating that resources concentrated by red squirrels varied with residency rates (Fig 2). Most (60.7%) of the variation in the eight original habitat features was captured by the first three principal components. The first principal component (25.4% of variation) described as- pects of forest density, with high loadings on percent canopy cover (factor weight = 0.81), north facing slopes (0.72), number of live trees >40 cm dbh (0.62), and low tree diversity (0.51). The second principal component (19.4%) described aspects of tree structure, with high loadings on number of large snags >40 cm dbh (0.83) and basal area (0.66). Lastly, the third principal component (15.8%) described aspects of the forest floor, including volume of downed logs (0.89) and flatness of slopes (0.58). Residency rates were associated with the first (r = 0.257, P = 0.010), second (r = 0.195, P = 0.053) and third principal component (r = 0.192, P = 0.055). Data Analysis To model the detection process, we assumed a behavioral response to trapping (Mb, based on our preliminary modeling) and explored the influence of mean low daily temperature and mean daily precipitation for each four-day trapping period (National Wildfire Coordinating Group 2012, from a weather station within our study area) and retained influential covariates (P < 0.10). To determine whether abundance of P. maniculatus and T. dorsalis was associated with a red squirrel being resident PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 5 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 doi:10.1371/journal.pone.0123633.g002 Influence of Resources Concentrated by Red Squirrels on Vertebrates We captured six species of small mammals during 11,600 trap nights (Table 1). Species rich- ness of small mammals was not associated with residency rate of red squirrels (t99 = 0.94, P = 0.350, r² = 0.009). After accounting for habitat features, species richness of small mammals was not influenced by a red squirrel being resident during surveys or resources concentrated by red squirrels (Tables 2 and 3). We observed 8 species of medium and large mammals in 1722 camera days in addition to red squirrels (Table 1). Species richness of medium and large mam- mals increased as residency rates of red squirrels increased (0.98 ± 0.32 species, t99 = 3.34, P = 0.001, r² = 0.102). After accounting for forest characteristics, species richness of medium and large mammals increased when a red squirrel was resident during surveys (Tables 2 and 3). We detected 25 species of birds, including four that we classified as ground foragers (Table 1). Species richness of birds increased slightly as residency rate of red squirrels increased 6 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Influence of Red Squirrel Middens on Abundance of Vertebrates Fig 2. Relationship between red squirrel residency rates and resources concentrated by red squirrels. Relationship between resources concentrated by red squirrels, red squirrel residency during surveys, and residency rate (proportion of quarterly surveys during the previous five years) of red squirrels at middens, Mt. Graham, Graham Co. Arizona, 2011–2012. doi:10.1371/journal.pone.0123633.g002 Fig 2. Relationship between red squirrel residency rates and resources concentrated by red squirrels. Relationship between resources concentrated by red squirrels, red squirrel residency during surveys, and residency rate (proportion of quarterly surveys during the previous five years) of red squirrels at middens, Mt. Graham, Graham Co. Arizona, 2011–2012. doi:10.1371/journal.pone.0123633.g002 (1.29 ± 0.77 species, t99 = 1.65, P = 0.101, r² = 0.027). After accounting for habitat features, spe- cies richness of birds was not influenced by a red squirrel being resident during surveys or re- sources concentrated by red squirrels (Table 2), but did increase with forest density (Table 3). After accounting for forest characteristics, abundance of P. maniculatus was lower in loca- tions with a red squirrel resident during surveys (-0.36 ± 0.12 individuals) and increased as the number of cached cones increased (0.08 ± 0.03 individuals/cone index unit; Tables 2 and 4). Detection probability of P. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Influence of Red Squirrel Middens on Abundance of Vertebrates Table 1. Proportion of sites surveyed at which mammals and birds were observed. Proportion of sites Residency Group Species χ2 P Small mammals Microtus longicaudus 0.32 0.01 0.929 Neotoma mexicana 0.59 0.60 0.439 Peromyscus boylii 0.07 1.82 0.178 Peromyscus maniculatus 1.00 Sorex monticolus 0.40 1.59 0.207 Tamias dorsalis 0.94 1.75 0.186 Medium and large mammals Lynx rufus 0.12 5.61 0.018 Mephitis mephitis 0.26 0.02 0.903 Odocoileus virginianus 0.09 1.62 0.202 Otospermophilus variegatus 0.53 0.01 0.926 Puma concolor 0.02 0.44 0.506 Sciurus aberti 0.09 0.06 0.802 Urocyon cinereoargenteus 0.04 1.37 0.241 Ursus americanus 0.26 4.22 0.040 Ground-foraging birds Catharus guttatus 0.74 0.00 0.989 Cyanocitta stelleri 0.19 2.73 0.098 Junco phaeonotus 0.86 0.46 0.498 Turdus migratorius 0.34 0.15 0.695 Chi-square and P-values from logistic regression models comparing proportion of sites surveyed at which mammal and bird species were observed across the gradient of residency rates for red squirrels (n = 100). Table 1. Proportion of sites surveyed at which mammals and birds were observed. Table 1. Proportion of sites surveyed at which mammals and birds were observed. Influence of Resources Concentrated by Red Squirrels on Vertebrates maniculatus changed in response to initial capture and with precipi- tation (Table 4). After accounting for forest characteristics, abundance of T. dorsalis was lower i l i i h d i l id d i ( 0 74 0 17 i di id l ) d i d (1.29 ± 0.77 species, t99 = 1.65, P = 0.101, r² = 0.027). After accounting for habitat features, spe- cies richness of birds was not influenced by a red squirrel being resident during surveys or re- sources concentrated by red squirrels (Table 2), but did increase with forest density (Table 3). After accounting for forest characteristics, abundance of P. maniculatus was lower in loca- tions with a red squirrel resident during surveys (-0.36 ± 0.12 individuals) and increased as the number of cached cones increased (0.08 ± 0.03 individuals/cone index unit; Tables 2 and 4). Detection probability of P. maniculatus changed in response to initial capture and with precipi- tation (Table 4). After accounting for forest characteristics, abundance of T. dorsalis was lower in locations with a red squirrel resident during surveys (-0.74 ± 0.17 individuals), and increased as the number of cached cones increased (0.11 ± 0.04 individuals/cone-index unit; Tables 2 and 4). Detection probability of T. dorsalis changed with precipitation (Table 4). After accounting for habitat features, abundance of C. guttatus was higher in locations with a red squirrel resident during surveys (0.66 ± 0.26 individuals; Tables 2 and 4). Abundances of C. stelleri and J. phaeonotus were not influenced by a red squirrel being resident during surveys or resources concentrated by red squirrels (Tables 2 and 4). Abundance of T. migratorius was higher in locations where a red squirrel was resident during survey (1.19 ± 0.47 individuals/ cone-index unit) and lower at locations with larger cone scale piles (-0.48 ± 0.19 m3; Tables 2 and 4). Detection probability of C. stelleri was influenced by low temperature and detection of T. migratorius was influenced by precipitation (Table 4). For C. guttatus, C. stelleri, and T. migratorius, a hazard-rate detection function fit best, whereas for J. phaeonotus, a half- normal detection function fit best. 7 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Group Chi-square and P-values from logistic regression models comparing proportion of sites surveyed at which mammal and bird species were observed across the gradient of residency rates for red squirrels (n = 100). doi:10.1371/journal.pone.0123633.t001 doi:10.1371/journal.pone.0123633.t002 Effects of red squirrels on vertebrate communities Although species classified as keystone modifiers often are thought to explain changes in com- munity-scale attributes, such as species diversity [43], we found that the largest effects of red Table 2. Estimates ± standard errors and test statistics with associated P-values from models of species richness and abundance contrasting sites where red squirrel were and were not resident during surveys. R d i l Table 2. Estimates ± standard errors and test statistics with associated P-values from models of species richness and abundance contrasting sites where red squirrel were and were not resident during surveys. Red squirrel Metric Group Resident Not Resident χ2 P Richness Small mammals 3.40 ± 0.17 3.28 ± 0.12 0.38 0.944 Medium and large mammals 2.60 ± 0.26 1.35 ± 0.11 14.96 0.002 Birds 7.90 ± 0.67 6.78 ± 0.30 2.52 0.471 Abundance Peromyscus maniculatus 17.32 ± 4.11 24.8 ± 5.36 11.43 0.010 Tamias dorsalis 7.02 ± 1.78 14.68 ± 1.78 18.79 < 0.001 Catharus guttatus 4.63 ± 1.04 2.39 ± 0.33 8.45 0.038 Cyanocitta stelleri 0.30 ± 0.20 0.40 ± 0.14 1.43 0.698 Junco phaeonotus 49.05 ± 9.99 55.98 ± 5.68 3.31 0.346 Turdus migratorius 4.34 ± 1.58 1.32 ± 0.29 10.13 0.017 Species richness of three taxonomic groups and abundance of two small mammal and four bird species in areas with and without a red squirrel resident during surveys. Chi-squared statistics and P-value (all with degrees of freedom = 3) are from drop-in-deviance tests comparing models with and without a resident red squirrel. s and test statistics with associated P-values from models of species richness and abundance contrasting ere not resident during surveys. Table 2. Estimates ± standard errors and test statistics with associated P-values from models of species richness and abundance contrasting sites where red squirrel were and were not resident during surveys. Table 2. Estimates ± standard errors and test statistics with associated P-values from models of species ric sites where red squirrel were and were not resident during surveys. Species richness of three taxonomic groups and abundance of two small mammal and four bird species in areas with and without a red squirrel resident during surveys. Chi-squared statistics and P-value (all with degrees of freedom = 3) are from drop-in-deviance tests comparing models with and without a resident red squirrel. Influence of Red Squirrel Middens on Abundance of Vertebrates Table 3. Parameter estimates, standard errors, and P-values from models assessing the importance of a red squirrel being resident and resources concentrated by red squirrels during surveys to species richness of mammals and birds. Group Parameter Estimate SE P Small mammals Intercept 1.17 0.07 < 0.001 Squirrel resident during surveys -0.01 0.19 0.952 Volume of cone-scale pile (m3) < 0.01 0.03 0.951 Index of cached cones 0.02 0.05 0.286 Medium and large mammals Intercept 0.26 0.10 0.013 Squirrel resident during surveys 0.49 0.34 0.043 Volume of cone-scale pile (m3) 0.03 0.03 0.330 Index of cached cones 0.02 0.06 0.797 Birds Intercept 1.92 0.05 < 0.001 Forest density 0.14 0.04 < 0.001 Tree structure 0.05 0.05 0.164 Squirrel resident during surveys 0.12 0.12 0.324 Volume of cone-scale pile (m3) -0.02 0.02 0.300 Index of cached cones 0.03 0.03 0.428 Variables without units are either indicator variables or were standardized for analysis. doi:10.1371/journal.pone.0123633.t003 Table 3. Parameter estimates, standard errors, and P-values from models assessing the importance of a red squirrel being resident and resources concentrated by red squirrels during surveys to species richness of mammals and birds. doi:10.1371/journal.pone.0123633.t003 squirrels on vertebrates was at the population scale. Importantly, changes in species diversity attributed to keystone species are rarely disentangled from overlapping habitat preferences shared by other species. After accounting for the influence of vegetation and landscape charac- teristics, a red squirrel being resident during surveys and resources concentrated by red squir- rels explained little of the variation in species richness for small mammals and birds. However, we observed an increase in species richness of medium and large mammals, principally meso- carnivores and predators, as residency rates of red squirrel middens increased, suggesting heightened activity of animals at red squirrel middens. After removing the effects of vegetation and landscape characteristics, species richness of medium and large mammals increased at lo- cations where a red squirrel was resident during survey. Bobcats may be attracted to middens due to increased prey [44] and black bears excavate middens in search of cones [45]. Mammals may also use middens for thermoregulatory properties, as marten use middens as resting sites in winter [46]. Species richness of birds varied with local habitat features, but was not influenced positively by a red squirrel being resident during surveys or resources concentrated by red squirrel. Metric Species richness of three taxonomic groups and abundance of two small mammal and four bird species in areas with and without a red squirrel resident during surveys. Chi-squared statistics and P-value (all with degrees of freedom = 3) are from drop-in-deviance tests comparing models with and without a resident red squirrel. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 8 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Mid- dens with high residency rates had high canopy cover and basal area, and squirrels likely select and persist at these sites for protection from avian predators [44], presence of large amounts of cone-bearing trees [47], and moderate temperatures with higher humidity that aid thermoreg- ulation [17] and cone storage [24]. Similarly, areas where the forest canopy is closed likely pro- vide protection from predators, a moderate microclimate, and nesting locations for a variety of species [48–49]. However, as documented in our photographs, avian ground foragers used the surface of the cone-scale pile at locations where a red squirrel was resident during surveys, per- haps to forage for seeds or insects [39]. We recommend further study on the abundance and diversity of arthropods in cone-scale piles to understand whether middens are a hotspot for ar- thropod diversity as found at structures created by other keystone modifiers such as packrats and kangaroo rats [9,14]. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 9 / 14 Influence of Red Squirrel Middens on Abundance of Vertebrates Table 4. Parameter estimates from models assessing the importance of resources concentrated by red squirrels and a red squirrel being resident during surveys to abundance of mammals and birds. Group Process Parameter Estimate SE P Peromyscus maniculatus Abundance Intercept 3.14 0.22 < 0.001 Tree structure 0.06 0.04 0.120 Forest ﬂoor 0.16 0.04 < 0.001 Squirrel resident during surveys -0.36 0.12 0.002 Volume of cone-scale pile (m3) 0.01 0.02 0.546 Cached cone index 0.08 0.03 0.011 Detection Initial capture probability -2.07 0.29 < 0.001 Recapture probability -1.31 0.07 < 0.001 Precipitation (mm) -17.24 7.21 0.017 Tamias dorsalis Abundance Intercept 2.37 0.17 < 0.001 Forest density 0.17 0.06 0.006 Midden 0.23 0.16 0.144 Squirrel resident during surveys -0.74 0.17 < 0.001 Volume of cone-scale pile (m3) 0.04 0.03 0.137 Cached cone index 0.11 0.04 0.013 Detection Intercept -2.46 0.14 < 0.001 Precipitation (mm) 0.90 0.36 0.014 Catharus guttatus Abundance Intercept 0.87 0.14 < 0.001 Forest density 0.17 0.09 0.052 Squirrel resident during surveys 0.66 0.26 0.010 Volume of cone-scale pile (m3) -0.12 0.11 0.257 Index of cached cones 0.00 0.07 0.996 Detection Intercept 4.21 0.65 < 0.001 Scale 2.00 6.99 0.775 Cyanocitta stelleri Abundance Intercept -0.97 0.37 0.009 Tree structure 0.81 0.30 0.007 Squirrel resident during surveys -0.28 0.70 0.689 Volume of cone-scale pile (m3) 0.27 0.32 0.385 Index of cached cones 0.06 0.19 0.749 Detection Intercept 4.08 0.35 < 0.001 Low temperature -0.10 0.08 0.196 Scale 3.90 26.4 0.882 Junco phaeonotus Abundance Intercept 4.06 0.10 0.000 Squirrel resident during surveys -0.13 0.22 0.555 Volume of cone-scale pile (m3) -0.02 0.08 0.853 Index of cached cones -0.05 0.06 0.405 Detection Intercept 2.72 0.03 0.000 Turdus migratorius Abundance Intercept 0.36 0.21 0.081 Squirrel resident during surveys 1.19 0.47 0.011 Volume of cone-scale pile (m3) -0.48 0.19 0.011 Index of cached cones -0.12 0.13 0.341 Detection Intercept 3.99 0.10 0.000 Precipitation (mm) -3.78 1.57 0.016 Scale 2.10 0.64 0.001 Parameter estimates and standard errors for models of abundance of mammals observed at > 90% of sites and ground foraging birds. Variables without units are either indicator variables or were standardized for analysis. assessing the importance of resources concentrated by red squirrels and a red squirrel being resident and birds Parameter estimates and standard errors for models of abundance of mammals observed at > 90% of sites and ground foraging birds. Conclusions By producing a midden, red squirrels modify the distribution of resources and influence local forest structure and microclimate in ways that can persist for decades when occupied serially by residents [24, 33]. Although red squirrel middens support higher levels of vertebrate diversi- ty and activity, most of their effect is due to vegetation characteristics associated with sites se- lected for construction of middens [19, 20]. Middens provide a visible structure that indicates the availability of resources that we found to be associated with increased occurrence and abun- dance of numerous other species. Knowledge of the important function of middens in influenc- ing patterns in animal diversity, and potentially larderhoards of other species [15], is critical to our understanding of ecosystem function. Although species richness did not increase directly as a result of red squirrel middens for all taxa, our finding that red squirrels are associated with high levels of mammal and bird diversity due to a common affinity for key forest characteristics is important. Middens serve as conspic- uous indicators of diversity in forests. The Mt. Graham red squirrel persists at the southern- most extent of the distribution of red squirrels [29], an area with warmer, drier forests than in much of the rest of the species range. At the southern end of the range, forest disturbances, such as loss of canopy cover from insect infestations or fire, which are likely to become more frequent with climate change [4], threaten to alter midden microclimate to the detriment of food storage [22]. With predicted global change creating warmer, drier climates, and threaten- ing montane diversity [51], such challenges in southern forests may portend the future of forest communities at northern latitudes. Our results suggest that although red squirrels are associated with increased species richness their primary impact occurs at the population scale for several common species. Abundance can be influenced through multiple pathways, including indirect facilitation as has been ob- served in kangaroo rats [9, 52]. Although mechanisms for these processes are unknown, the in- fluence of red squirrels on multiple dominant species within the forest suggests a subtle impact with potential links to pathways of influence observed in many ecosystem engineers [6, 53]. Variables without units are either indicator variables or were standardized for analysis. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 10 / 14 Influence of Red Squirrel Middens on Abundance of Vertebrates Effects of red squirrels on species abundances Resources concentrated by red squirrels had their largest effects at the population scale for a number of common small mammals and birds. We detected P. maniculatus at all locations and T. dorsalis at all but six locations. These species are generalists, and their diet may include seeds of conifers, shrubs, grasses and forbs as well as insects, snails, fungi, bones, leaves or bark [29]. Abundances of P. maniculatus and T. dorsalis were associated positively with food resources at middens. P. maniculatus live and nest in logs and under debris piles [29], which may result in an association of P. maniculatus with the food resources of the midden cone-scale pile. This is supported by photographs of these species on the cone-scale pile, including images of T. dorsa- lis holding conifer cones that may have been taken from the midden. T. dorsalis also tunnels and creates dens and nests in many substrates [50], which may be facilitated by the microcli- mate and structure of the cone-scale pile. Abundances of T. migratorius and C. guttatus were associated positively with a red squirrel being resident during surveys. Ground-foraging birds were the only birds that we detected with cameras at the cone-scale pile. The diet of these birds is composed primarily of seeds and inver- tebrates [39], which may be higher with a resident red squirrel dropping seeds while feeding [24], and might explain detections of these birds on the surface of the cone-scale pile. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Acknowledgments We thank W. M. Block, United States Department of Agriculture Forest Service and the Arizona Agriculture Experiment Stations for supporting this research and R. W. Mannan for suggestions and contributions throughout the project. M. Altemus, H. Chen, T. Crimmins, J. Derbridge, K. Gerst, R. Jessen, J. Kellerman, A. Kilanowski, M. Merrick, and A. Rosemartin provided comments on earlier drafts. S. Barnett, A. Grajal-Puche, V. Greer, M. Merrick, B. Raschke, V. Roundtree, S. Snedecker, A. Williams and especially L. McHugh assisted in the field. M. Borgstrom provided statistical advice. Field efforts were conducted under permits graciously provided by the United States Department of Agriculture Forest Service’s Coronado National Forest, Arizona Game and Fish Department, and United States Fish and Wildlife Service. S1 Dataset. Data used in analyses. (XLSX) S1 Dataset. Data used in analyses. (XLSX) S1 Dataset. Data used in analyses. (XLSX) Author Contributions Conceived and designed the experiments: EEP JLK RJS. Performed the experiments: EEP. Ana- lyzed the data: EEP RJS. Contributed reagents/materials/analysis tools: EEP JLK. Wrote the paper: EEP JLK RJS. Conceived and designed the experiments: EEP JLK RJS. Performed the experiments: EEP. Ana- lyzed the data: EEP RJS. Contributed reagents/materials/analysis tools: EEP JLK. Wrote the paper: EEP JLK RJS. Conclusions In- creased knowledge of the interaction strength between other ecosystem modifiers or larder- hoarders and their environments may help to inform decisions related to forest management and restoration and offer insight on the conservation value of these species. 11 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Influence of Red Squirrel Middens on Abundance of Vertebrates Supporting Information S1 Dataset. Data used in analyses. (XLSX) PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 References 1. Power M, Tilman D, Estes J, Menge B. Challenges in the quest for keystones. Bioscience. 1996; 46: 609–620. 2. Cardinale BJ, Srivastava DS, Duffy JE, Wright JP, Downing A, Sankaran M, et al. Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature. 2006; 443: 989–92. PMID: 17066035 3. Kotliar NB. Application of the new keystone-species concept to prairie dogs: how well does it work? Conserv Biol. 2000; 14: 1715–1721. 4. Parmesan C. Ecological and evolutionary responses to recent climate change. Annu Rev Ecol Syst. 2006; 37: 637–669. 5. Paine RT. A note on trophic complexity and community stability. The Am Nat. 1969; 103: 91–93. 6. Mills LS, Soulé ME, Doak DF. The keystone-species concept in ecology and conservation. Bioscience. 1993; 43: 219–224. 7. Naiman RJ, Melillo JM, Hobbie JE. Ecosystem alteration of boreal forest streams by beaver (Castor canadensis). Ecology. 1986; 67: 1254–1269. 8. Kotliar NB, Baker BW, Whicker AD, Plumb G. A critical review of assumptions about the prairie dog as a keystone species. Environ Manage. 1999; 24: 177–192. PMID: 10384028 9. Prugh LR, Brashares JS. Partitioning the effects of an ecosystem engineer: kangaroo rats control com- munity structure via multiple pathways. J Anim Ecol. 2012; 81: 667–78. doi: 10.1111/j.1365-2656.2011. 01930.x PMID: 22098534 10. Estes JA, Palmisano JF. Sea otters: their role in structuring nearshore communities. Science. 1974; 185: 1058–1060. PMID: 17738247 11. Knapp AK, Blair JM, Briggs JM, Collins SL. The keystone role of bison in North American tallgrass prai- rie. Bioscience. 1999; 49: 39–50. 12. Jones C, Lawton J, Shachak M. Organisms as ecosystem engineers. Oikos. 1994; 69: 373–386. 13. Platt WJ. The colonization and formation of equilibrium plant apecies associations on badger distur- bances in a tall-grass prairie. Ecol Monogr. 1975; 45: 285–305. 14. Whitford WG, Steinberger Y. Pack rats (Neotoma spp.): Keystone ecological engineers? J Arid Envi- ron. 2010; 74: 1450–1455. 15. Vander Wall SB. Food hoarding in animals. Chicago, IL: The University of Chicago Press; 1990. PMID: 17818287 16. Steele MA. Tamiasciurus hudsonicus. Mammal Species. 1998; 586: 1–9. 16. Steele MA. Tamiasciurus hudsonicus. Mammal Species. 1998; 586: 1–9. 12 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Influence of Red Squirrel Middens on Abundance of Vertebrates 17. Gurnell J. The natural history of squirrels. London: Christopher Helm; 1987. 18. Gurnell J. References Home range, territoriality, caching behaviour and food supply of the red squirrel (Tamias- ciurus hudsonicus fremonti) in a subalpine lodgepole pine forest. Anim Behav. 1984; 32: 1119–1131. 19. Merrick MJ, Bertelsen SR, Koprowski JL. Characteristics of Mount Graham red squirrel nest sites in a mixed conifer forest. J Wildlife Manage. 2007; 71: 1958–1963. 20. Smith AA, Mannan RW. Distinguishing characteristics of Mount Graham red squirrel midden sites. J Wildlife Manage. 1994; 58: 437–445. 21. Fitzpatrick LA, Froehlich GF, Johnson TB, Smith RA, Spicer RB. Mount Graham red squirrel recovery plan. U.S. Fish & Wildlife Service; 1993. 22. Zugmeyer CA, Koprowski JL. Habitat selection is unaltered after severe insect infestation: concerns for forest-dependent species. J Mammal. 2009; 90: 175–182. 23. Patton DR, Vahle JR. Cache and nest characteristics of the red squirrel in an Arizona mixed-coniferous forest. West J Appl For. 1986; 1: 48–51. 24. Finley RB. Cone caches and middens of Tamiasciurus in the Rocky Mountain region. Misc. Publ. Uni- versity of Kansas Museum Natural History. 1969; 51: 233–273. 25. Edelman AJ, Koprowski JL, Edelman JL. Kleptoparasitic behavior and species richness at Mt. Graham red squirrel middens. USDA Forest Service Proceedings. 2005; RMRS-P-36: 395–398. 26. Pearson DE, Ruggiero LF. Test of the prey-base hypothesis to explain use of red squirrel midden sites by American martens. Can J Zool. 2001; 79: 1372–1379. 27. Sherburne SS. Squirrel middens influence marten (Martes americana) use of subnivean access points. Am Midl Nat. 1993; 129: 204–207. 28. Mattson DJ, Reinhart DP. Excavation of red squirrel middens by grizzly bears in the whitebark pine zone. J Appl Ecol. 1997; 34: 926–940. 29. Hoffmeister DF. Mammals of Arizona. Tucson, AZ: University of Arizona Press; 1986. 30. Fauteux D, Imbeau L, Drapeau P, Mazerolle MJ. Small mammal responses to coarse woody debris dis- tribution at different spatial scales in managed and unmanaged boreal forests. Forest Ecol Manag. 2012; 266: 194–205. 31. Lee SD. Population dynamics and demography of deermice (Peromyscus maniculatus) in heteroge- neous habitat: role of coarse woody debris. Pol J Ecol. 2004; 52: 55–62. 32. Amaranthus M, Trappe JM, Bednar L, Arthur D. Hypogeous fungal production in mature Douglas-fir for- est fragments and surrounding plantations and its relation to coarse woody debris and animal mycoph- agy. Can J Forest Res. 1994; 24: 2157–2165. 33. Sanderson HR, Koprowski JL. The last refuge of the Mt. Graham red squirrel. Tucson, AZ: University of Arizona Press; 2009. 34. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 References Lemmon PE. A spherical densiometer for estimating forest overstory density. Forest Sci. 1956; 2: 314–321. 35. Anderson DR, Burnham KP, White GC, Otis DL. Density estimation of small-mammal populations using a trapping web and distance sampling methods. Ecology. 1983; 64: 674–680. 36. McDiarmid RW, Wilson DE. Data standards. In: Wilson DE, Cole FR, Nichils JD, Rudran R, Foster MS, editors. Measuring and monitoring biological diversity: standard methods for mammals. Washington, DC: Smithsonian Institution Press; 1996. pp 55–60. 37. Sikes RS, Gannon WL. Guidelines of the American Society of Mammalogists for the use of wild mam- mals in research. J Mammal. 2011; 92: 235–253. 38. Ralph CJ, Geupel GR, Pyle P, Martin TE, DeSante DF. Handbook of field methods for monitoring land- birds. Albany: U.S. Department of Agriculture, Forest Service, Pacific Southwest Research Station. Gen. Tech. Rep. PSW-GTR-144; 1993. 39. Elphick C, Dunning JB, Sibley DA. The Sibley guide to bird life & behavior. New York: Alfred A. Knopf; 2001. 40. Hastings KK, Sydeman WJ. Population status, seasonal variation in abundance, and long-term popula- tion trends of Steller sea lions (Eumetopias jubatus) at the South Farallon Islands, California. Fish B-NOAA. 2002; 100: 51–62. 41. Ramsey FL, Schafer DW. The statistical sleuth: a course in methods of data analysis. Belmont: Brooks/Cole; 2002. 42. Fiske IJ, Chandler RB. unmarked: an R package for fitting hierarchical models of wildlife occurrence and abundance. J Stat Softw. 2011; 43: 1–23. PMID: 22003319 43. Cottee-Jones HEW, Whittaker RJ. The keystone species concept: a critical appraisal. Front Biogeogr. 2012; 4: 117–127. 13 / 14 PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 Influence of Red Squirrel Middens on Abundance of Vertebrates 44. Schauffert CA, Koprowski JL, Greer VL, Alanen MI, Hutton KA, Young PA. Interactions between preda- tors and Mt. Graham red squirrels (Tamiasciurus hudsonicus grahamensis). Southwest Nat. 2002; 47: 498–501. 45. Kendall KC. Use of pine nuts by grizzly and black bears in the Yellowstone area. Int C Bear. 1983; 5: 166–173. 46. Buskirk SW. Seasonal use of resting sites by marten in south-central Alaska. J Wildlife Manage. 1984; 48: 950–953. 47. Koprowski JL, King SRB, Merrick MJ. Expanded home ranges in a peripheral population: space use by endangered Mt. Graham red squirrels. Endangered Species Res. 2008; 3: 227–232. 48. Azeria ET, Fortin D, Lemaître J, Janssen P, Hébert C, Darveau M, et al. PLOS ONE \| DOI:10.1371/journal.pone.0123633 April 29, 2015 References Fine-scale structure and cross- taxon congruence of bird and beetle assemblages in an old-growth boreal forest mosaic. Global Ecol Biogeogr. 2009; 18: 333–345. 49. Carey AB, Johnson ML. Small mammals in managed, naturally young, and old-growth forests. Ecol Appl. 1995; 5: 336–352. 50. Hart EB. Tamias dorsalis. Mammalian Species. 1992; 399: 1–6. 51. McCain CM, Colwell RK. Assessing the threat to montane biodiversity from discordant shifts in temper- ature and precipitation in a changing climate. Ecol Lett. 2011; 14: 1236–45. doi: 10.1111/j.1461-0248. 2011.01695.x PMID: 21981631 52. Edelman AJ. Positive interactions between desert granivores: localized facilitation of harvester ants by kangaroo rats. PLOS ONE. 2012; doi: 10.1371/journal.pone.0030914 53. Romero GQ, Gonçalves-Souza T, Vieira C, Koricheva J. Ecosystem engineering effects on species di- versity across ecosystems: a meta-analysis. Biol Rev. 2014. doi: 10.1111/brv.12138 14 / 14
https://openalex.org/W4200296120	https://journals.us.edu.pl/index.php/TPDJP/article/download/12712/9875	Polish	null	Wstęp	Z Teorii i Praktyki Dydaktycznej Języka Polskiego	2,021	cc-by-sa	808	„Z Teorii i Praktyki Dydaktycznej Języka Polskiego” 2021, t. 30, s. 7—9 ISSN 2353‍‑9577 https://doi.org/10.31261/TPDJP.2021.30.01 „Z Teorii i Praktyki Dydaktycznej Języka Polskiego” 2021, t. 30, s. 7—9 ISSN 2353‍‑9577 https://doi.org/10.31261/TPDJP.2021.30.01 Wstęp Nie ma ważniejszej dziedziny życia społecznego niż edukacja, zwłaszcza dziś, gdy kryzysy natury społecznej, ekonomicznej czy klimatycznej mają cha- rakter globalny i zagrażają utrwalonym zasadom życia społecznego. Edukacja może oczywiście być realizowana jako misja medialna, polityczna czy społecz- na, ale warto podkreślić, że ma ją realizować szkoła. Istnieje ona jako jedna z wielu instytucji powołanych do kształcenia i wychowania młodego pokolenia, co ważne organizacja i struktura szkoły, wyznaczane jej cele i zadania mają wielki wpływ na świadomość, dojrzałość i postawy społeczeństwa. W XXX tomie czasopisma „Z Teorii i Praktyki Dydaktycznej Języka Pol- skiego” przekonujemy o konieczności spojrzenia na zagadnienia społeczne z perspektywy nowej humanistyki. „Nowej” — czyli jakiej? Najogólniej moż- na by określić ją jako tę, która oświetla problemy związane z życiem wszyst- kich istot naszej planety, humanistyki pokazującej człowieka odpowiedzialnego za harmonijny i bezpieczny rozwój wszystkich bytów, a także rezygnującego z jakiejkolwiek dominacji nad kim-/czymkolwiek czy kolonizacji kogo-/czego- kolwiek. Wizja tak zarysowana niesie oczywiście w sobie pierwiastek utopii i w refleksji humanistycznej odczytać można jej świadomość. W sytuacji jednak, gdy zaczynamy zdawać sobie sprawę z tego, że dotychczasowa działalność czło- wieka, jego beztroska w wykorzystywaniu przyrody doprowadziły do zagrożeń pandemicznych, klimatycznych i ekologicznych, trzeba szukać sposobów na za- trzymanie nadciagających katastrof. Nowa humanistyka, rozbudowując reflek- sję literacką i językową o komponenty społeczny, bioetyczny, ekologiczny, nie tylko wprowadza nowe spojrzenie na stare zagadnienia literackie, lecz także odkrywa w nich aktualne problemy. Propozycje rozważania problematyki społecznej w świetle nowej huma- nistyki zamieszczone w artykułach XXX tomu „Z Teorii i Praktyki Dydak- tycznej Języka Polskiego” zmierzają do pokazania, jak można kształcić tole- 8 Wstęp rancję, wrażliwość, empatię w procesie odbioru tekstów literackich i rozmowy o nich z uczniami. Autorzy postulują skupienie uwagi na literaturze współczes- nej. W procesie czytania i dyskusji młodzi ludzie mogą nauczyć się otwartego wyrażania swoich poglądów, emocji, zdobędą umiejętność interpretacji nie tyl- ko utworów literackich, lecz także zjawisk społecznych, kulturowych, politycz- nych. Wprowadzenie do edukacji szerokiego spektrum kwestii związanych z nową humanistyką z jednej strony umożliwi młodym ludziom radzenie sobie z ich własnymi lękami, a z drugiej może przyczynić się do zdobywania przez młodzież takiej świadomości, która stanie się podstawą dla stworze- nia w przyszłości odpowiedzialnego społeczeństwa obywatelskiego. 1 Artykuł Magdaleny Ochwat Edukacja polonistyczna wobec kryzysu klimatycznego — roz- poznania i rekomendacje dydaktyczne, zamieszczony w czasopiśmie „Annales Universitatis Ma- riae Curie-Sklodowska. Sectio N — Educatio Nova” został wyróżniony Nagrodą im. prof. Roma- na Czerneckiego. Nagrodę wręczono autorce 3 listopada 2021 roku. Wstęp Idea kształcenia mądrego, świadomego, odpowiedzialnego za swe wybory społeczeństwa artykułowana jest w wielu współczesnych publikacjach filozo- ficznych i socjologicznych, wystarczy wymienić choćby te Marty Nussbaum, Hannah Arendt, Zygmunta Baumana, a z bliskich naszemu środowisku nauko- wemu takich uczonych, jak Tadeusz Sławek czy Ryszard Koziołek. Aby idea tak rozumianego wychowania obywatelskiego mogła być realizo- wana, konieczna jest wspólnota działania szkoły i akademii. Kwestię tę także pokazują autorzy zamieszczonych w tomie artykułów. Humanistyczna reflek- sja naukowa zrodzona z badań akademickich, artykułowana w pracach Ry- szarda Nycza, Przemysława Czaplińskiego, Julii Fiedorczuk czy Urszuli Za- jączkowskiej, przebija się do szkolnej polonistyki. Dzięki pracom naukowym takich badaczy dydaktyki szkolnej, jak Maria Kwiatkowska-Ratajczak, Krzysz- tof Koc, Anna Janus-Sitarz, Grażyna Tomaszewska, Dariusz Szczukowski, Małgorzata Wójcik-Dudek czy Magdalena Ochwat1, idee nowego myślenia humanistycznego przekładane są na propozycje rozwiązań w obrębie dydak- tyki szkolnej, te zaś są publikowane i stanowią inspirację dla poszukujących nauczycieli. W XXX tomie „Z Teorii i Praktyki Dydaktycznej Języka Polskiego” po- kazano także ważność kształtowania świadomości językowej uczniów w wielu perspektywach. Z jednej strony w perspektywie nauczania języka ojczystego, a z drugiej — w perspektywie nauczania języka jako drugiego — odziedziczo- nego, obcego. Kwestie językowe w kontekście nowej humanistyki coraz częściej stanowią świadectwo artykułowania określonej świadomości etycznej, estetycz- nej czy wreszcie ekologicznej. O etyce słowa już w latach dziewięćdziesiątych ubiegłego wieku pisała Jadwiga Puzynina, kładąc podwaliny pod badania doty- czące języka wartości. Badania te włączone w refleksję nowej humanistyki jesz- cze silniej podkreślają nierozerwalne związki języka i literatury, ale także ujaw- 9 Wstęp niają ukryte czy wręcz ukrywane intencje mówiących (piszących). Kształcenie językowej świadomości, dyskutowanie o języku nie tylko jako o tworzywie literackim, ale i o budulcu wypowiedzi mówionej oraz pisanej, kreatorze obrazu świata wydaje się zatem ważnym elementem edukacji obywatelskiej. Natomiast porównanie badań i zachowań językowych w szkolnej rzeczywistości polskiej i innych krajów, zwłaszcza sąsiednich (słowiańskich), dostarcza ciekawych obserwacji i wniosków dla budowania procesu dydaktycznego. Jubileuszowe wydanie „Z Teorii i Praktyki Dydaktycznej Języka Polskiego” stanowi zwrot w prezentacji tematów — rozszerza je o zagadnienia społeczne, wprowadza konteksty międzynarodowe, pokazuje zainteresowanie polszczyzną w innych obszarach językowych, a także sygnalizuje perspektywę porównaw- czą w nauczaniu języka ojczystego w obszarze tej samej, słowiańskiej grupy. Ewa Jaskółowa orcid.org/0000-0002-4477-9963 Uniwersytet Śląski w Katowicach Ivana Dobrotová orcid.org/0000-0002-1832-8376 Univerzita Palackého v Olomouci, Česká republika
https://openalex.org/W2033988165	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0089821&type=printable	English	null	Zscan4 Is Regulated by PI3-Kinase and DNA-Damaging Agents and Directly Interacts with the Transcriptional Repressors LSD1 and CtBP2 in Mouse Embryonic Stem Cells	PloS one	2,014	cc-by	15,519	Abstract The Zscan4 family of genes, encoding SCAN-domain and zinc finger-containing proteins, has been implicated in the control of early mammalian embryogenesis as well as the regulation of pluripotency and maintenance of genome integrity in mouse embryonic stem cells. However, many features of this enigmatic family of genes are poorly understood. Here we show that undifferentiated mouse embryonic stem cell (ESC) lines simultaneously express multiple members of the Zscan4 gene family, with Zscan4c, Zscan4f and Zscan4-ps2 consistently being the most abundant. Despite this, between only 0.1 and 0.7% of undifferentiated mouse pluripotent stem cells express Zscan4 protein at a given time, consistent with a very restricted pattern of Zscan4 transcripts reported previously. Herein we demonstrate that Zscan4 expression is regulated by the p110a catalytic isoform of phosphoinositide 3-kinases and is induced following exposure to a sub-class of DNA-damage- inducing agents, including Zeocin and Cisplatin. Furthermore, we observe that Zscan4 protein expression peaks during the G2 phase of the cell cycle, suggesting that it may play a critical role at this checkpoint. Studies with GAL4-fusion proteins suggest a role for Zscan4 in transcriptional regulation, further supported by the fact that protein interaction analyses demonstrate that Zscan4 interacts with both LSD1 and CtBP2 in ESC nuclei. This study advances and extends our understanding of Zscan4 expression, regulation and mechanism of action. Based on our data we propose that Zscan4 may regulate gene transcription in mouse ES cells through interaction with LSD1 and CtBP2. Citation: Storm MP, Kumpfmueller B, Bone HK, Buchholz M, Sanchez Ripoll Y, et al. (2014) Zscan4 Is Regulated by PI3-Kinase and DNA-Damaging Agents and Directly Interacts with the Transcriptional Repressors LSD1 and CtBP2 in Mouse Embryonic Stem Cells. PLoS ONE 9(3): e89821. doi:10.1371/journal.pone.0089821 Editor: Michael Kyba, University of Minnesota, United States of America Received September 1, 2013; Accepted January 23, 2014; Published March 3, 2014 Received September 1, 2013; Accepted January 23, 2014; Published March 3, 2014 Copyright: 2014 Storm et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: Funded by grant G0801108 from The Medical Research Council (http://www.mrc.ac.uk/index.htm). BK and MB were supported by the Marie Curie Early Stage Training programme MEST-CT-2005-019822. YSR was the recipient of an MRC Capacity Building Studentship. Abstract The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Competing Interests: The authors have declared that no competing interests exist. . These authors contributed equally to this work. ¤ Current address: Biotechnology and Biological Sciences Research Council, Polaris House, Swindon, United Kingdom via the Jak-Stat3 and Smad-Id pathways respectively [5,6,7]. Inhibition of glycogen synthase kinase 3 (Gsk-3), which mimics both activation of the Wnt pathway and growth factor-induced PI3K signaling, can enhance mouse ESC self-renewal [8,9] and assist in maintaining the ‘ground state’ of mouse ESC pluripotency [10,11,12]. Inhibition of MAPK signaling, in addition to Gsk-3 inhibition (referred to as 2i conditions) is sufficient to maintain self- renewal of mouse ESCs in the absence of additional exogenous factors [10]. Phosphoinoside 3-kinase (PI3K) signaling has also been implicated in the maintenance of both mouse [13,14,15] and human ESC [16] pluripotency. Zscan4 Is Regulated by PI3-Kinase and DNA-Damaging Agents and Directly Interacts with the Transcriptional Repressors LSD1 and CtBP2 in Mouse Embryonic Stem Cells Michael P. Storm1., Benjamin Kumpfmueller1,3., Heather K. Bone1, Michael Buchholz2, Yolanda Sanchez Ripoll1, Julian B. Chaudhuri3, Hitoshi Niwa4, David Tosh2, Melanie J. Welham1¤ 1 Centre for Regenerative Medicine and Departments of Pharmacy & Pharmacology University of Bath Bath United Kingdom 2 Department of Biology and Biochemistry Michael P. Storm1., Benjamin Kumpfmueller1,3., Heather K. Bone1, Michael Buchh Ripoll1, Julian B. Chaudhuri3, Hitoshi Niwa4, David Tosh2, Melanie J. Welham1¤ p 1 Centre for Regenerative Medicine and Departments of Pharmacy & Pharmacology, University of Bath, Bath, United Kingdom, 2 Department of Biology and Biochemistry, University of Bath, Bath, United Kingdom, 3 Department of Chemical Engineering, University of Bath, Bath, United Kingdom, 4 RIKEN Centre for Developmental Biology, Kobe, Hyogo, Japan ation: Storm MP, Kumpfmueller B, Bone HK, Buchholz M, Sanchez Ripoll Y, et al. (2014) Zscan4 Is Regulated by PI3-Kinase and DNA- ectly Interacts with the Transcriptional Repressors LSD1 and CtBP2 in Mouse Embryonic Stem Cells. PLoS ONE 9(3): e89821. doi:10.1371 Generation of GFP-Zscan4c and Zscan4c-V5-His mouse ES cell lines Generation of GFP-Zscan4c and Zscan4c-V5-His mouse ES cell lines An eGFP-Zscan4c fusion protein was generated by amplifying the Zscan4c coding sequence with the primers: forward (59-T- TATGGCTTCACAGCAGGCA) and reverse (59- TTAATTG- CGGCCGCTCAGTCAGATCTGTGGTAAT). The resulting product was cloned into SmaI/NotI digested pTRE-Tight (Clontech) which contained the coding sequence for eGFP. This generated an in-frame N-terminal fusion between eGFP and Zscan4c. To generate ESCs with eGFP-Zscan4c under the control of Tet-on regulation, R1 mESCs containing the Tet-On Advanced Transactivator (Clontech; a kind gift of Giusi Manfredi, University of Bath) were co-transfected by electroporation with pTRE-eGFP- Zscan4c and a vector conferring resistance to Hygromycin B. Clones were selected in 300 mg/ml Hygromycin B, expanded, screened for expression and further characterised (Supplemental Figure S1). The karyotype of clones used for further analyses was checked and confirmed as normal (data not shown). To generate ESCs expressing an inducible c-terminal V5-His epitope tagged version of Zscan4c under the regulation of the Tet-off system, Zscan4c-V5-His was PCR amplified from a pcDNA3.1vector containing the required insert [24] using the primers forward (59- AATGTCGACTAATACGACTCACTATAGGG) and reverse (59-AATGTCGACTAGAAGGCACAGTCGAGG). This prod- uct was blunt-end ligated into the Tet-off response plasmid pUHD10-3 [40] to create the vector pTet-off-Zscan4c-His-V5. To generate inducible Zscan4c-Tet-off ESC lines, linearized pTet-off- Zscan4c-His-V5 was electroporated as previously described [14] into E14tg2A (clone R63) ESCs which constitutively express the tetracycline transactivator (tTA) driven by the CAG promoter [41]. Transfectants were selected in G418 and surviving clones expanded and screened by immunoblotting for the induction of Zscan4c expression after withdrawal of tetracycline. Clones with the highest degree of inducible expression were used in this study and further characterised (Supplemental Figure S2). Clones were routinely cultured in the presence of 1 mg/ml tetracycline to maintain transgene expression switched off, tetracycline was removed to induce Zscan4c-V5-His expression. Recently, members of the Zscan4 family of zinc finger proteins have been identified as important contributors to the maintenance of the pluripotent state of mouse ESCs [24,25]. The Zscan4 family comprises 9 very closely related gene paralogues located on mouse chromosome 7 [24,26]. Full-length Zscan4 proteins comprise an N-terminal SCAN domain of approximately 160 residues and 4 zinc finger motifs, located towards the C-terminus [24,26]. SCAN domains can mediate dimerisation of other Zinc finger-containing proteins [27,28,29], although to date this has not been demon- strated for the Zscan4 family. Generation of GFP-Zscan4c and Zscan4c-V5-His mouse ES cell lines In terms of the functional role of Zscan4 family members, Falco et al., first identified Zscan4d as a gene upregulated during zygotic genome activation in mouse 2-cell stage embryos [26]. Blockade of Zscan4d expression during pre-implantation development led to a delay in 2-cell to 4-cell progression and blastocysts that did develop were unable to implant or proliferate in culture [26]. Interestingly, in mouse ESCs Zscan4 mRNA exhibits a very unusual pattern of expression, restricted to only a small proportion of cells [26,30]. Studies have suggested that Zfp206 and Smad4 play positive and negative roles respectively in regulating Zscan4 expression [31,32]. We first identified Zscan4c as a gene whose expression is down- regulated in mouse ESCs following inhibition of PI3K signaling [24] and demonstrated that knock-down of Zscan4 family members decreased self-renewal of mESCs, consistent with a role for Zscan4 proteins in maintenance of the ESC state [24]. More recently it has been reported that Zscan4 plays a key role in maintaining the stability and integrity of the ESC genome [25]. Furthermore its transient expression can enhance both the efficiency of generation of induced pluripotent stem cells [33] and genome stability during the reprogramming phase [34]. More recently, Zscan4 has also been shown to be capable of restoring developmental pluripotency to embryonic stem cells [35]. These findings suggest that the Zscan4 family plays multiple roles in maintenance of the ESC state. To increase our understanding of the biology of the Zscan4 family of proteins we have investigated the patterns and regulation of expression of Zscan4 genes and proteins in mouse ESCs and probed the mechanisms of action of Zscan4. Here we report that multiple Zscan4 genes are expressed in different ESC lines, including genes previously proposed as pseudogenes. In multiple pluripotent mouse stem cell lines Zscan4 protein expression was consistently found to be highly restricted, with between 0.1 and 0.7% Zscan4-positive cells detected. We also show that Zscan4 protein levels are modulated by PI3K and Gsk-3-dependent signaling, in response to DNA damage and during the G2 phase of the cell cycle. Consistent with a role as a potential regulator of transcription, we demonstrate that Zscan4c modulates transcrip- tion in a heterologous system and that Zscan4 proteins are located in the nucleus, where they interact with components of co- repressor complexes, including LSD1 and CtBP2. Generation of GFP-Zscan4c and Zscan4c-V5-His mouse ES cell lines Based on our current understanding of Zscan4 biology we believe that this gene family represents an intriguing new paradigm for the control of the ESC state. Generation of mouse ES cell lines expressing myristyolated p110a y y p A version of p110a containing an in-fame N-terminal N- myristoyl transferase recognition peptide sequence, MGSSKSKPK, was amplified by PCR from the plasmid pPyCAG-myrp110a-IP [42] and subcloned into the piggyBac vector pPBCAGcHAIN to generate pPB-myr-p110a. One day prior to transfection, 26104 OCRG9 mouse ESCs (Rex1-GFP/ Oct3/4-CFP double knock-in ES cells)/well were plated into 12- well cell culture trays. The following day, 1 mg of pPB-myr-p110a and 1 mg transposase expressing helper plasmid (pCAG-PBase) were mixed in 25 ml of GMEM without serum. 25 ml of diluted Lipofectamine2000 (2 ml Lipofectamine2000 plus 23 ml GMEM) was added to the DNA mix and incubated at room temperature for 10 min. 450 ml of GMEM plus serum was then added and the mixture applied to prepared cells. After 3 hours, the medium was replaced with fresh medium. Selection with G418 was initiated after a further 24 h and colonies generated after 6 days in selection were picked and expanded. Materials and Methods Culture of mouse pluripotent cells Introduction Embryonic stem cells (ESCs) self-renew and are pluripotent, meaning they can differentiate into all cells comprising an adult organism [1]. These properties have made ESCs an attractive source of differentiated cell types for use in both drug discovery and regenerative medicine. While the potential of ESCs has been widely recognized, it is imperative that the mechanisms regulating their self-renewal, pluripotency and stability are better understood, to ensure their efficacy and safety. The extrinsic factors, signaling pathways and transcription factor networks that contribute to maintenance of mouse ESC self- renewal and pluripotency, referred to as the ‘ESC state’, have been extensively studied [1,2,3,4]. Leukaemia inhibitory factor (LIF) and Bone morphogenetic protein 4 (BMP4) are the key cytokines required for maintenance of ESC self-renewal in culture, acting Oct4, Sox2 and Nanog are amongst the most important transcription factors that contribute to regulation of ESC pluripotency, often referred to as the ‘core transcription factors’ or ‘master regulators’ [1,17]. Other transcription factors work in concert with these core factors and include Zfx [18], Klf 2 & 4 March 2014 \| Volume 9 \| Issue 3 \| e89821 1 PLOS ONE \| www.plosone.org Regulation of Expression and Mode of Action of Zscan4 [19], c-Myc [20], Esrrb [21] and Tbx3 [13,21]. In addition, epigenetic regulation has been proposed to play an important role in control of the ESC state [1,22,23]. Generation of transcriptional reporters p p A synthetic GAL4-UAS promoter, containing five GAL4 bindings sites (59-CGGAGTACTGTCCTCCG-39), was cloned into the pGL4.26 luciferase reporter (Promega), using the synthetic sequences: forward 59- CCGGAGTACTGTCCTCCGTACGG- AGTACTGTCCTCCGTATGCCGGAGTACTGTCCTCCGA- TCGGAGTACTGTCCTCCGTATGCCGGAGTACTGTCCT- CCGC & reverse 59-TCGAGCGGAGGACAGTACTCCGG- CATACGGAGGACAGTACTCCGATCGGAGGACAGTACT- CCGGCATACGGAGGACAGTACTCCGTACGGAGGACAG- TACTCCGGGTAC, as shown in Supplemental Figure S3. The trans-activator reporter plasmid, pFA-CMV (Stratagene), was used to create a series of Zscan4c-Gal4 fusion proteins using the primer sequences listed in Supplemental Table S2. A Nanog Gal4-fusion was also generated. After sequence verification of the constructs, transcriptional activity was tested in heterologous Hek293 cells. For transfection, cells were plated at 60,000 cells/well in a 96-well dish and incubated for 24 h in 100 ml DMEM plus 10% (v/v) FBS, which was subsequently replaced with 50 ml of the same medium. Following this, 100 ng pGL4.26 GAL4-UAS, 20 ng pFA-CMV SCAN (or SCAN+ or SCAN+ZnF) and 20 ng phRL- TK (Renilla plasmid to control for transfection efficiency) were combined in 12.5 ml Optimem and incubated at room tempera- ture for 5 min. The DNA was then combined with diluted Lipofectamine 2000 (0.5 ml in 12.5 ml Opti-Mem) and incubated for 20 min before addition to cells. After 24 h luciferase detection was carried out using Dual-Glo Luciferase Assay System (Promega – E2920). After subtraction of background fluorescence for Renilla and Firefly luciferase, the ratio of reporter (firefly) luminescence to control (Renilla) luminescence was calculated and values were normalized relative to Gal DBD response alone – this value represents the Relative Response Ratio. Cytosolic or nuclear cell extracts were prepared as described previously [44] and protein concentrations determined. To generate total cell extracts RIPA lysis buffer (150 mM NaCl, 50 mM TrisHCl pH8, 1% (v/v) NP40, 0.5%(w/v) Na Deoxyco- late, 0.1% (w/v) SDS, 25 U/ml Benzonase, 1 mM sodium vanadate, 1 mM sodium molybdate, 10 mM sodium fluoride, 40 mg/ml PMSF, 0.7 mg/ml Pepstatin , 10 mg/ml Aprotinin, 10 mg/ml Leupeptin, 10 mg/ml Soyabean trypsin inhibitor) was used. For immunoprecipitation, equal amounts of protein (cytosolic or nuclear extracts) were pre-cleared with protein A sepharose beads and then precipitated with either Nanotrap beads (30 ml of 50% (v/v) slurry, Chromatek), pre-immune serum (5 ml), anti-panZscan4 antiserum (5 ml), anti-LSD-1 (2 mg per sample, Abcam, ab37165), anti-CtBP2 (2 mg per sample, BD Transduction Labs, Cat No. Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 University of Bath, UK). For cell cycle analyses by flow cytometry, cells were trypsinised and fixed with 70% (v/v) ice-cold ethanol. After washing with PBS and re-hydration for 10 min, cells were incubated with 7-AAD (0.5 mg/ml) for 1.5 h at 4uC. A minimum of 10,000 cell events were acquired using FACS Canto flow cytometer (Becton-Dickinson) and analysed using FACS Diva software. Alternatively, live cells were stained with Nuclear-ID Red according to the manufacturer’s instructions (Enzo Life Sciences) and a minimum of 140,000 cell events were analysed as above. Cloning and sequencing of Zscan4 transcripts RNA was isolated and RT-PCR performed as above using the following primers (Forward 59-ACAATGGCTTCACAGCAGG, Reverse 59-ACGATGGTAAGTGGATGATTGG) and limited amplification (22 cycles). PCR products were TOPO cloned into pcDNA3.1 (Invitrogen). Clones were picked, plasmid DNA extracted and 96 insert verified clones were sequenced using the primer 59-CATCCTAGAACATTCTTCACAC. All primers used bind to all 9 Zscan4 paralogs with equal efficiency. Sequences were analysed using Sequencher (GeneCodes). Affinity purification of Zscan4 interacting proteins GFP-Trap-A beads (Chromotek) provide a means of Affinity purification of Zscan4 interacting proteins GFP-Trap-A beads (Chromotek) provide a means of high affinity purification in a single step, facilitating rapid precipitation and the potential to capture less stable interactions. Following induction of eGFP-Zscan4 expression by addition of Dox to GFP- Zscan4c ESCs for 24 h, cells were harvested, washed in PBS and cell pellets resuspended in cytosolic extraction buffer (20 mM HEPES, 10 nM KCl, 1 mM EDTA, 10% (v/v) Glycerol, 1 mM PMSF, 1 mg/ml Aprotinin, 1 mg/ml leupeptin, 1 mg/ml pepstatin, 5 mg/ml antipain, 157 mg/ml benzamidine, 5 mM b-glycerophos- phate, 5 mM sodium fluoride, 1 mM sodium orthovanadate, pH 7.9). After a 30 min incubation at 4uC, the suspension was drawn through a 27G needle ten times and centrifuged at 1630006 g for 5 min at 4uC. The supernatant was removed and designated ‘cytosol’. The remaining nuclear pellets were washed Generation of transcriptional reporters 612044) or anti-V5 epitope (2 mg per sample, Abcam ab27671) antibodies and immune complexes captured on protein A or G sepharose beads, prior to extensive washing with nuclear extraction buffer and boiling in Laemmli buffer. For immunoblotting, 20 mg of each protein sample or the entire immunoprecipitate was separated by SDS-PAGE and transferred to nitrocellulose as previously described [45]. Immu- noblotting was carried out with the following primary rabbit polyclonal antibodies: 1:4000 anti-panZscan4 (Merck Millipore, AB4340); 1:2000 LSD-1 (Abcam, ab37165); 1:5000 GFP (MBL, 598), or mouse monoclonal antibodies at 1:2000 anti-CtBP2 (BD Transduction Laboratories, 612044), 1:5000 anti-V5 epitope (Abcam, ab27671), 1:1000 c-H2AX (Merck Millipore clone JBW301) or 1:20000 anti-GAPDH (Ambion, AM4300). Anti- mouse or anti-rabbit secondary antibodies conjugated to horse- radish peroxidase (DAKO) were used for detection and blots were developed using ECL Prime according to the manufacturer’s instructions (GE Healthcare). Protein relative quantification was carried out using ImageQuant RT-ECL imager and analysed using ImageQuant TL software (GE Healthcare). Blots were stripped and reprobed as previously described [45]. RNA isolation and RT-PCR The mouse ES cell lines E14tg2a [36], E14 Gsk-3 double knock- out (DKO; [37]), CCE, CGR8, IOUD2, R1 and ZE3-MC-1 [25] and the mouse iPS cell line [38] were cultured as described previously [8,14,39]. RNA was isolated and RT-PCR performed as previously described [43]. The list of primers used in this study are summarised in Table S1. March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 2 Regulation of Expression and Mode of Action of Zscan4 Generation of mouse pan-Zscan4 antibodies Polyclonal rabbit antisera were generated against a Zscan4- specific peptide (GVPQDSTRASQGTSTC, amino acids 322– 337; anti-pan Zscan4) by Millipore/Merck according to company procedures. Specificity for Zscan4 was confirmed by immunoblot- ting in the presence and absence of blocking peptide and immunoprecipitation (Figure S4). Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 twice with ice cold cytosolic extraction buffer before being resuspended in nuclear extraction buffer (20 mM HEPES, 10 nM KCl, 400 mM NaCl, 1 mM EDTA, 20% (v/v) Glycerol, 1 mM PMSF, 1 mg/ml Aprotinin, 1 mg/ml leupeptin, 1 mg/ml pepstatin, 5 mg/ml antipain, 157 mg/ml benzamidine, 5 mM b- glycerophosphate, 5 mM sodium fluoride, 1 mM sodium ortho- vanadate, pH 7.9). Following rotation at 4uC for 1 h samples were centrifuged at 1630006 g for 15 min. The supernatant was collected and designated as the nuclear extract. Cytosolic and nuclear fractions were cleared by centrifugation at 60,0006 g for 30 minutes at 4uC. A subsequent buffer exchange to immunopre- cipitation (IP) buffer (50 mM Na2HPO4, 150 mM NaCl, pH 7.4) was performed using Amicon centrifugal filters with a 10 kDa cut- off according to the manufacturer’s protocol. Prior to immuno- precipitation, lysates were pre-cleared with hydrated sepharose beads for 1 h at 4uC on a rolling shaker. IPs were performed with 500 ml GFP-Trap-A beads (Chromotek) for 1–3 hours on a rolling shaker at 4uC, followed by three washes with IP buffer. Bound protein was eluted with 200 mM glycine at pH 2.5 and 1 M Tris- base (pH 10.4) was added for neutralization. The eluates were centrifuged at 1000 rpm for 1 minute and the supernatant transferred to Amicon centrifugal filters with a 3 kDa cut-off for concentration, according to the manufacturer’s recommendations. SDS loading buffer was added and samples were boiled for 5 minutes. Proteins purified from uninduced and induced-derived cell extracts were separated by large format SDS-PAGE, stained with Coomassie blue and proteins enriched in the samples prepared from induced cells were excised from the gel and submitted for protein sequencing at The University of Bristol Proteomics Facility (http://www.bristol.ac.uk/biochemistry/ proteomics/services.html). Automated in-gel tryptic digestion was performed using a ProGest unit and resulting peptides analysed by reverse-phase LC MSMS using an LTA-Orbitrap Velos mass spectrometer. The mass spectral data from each fraction was combined prior to database searching using Mascot to identify the proteins present in the sample. Protein scores greater than 64 were classed as significant (p,0.05). In addition, MSMS analyses generated sequence data for a number of peptides. Following these analyses and comparison to the mouse protein sequence database, sequences were returned for 3 out of 4 proteins. Results Members of the Zscan4 family of genes have been implicated in the regulation of pluripotency and genome stability, both functions that impact on maintenance of the stem cell state of mouse ESCs [24,25]. Two enigmatic features of the Zscan4 family are the existence of 9 highly related gene paralogues and a very restricted pattern of expression. These features suggest Zscan4 family members may regulate the ESC state by novel mechanisms. Our aim was to investigate the underlying basis of the restricted expression pattern observed and further examine the mechanism of action of Zscan4 in ESCs. Regulation of Expression and Mode of Action of Zscan4 ensuring that the most relevant are studied further, we investigated the patterns of expression of individual Zscan4 genes in two distinct mouse ESC lines. The very high sequence homology of Zscan4 paralogues precluded a PCR-based approach to quantify expression of individual genes and instead ESC cDNA was subjected to a limited number of amplification cycles, using primers that hybridise equally to transcripts from all 9 Zscan4 genes, prior to direct cloning into TOPO vectors. 96 clones were selected at random for DNA sequencing and each individual sequence returned was assigned to one of the 9 Zscan4 genes based on information in the most recent build of the mouse genome on NCBI (NCBIm37), as described in Supplemental Figure S5. Figure 1A shows the relative expression of each Zscan4 transcript in E14 and CCE ESCs, compared to the 129.3 ESC line reported previously [26]. Zscan4c, 4f and ps2 were the most abundant transcripts detected in E14 and CCE ESCs. The most striking difference observed was in the expression of Zscan4-ps2 in both E14 and CCE, compared to its absence in 129.3. Interestingly, phylogenetic analyses of the 9 Zscan4 paralogues (Figure 1B), suggests that Zscan4c, 4f and ps2 are the most closely related family members. In view of this, it is interesting that these 3 genes should consistently be the most abundant transcripts detected, suggesting they constitute the most functionally relevant Zscan4 genes in ESCs. We also detected expression of Zscan4-ps1 and Zscan4-ps3, suggesting their designation as pseudogenes, based on lack of expression, is no longer correct in the light of this new analysis. Restricted patterns of Zscan4 expression are a feature of multiple pluripotent mouse stem cell lines To date, whole mount in situ hybridization of ESC colonies [26] and Zscan4c-based promoter-reporter systems [25] have suggested that expression of Zscan4 transcripts is restricted to between 3 and 5% of cells in a population of undifferentiated mouse ESCs at any one time. We wanted to investigate this highly unusual pattern of expression further and, in particular, determine whether Zscan4 protein shows a similarly restricted pattern of expression to Zscan4 RNAs. To achieve this aim we generated a pan-Zscan4 anti- peptide polyclonal antibody, which detects Zscan4 in immunoblots and is specifically blocked by the immunizing peptide (Supple- mental Figure S4A). We initially used ZE3-MC1 ESCs, that express Emerald-GFP under the control of a 2.6 kb fragment of the Zscan4c promoter [25], to measure the correlation between Zscan4 transcriptional reporter and protein expression. As shown in Figure 2A (i) a robust correlation between detection of Emerald- GFP and staining with the Zscan4 antibody was observed, with Zscan4 protein being localized primarily in the nucleus (Figure 2A(ii)). Having established concordance between the Zscan4 reporter and protein expression, we analysed the proportion of Zscan4 positive cells present in populations of undifferentiated mouse ESC lines and a mouse iPSC line. Our data, shown in Figure 2B, demonstrate that the proportion of cells staining positive for Zscan4 protein is between 0.1 and 0.7% (average of 0.33%) lower than the estimated number of cells expressing Zscan4 transcripts reported previously [25,26]. Even with the ZE3-MC1 reporter cell line generated and used by Zalzman and colleagues [25], we consistently detected 10-fold fewer Zscan4 positive cells than reported previously and we obtained similar results when the expression was measured using flow cytometry (average of 0.43% positive cells, S.E.M 0.03, n = 3, events = 140,000–478,000). Despite these differences, our studies clearly demonstrate that Zscan4 protein expression, like Zscan4 transcripts, exhibits a highly restricted pattern of expression in a range of undifferentiated mouse pluripotent cells, including iPSCs. Immunofluorescence and flow cytometry Cells were fixed with 4% (w/v) paraformaldehyde for 20 min, permeablised in 1% Triton X-100 for 20 min, blocked for 1 h in 1% blocking reagent (Roche), washed and incubated with 1:500 dilution of anti-Zscan4 antisera. After washing, a 1:100 dilution of anti-rabbit-Texas Red antibody (Vector Laboratories) was added. After a second round of washing DAPI was added for 20 min. Samples were again washed and mounted under a coverslip with fluorescence mounting medium (DAKO). Fluorescence was visualised using Leica DMI 4000B fluorescence microscope or Zeiss LSM Meta Confocal Microscope (Bioimaging Suite, March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 3 Zscan4 gene expression is regulated by the p110a isoform of phosphoinositide-3 kinases p p We originally identified Zscan4c as a gene that was rapidly down-regulated upon inhibition of PI3K signaling in mouse ESCs [24]. In this earlier study we had used the broad selectively PI3K inhibitor, LY294002, which targets most PI3K isoforms. Owing to the significant interest in the PI3K signaling pathway as a drug target, small molecule inhibitors, which show selectivity for specific PI3K isoforms, have become available [47]. Mouse ESCs express three of the class 1 PI3K catalytic isoforms, p110a, p110b and p110d [48]. Therefore, we investigated the coupling of specific p110 PI3K catalytic subunit isoforms to regulation of Zscan4 expression, initially using qRT-PCR. As shown in Figure 3A, inhibition with LY294002 led to a reduction of approximately 80% in Zscan4 transcripts, consistent with previous results [24]. Inhibition of p110b isoform with either TGX121 (Fig. 3A(i)) or TGX221 (Fig. 3A(ii)) or inhibition of the p110d isoform with IC87114 (Fig. 3A(i)) did not alter Zscan4 RNA levels. In contrast, inhibition of p110a with PIK75 (Fig. 3A (ii)), which is among the most selective p110a inhibitors available to date [49], led to a reduction in Zscan4 expression to the same extent to that observed with LY294002. An un-related p110a inhibitor, compound 15e, also led to a reduction in Zscan4 expression, while the mTOR inhibitor, Rapamycin, had no effect (Supplemental Figure S6). Treatment with LY294002 or PIK-75 also resulted in a significant (3–4-fold) reduction in the number of Zscan4 positive cells present within the treated population (Fig. 3B) and these data were further confirmed by immunoblot analyses (data not shown). To further investigate the regulation of Zscan4 expression by the p110a PI3K isoform, we over-expressed a myristyolated version of p110a (myr- p110a) under the control of the CAG promoter [50] in OCRG9 mESCs. Zscan4 expression was analysed by qRT-PCR in three independent clones over-expressing myr-p110a. Consistent with our loss of function data, expression of p110a led to a significant up-regulation in Zscan4 expression (Figure 3D) and this was observed in cells grown with or without LIF for 4 days. Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 In further analyses we examined whether there was any correlation between Zscan4 protein expression and the master regulator, Nanog, known to exhibit fluctuations in expression [46]. Using an ESC line expressing GFP under the control of the endogenous Nanog promoter (Nanog-GFP; [46]) no clear correlation between Zscan4 expression and levels of Nanog was observed (Figure 2C). Figure 1. Expression patterns of Zscan4 gene paralogues in mouse ESC lines. A. The mouse ESC lines E14 or CCE were cultured in the presence of KO serum replacement and LIF, RNA was extracted, limited amplification by RT-PCR was performed and fragments cloned. 96 independent clones were sequenced and categorised as represent- ing one of the 9 mouse Zscan4 paralogues present in the NCBIm37 database based on SNPs. The relative expression of each Zscan4 paralogue, as a percentage of the total number of Zscan4 gene sequences obtained for E14 and CCE ESCs, are shown in comparison with the 129.3 ESC line, data derived from that presented in [26]. B. The suggested evolution of Zscan4 genes in the mouse based on sequence variation. Dendrogram (Unweighted Pair-Groups Method using Aver- ages) showing genetic relationship between Zscan4 paralogs based on Nei’s [1972] original distance with 1000 bootstrap replications (confidence values given at each node). The dendrogram was produced using Tools for population genetic analysis (www.marksgeneticsof tware.net). doi:10 1371/journal pone 0089821 g001 Multiple Zscan4 gene transcripts are expressed in self- renewing mouse ESCs Individual Zscan4 genes have been reported to be differentially expressed in 2-cell stage mouse embryos and mouse ESCs, with three members, Zscan4-ps1, Zscan4-ps2 and Zscan4-ps3, pro- posed as pseudogenes [26]. In contrast, we have previously detected Zscan4-ps2 transcripts in undifferentiated E14 ESCs [24]. Given this discrepancy, and the importance of knowing which Zscan4 genes are most highly expressed in ESC, thereby March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 4 Involvement of Gsk-3-dependent signaling in regulation of Zscan4 expression Expression of the Emerald reporter is shown (Green) and cells were counterstained with DAPI (Blue). (i) Scale bars = 100 mm. (ii) Scale bars = 5 mm. B. The mouse ESC and iPSC lines indicated were cultured in the presence of serum replacement plus LIF and subsequently fixed and immunostained with the pan- Zscan4 antibody, while nuclei were counter-stained with DAPI. The proportion of Zscan4 positive cells present were determined by counting several thousand cells for each cell line sample. Data represent average of three independent replicates 6 SEM, unless otherwise indicated. C. A colony of Nanog-GFP reporter cells stained with the anti-Zscan4 antibody. (i) The GFP reporter (Green), Zscan4 (Red) and nuclei (DAPI, Blue) are shown, along with a merged image. (ii) Nanog-GFP cells were stained for Zscan4 protein. The levels of Nanog-GFP and Zscan4 protein expression were calculated using Cell-P (Olympus) using fluorescent images captured using an Olympus IX51 epifluorescence microscope. Mean fluorescence intensity for GFP (Nanog) was compared between all cells, Zscan4 expressing (+ve) and non-expressing (2ve) cells. doi:10.1371/journal.pone.0089821.g002 Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 Figure 2. Zscan4 protein expression is highly restricted in mouse ESC and iPSC lines. A. The ZE3-MC1 Zscan4 reporter line [25] was cultured in serum replacement plus LIF, cells were fixed and immunostaining was carried out with pan-Zscan4 antibodies (Red). Expression of the Emerald reporter is shown (Green) and cells were counterstained with DAPI (Blue). (i) Scale bars = 100 mm. (ii) Scale bars = 5 mm. B. The mouse ESC and iPSC lines indicated were cultured in the presence of serum replacement plus LIF and subsequently fixed and immunostained with the pan- Zscan4 antibody, while nuclei were counter-stained with DAPI. The proportion of Zscan4 positive cells present were determined by counting several thousand cells for each cell line sample. Data represent average of three independent replicates 6 SEM, unless otherwise indicated. C. A colony of Nanog-GFP reporter cells stained with the anti-Zscan4 antibody. (i) The GFP reporter (Green), Zscan4 (Red) and nuclei (DAPI, Blue) are shown, along with a merged image. (ii) Nanog-GFP cells were stained for Zscan4 protein. The levels of Nanog-GFP and Zscan4 protein expression were calculated using Cell-P (Olympus) using fluorescent images captured using an Olympus IX51 epifluorescence microscope. Involvement of Gsk-3-dependent signaling in regulation of Zscan4 expression We next investigated whether other pathways implicated in the coordination of ESC self-renewal and pluripotency also control Zscan4 expression. Inhibition of Gsk-3 has been implicated in the maintenance of ESC self-renewal [8,9], as well as the ground state of ESC pluripotency [10] and our previous work has demonstrated that short-term treatment with the Gsk-3 inhibitor BIO leads to a small, but significant increase in Zscan4 RNA expression [24]. Here we investigated Zscan4 expression in ESCs in which all 4 Gsk-3 alleles have been knocked out (DKO; [37]) compared to their wild-type parental cells incubated with or without the more selective Gsk-3 inhibitor, 1 m [8]. Interestingly, levels of Zscan4 protein were much higher in Gsk-3 double knock-out ESCs, both in the presence and absence of LIF (Fig. 4A and C), when compared to WT ESCs, while treatment of WT ESCs with 1 m for 24 or 48 h had little influence on Zscan4 expression (see Fig. 4A and B). These data suggest long-term inactivation of Gsk-3 leads to Figure 1. Expression patterns of Zscan4 gene paralogues in mouse ESC lines. A. The mouse ESC lines E14 or CCE were cultured in the presence of KO serum replacement and LIF, RNA was extracted, limited amplification by RT-PCR was performed and fragments cloned. 96 independent clones were sequenced and categorised as represent- ing one of the 9 mouse Zscan4 paralogues present in the NCBIm37 database based on SNPs. The relative expression of each Zscan4 paralogue, as a percentage of the total number of Zscan4 gene sequences obtained for E14 and CCE ESCs, are shown in comparison with the 129.3 ESC line, data derived from that presented in [26]. B. The suggested evolution of Zscan4 genes in the mouse based on sequence variation. Dendrogram (Unweighted Pair-Groups Method using Aver- ages) showing genetic relationship between Zscan4 paralogs based on Nei’s [1972] original distance with 1000 bootstrap replications (confidence values given at each node). The dendrogram was produced using Tools for population genetic analysis (www.marksgeneticsof tware.net). doi:10.1371/journal.pone.0089821.g001 March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 5 Figure 2. Zscan4 protein expression is highly restricted in mouse ESC and iPSC lines. A. The ZE3-MC1 Zscan4 reporter line [25] was cultured in serum replacement plus LIF, cells were fixed and immunostaining was carried out with pan-Zscan4 antibodies (Red). Zscan4 expression is induced in response to DNA damage da age Zalzman et. al., [25] reported that Zscan4 is key to maintaining the genomic integrity of mouse ESCs. We predicted that if this were the case, then Zscan4 expression may be sensitive to the effects of DNA damage-inducing agents, so we examined whether Zeocin, which induces double strand breaks, or Cisplatin, an alkylating-like agent that induces DNA cross-links leading to double strand breaks via stalled replication forks, affect expression of Zscan4. As shown in Fig. 5A, a 12 h treatment with Zeocin followed by wash-out and incubation for a further 8 and 24 h led to a dose-dependent increase in Zscan4 expression (Fig. 5A (i) and (ii)). In the case of Cisplatin treatment (12 h), a notable enhancement of Zscan4 expression was detected 24 h following wash-out (Figure 5B (i) and (ii)). Induction of cH2AX and the proportion of cells in G2/M phase of the cell cycle were measured to assess the cellular response to these agents (Fig. 5 A and B, panels (i) and (iii)), which showed enhanced cH2AX phosphory- lation and accumulation of cells in G2/M phase. Not only were levels of Zscan4 protein enhanced following exposure to Zeocin, but treatment with Zeocin also led to an increase in the proportion of ESCs expressing detectable Zscan4 protein (Fig. 5C(i)), a response partly attenuated by the PI3K inhibitor LY294002 (Fig. 5C(ii)). While the absolute frequency of Zscan4 positive cells varied, due to variation in levels in control populations, we consistently observed a 4–10-fold increase in Zscan4 positive cells following exposure to Zeocin. Phosphorylation of the histone H2AX (to generate cH2AX) is an indicator of DNA damage- induced DNA double strand break and basal levels are known to be elevated in mouse ESCs [51]. However, cH2AX levels are further elevated in ESCs in response DNA damage (see Fig. 5A(i) and B(i)), so we examined the relationship between Zscan4 expression and levels of cH2AX. We found levels of cH2AX were higher in Zscan4 positive cells both under basal (control) conditions and also following treatment with Zeocin (Fig. 5D), consistent with DNA damage corresponding to enhanced Zscan4 expression. Figure 3. Zscan4 gene expression is regulated by the p110a isoform of PI3Ks. A. Regulation of Expression and Mode of Action of Zscan4 Having established that long-term inhibition of Gsk-3 results in enhanced levels of Zscan4 protein we wondered whether Gsk-3 plays a role in regulating the stability of Zscan4 protein, as it does for other regulators of pluripotency, including ß-catenin and c-Myc. To assess protein half-life, WT ESCs incubated with or without 1 m (Gsk-3 inhibitor) (Figure 4B) or DKO (Figure 4C) ESCs were treated with cycloheximide for 1 to 6 h. These data suggest that Zscan4 protein has a half-life of between 3–6 h, which is not notably prolonged upon disruption of Gsk-3 activity. Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 Mean values are shown with standard deviations (n = 4). , p,0.005, , p,0.0005 in a Student’s t-test. B. ESCs were cultured in LIF and treated with either DMSO as a control or with 5 mM LY294002 (LY), 10 nM PIK75, 50 nM TGX221 (TGX), or 5 mM IC87114 for 48 h prior to immunostaining for Zscan4. The mean percentage of Zscan4 positive cells with SEM are shown. , p,0.05, , p,0.005 following an ANOVA and Tukey’s post-hoc test. C. ESC clones over-expressing myristoylated p110a catalytic subunit of PI3Ks were cultured in the presence or absence of LIF. As a control parental OCRG9 ESCs were grown in presence and absence of LIF for 4 days. Expression of Zscan4 was analysed by qRT-PCR and Zscan4 expression normalised to levels of GAPDH. The averages and SEM of triplicate samples from each of three independent biological replicates are shown: , p,0.0005, in a Student’s t-test. Figure 3. Zscan4 gene expression is regulated by the p1 isoform of PI3Ks. A. (i) and (ii) E14 ESCs were cultured in presence of LIF, with the addition of DMSO (control) or the inhib indicated: 5 mM LY294002 (broad spectrum PI3K inhibitor), 10 TGX121 (p110b), 5 mM IC87114 (p110d), 25 nM PIK75 (p110a), 100 TGX221 (p110b), for 48 h. RNA was extracted and levels of Zsc expression analysed by qRT-PCR and normalised to levels of b-a doi:10.1371/journal.pone.0089821.g003 doi:10.1371/journal.pone.0089821.g003 up-regulation of Zscan4 expression and consistent with this we find that culture of WT ESCs in 1 m for 14 days leads to a corresponding increase in Zscan4 RNA expression (data not shown). Having established that long-term inhibition of Gsk-3 results in enhanced levels of Zscan4 protein we wondered whether Gsk-3 plays a role in regulating the stability of Zscan4 protein, as it does for other regulators of pluripotency, including ß-catenin and c-Myc. To assess protein half-life, WT ESCs incubated with or without 1 m (Gsk-3 inhibitor) (Figure 4B) or DKO (Figure 4C) ESCs were treated with cycloheximide for 1 to 6 h. These data suggest that Zscan4 protein has a half-life of between 3–6 h, which is not notably prolonged upon disruption of Gsk-3 activity. up-regulation of Zscan4 expression and consistent with this we find that culture of WT ESCs in 1 m for 14 days leads to a corresponding increase in Zscan4 RNA expression (data not shown). Involvement of Gsk-3-dependent signaling in regulation of Zscan4 expression Mean fluorescence intensity for GFP (Nanog) was compared between all cells, Zscan4 expressing (+ve) and non-expressing (2ve) cells. doi:10.1371/journal.pone.0089821.g002 March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 6 Zscan4 expression is elevated in late S-phase/early G2 of the cell cycle Zscan4 protein. As shown in Figure 6B(i–iii) approximately 90% of cells were arrested in G2/M by Nocodazole treatment (100 ng/ ml). 3 h after release many cells had re-entered G1 (Fig. 6B(ii)), after 6–9 h cells had started to transit through S-phase (Fig. 6B(iii)) and by 12 h were entering G2/M, all consistent with the reported pattern of the ESC cell cycle [52]. Interestingly, Nocodazole treatment alone did not lead to an increase in Zscan4 expression and instead Zscan4 protein levels remained at a constant level until 12 h, when they rose by approximately 4-fold, with the rise continuing to 18 h (Figure 6B(iv)). These data indicate that Zscan4 expression increases as cells transit from late S-phase and into G2 and implicate a selective role for Zscan4 during the G2 checkpoint. Zscan4 protein. As shown in Figure 6B(i–iii) approximately 90% of cells were arrested in G2/M by Nocodazole treatment (100 ng/ ml). 3 h after release many cells had re-entered G1 (Fig. 6B(ii)), after 6–9 h cells had started to transit through S-phase (Fig. 6B(iii)) and by 12 h were entering G2/M, all consistent with the reported pattern of the ESC cell cycle [52]. Interestingly, Nocodazole treatment alone did not lead to an increase in Zscan4 expression and instead Zscan4 protein levels remained at a constant level until 12 h, when they rose by approximately 4-fold, with the rise continuing to 18 h (Figure 6B(iv)). These data indicate that Zscan4 expression increases as cells transit from late S-phase and into G2 and implicate a selective role for Zscan4 during the G2 checkpoint. During our analyses of Zscan4 protein expression by immuno- staining we noted that positive cells often appeared in close proximity to each other (see Figure 2). Enumerating the proportion of Zscan4 positive cells appearing as single cells, doublets, triplets or quadruplicates (shown in Figure 6A(i)), revealed that approximately 45% of Zscan4 positive cells occur in close proximity to at least one other Zscan4 positive cell. This prompted us to investigate whether Zscan4 proteins were expressed preferentially by cells just prior to or during mitosis. Using the ZE3-MC1 reporter line we investigated the cell cycle distribution profiles of Zscan4 positive and negative cells. As shown in Figure 6A(ii), of all Zscan4 positive cells, approximately 45% were in G2/M phase of the cell cycle, very consistent with our proximity data. Zscan4 expression is induced in response to DNA damage (i) and (ii) E14 ESCs were cultured in the presence of LIF, with the addition of DMSO (control) or the inhibitors indicated: 5 mM LY294002 (broad spectrum PI3K inhibitor), 10 mM TGX121 (p110b), 5 mM IC87114 (p110d), 25 nM PIK75 (p110a), 100 nM TGX221 (p110b), for 48 h. RNA was extracted and levels of Zscan4 expression analysed by qRT-PCR and normalised to levels of b-actin. March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 7 Regulation of Expression and Mode of Action of Zscan4 Figure 4. Long-term inhibition of Gsk-3 signalling leads to up-regulation of Zscan4 protein expression. A. Wild-type (WT) and Gsk-3 double knock-out (DKO) ESCs cells were grown in the presence or absence of LIF for the times indicated. WT ESCs cultured in the absence of LIF were also incubated with 2 mM 1 m. Protein and RNA were extracted at the times indicated and immunoblotting performed with the indicated antibodies. B. WT mESCs (CTL) cultured for 24 h in N2B27 plus LIF and BMP4 with or without 2 mM 1 m, or C. WT or DKO ESCs cultured in N2B27 plus LIF and BMP4, were incubated with Cycloheximide (CHX) to halt protein synthesis. (i) Protein samples were extracted after 0, 1, 3 and 6 hours CHX treatment and immunoblotting performed with the indicated anti-Zscan4 or GAPDH antibodies. (ii) Zscan4 protein levels were normalised to GAPDH and a value of 100 was given to the t = 0 samples, to allow direct comparison of half-life values between treatments. The graphs show the average and S.E.M of triplicate experiments. doi:10.1371/journal.pone.0089821.g004 Figure 4. Long-term inhibition of Gsk-3 signalling leads to up-regulation of Zscan4 protein expression. A. Wild-type (WT) and Gsk-3 double knock-out (DKO) ESCs cells were grown in the presence or absence of LIF for the times indicated. WT ESCs cultured in the absence of LIF were also incubated with 2 mM 1 m. Protein and RNA were extracted at the times indicated and immunoblotting performed with the indicated antibodies. B. WT mESCs (CTL) cultured for 24 h in N2B27 plus LIF and BMP4 with or without 2 mM 1 m, or C. WT or DKO ESCs cultured in N2B27 plus LIF and BMP4, were incubated with Cycloheximide (CHX) to halt protein synthesis. Zscan4 expression is elevated in late S-phase/early G2 of the cell cycle Furthermore, the mean fluorescence intensity of Emerald-GFP was highest in cells in G2/M (Figure 6A(iii)), consistent with Zscan4 expression being elevated in cells during these cell cycle phases. To examine this correlation further, we induced mitotic arrest of ESCs using Nocodazole and after release followed both progression through the cell cycle and expression of Zscan4 expression is induced in response to DNA damage (i) Protein samples were extracted after 0, 1, 3 and 6 hours CHX treatment and immunoblotting performed with the indicated anti-Zscan4 or GAPDH antibodies. (ii) Zscan4 protein levels were normalised to GAPDH and a value of 100 was given to the t = 0 samples, to allow direct comparison of half-life values between treatments. The graphs show the average and S.E.M of triplicate experiments. doi:10.1371/journal.pone.0089821.g004 Regulation of Expression and Mode of Action of Zscan4 These data are consistent with Zscan4c playing a direct role in regulating transcription and led us to undertake studies to define Zscan4-interacting proteins in the nucleus of ESCs, which could assist in further definition of the mode of action of Zscan4. regulators [53] and while over-expression of Zscan4 has been shown to lead to changes in expression of ,1000 genes [54], there is no evidence that it acts directly. To examine the ability of Zscan4 to act as a regulator of transcription, we used the GAL4- UAS-luciferase reporter system and generated a series of GAL4- DNA binding domain-Zscan4c fusions, depicted in Fig. 7A. As a control for this assay, we also generated a GAL4-DBD-Nanog fusion. Transient transfection into heterologous HEK293 cells was performed with combinations of the GAL4-UAS-luciferase reporter (containing a synthetic GAL4-UAS promoter, see supplemental Figure S3), GAL4-DBD fusions and Renilla plasmid to control for transfection efficiency. As shown in Fig. 7B, compared to the GAL-DBD alone, fusion of the GAL-DBD with Nanog, the Zscan4c Scan domain alone or full-length Zscan4c, lead to significant decreases in luciferase activity, consistent with transcriptional repression. These data are consistent with Zscan4c playing a direct role in regulating transcription and led us to undertake studies to define Zscan4-interacting proteins in the nucleus of ESCs, which could assist in further definition of the mode of action of Zscan4. V5 epitope tag. Expression of Zscan4c-V5 is under the control of the Tet-off expression system and as shown in Figure 7D, probing LSD-1 and CtBP2 immunoprecipitates with either anti-V5 epitope or anti-pan Zscan4 antibodies, demonstrates the presence of Zscan4 in those precipitates prepared following Tet-removal, which induces Zscan4c-V5 expression. Interestingly, the Zscan4- V5 precipitating with LSD-1 migrates more slowly than that co- precipitating with CtBP2, suggesting that Zscan4 may be participating in distinct protein complexes and subjected to post- translational modification. This is also apparent when LSD-1 and CtBP2 precipitates are probed with the pan-Zscan4 antibody (Fig. 7D, lower panel) and in these samples there is also evidence for precipitation of endogenous Zscan4 in the +Tet samples which do not over-express Zscan4c-V5. Next we attempted to confirm these interactions in wild-type ESCs, expressing endogenous levels of Zscan4 proteins, a considerable challenge based on our demonstration that less than 1% of the ESC population are expressing Zscan4. As shown in Figure 7E, Zscan4 can be detected in both LSD-1 and CtBP2 precipitates prepared from WT E14 ESCs. Regulation of Expression and Mode of Action of Zscan4 Again, LSD-1-associated Zscan4 exhibited slower migration compared to that associated with CtBP2. It has been suggested that the SCAN domain mediates dimerisation [27,28,29]. To examine whether this is the case for Zscan4, we precipitated with anti-V5 epitope antibodies and examined if endogenous Zscan4 protein could be co-precipitated. Figure 7F shows that endogenous Zscan4 is present in anti-V5 precipitates, suggesting that exogenously expressed Zscan4c can form dimers with endogenous Zscan4 proteins. Taken together, using three different cell lines, these results confirm that Zscan4 proteins are able to form complexes with LSD-1 and CtBP2 and potentially exist as dimers within ESCs, providing important new insights into the mode of action of Zscan4. To facilitate purification of Zscan4c interacting proteins, we constructed an eGFP-Zscan4c fusion protein, expressed in R1 ESCs under the control of the pTRE-tight, Dox-inducible expression system (Supplemental Figure S1). Following induction of eGFP-Zscan4c for 24 h, cytosolic and nuclear extracts were prepared and affinity purification performed with GFP-TrapA beads. Induced and un-induced extracts were subjected to the same procedure and after affinity purification and extensive washing protein complexes were released and separated by SDS- PAGE. After staining with Coomassie Blue, protein bands of approximately 116, 100, 75 and 50 kDa, were consistently enriched in the material affinity purified from the induced samples. These proteins were excised and submitted for sequenc- ing at the University of Bristol Proteomics Facility. Following proteolytic digestions with trypsin, tandem mass spectrometry was used to generate peptide mass fingerprints and the Mascot search engine (Matrix Science) exploited to identify proteins. Following these analyses, the 116 kDa protein was identified as the lysine specific demethylase, LSD1, the 100 kDa and 75 kDa proteins as full-length and truncated versions of eGFP-Zscan4c fusion protein respectively and the 50 kDa protein as the transcriptional co- repressor, C-Terminal Binding Protein-2 (CtBP2). Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 Figure 5. Zscan4 expression is enhanced in response to DNA damage induction. E14 ESCs were plated in the presence of LIF for 4 h prior to application of A. Zeocin or B. Cisplatin for 12 h. After this cells were washed extensively and fresh media added. Following a further 8 or 24 h samples were taken and cells fixed for cell cycle analysis or protein extracted into RIPA buffer. (i) Levels of H2AX phosphorylation at S139 (cH2AX) and Zscan4 protein expression were examined by immunoblotting. (ii) Zscan4 protein expression normalized to GAPDH is shown. (iii) The percentage of cells in G2/M phase of the cell cycle are shown for different samples. C. (i) The percentage of Zscan4 positive cells were determined after 24 h treatment with the indicated doses of Zeocin. (ii) The effect of inhibition of PI3Ks by LY294002 (LY, 5 mM) on the ability of Zeocin to induce Zscan4 expression was examined following 24 h treatment (12.5 mg/ml). In each case the mean and SEM are shown. , p,0.05, , p,0.005, , p,0.0005 following an ANOVA and Tukey’s post-hoc test. D. (i) E14 ESCs were co-stained for cH2AX and Zscan4 and the mean fluorescence intensity of cH2AX staining determined for Zscan4 positive and negative populations using Cell-P (Olympus). *, p,0.0005, in a Student’s t-test. Representative images are shown in (ii). doi:10 1371/journal pone 0089821 g005 doi:10.1371/journal.pone.0089821.g005 regulators [53] and while over-expression of Zscan4 has been shown to lead to changes in expression of ,1000 genes [54], there is no evidence that it acts directly. To examine the ability of Zscan4 to act as a regulator of transcription, we used the GAL4- UAS-luciferase reporter system and generated a series of GAL4- DNA binding domain-Zscan4c fusions, depicted in Fig. 7A. As a control for this assay, we also generated a GAL4-DBD-Nanog fusion. Transient transfection into heterologous HEK293 cells was performed with combinations of the GAL4-UAS-luciferase reporter (containing a synthetic GAL4-UAS promoter, see supplemental Figure S3), GAL4-DBD fusions and Renilla plasmid to control for transfection efficiency. As shown in Fig. 7B, compared to the GAL-DBD alone, fusion of the GAL-DBD with Nanog, the Zscan4c Scan domain alone or full-length Zscan4c, lead to significant decreases in luciferase activity, consistent with transcriptional repression. Zscan4 can modulate transcription and interacts with co- repressor complexes in undifferentiated ESCs To understand the mechanism of action of Zscan4 in regulation of ESC fate, particularly how this may relate to a specific role during G2, we investigated the ability of Zscan4 to act as a regulator of transcription. Other Scan-domain and zinc finger containing proteins have been shown to act as transcriptional March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 8 Regulation of Expression and Mode of Action of Zscan4 Re g p PLOS ONE \| www.plosone.org 9 March 2014 \| Volume 9 \| Issue 3 \| e89821 March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org PLOS ONE \| www.plosone.org 9 Discussion The cell cycle distribution of GFP negative and positive populations are presented, the mean and SEM are shown (n = 3). , p,0.05, , p,0.005, , p,0.0005, in a Student’s t-test. (iii) The mean fluorescence intensity of the GFP positive (Zscan4c expressing) cells distributed in the different cell cycle phases. In each case the mean and SEM are shown. , p,0.05, , p,0.005, , p,0.0005 following an ANOVA and Tukey’s post-hoc test. B. E14 ESCs were cultured in LIF and treated with 100 ng/ml Nocodazole for 12 h to induce mitotic arrest. Following release from the block, cell cycle analyses ((i), (ii) and (iii)) and immunoblotting, to detect Zscan4 protein expression (iv) were performed at the times indicated (n = 3). Mean and SEM are shown on the graphs. Zscan4 protein expression was normalised to GAPDH ((iv), lower panel). doi:10.1371/journal.pone.0089821.g006 Regulation of Expression and Mode of Action of Zscan4 Figure 6. Zscan4 expression is enhanced in late S/early G2 phase of the cell cycle. A. (i) The proportion of Zscan4 positive cells appearing as individual single cells or in proximity with one, two or three other Zscan4 positive cells were determined. Mean and SEM are shown. Average frequency of Zscan4 positive cells was 0.4%. (ii) ZE3-MC1 ESCs were grown in the presence of LIF for 48 h, trypsinised and Nuclear-ID Red (Enzo) used to stain cellular DNA. Cells were analysed by flow cytometry and G1, S and G2/M gates were assigned to the population based on DNA content and using a Nocodazole treated sample as a reference. The cell cycle distribution of GFP negative and positive populations are presented, the mean and SEM are shown (n = 3). , p,0.05, , p,0.005, , p,0.0005, in a Student’s t-test. (iii) The mean fluorescence intensity of the GFP positive (Zscan4c expressing) cells distributed in the different cell cycle phases. In each case the mean and SEM are shown. , p,0.05, , p,0.005, , p,0.0005 following an ANOVA and Tukey’s post-hoc test. B. E14 ESCs were cultured in LIF and treated with 100 ng/ml Nocodazole for 12 h to induce mitotic arrest. Following release from the block, cell cycle analyses ((i), (ii) and (iii)) and immunoblotting, to detect Zscan4 protein expression (iv) were performed at the times indicated (n = 3). Mean and SEM are shown on the graphs. Zscan4 protein expression was normalised to GAPDH ((iv), lower panel). Discussion The Zscan4 family of genes have been implicated in the control of early development [26], ESC pluripotency [24] and genome stability/integrity [25], but their regulation and mode of action remain enigmatic. Here we demonstrate that Zscan4c, 4f and ps2 are the most abundant and consistently expressed members of this family, but despite this, expression of Zscan4 protein is restricted to between only 0.1 and 0.7% of ESCs. Significantly, Zscan4 expression was found to be highest in cells transiting late S-phase/ early G2 of the cell cycle and we show that DNA damage-inducing agents lead to enhancement of Zscan4 expression, corresponding with accumulation of cells in G2/M. Together these data suggest that Zscan4 plays a selective role in ESCs during G2. Our demonstration that Zscan4 can directly regulate transcription and exists in complexes with LSD-1 and CtBP2 sheds new and important light on its mode of action. These findings lead us to suggest that the Zscan4 family of genes represents an intriguing paradigm for factors that regulate the ESC state. To confirm these Zscan4 protein-protein interactions we first assessed interaction of Zscan4c with LSD-1 and CtBP2 in the eGFP-Zscan4c fusion expressing ESCs. As shown in Figure 7C, the GFP-TrapA beads precipitate the GFP-Zscan4c fusion protein and co-precipitate both LSD-1 and CtBP2 in samples prepared following Dox-induction of eGFP-Zscan4c expression. Further- more, following immunoprecipitation of LSD-1, eGFP-Zscan4c can be detected in Dox-induced samples. CtBP2 has been reported to form complexes with LSD-1 and we confirm this in mESCs, as CtBP2 is also present in LSD-1 precipitates. We also examined co- precipitation with a version of Zscan4c containing a C-terminal March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 10 Regulation of Expression and Mode of Action of Zscan4 Figure 6. Zscan4 expression is enhanced in late S/early G2 phase of the cell cycle. A. (i) The proportion of Zscan4 positive cells appearing as individual single cells or in proximity with one, two or three other Zscan4 positive cells were determined. Mean and SEM are shown. Average frequency of Zscan4 positive cells was 0.4%. (ii) ZE3-MC1 ESCs were grown in the presence of LIF for 48 h, trypsinised and Nuclear-ID Red (Enzo) used to stain cellular DNA. Cells were analysed by flow cytometry and G1, S and G2/M gates were assigned to the population based on DNA content and using a Nocodazole treated sample as a reference. Discussion Precipitates were separated through 7.5% polyacrylamide gels prior to immunoblotting. Immunoblotting was first performed with anti-Zscan4 antibodies. The blots were then stripped and reprobed with anti-V5 epitope antibodies. Positions of precipitated proteins are indicated. E. Nuclear extracts (Nuc) were prepared from E14 ESCs and precipitates from 200 mg of protein per sample generated using anti-Zscan4, anti-LSD-1 or anti-CtBP2 antibodies. Precipitates were separated through a 6.5% polyacrylamide gel prior to immunoblotting and the lanes between the samples were left blank to avoid the potential of any bleed-through. Immunoblotting was performed with anti-Zscan4 antibodies. Positions of precipitated proteins are indicated. F. ESCs engineered to express a C-Terminally V5 epitope-tagged version of Zscan4c under the control of the Tet-off expression system were cultured in the presence (+) or absence (2) of Tet for 72 h. Nuclear extracts were prepared (Nuc) and immunoprecipitates prepared from 80 mg of protein per sample using anti-V5 epitope antibodies. Precipitates were separated through a 7 5% polyacrylamide gel prior to immunoblotting Immunoblotting was performed with anti-Zscan4 antibodies Positions of precipitated Regulation of Expression and Mode of Action of Zscan4 Figure 7. Zscan4 acts as a transcriptional regulator and interacts with LSD-1 and CtBP2. A. Schematic showing the structure of GAL4-DNA binding domain fusions with Zscan4c. B. Plasmids expressing the GAL-4-DNA binding domain fusions indicated were co-transfected into HEK293 cells along with pGL4-GAL4-UAS plasmid and the pRenilla plasmid (to assess transfection efficiency). 48 h after transfection luciferase activity was determined using Dual-Glo system, according to the manufacturer’s recommendations (Promega). Mean and SEM of 3 independent experiments are shown. , p,0.05, *, p,0.005 following an ANOVA and Tukey’s post-hoc test. C. ESCs engineered to express an N-Terminal eGFP-Zscan4c fusion protein under the control of the pTRE tight Dox-inducible expression system were cultured in the absence (2) or presence (+) of Dox for 72 h. Nuclear extracts were prepared (Nuc) and immunoprecipitates prepared from 80 mg of protein per sample using either GFP-Trap-A beads or anti-LSD- 1 antibodies. Precipitates were divided into two aliquots and separated through 10% polyacrylamide on duplicate gels, prior to immunoblotting. Immunoblotting was performed with anti-GFP, anti-CtBP2 or anti-Zscan4 antibodies, sequentially with one immunoblot. The duplicate immunoblot was probed with anti-LSD-1 antibodies. Positions of precipitated proteins are indicated. D. Discussion doi:10 1371/journal pone 0089821 g006 doi:10.1371/journal.pone.0089821.g006 doi:10.1371/journal.pone.0089821.g006 0.7%, 10-fold lower than the 5% previously reported [25]. Interestingly, we found this level of expression to be consistent across a number of mouse ESC lines and also observed with an iPSC line. We do not think the difference in the proportion of Zscan4 positive cells detected in our study versus previous studies The very restricted pattern of expression reported for the Zscan4 family is highly distinctive [25,26], although one caveat to these previous reports has been that they only measured RNA or reporter levels. Our data indicate that the proportion of ESCs within a population expressing Zscan4 protein is between 0.1 and March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org March 2014 \| Volume 9 \| Issue 3 \| e89821 11 Regulation of Expression and Mode of Action of Zscan4 Figure 7. Zscan4 acts as a transcriptional regulator and interacts with LSD-1 and CtBP2. A. Schematic showing the structure of GAL4-DNA binding domain fusions with Zscan4c. B. Plasmids expressing the GAL-4-DNA binding domain fusions indicated were co-transfected into HEK293 cells along with pGL4-GAL4-UAS plasmid and the pRenilla plasmid (to assess transfection efficiency). 48 h after transfection luciferase activity was determined using Dual-Glo system, according to the manufacturer’s recommendations (Promega). Mean and SEM of 3 independent experiments are shown. , p,0.05, **, p,0.005 following an ANOVA and Tukey’s post-hoc test. C. ESCs engineered to express an N-Terminal eGFP-Zscan4c fusion protein under the control of the pTRE tight Dox-inducible expression system were cultured in the absence (2) or presence (+) of Dox for 72 h. Nuclear extracts were prepared (Nuc) and immunoprecipitates prepared from 80 mg of protein per sample using either GFP-Trap-A beads or anti-LSD- 1 antibodies. Precipitates were divided into two aliquots and separated through 10% polyacrylamide on duplicate gels, prior to immunoblotting. Immunoblotting was performed with anti-GFP, anti-CtBP2 or anti-Zscan4 antibodies, sequentially with one immunoblot. The duplicate immunoblot was probed with anti-LSD-1 antibodies. Positions of precipitated proteins are indicated. D. ESCs engineered to express a C-Terminally V5 epitope- tagged version of Zscan4c under the control of the Tet-off expression system were cultured in the presence (+) or absence (2) of Tet for 72 h. Nuclear extracts were made (Nuc) and immunoprecipitates prepared from 80 mg of protein per sample using either anti-LSD-1 or anti-CtBP2 antibodies or protein-A sepharose (PAS) alone as a control. Regulation of Expression and Mode of Action of Zscan4 Given the report that Zscan4 plays a critical role in telomere elongation and genome integrity in mouse ESCs [25] our demonstration that agents that damage DNA lead to an enhancement in Zscan4 expression is consistent with a role for Zscan4 in genome surveillance and maintenance. The fact that members of the Zscan4 family contain zinc finger domains has led to the suggestion that they act as transcription factors [24], as do other members of the SCAN-domain containing family [53]. Over-expression of Zscan4 in mouse ESCs has been reported to result in the change in expression of over 1000 genes [54], while inclusion of Zscan4 as a reprogramming factor for mouse embryo fibroblasts results in the transient induction of pre-implantation specific genes [33]. Despite these reports, evidence to support the ability of Zscan4 to directly regulate transcription has been lacking. Using a heterologous reporter system, we demonstrate that Zscan4 can act as a transcriptional regulator. Both full-length Zscan4 and the SCAN domain alone were able to repress GAL4 binding domain- mediated reporter expression (Figure 7B). This could be due to the ability of the SCAN domain to promote dimerisation, as demonstrated for other SCAN-domain-containing proteins [28,29]. Importantly, we provide the first evidence that Zscan4 may act as a dimer, since we show that a V5-epitope-tagged version of Zscan4 co-precipiates endogenous Zscan4 proteins (Figure 7F). To gain further insight into the mode of action of Zscan4, we investigated whether Zscan4 interacted with other proteins within ESC nuclei. Our analyses revealed two major interacting proteins -LSD-1 and CtBP2. LSD1 has been reported to specifically demethylate mono and di-methyl histone H3 at K4 and K9 and thus acts as a co-repressor of transcription. CtBP2 belongs to a class of transcriptional co-repressors that bind to transcription factors via a PXDLS motif and has been shown to interact with a range of Zinc finger containing transcriptional regulators [57,58]. Zscan4c exhibits a PXDLS motif at position 263–268, as do Zscan4 family members, suggesting that each of them can bind CtBP2. Importantly, both LSD-1 and CtBP2 have been reported to be components of the CoREST-CtBP repressor complex [58,59]. Thus, the identification of these two proteins as Zscan4c interacting partners in ESCs is consistent with the fact that Zscan4c is a putative regulator of transcription, possibly involved in repression of transcription. Regulation of Expression and Mode of Action of Zscan4 It has been suggested that the fact Zscan4 over-expression leads to minor changes in the transcriptome during early reprogramming events, but major differences in iPSC outcome, indicates it acts differently to many other factors and may be involved in epigenetic regulation or chromatin remodeling [33]. Our finding that Zscan4 interacts with LSD-1 is consistent with this latter possibility. In future studies it will be interesting to determine the genes that are directly targeted by Zscan4 in ESC and iPSCs and compare these to known targets of LSD-1 and CtBP2. q g p Based on the evidence published to date [24,25,26,33,35,55], it seems likely that Zscan4 plays multiple roles in pluripotent cells and we were particularly interested to discover that Zscan4 expression varied at different stages of the cell cycle. Initially our observation that 45% of Zscan4 positive cells were found in close proximity to another Zscan4 positive cell led us to consider that Zscan4 expression may be enriched in cells undergoing mitosis. However, when ESCs were treated with Nocodazole, which leads to arrest in pro-metaphase, Zscan4 expression was not elevated. Instead, upon release and progression through the cell cycle, increases in Zscan4 protein levels corresponded with the stage at which cells transited late S-phase and entered early G2. Further support for a role of Zscan4 in G2 phase of the cell cycle comes from our studies examining regulation of Zscan4 expression in response to DNA damaging agents. Both Zeocin and Cisplatin induce double-stranded DNA breaks, as measured by an increase in cH2AX, which lead to cell cycle arrest in G2/M [56]. We detected the highest levels of Zscan4 induction by Zeocin at the point when the number of cells in G2/M had increased from 40 to 65%. In the case of Cisplatin, the most significant elevation in Zscan4 expression was observed with the highest dose of Cisplatin we used following a 24 h recovery phase. At this same time point, the proportion of cells in G2/M had increased from 40 to 57%. While we had initially assessed cell cycle profiles of these cells to gauge the ESC response to DNA damage-inducing agents, these data provide additional evidence to support the fact that Zscan4 expression is preferentially increased during G2-phase of the cell cycle. Regulation of Expression and Mode of Action of Zscan4 In addition, this is the same reporter line used by the Ko group [25] but we record fewer Zscan4 positive cells than previously reported. In view of this, it may be that Zscan4 expression is influenced by variation in culture conditions, although when we evaluated this no clear factors emerged. Alternatively, we may have set more stringent thresholds for detection of Zscan4 protein expression, which would have decreased the number of positive cells scored. Nevertheless, it is clear from previous work and the new data that we report here that Zscan4 is expressed by only a small proportion of ESCs in culture. Such a restricted pattern of expression is quite unprec- edented in the ESC field and leads to the question of how Zscan4 expression is regulated. Of relevance here is a recent study that has reported the identification of a rare and transient population of cells within ESC cultures that share properties and expression profiles with two-cell stage embryos [55]. This is interesting from the point of view of the Zscan4 family, since Zscan4d was first identified as a gene whose expression was elevated during zygotic genome activation and highly expressed in 2-cell stage embryos [26]. Macfarlan et al., report that many of the ‘2C’ transcripts detected in this transient sub-population of ESCs are initiated from the LTRs of endogenous retroviruses, including those of Zscan4 [55]. Thus, it is possible that activation of these endogenous retroviral elements play a role in establishing the highly restricted pattern of expression of Zscan4 proteins within ESC populations. Furthermore, it is tempting to speculate that Zscan4 is a marker of this rare and transient ‘2C’ stem cell population and given that RNAi knock-down of Zscan4 leads to arrest of 2-cell embryo development [26], it could play an important functional role in this rare cell population. However, alternative explanations for the restricted pattern of expression of Zscan4 also need to be considered and explored, including auto- regulatory or oscillatory-based mechanisms. Further work will be required to distinguish between these possibilities. Zscan4 expression do not follow the profile of cH2AX detection, indicating Zscan4 up-regulation is not an early response to DNA damage. Discussion ESCs engineered to express a C-Terminally V5 epitope- tagged version of Zscan4c under the control of the Tet-off expression system were cultured in the presence (+) or absence (2) of Tet for 72 h. Nuclear extracts were made (Nuc) and immunoprecipitates prepared from 80 mg of protein per sample using either anti-LSD-1 or anti-CtBP2 antibodies or protein-A sepharose (PAS) alone as a control. Precipitates were separated through 7.5% polyacrylamide gels prior to immunoblotting. Immunoblotting was first performed with anti-Zscan4 antibodies. The blots were then stripped and reprobed with anti-V5 epitope antibodies. Positions of precipitated proteins are indicated. E. Nuclear extracts (Nuc) were prepared from E14 ESCs and precipitates from 200 mg of protein per sample generated using anti-Zscan4, anti-LSD-1 or anti-CtBP2 antibodies. Precipitates were separated through a 6.5% polyacrylamide gel prior to immunoblotting and the lanes between the samples were left blank to avoid the potential of any bleed-through. Immunoblotting was performed with anti-Zscan4 antibodies. Positions of precipitated proteins are indicated. F. ESCs engineered to express a C-Terminally V5 epitope-tagged version of Zscan4c under the control of the Tet-off expression system were cultured in the presence (+) or absence (2) of Tet for 72 h. Nuclear extracts were prepared (Nuc) and immunoprecipitates prepared from 80 mg of protein per sample using anti-V5 epitope antibodies. Precipitates were separated through a 7.5% polyacrylamide gel prior to immunoblotting. Immunoblotting was performed with anti-Zscan4 antibodies. Positions of precipitated proteins are indicated. doi:10.1371/journal.pone.0089821.g007 March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 12 Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 is due to post-transcriptional regulation, since we saw a robust correlation between reporter expression in the ZE3-MC1 line and detection of Zscan4 protein. In addition, this is the same reporter line used by the Ko group [25] but we record fewer Zscan4 positive cells than previously reported. In view of this, it may be that Zscan4 expression is influenced by variation in culture conditions, although when we evaluated this no clear factors emerged. Alternatively, we may have set more stringent thresholds for detection of Zscan4 protein expression, which would have decreased the number of positive cells scored. Nevertheless, it is clear from previous work and the new data that we report here that Zscan4 is expressed by only a small proportion of ESCs in culture. Such a restricted pattern of expression is quite unprec- edented in the ESC field and leads to the question of how Zscan4 expression is regulated. Of relevance here is a recent study that has reported the identification of a rare and transient population of cells within ESC cultures that share properties and expression profiles with two-cell stage embryos [55]. This is interesting from the point of view of the Zscan4 family, since Zscan4d was first identified as a gene whose expression was elevated during zygotic genome activation and highly expressed in 2-cell stage embryos [26]. Macfarlan et al., report that many of the ‘2C’ transcripts detected in this transient sub-population of ESCs are initiated from the LTRs of endogenous retroviruses, including those of Zscan4 [55]. Thus, it is possible that activation of these endogenous retroviral elements play a role in establishing the highly restricted pattern of expression of Zscan4 proteins within ESC populations. Furthermore, it is tempting to speculate that Zscan4 is a marker of this rare and transient ‘2C’ stem cell population and given that RNAi knock-down of Zscan4 leads to arrest of 2-cell embryo development [26], it could play an important functional role in this rare cell population. However, alternative explanations for the restricted pattern of expression of Zscan4 also need to be considered and explored, including auto- regulatory or oscillatory-based mechanisms. Further work will be required to distinguish between these possibilities. is due to post-transcriptional regulation, since we saw a robust correlation between reporter expression in the ZE3-MC1 line and detection of Zscan4 protein. (TIF) Figure S5 Assigning Zscan4 gene transcripts. The transcript sequence from each Zscan4 gene was obtained from the NCBI database. These were aligned and distinguished based on the identification of SNPs – highlighted in black in the schematic. Sequenced Zscan4 cDNAs derived from E14tg2a or CCE cell lines (numbers assigned based on position in 96-well tray, e.g. E11, C07 as shown) were edited and added to this alignment. Sequences were assigned to a Zscan4 gene based on the distinguishing SNPs each contained. As an example, the 5 SNPs used to identify Zscan4a transcripts are shown. Sequence data were edited and aligned using Sequencher (GeneCodes). (TIF) 1. Young RA (2011) Control of the Embryonic Stem Cell State. Cell 144: 940–954. 2. Boiani M, Schoeler HR (2005) Regulatory networks in embryo-derived pluripotent stem cells. Nat Rev Mol Cell Biol 6: 872–881. Supporting Information Anti-V5 antibody was used to detect Zscan4c-V5 protein (red). Cell nuclei were counter-stained with DAPI (blue). (ii) Zscan4c-V5-His-Tet-off inducible mESC (clone 45) were grown in the presence and absence of Tet. Cytosolic and nuclear proteins were separated and immunoblotting performed with anti-V5, anti-TBP (predominantly nuclear) and anti-GAPDH (predominantly cytosolic) antibodies. Table S1 Primer sequences used for quantitative RT-PCR. (DOCX) Table S2 Primer sequences used for construction of Gal4- Zscan4 fusion transcriptional reporters. (DOCX) Author Contributions Conceived and designed the experiments: MPS BK HKB MB YSR JBC HN DT MJW. Performed the experiments: MPS BK HKB MB YSR MJW. Analyzed the data: MPS BK HKB MB YSR HN MJW. Contributed reagents/materials/analysis tools: HN DT. Wrote the paper: MPS BK HKB MJW. Conceived and designed the experiments: MPS BK HKB MB YSR JBC HN DT MJW. Performed the experiments: MPS BK HKB MB YSR MJW. Analyzed the data: MPS BK HKB MB YSR HN MJW. Contributed reagents/materials/analysis tools: HN DT. Wrote the paper: MPS BK HKB MJW. 5. Boeuf H, Hauss C, Graeve FD, Baran N, Kedinger C (1997) Leukemia inhibitory factor-dependent transcriptional activation in embryonic stem cells. J Cell Biol 138: 1207–1217. Regulation of Expression and Mode of Action of Zscan4 Furthermore, it is worth noting that temporal changes in Our expression data indicate that three Zscan4 genes are most abundantly expressed in ESCs, namely Zscan4c, 4f and ps2. Zscan4-ps2 has previously been characterized as a pseudogene, primarily based on the fact that in a previous study no ps2 transcripts were detected in mouse ESCs [26], although in contrast our data indicate Zscan4-ps2 transcripts can be detected in mouse ESCs. Given the close relationship between Zscan4c, 4f and ps2 (Fig. 1B), it is intriguing that these should consistently be the most highly expressed in different ESCs lines, but argues for a predominant role of their gene products. Also worth noting is the fact that the p110a isoform of PI3Ks plays the predominant role in regulating Zscan4 expression, supported by both loss and March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 13 Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 gain of function studies (Figure 3). We have previously implicated p110a PI3K in regulation of ESC proliferation [48,60], which complements with the findings we report here relating to Zscan4 expression and the cell cycle. PI3K signaling is one of the pathways that regulate activity of Gsk-3. Surprisingly, given its role in the ground state of ESC pluripotency, short-term inhibition of Gsk-3-dependent signaling did not lead to increased Zscan4 protein expression, whereas a sustained lack of Gsk-3 activity did. However, no significant increase in the proportion of Zscan4 positive cells was detected in Gsk-3 double knock-out ESCs (data not shown). Figure S3 The sequence (59-39) of the synthetic GAL4-UAS promoter used to drive luciferase expression in this study. The strands were mixed, heated to 95uC and cooled before being ligated into pGL4.26 (Promega). (TIF) gain of function studies (Figure 3). We have previously implicated p110a PI3K in regulation of ESC proliferation [48,60], which complements with the findings we report here relating to Zscan4 expression and the cell cycle. PI3K signaling is one of the pathways that regulate activity of Gsk-3. Surprisingly, given its role in the ground state of ESC pluripotency, short-term inhibition of Gsk-3-dependent signaling did not lead to increased Zscan4 protein expression, whereas a sustained lack of Gsk-3 activity did. However, no significant increase in the proportion of Zscan4 positive cells was detected in Gsk-3 double knock-out ESCs (data not shown). Figure S4 Zscan4 antibody specificity. ESCs engineered to express a C-terminally V5 epitope-tagged version of Zscan4c under the control of the Tet-off expression system were cultured in the presence (+) or absence (2) of Tet for 48 h when nuclear (Nuc) and cytoplasmic (Cyto) protein extracts were prepared. A. Cytosolic and nuclear extracts were separated by SDS-PAGE and duplicate blots prepared. A 1:2000 dilution of anti-Zscan4 anti-peptide antibody was incubated with or without 1 mg/ml blocking peptide for 1 h prior to being used for immunoblotting. B. Immunoprecipitates were prepared from (i) 1 mg of cytosolic or (ii) 80 mg of nuclear protein extract per sample using either protein-A sepharose (PAS) bead alone or together with 2 mg anti- Zscan4 antibody. Precipitates were separated through 7.5% polyacrylamide gels prior to immunoblotting. Immunoblotting was performed with anti-V5 epitope (upper panels) or anti-Zscan4 (lower panels) antibodies. Positions of precipitated proteins are indicated by the arrows. (TIF) Figure S4 Zscan4 antibody specificity. 4. Silva J, Smith A (2008) Capturing pluripotency. Cell 132: 532–536. 5. Boeuf H, Hauss C, Graeve FD, Baran N, Kedinger C (1997) Leukemia inhibitory factor dependent transcriptional activation in embryonic stem cells Supporting Information Figure S1 Establishment and characterisation of eGFP- Zscan4c-Tet-on inducible ESC lines. A. Schematic of the GFP-Zscan4c fusion protein generated. B. Following induction with 1 mg/ml doxycycline for the times indicated (in hours, h), nuclear lysates were prepared from eGFP-Zscan4c ESCs and immunoblotted with an anti-GFP antibody. Blots were stripped and re-probed with an anti-TBP antibody to assess the equality of protein loading. C. Quantitative RT-PCR was used to confirm induction of Zscan4c expression normalised to b-actin levels. D. Fluorescent images showing expression of eGFP-Zscan4c (green) after addition of doxycycline for the times indicated. (TIF) Figure S6 Regulation of Zscan4 by the p110a catalytic subunit of PI3K. E14tg2a cells were treated with either A. 5 mM LY294002 or 600 nM Compound 15e or B. 5 mM LY294002, 10 mM TGX-121, 5 mM IC87114 or 1 nM Rapamycin. RNA was extracted 48 h after inhibitor treatment, quantitative RT-PCR was performed and Zscan4 expression normalised relative to b- actin levels. Graphs show standard deviation and are representa- tive of three experimental repeats. Figure S2 Establishment and characterisation of Zscan4c-V5-His Tet-off inducible ESC lines. A. Immuno- blots showing induction of Zscan4c-V5-His protein (clone 43) upon withdrawal of tetracycline (tet) detected with anti-V5 antibody. (i) Expression 24 h after induction in the presence of different concentrations of LIF (indicated in U/ml). (ii) Expression of Zscan4c-V5-His was maintained over a time-course of 120 h. Immunoblots were stripped and re-probed with Oct-4 antibody. B. (i) Localisation of Zscan4c-V5-His protein was assessed by immunostaining of Tet-off inducible cell lines grown in the absence of tetracycline. Anti-V5 antibody was used to detect Zscan4c-V5 protein (red). Cell nuclei were counter-stained with DAPI (blue). (ii) Zscan4c-V5-His-Tet-off inducible mESC (clone 45) were grown in the presence and absence of Tet. Cytosolic and nuclear proteins were separated and immunoblotting performed with anti-V5, anti-TBP (predominantly nuclear) and anti-GAPDH (predominantly cytosolic) antibodies. Figure S2 Establishment and characterisation of Zscan4c-V5-His Tet-off inducible ESC lines. A. Immuno- blots showing induction of Zscan4c-V5-His protein (clone 43) upon withdrawal of tetracycline (tet) detected with anti-V5 antibody. (i) Expression 24 h after induction in the presence of different concentrations of LIF (indicated in U/ml). (ii) Expression of Zscan4c-V5-His was maintained over a time-course of 120 h. Immunoblots were stripped and re-probed with Oct-4 antibody. B. (i) Localisation of Zscan4c-V5-His protein was assessed by immunostaining of Tet-off inducible cell lines grown in the absence of tetracycline. 4. Silva J, Smith A (2008) Capturing pluripotency. Cell 132: 532–536. 5. Boeuf H, Hauss C, Graeve FD, Baran N, Kedinger C (1997) Leukemia inhibitory factor-dependent transcriptional activation in embryonic stem cells. J Cell Biol 138: 1207–1217. 3. Pera MF, Tam PP (2010) Nature 465: 713–720. . Silva J, Smith A (2008) Capturing pluripotency. Cell 132: 532–536 Regulation of Expression and Mode of Action of Zscan4 ESCs engineered to express a C-terminally V5 epitope-tagged version of Zscan4c under the control of the Tet-off expression system were cultured in the presence (+) or absence (2) of Tet for 48 h when nuclear (Nuc) and cytoplasmic (Cyto) protein extracts were prepared. A. Cytosolic and nuclear extracts were separated by SDS-PAGE and duplicate blots prepared. A 1:2000 dilution of anti-Zscan4 anti-peptide antibody was incubated with or without 1 mg/ml blocking peptide for 1 h prior to being used for immunoblotting. B. Immunoprecipitates were prepared from (i) 1 mg of cytosolic or (ii) 80 mg of nuclear protein extract per sample using either protein-A sepharose (PAS) bead alone or together with 2 mg anti- Zscan4 antibody. Precipitates were separated through 7.5% polyacrylamide gels prior to immunoblotting. Immunoblotting was performed with anti-V5 epitope (upper panels) or anti-Zscan4 (lower panels) antibodies. Positions of precipitated proteins are indicated by the arrows. (TIF) Based on previously published studies and the results we present here, we would like to suggest that the Zscan4 family represents an intriguing paradigm for regulators of the for ESC state. Multiple, closely related Zscan4 genes are co-expressed by ESCs, but despite this Zscan4 protein is detected in less than 1% of ESCs. Zscan4 expression is higher in cells in late S/early G2-phase of the cell cycle, consistent with it playing a role during the G2 checkpoint and genome stability. Related to this we have demonstrated that Zscan4 acts as a regulator of transcription and interacts with LSD- 1 and CtBP2, suggesting it may have a role in transcriptional repression. While many questions remain to be answered about this enigmatic family of zinc finger proteins, our study provides new and important insight into the regulation and action of Zscan4 proteins. 1. Young RA (2011) Control of the Embryonic Stem Cell State. Cell 144: 940–954. Regulation of Expression and Mode of Action of Zscan4 y g 53. Edelstein LC, Collins T (2005) The SCAN domain family of zinc finger transcription factors. Gene 359: 1–17. 25. Zalzman M, Falco G, Sharova LV, Nishiyama A, Thomas M, et al. (2010) Zscan4 regulates telomere elongation and genomic stability in ES cells. Nature 464: 858–863. 54. Nishiyama A, Xin LAA, Sharov M, Thomas G, Mowrer E, et al. (2009) Uncovering early response of gene regulatory networks in ESCs by systematic induction of transcription factors. Cell Stem Cell 5: 420–433. 26. Falco G, Lee S-L, Stanghellini I, Bassey UC, Hamatani T, et al. (2007) Zscan4: A novel gene expressed exclusively in late 2-cell embryos and embryonic stem cells. Developmental Biology 307: 539–550. 55. Macfarlan TS, Gifford WD, Driscoll S, Lettier K, Rowe HM, et al. (2012) Embryonic stem cell potency fluctuates with endogenous retrovirus activity. Nature doi:10.1038/nature11244. 27. Schumacher C, Wang HG, Honer C, Ding W, Koehn J, et al. (2000) The SCAN domain mediates selective oligomerisation. J Biol Chem 275: 17173–17179. 56. Nagaria P, Robert C, Rassool FV (2013) DNA double-strand break response in stem cells: mechanisms to maintain genomic integrity. Biochem Biophys Acta 1830: 2345–2353. 28. Stone JR, Maki JL, Blacklow SC, Collins T (2002) The SCAN domain of ZNF174 is a dimer. J Biol Chem 277: 5448–5452. 29. Williams AJ, Blacklow SC, Collins T (1999) The zinc finger-associated SCAN box is a conserved oligomerisation domain. Mol Cell Biol 19: 8526–8535. 57. Schaeper U, Subramanian T, Lim L, Boyd JM, Chinnadurai G (1998) Interaction between a cellular protein that binds to the C-terminal region of adenovirus E1A (CtBP) and a novel cellular protein is disrupted by E1A through a conserved PLDLS motif. J Biol Chem 273: 8549–8552. 30. Carter MG, Stagg CA, Falco G, Yoshikawa T, Bassey UC, et al. (2008) An in situ hybridzation-based screen for heterogeneously expressed genes in mouse ES cells. Gene Expression Patterns 8: 181–198. a conserved PLDLS motif. J Biol Chem 273: 8549–8552. 58. Chinnadurai G (2007) Transcriptional regulation by C-terminal binding proteins. Int J Biochem Cell Biol 39: 1593–1607. 31. Costello I, Biondi CA, Taylor JM, Bikoff EK, Robertson EJ (2009) Smad4- dependent pathways control basement membrane deposition and endodermal cell migration at early stages of mouse development. BMC Developmental Biology 9: 54. 59. Marmorstein R, Trievel RC (2009) Histone modifying enzymes: structures, mechanisms, and specificities. Biochem Biophys Acta 1789. mechanisms, and specificities. Biochem Biophys Acta 1789. 60. Regulation of Expression and Mode of Action of Zscan4 (2011) Differential requirement for the dual functions of beta-catenin in embryonic stem cell self-renewal and germ layer formation. Nat Cell Biol 13: 753–761. 39. Orlando L, Sanchez-Ripoll Y, Foster J, Bone H, Giachino C, et al. (2012) Differential Coupling of Self-Renewal Signaling Pathways in Murine Induced Pluripotent Stem Cells. PloS ONE 7: e30234. 12. Wray J, Kalkan T, Gomez-Lopez S, Eckardt D, Cook A, et al. (2011) Inhibition of glycogen synthase kinase-3 alleviates Tcf3 repression of the pluripotency network and increases embryonic stem cell resistance to differentiation. Nat Cell Biol 13: 838–845. p 40. Gossen M, Bujard H (1992) Tight control of gene-expression in mammalian- cells by tetracycline- responsive promoters. Proc Natl Acad Sci USA 89: 5547– 5551. 41. Era T, Witte ON (2000) Regulated expression of P210 Bcr-Abl during embryonic stem cell differentiation stimulates multipotential progenitor expansion and myeloid cell fate. Proc Natl Acad Sci U S A 97: 1737–1742. 13. Niwa H, Ogawa K, Shimosato D, Adachi K (2009) A parallel circuit of LIF signalling pathways maintains pluripotency of mouse ES cells. Nature 460: 118– 122. 42. Takahashi K, Mitsui K, Yamanaka S (2003) Role of ERas in promoting tumour- like properties in mouse embryonic stem cells. Nature 423: 541–545. 14. Paling NR, Wheadon H, Bone HK, Welham MJ (2004) Regulation of embryonic stem cell self-renewal by phosphoinositide 3-kinase-dependent signaling. J Biol Chem 279: 48063–48070. like properties in mouse embryonic stem cells. Nature 423: 541–545. 43. Storm MP, Bone HK, Beck CG, Bourillot PY, Schreiber V, et al. (2007) Regulation of Nanog expression by phosphoinositide 3-kinase-dependent signaling in murine embryonic stem cells. J Biol Chem 282: 6265–6273. 15. Watanabe S, Umehara H, Murayama K, Okabe M, Kimura T, et al. (2006) Activation of Akt signaling is sufficient to maintain pluripotency in mouse and primate embryonic stem cells. Oncogene 25: 2697–2707. 44. Craddock BL, Hobbs J, Edmead CE, Welham MJ (2001) Phosphoinositide 3- kinase-dependent regulation of interleukin-3-induced proliferation: involvement of mitogen-activated protein kinases, SHP2 and Gab2. J Biol Chem 276: 24274– 24283. 16. Singh AM, Reynolds D, Cliff T, Ohtsuka S, Mattheyses AL, et al. (2012) Signaling network crosstalk in human pluripotent cells: A Smad2/3-regualted switch that controls the balance between self-renewal and differentiation. Cell Stem Cell 10: 3112–3326. 45. Welham MJ, Duronio V, Leslie KB, Bowtell D, Schrader JW (1994) Multiple hemopoietins, with the exception of interleukin-4, induce modification of Shc and mSos1, but not their translocation. Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 Regulation of Expression and Mode of Action of Zscan4 6. Niwa H, Burdon T, Chambers I, Smith A (1998) Self-renewal of pluripotent embryonic stem cells is mediated via activation of STAT3. Genes Dev 12: 2048– 2060. 33. Hirata T, Amano T, Nakatake Y, Amano M, Piao Y, et al. (2012) Zscan4 transiently reactivates early embryonic genes during the generation of induced pluripotent stem cells. Scientific Reports 2:208: 1–11. p p p 34. Jiang J, Lv W, Ye X, Wang L, Zhang M, et al. (2013) Zscan4 promotes genomic stability during reprogramming and dramatically improves the quality of iPS cells as demonstrated by tetraploid complementation. Cell Res 23: 92–106. 7. Ying QL, Nichols J, Chambers I, Smith A (2003) BMP induction of Id proteins suppresses differentiation and sustains embryonic stem cell self-renewal in collaboration with STAT3. Cell 115: 281–292. 35. Amano T, Hirata K, Geppino Falco, Manuela Monti, Lioudmila V . Sharova, et al. (2013) Zscan4 restores the developmental potency of embryonic stem cells. Nature Communications 4: 1966. 8. Bone HK, Damiano T, Bartlett S, Perry A, Letchford J, et al. (2009) Involvement of glycogen synthase kinase-3 in regulation of murine embryonic stem cell self-renewal revealed by a series of bisindolylmaleimides. Chemistry and Biology 16: 15–27. 36. Smith AG, Hooper ML (1987) Buffalo rat liver cells produce a diffusible activity which inhibits the differentiation of murine embryonal carcinoma and embryonic stem cells. Dev Biol 121: 1–9. 9. Sato N, Meijer L, Skaltsounis L, Greengard P, Brivanlou AH (2004) Maintenance of pluripotency in human and mouse embryonic stem cells through activation of Wnt signaling by a pharmacological GSK-3-specific inhibitor. Nat Med 10: 55–63. 37. Doble B, Patel S, Wood G, Kockeritz L, Woodgett J (2007) Functional redundancy of GSK-3a and GSK-3b in Wnt/b-catenin signalling shown by using an allelic series of embryonic stem cell lines. Dev Cell 12: 957–971. 10. Ying Q-L, Wray J, Nichols J, Batlle-Morera L, Doble B, et al. (2008) The ground state of embryonic stem cell self-renewal. Nature 453: 519–523. 38. Okita K, Ichisaka T, Yamanaka S (2007) Generation of germline-competent induced pluripotent stem cells. Nature 448: 313–317. 38. Okita K, Ichisaka T, Yamanaka S (2007) Generation induced pluripotent stem cells. Nature 448: 313–317. 11. Lyashenko N, Winter M, Migliorini D, Biechele T, Moon RT, et al. Regulation of Expression and Mode of Action of Zscan4 J Biol Chem 269: 21165–21176. 17. Niwa H (2007) How is pluripotency determined and maintained? Development 134: 635–646. 46. Chambers I, Silva J, Colby D, Nichols J, Nijmeijer B, et al. (2007) Nanog safeguards pluripotency and mediates germline development. Nature 450: 1230– U1238. 18. Galan-Caridad JM, Harel S, Arenzana TL, Hou ZE, BDoetsch FK, et al. (2007) Zfx controls the self-renewal of embryonic and hemopoietic stem cells. Cell 129: 345–357. 47. Crabbe T, Welham MJ, Ward SG (2007) The PI3K inhibitor arsenal: choose your weapon! Trends Biochem Sci 32: 450–456. 19. Jiang JM, Chan YS, Loh YH, Cai J, Tong GQ, et al. (2008) A core Klf circuitry regulates self-renewal of embryonic stem cells. Nature Cell Biology 10: 353– U103. 48. Kingham E, Welham MJ (2009) Distinct roles for class IA PI3K catalytic subunit isoforms in the regulation of murine embryonic stem cell behaviour. J Cell Sci 122. 20. Cartwright P, McLean C, Sheppard A, Rivett D, Jones K, et al. (2005) LIF/ STAT3 controls ES cell self-renewal and pluripotency by a Myc-dependent mechanism. Development 132: 885–896. 49. Knight ZA, Gonzalez B, Feldman ME, Zunder ER, Goldenberg DD, et al. (2006) A Pharmacological Map of the PI3-K Family Defines a Role for p110[alpha] in Insulin Signaling. Cell 125: 733–747. 21. Ivanova N, Dobrin R, Lu R, Kotenko I, Lervorse J, et al. (2006) Dissecting self- renewal in stem cells iwth RNA interference. Nature 442: 533–538. 50. Niwa H, Yamamura K, Miyazaki J (1991) Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108: 193–199. 22. Boyer LA, Mathur D, Jaenisch R (2006) Molecular control of pluripotency. Current Opinion in Genetics & Development 16: 455–462. 51. Turinetto V, Orlando L, Sanchez Ripoll Y, Kumpfmueller B, Storm MP, et al. (2012) High Basal gH2AX Levels Sustain Self-Renewal of Mouse Embryonic and Induced Pluripotent Stem Cells. Stem Cells 30: 1414–1423. 23. Boyer LA PK, Zeitlinger J, Brambrink T, Medeiros LA, Lee TI, et al. (2006) Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 441: 349–353. 52. Savatier P, Huang S, Szekely L, Wiman KG, Samarut J (1994) Contrasting Patterns of Retinoblastoma Protein Expression in Mouse Embryonic Stem-Cells and Embryonic Fibroblasts. Oncogene 9: 809–818. 24. Storm MP, Kumpfmueller B, Thompson B, Kolde R, Vilo J, et al. (2009) Characterisation of the PI3K-dependent transcriptome in murine ES cells: Identification of novel regulators of pluripotency. Stem Cells 27: 764–775. References March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 14 Regulation of Expression and Mode of Action of Zscan4 Welham MJ, Kingham E, Sanchez-Ripoll Y, Kumpfmueller BK, Storm MP, et al. (2011) Controlling embryonic stem cell proliferation and pluripotency: the role of PI3K- and GSK-3-dependent signaling. Biochem Soc Trans 39: 674– 678. 32. Zhang W, Walker E, Tamplin OJ, Rossant J, Stanford WL, et al. (2006) Zfp206 regualtes ES cell gene expression and differentiation. Nucleic Acids Research 34: 4780–4790. March 2014 \| Volume 9 \| Issue 3 \| e89821 PLOS ONE \| www.plosone.org 15
https://openalex.org/W2469195175	https://scholarlypublications.universiteitleiden.nl/access/item%3A2861591/view	English	null	An Expanding Culture of Control? The Municipal Administrative Sanctions Act in Belgium	European journal on criminal policy and research	2,016	cc-by	11,378	An expanding culture of control? The municipal administrative sanctions Act in Belgium Devroe, E.; Bruinsma, G.; Vander Beken, T. Citation Devroe, E., Bruinsma, G., & Vander Beken, T. (2016). An expanding culture of control? The municipal administrative sanctions Act in Belgium. European Journal On Criminal Policy And Research, 23(1), 59-76. doi:10.1007/s10610-016-9321-2 Citation Devroe, E., Bruinsma, G., & Vander Beken, T. (2016). An expanding culture of control? The municipal administrative sanctions Act in Belgium. European Journal On Criminal Policy And Research, 23(1), 59-76. doi:10.1007/s10610-016-9321-2 Citation Version: License: Downloaded from: Publisher's Version Creative Commons CC BY 4.0 license https://hdl.handle.net/1887/42121 Version: License: Note: To cite this publication please use the final published version (if applicable). Eur J Crim Policy Res DOI 10.1007/s10610-016-9321-2 An Expanding Culture of Control? The Municipal Administrative Sanctions Act in Belgium Elke Devroe1 & Gerben Bruinsma2 & Tom Vander Beken3 # The Author(s) 2016. This article is published with open access at Springerlink.com Abstract This article provides an in-depth study of the Act on Municipal Administrative Sanctions 1999 (MAS), which is the first major piece of legislation regulating antisocial behaviour in Belgium. MAS provides municipalities with an instrument to sanction antisocial behaviour and conduct perceived to disturb public order. The article uses Garland’s (2001) thesis on the culture of control to analyse whether MAS has led to increased government control and the exclusion of significant groups of the population. The research is based on a multiple case study in which the application of MAS was analysed over a 25-year period of security policies in Belgium (1985–2010). The Act’s implementation was studied in the two Belgian cities of Antwerp and Liège in order to consider the influence of the Flemish government and the Walloon government, respectively, in this policy area. The article uses insights from this comparison to revisit the culture of control thesis and its limitations in understanding the political competition that exists over the formulation of policies on antisocial behaviour. Keywords Policing public space . Incivilities . Police . Administrative sanctions . Culture of control . Antisocial behaviour * Elke Devroe e.devroe@fgga.leidenuniv.nl Introduction In 1999 Belgium introduced the Act on Municipal Administrative Sanctions (MAS), which enabled municipalities to control antisocial behaviour through the imposition of administrative, * Elke Devroe e.devroe@fgga.leidenuniv.nl 1 Institute of Security and Public Affairs, Department of Governance and Global Affairs, Leiden University, Leiden, The Netherlands 1 Institute of Security and Public Affairs, Department of Governance and Global Affairs, Leiden University, Leiden, The Netherlands 2 Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), Amsterdam, The Netherlands 2 Netherlands Institute for the Study of Crime and Law Enforcement (NSCR), Amsterdam, The Netherlands 3 Department of Criminology, Criminal Law and Social Law, Ghent University, Gent, Belgium 3 Department of Criminology, Criminal Law and Social Law, Ghent University, Gent, Belgium E. Devroe et al. rather than criminal, sanctions. To some extent, this framework is comparable to legal initiatives in other countries (see, e.g. Devroe 2012b; Di Ronco and Peršak 2014; Crawford 2008, 2009; De Maillard 2002; Melossi and Selmini 2000; Frevel 2015). This article discusses the origins and development of this approach during the period 1985–2010. The more recent Municipal Administrative Sanctions Act of 2014 is therefore not included within the scope of this research.1 In particular, the article considers the political context of the introduction of MAS, its objectives, underlying policy presumptions and moral entrepreneurs. Furthermore, the implementation of MAS is examined in two case studies, namely, the municipalities of Antwerp and Liège. This article seeks to answer the question: BCan the development of law on antisocial behaviour in Belgium be explained in terms of Garland’s culture of control thesis?^ Garland’s thesis on the sociology of punishment in the Anglo American context, as presented in the book Culture of Control (2001), is used as a starting point for interpreting the origins and subsequent development of policy responses to antisocial behaviour in Belgium, given the influence of this thesis on other accounts of contemporary crime control in Western European countries (Boutellier 2007; Daems 2009; Snacken 2007; Swaaningen 2005). Garland’s (2001) work initiated the debate on the relative importance of social forces associated with the onset of late or postmodernity and of local factors in determining penal policy, also framed as the punitive turn (Daems 2008), penal regression (Radzinowicz 1991), pénalisation du social (Mary 1997, 2005), penal temptation (Wacquant 1998, 2004), new punitiveness (Pratt et al. 2005), actuarial justice (Feeley and Simon 1992, 1994) and risk justice and management (Swaaningen 1996). Introduction In this article Garland’s thesis is used to draw out the broader questions and themes implied by the introduction of MAS, specifically whether the political imperative for authorities to appear tough and present a punitive display, whilst pursuing more pragmatic measures for managing crime and civil unrest works as an explanation of the origins and subsequent application of MAS. The article therefore outlines the potential significance of Garland’s thesis and how it was used to design a multiple case study of the application of MAS in Antwerp and Liège. The results of the case study are then presented, followed by concluding thoughts on the explanatory value and generalizability of Garland’s thesis. 1 See: Devroe 2016; Cops and Pleysier 2014; Bauwens et al. 2016. David Garland’s culture of control We analysed the social and political context of MAS, the Belgian law on antisocial behaviour. Did MAS originate in a fear of crime, as suggested in Garland’s thesis? 1. We analysed the social and political context of MAS, the Belgian law on antisocial behaviour. Did MAS originate in a fear of crime, as suggested in Garland’s thesis? 2. We focused on the moral entrepreneurs of MAS in Belgium: Were they the professional groups who once dominated the policy-making processes but are increasingly disenfranchised as policy comes to be formulated by political action committees and political advisers (Garland 2001, p. 13)? Was MAS the result of existing ideological and reforming political powers or a product of new institutional arrangements developed as problem-solving devices to deal with Bthe practical experience of government agencies and their constituencies^ (Garland 2001, p. 72)? 3. We studied the policy presumptions underlying the regulation of disorder that could be situated between an expressive/symbolic dimension of the criminology of the other (acting out) or denial—both control dimensions (Garland 2001). Are criminologies of everyday life (rational choice, routine activity or opportunity theories) present in the policy assumptions of Belgian policy makers? 4. We wanted to learn more about the objectives of MAS: Are they the result of a new and urgent emphasis upon the need for security (Garland 2001, p. 13), or are they the result of Bmodest improvements at the margin, the better management of risks and resources, reduction of the fear of crime, reduction of criminal justice expenditure, greater support for crime victims, more expressive penal measures [that] have become the new policy objectives, as policy analysts deem it more realistic to deal with the effects of crime rather than address the thing itself^ (Garland 2001, p. 109)? 5. We wanted to understand the implementation of MAS at the local level and the social reactions of local institutions towards antisocial behaviour: Which different antisocial behav- iour measures were taken, and how were they implemented? We sought to analyse Garland’s statement that Bmeasures of expulsion and exclusion go hand in hand with the development of a network of preventative partnerships at local levels by patient, ongoing, low-key efforts to build up the internal controls of neighbourhoods and to encourage communities to police themselves^ (Garland 2001, p. 17), the latter being an element in the adaptive strategy (Swaaningen 2004). David Garland’s culture of control As this article explores governmental regulation of disorder, Garland’s analysis of govern- mental response to late modern issues of insecurity regarding crime is highly relevant. Central to his culture of control thesis is an analysis of governmental response to disorder and, in particular, the policy predicament. This can be observed as a contradictory mix of measures (see also O’Malley 1999), with on the one hand punitive segregation (control and exclusionary measures) and on the other preventative partnerships and increased informal control by the community. This policy of bifurcation contains no polarization (meaning a sensu stricto governmental response) but a simultaneous performance of contradictory tendencies. Garland identifies the crime-prone society as a main cause of this ambivalent reaction, clustering two basic strategies: the adaptive strategy and the sovereign state strategy. In this context Garland adheres to the fear of crime thesis: the adaptive strategy supposes that the government has to An Expanding Culture of Control? adapt to the reality of high crime rates and manage them; the sovereign state strategy denies that the government has lost its grip and constitutes the core of the control theory, with reactions of denial and acting out. Furthermore, he states that the emergence of a culture of control depends on the context of the crime situation and political context, policy assump- tions, moral entrepreneurs (advocates, stakeholders) and objectives and criminological theories behind late modern law enforcement. He identifies the loss of penal welfarism in favour of retribution and deterrence as one of the main causes. The welfare mode became more risk prone and offence oriented, and control theories such as the criminologies of everyday life and the criminology of the other began to dominate criminological discourse. Last but not least, the economic style of reasoning and the neoliberal way of thinking are, according to Garland, the main causes of an expanding culture of control. Despite the fact that Garland does not fully expand on the consequences of his thesis for local security governance, his analysis of preventative partnerships is important for framing the empir- ical results in this article. He places these partnerships within the control framework (Daems 2009) and depicts them as a mix of informal social controls embedded in everyday activities and the interactions of civic society. We conclude this section with five research questions that are addressed in the article: 1. Methods and data collection The research is a multiple case-study design with triangulation of methods. We wanted to obtain an insight into how and why MAS emerged and was implemented (Orum and Feagin 1991), and a detailed understanding of the institutional political contexts, cultures, historical traditions and governmental arrangements (Newburn et al. 2002). As Belgium has a very specific constitutional/legal setting, it is important to examine general theses about crime control, such as Garland’s, with reference to the control policies that apply in the different regions of Belgium [the political specificity of the environment (Mcara 2005)], on the presumption that the politics of this regional, subnational government will help to explain the origins and development of these policies. Belgium has about 11 million residents, multiple language regions (60 % Flemish speakers, 40 % French speakers and a small proportion of German speakers) and five large municipal- ities.2 Since the state reform of 1970, Belgium consists of three linguistic regions, 11 provinces, 27 judicial districts and 589 municipalities. As a consequence of the state reform in 1988, Belgium became a federal state. Given this historical evolution of the Belgian federal state and its division into and enmities between Flemish/Dutch-, German- and French- speaking community councils and the regional governments of Flanders, Wallonia and Brussels-Capital, these regional authorities have developed different approaches to local security (Devroe 2013). The Flemish Region consists of 308 municipalities, the Brussels- Capital Region 19 and the Walloon Region 262, each steered by the mayor who directs the (local) police to act in close contact with the citizens and to practise community policing (Devroe 2013). A key consequence of the 1988 federal constitution has been a power shift from national to regional and municipal authorities (Devroe 2013). Only a few policy domains remained exclusively within federal government competence, such as the maintenance of public order (police) and law enforcement (justice). Competencies relating to social policy— more specifically, education, employment, health, living conditions, culture, housing, spatial planning—became the responsibility of the regional governments. Because the causes of crime are often situated in regional policy domains, we included regional policy issues in our analysis. Whilst federal and regional authorities retain responsibility for formulating stated goals on security policy, policy implementation (policing in action) (Wodak 2011) became the responsibility of the municipalities (Devroe and Ponsaers 2005). David Garland’s culture of control This last question will be answered by analysing the local case studies in Antwerp and Liège. E. Devroe et al. 2 Antwerp and Ghent (Flemish speaking), Charleroi and Liège (French speaking), and Brussels (both languages). 3 Renewed by the New Communal Law (Nieuwe Gemeentewet), B.S., 3 September 1988. p ( p g) g ( p g) ( 3 Renewed by the New Communal Law (Nieuwe Gemeentewet), B.S., 3 September 1988. ntwerp and Ghent (Flemish speaking), Charleroi and Liège (French speaking), and Brussels (both languages). enewed by the New Communal Law (Nieuwe Gemeentewet), B.S., 3 September 1988. 2 Antwerp and Ghent (Flemish speaking), Charleroi and Liège (French speaking), and Brussels (both l 3 Methods and data collection To draw upon a classic distinction between steering and rowing made in the study of governance (Osborne and Gaebler 1993), regional authorities steer social policy responses (policy formulation, e.g. agenda setting, allocation of resources) to quality of life issues; federal authorities steer law enforcement and maintenance of public order; municipal authorities are obliged to row (i.e. implement) both criminal justice and social policy approaches. Consequently, it is at the municipal level that these competing federal and regional responses are negotiated and then executed (Devroe 2015b). In Belgium, the mayor is an independent political figure in each municipality, and Belgian mayors are—like their French counterparts—in charge of and responsible for public order, public peace, security and public health. The mayor can establish general police regulations in his/her municipality and recruit the public police (called admin- istrative police) to determine and record violations. The Communal Act of 30 April 18363 created an extensive legislative municipal framework to tackle problems of public disorder An Expanding Culture of Control? under the patronage of the federal government. It is important to mention here that from 1991 until 2014 the Socialist Party ruled in a federal coalition. In the period covering MAS’s genesis (1988–1999), all ministers of interior (federal), municipalities (federal) and cities in Flanders and Wallonia (regional ministers) were socialists. Data processing for this study was oriented around two main topics: (1) the genesis and (2) the implementation of MAS. In case study A, the genesis was analysed at a federal level over a 25-year period (1985–2010). Data collection was organized around the five research questions (supra). Political reactions to increasing problems of urban antisocial behaviour were analysed in a broader historical context of social and governmental changes in security policy over the study period. This historical reconstruction involved 26 semistructured face-to-face interviews4 (Fontana and Frey 1994), with experts (Aberbach and Rockman 2002) from different policy levels. All ministers of interior and justice who served during the research period were interviewed. In 1999, during the presidency of Prime Minister Verhofstadt I, the position of Government Commissioner of Metropolises was introduced. In 2000 this function was transformed into the Minister of Metropolises, which was installed to steer and finance renovation and social cohesion projects in the five major cities in Belgium. These successive ministers were also interviewed, along with representatives of their cabinets, members of parliament, political advisors, governors and civil servants. 4 All interviews were completely transcribed, and the 72 interviews have been kept on tape in archives. A list of names of all respondents and their functions has been published (Devroe 2012a). A selection of quotes from the interviewees is used in this article. 5 A list of the documents has been published (Devroe 2012a), and all documents used are stored in archives. 6 This information was used in the analysis, and the source was referred to as Bconfidential document’. Methods and data collection We gathered significant informa- tion on how the perceptions of these policy experts led to certain policy assumptions, motivations, basic principles, statements and, finally, actions. We chose a semistructured style of interviewing in order to gather a diversity of significant answers at a sophisticated level. 5 In addition to interviews, an extensive document analysis5 (N: 432) of official and confidential policy documents was carried out. Analysis included governmental programmes and policy plans; national and local security plans; reports of the Council of Ministers; the Cabinet of the Ministers of the Interior, Justice and Metropolises; statistical reports; annual reports; elective programmes; reports and informal notes of regional councils; preparative documents for Acts; personal letters between ministers; and so on. Thanks to ministers’ and cabinet members' assistance, we were able to gather exclusive, confidential6 political material that had not been revealed before (e.g. minutes of meetings in parliamentary committees, informal and formal letters between ministers, minutes of intercabinet meetings). In order to study the implementation processes of the federal MAS and possible diverging policy instructions from regions at the metropolitan level, two nested case studies (based on differences) in Antwerp (Flemish speaking) (case study B) and Liège (French speaking) (case study C) were added to the analysis. Both cities implemented MAS in 2005, so the relevant period for an analysis of the uses of MAS in these cities is 2005–2012. We chose these particular cities in order to identify any intraregional differences originating in linguistic political influences on the one hand (both cities’ political representatives are currently left wing, but 25 % of voters chose the extreme right-wing party Vlaams Blok in Antwerp in 1991; voters in Liège have consistently voted for left-wing representatives) and political culture (Flemish or Walloon) on the other. Although the population size of Antwerp and Liège varies greatly, they both face a massive inflow of commuters (work and tourism) on a daily basis. In both cases, we focused on policy E. Devroe et al. makers from the particular institutions responsible for shaping and conceptualizing antisocial behaviour strategies and civil servants who are responsible for implementation of MAS. This approach led us to actors in prevention and social welfare and to institutions of law enforce- ment (police, justice). 7 All interviews were completely transcribed and kept in the archives. Names of respondents and a list of all documents used in the analysis have been published (Devroe 2012a). 8 Called the BOctopus agreement’ because suddenly the eight political parties all agreed on the Police Reform Act (7 December 1998, B.S. 5 January, 1999). Methods and data collection In each city, we held 23 face-to-face interviews7 (in Dutch in Antwerp; in French in Liège) with mayors, public prosecutors, police constables and chief commis- sioners, social workers, wardens in public transport, city units responsible for the prosecution and sanctioning of antisocial behaviour, members of the local council (college of alderman), city wardens, city coachers, mediators, street-based social workers, those piloting neighbourhood projects, etc. These interviews were complemented with an extensive docu- ment analysis for each city, including municipal statistics on antisocial behaviour, policy plans of majors, local security plans, police plans, antisocial behaviour plans, minutes of the city council, activity reports of institutions, reports of city wardens, and so on. MAS genesis Regarding the social and political context of MAS (question 1), most respondents referred to various events and scandals that shook Belgium’s political institutions and challenged their authority to cope with issues of crime and safety in the late 1980s. The 1980s were described as a black page in Belgian history, marked by the extremely violent raids and killings by the Nivelles Gang (Dupont and Ponsaers 1988), terrorist attacks by the communist Cellules Communistes Combattantes and the deaths of 39 football supporters at the 1985 European Cup Final in the Heysel Stadium. In addition to those events, many other incidents during that period cast a shadow over police and the justice system. Respondents also noted a growing concern about the influx of migrants in the larger cities, a rapidly changing demographic situation in the capital cities and the political success of the extreme rightwing party (Vlaams Blok) in Flanders. On 24 November, 1991, Vlaams Blok obtained 6.6 % of the votes, or 12 of 212 seats in Parliament. This day is now referred to as Black Sunday, a day on which more than 100,000 voters formally expressed their lack of trust in the existing political system (Huyse 1996). Considered a state legitimacy problem, this electoral shift stimulated political awareness of problems of social disorder and minor crimes in deprived neighbourhoods (Van Limbergen 1995). The last wave of incidents occurred in the mid to late 1990s with the 1996 arrest of Marc Dutroux, a paedophile who imprisoned and murdered children. His arrest was accompanied by a massive public expression of solidarity with the children’s parents. The White Balloon March in Brussels in 1996, attended by 300,000 people, was a public statement against the incompetence of police and the justice system. In April 1998 Dutroux briefly escaped during a prison transfer, leading to moral panic in the country and resulting in the unavoidable resignation of the Ministers of Interior and Justice. According to many respon- dents, the police reform that was quickly voted in after this escape8 was considered by all political parties to be necessary to the recovery of governmental legitimacy. An Expanding Culture of Control? A second stimulus for MAS’s genesis was the state reform of 1988 (supra). MAS genesis Following this reform, the Minister of Interior, Louis Tobback, called the Cabinet of the Interior Bthe preservation room for grumpy old men with nothing more to decide.^ The election programme of the Socialist Party9 showed, following state reform, that socialist politicians did not want to be constrained by the party's very narrow range of policies, which stressed repression and public order. They expressed a strong belief in the need for tailor-made support for major municipal areas, and wanted to formulate and implement new policy measures that tackled the causes of crime, even though such tasks had, since 1998, been allocated to the regions and so were officially outside their political sphere. Furthermore, the electoral shift to the extreme right led to a political awareness among socialists that something had to be done to equip cities with the right tools to tackle the deterioration and collapse of social cohesion. Prevention measures (situational, but also social) became key strategies in tackling crime and feelings of insecurity. The federal government financed many social projects, including drug rehabilitation programmes, situational prevention projects such as street cleaning and litter removal, neighbourhood centres and street-based social workers. Policy makers at that time were convinced that punitive answers to minor crimes could only exacerbate the issues at hand (De Witte 1988). Political assumptions were closer to those of the Broken Windows theory (Wilson and Kelling 1989), transforming cities in attractive and clean places, than to deterrence and punishment that was oriented towards individuals (mostly young people), as had been the case after the murder of Jamie Bulger in the UK. Belgian laws were not obviously directed at offensive behaviour (especially by young people), which has featured so significantly in policy in the UK. Because antisocial behaviour is legally considered to be a part of public order problems, the debate on antisocial behaviour that characterized UK debates was not seen in Belgian political debates. The Belgian law is called Municipal Administratrive Sanctions, which reflects its focus on public nuisances (trespass, mess/litter, environmental harms—with the mayor being responsible for maintaining health, peace and order in the city) rather than possessing the almost subcriminal character that antisocial behaviour laws have acquired in the UK (Devroe 2016). Document analysis and interviews clearly reveal that socialists’ reactions to police and judicial malaise was inspired by the French politician Bonnemaison,10 to take a preventative approach. 9 De Witte (1988) BSecurity and petty crime’ (Veiligheid en kleine criminaliteit), an electoral document for the municipal elections. 10 A 1982 report Face a la délinquence, prévention, répression, solidarité, written by Gilbert Bonnemaison and 26 municipal mayors, explored crime in metropolitan areas and prevention measures for young people. This report led to the establishment of prevention councils and prevention contracts in France and was the example for the Belgian prevention policy contracts. 11 Minister of the Interior at that time. 12 Directive of the Minister of the Interior of 12 November 1990 to the mayors on the prevention of crime, and Directive of the Minister of the Interior of 30 November 1990 to the mayors on the security and prevention contracts. 12 Directive of the Minister of the Interior of 12 November 1990 to the mayors on the prevention of crime, Directive of the Minister of the Interior of 30 November 1990 to the mayors on the security and prevent contracts. e Witte (1988) BSecurity and petty crime’ (Veiligheid en kleine criminaliteit), an electoral document for the nicipal elections. MAS genesis The federal government approved of the intentions of Louis Tobback,11 transcribed in the ministerial declaration BSecurity of the citizen, police and security^ of June 1992. One of the main propositions in this document was extensive funding for crime prevention12 (Glorie 1997, 2004; Willekens 2008). In order to get a new grip on city governance after the state reform, city contracting (like in France; see Roché 2005), was installed. After consulting the five mayors of the major cities (all socialists) about the amount of federal funding, so-called prevention contracts were signed between the Minister of the Interior and those mayors. The investment consisted mainly of personnel: prevention workers were installed who introduced E. Devroe et al. unemployed, low-skilled citizens into entry-level jobs, such as football coaches, stewards, city monitors, park guards, parking-lot guards and other prevention personnel. This was the first occurrence of a cohort of civil surveillance wardens in public spaces executing tasks such as warning, helping citizens with enquiries and noting down complaints about public infrastruc- ture (Devroe 2015a). As most of these workers were located in the same building as the public police housed within police offices, some were—like the public police—employees of the local authority. One year prior to voting on MAS in 1998, a total of 123 million euro was invested in the cities, installing a fragmented body of surveillance wardens on public streets and squares, helping police to keep peace and order. In 2000, the Minister of Metropolises (also socialist) followed his colleague at the interior and invested a significant amount of funding in the five largest cities, more specifically in social cohesion projects,13 resulting in an increasing body of social workers such as street-based workers and mediators in public spaces. Confidential letters between Ministers of the Interior and of Metropolises14 indicate that the positioning of civil wardens in public spaces—an important aspect of the prevention con- tracts—was the essence of what would become MAS the year after. This clearly illustrates the fact that MAS originated as a preventative and not a criminal law enforcement measure (as had the Antisocial Behaviour Orders (ASBOs) in the UK; Matthews et al. 2007; Field 2003). We specify this last statement in the answers to our second and third queries—namely, moral entrepreneurs, policy assumptions and MAS objectives—which we analyse in the following paragraphs. 13 Act of 17 July 2000 on the requirements municipal authorities have to meet in order to get funding stipulated in contracts with the Ministry of Metropolises, B.S. 8 August 2000. 14 The Department of Metropolises was socialist for the entire period, from its establishment in 1999 until 2008 (Ch. Piqué, J. Vande Lanotte: 1999–2003, M. Arena: 2003–2004 and Ch. Dupont 2004–2008). 15 To give a concrete example, when the Socialist Party no longer occupied the position of Minister of Interior, the socialist (French) Minister of Justice (from 2003–2007) became—together with the Flemish socialist Minister of Finance and the French Minister of Metropolises—moral entrepreneurs for the successive MAS Acts of 2004– 2005. MAS genesis Analysis of the extensive body of policy documents dating from the crucial period before the voting for MAS (1999–2005) and the interviews leads to the conclusion that the socialist body of thought (the belief in the feasibility of society, in state engineering) was transferred across key positions, governmental levels (federal and regional) and even across linguistic tensions.15 The in-depth political reconstruction of communication between different stake- holders shows that the sole moral entrepreneurs were the socialists (politicians and cabinet members), installing MAS as a last resort following the preventive financial injection that political party officials had implemented from the early 1990s onwards. Their main objective was to install formal and informal surveillance in public spaces in order to address public feelings of insecurity and to establish what could be called a city code of conduct by implementing MAS. There were no theoretical assumptions (our third question) underlying this decision: the approach to antisocial behaviour arose from the pragmatic experience of politicians, called a rational-within-limits view (Garland 2004). Only the Minister of Metropolises asserted that he had analysed a number of French scientific studies on urban crime problems, such as Bonnemaison. Though the installation of a body of civil city wardens could refer to capable guardianship (Felson 2002) and so fit in the opportunity theory, the MAS approach could not be called a demoralizing effect of situational crime prevention where the new approach does An Expanding Culture of Control? not look to the revival of social controls and self-controls. Instead, it seeks to make a new style of ‘situational engineering’ succeed where ‘social engineering’ failed^ (Garland 1997, p. 189). The objectives (question 4) of MAS were—after 10 years of federal investment in crime prevention and a penal system that completely ignored sanctioning municipal violations—to counteract impunity by providing municipalities with a legal instrument to sanction incivilities. By broadening city council competences to sanctioning, municipalities were put in control of the entire chain of enforcement policy: establish regulations, prosecute and punish (Van Heddeghem 2006). As the Flemish government had competence over sanctions and justice but wanted to decrease crime and disorder, the Flemish Minister of Cities (socialist) also invested in contracting (Social Impulse Funding, Sociaal Impulsfonds) in order to install social policies of city renovation and social cohesion projects in exchange for (Flemish) funding. 16 Decree of the Walloon Government of 15 May 2003 on prevention and proximity in cities and communes in Wallonia, Ministerial Decision, 4 February 2004: 6701. 17 Act of 13 May 1999 on Municipal Administrative Sanctions, B.S. 10 June 1999. MAS genesis The following quote illustrates the intention: As we had no competences on problems of crime, and we wanted to reorganize crime- prone cities into viable, agreeable spaces, we tried to invest in citizens’ participation. Of course this was done to avoid a conflict between our official (narrow) competencies and our own policy goals. This Flemish money was only meant for the so-called deprived neighbourhoods. So you were obliged to work in certain bad neighbourhoods, which were overpopulated, with a great mix of cultures and a lot of integration problems and sociodemographically and economically floating beneath the societal surface. And those you have to renovate, so you organize a competition to find out what people think it is important to have in public space (civil servant of the Flemish cabinet). In 2003, almost 7 years after the Flemish investments, Wallonia16 was also able to provide funding for Walloon cities following the signing of contracts between the cities and the Walloon executive that guaranteed the use of prevention-based approaches. This gave a great boost to Flanders, providing infrastructure improvements that were supported by local citizens, and contributed to a reduction in feelings of insecurity (Devroe 2012b). Therefore, if MAS is viewed in isolation, its objectives can be considered to be Bmodest improvements at the margin, the better management of risks and resources, reduction of the fear of crime, reduction of criminal justice expenditure, greater support for crime’s victims^ (Garland 2001, p. 109). We now elaborate on MAS itself. Regulating antisocial behaviour: MAS Tackling antisocial behaviour is a municipal matter in Belgium, directed by the mayor and executed by the public police. This practice, together with the strict division between admin- istrative (public order) and judicial police, is particular to the Belgian constitutional setting stemming from a history of local autonomy (Devroe and Ponsaers 2013). MAS17 provides local councils with an appropriate instrument enabling them to swiftly and forcefully tackle antisocial behaviour on municipal land. As the Belgian legislator considers antisocial behav- iour to be an element of disturbing public order, MAS was embedded in the existing Municipal Act of 1999, creating three innovations: (1) the ability to close down establishments and E. Devroe et al. suspend licences, (2) enlargement of the notion of public order to antisocial behaviour18 and (3) competence for the mayor and alderman to sanction (with fines up to 340 euro) infringe- ments of municipal regulations. MAS combines administrative measures (briefly, measures taken by the mayor, such as closing down bars and discotheques, withdrawal of licences, issuing fines) and administrative sanctions imposed by the city council (mainly fines). Before the amendment in 2004, municipal authorities could only react to the types of behaviour for which neither the federal nor the regional parliament has provided penalties. In other words, municipalities could only respond to new types of antisocial behaviour that had not already been addressed by existing legislation.19 In practical terms, local administrative sanction legislation proved to be an Bempty barrel^ for municipalities. As any overlap with existing provisions elsewhere was formally excluded, local councils did not have much room to manoeuvre (Vander Beken and Vandeviver 2014). Moreover, the Act of 1999 allowed municipalities to draft their own laws but not to appoint designated enforcement staff. Antisocial behaviour could only be established by a report made by police (Devroe 2008). Before MAS, sanctioning antisocial behaviour was the responsibility of the judicial authority (police courts), creating considerable congestion in the public prosecutor’s daily workload. For political reasons—namely the belief in tailor-made local governance with a strong emphasis on municipal autonomy—MAS was meticulously amended by the federal parliament in 2004, leading to two new acts in 2004. In the preparation of this act we noticed—for the first time—a strong voice (Minister of Justice L. Onkelinx, also socialist) underlining the need for manda- tory mediation before punishment. 18 Art. 135 par. 2 al 7 MAS. 19 This means not only federal legislation but also regional decrees on environmental issues, town and country planning, construction companies, residential inspection, etc. 20 Parliament, Chamber 1, 2002–2003, number 2366/1 and 2367/1, 3–5. 18 Art. 135 par. 2 al 7 MAS. MAS implementation in two cities Regarding the fifth research question on the mix of informal social controls embedded in the everyday activities within preventive partnerships at local levels, analysis of 46 interviews and numerous documents led to some specific results. Questions such as: BWhat are the rules of discourse and action that organise the diverse practices that made up this field?^ BHow are these rules and these practices related to those of other social domains, such as welfare, politics or the economy?^ BWhat role do these practices play in the governance of late modern society?^ (Garland 2001, p. 2) were analysed in both cities. The way in which MAS was implemented was reviewed—whether top down, meaning mandatory as imposed by city hall, or bottom up, originating in small initiatives within civic society. During the empirical search, attention was paid to the ways of thinking that underlie the programs (mentalities), the organisational forms used to implement those ways of thinking (institutions), the techniques used to turn mentalities into action (technologies) and, of course, results (practices) (Johnston and Shearing 2003, p. 13). Results show a great variety in both cities. In Antwerp, we noticed a very enthusiastic and immediate application of MAS, welcomed by city hall as an excellent opportunity to govern the city, with extra tools for the mayor to execute local security policy. Wardens were able to initiate antisocial behaviour reports, and the administrative staff had the authority to sanction. The local police code was completely overhauled, with administrative sanctions for many minor offences ranging from traffic offences; to antisocial violations at markets, restaurants, bars and sports events; to begging and drinking in public spaces; and hanging out, insulting and intimidating others. Under the socialist mayor P. Janssens, a special city department called Living Together was created, which initiated surveillance and control, street-based social work and educational programmes. The whole security chain, from prevention to mediation, to prosecution and sanctioning, was coordinated by one director in what was called an integrated approach. As had been found when assessing the objectives question, socialist ministers intended to treat MAS as a final option after preventive measures; handing out fines was seen as a final stage when prevention approaches had failed. Mediation is in place for young people. 21 Act dated 20 July 2005 establishing various provisions—matters pertaining to the civil service and urban policy—municipal administrative penalties (art. 21–22), B.L.G, 29 July 2005. An Expanding Culture of Control? An Expanding Culture of Control? The second series of criminal offences (the so-called more serious mixed violations) results in penalisation under traditional criminal justice, unless the public prosecutor’s office indicates that it would be more appropriate to impose a pecuniary fine. With regard to this series of criminal offences, the municipal civil servant can only impose an administrative pecuniary fine if the public prosecutor has advised him/her within a preset timespan that the violation shall not be pursued any further and that he/she feels it would be appropriate to impose an administrative penalty. If the public prosecutor fails to communicate his/her position, and the provided time period has expired, the local council is no longer at liberty to act in an administrative sense, which means the violation must be settled in a criminal justice context. It is up to each municipality to draft a list of violations that are punishable by MAS. The 2005 revision of the act21 made it possible for municipal officers to draft reports and sanction all municipalities’ regulations, in addition to the public police. Regulating antisocial behaviour: MAS These acts treat minors (aged 16–17) as eligible for sanctioning using MAS, but only after mandatory mediation (Vander Beken 2005). This revised act combines administrative sanctions with criminal procedures in a particular way. In order to hand over sanctioning competencies to the municipality, a series of violations were depenalized. These violations were no longer delimited by other legislation and could be integrated into municipal procedures. A second group of criminal offences are referred to as mixed violations, i.e. behaviour that qualifies as both a criminal offence and an offence written down in municipal regulations. These are not depenalized, but the municipal council is at liberty to issue a pecuniary fine of up to 250 euro for these offences in its byelaws, as the existing penalties provided under the penal code remain untouched when the public prosecutor decides to pursue a case. The penalties already included in the penal code are in turn divided into two separate procedural settlements in which the official police report is required to be sent to the public prosecutor’s office in all cases. With regard to the first series of criminal offences (the so-called lighter-mixed violations), administrative pecuniary fines can only be imposed when the public prosecutor does not communicate his position in the case to the civil servant within 1 month of receiving the police report. If the public prosecutor has allowed this term to expire without making his or her position known, offences under this first series of criminal offences can only and exclusively be punished administratively, with the local council at liberty to take up the prosecution of this offence. Of course, an individual cannot be punished twice for the same offence (non bis in idem).20 MAS implementation in two cities According to the mayor of Antwerp, the director of the Living Together department and civil servants working in city hall, the aim of mediation is to change the attitudes and conduct of young people who exhibit E. Devroe et al. antisocial behaviour by providing a variety of programmes and services (educational, aggres- sion management, sports facilities, assistance at school, etc.) in order to avoid sanctions (fines). Some respondents saw this as a positive early intervention; others saw it as subjecting the young people to intrusive control in the name of diversion or prevention, forcing them into welfare. In Antwerp, young people who are following the programme in order to avoid sanctions may be subject to net-widening effects, because what was considered normal adolescent behaviour of short duration before MAS is now labelled as a problem that requires special attention and care, and—if a young person does not complete a prevention pro- gramme—is regarded as worthy of a fine. antisocial behaviour by providing a variety of programmes and services (educational, aggres- sion management, sports facilities, assistance at school, etc.) in order to avoid sanctions (fines). Some respondents saw this as a positive early intervention; others saw it as subjecting the young people to intrusive control in the name of diversion or prevention, forcing them into welfare. In Antwerp, young people who are following the programme in order to avoid sanctions may be subject to net-widening effects, because what was considered normal adolescent behaviour of short duration before MAS is now labelled as a problem that requires special attention and care, and—if a young person does not complete a prevention pro- gramme—is regarded as worthy of a fine. Antwerp uses the bargaining model of contracting that is widely used in Anglo Saxon countries, showing characteristics of a multiagency approach with fixed horizontal networks. The preventative partnerships are a core aspect of antisocial behaviour governance, linking actors from different backgrounds and policy domains in one network, and considering the creation of personal dossiers of young offenders to be a policy priority for the whole city. These networks are not mandatory in the way that Crime and Disorder Reduction Partnerships are in the UK (Edwards and Prins 2014) but can be understood within a broader tendency towards horizontal security governance instead of top-down government. In Liège we found a completely different interpretation of MAS. MAS implementation in two cities As policy decisions there are based on the Canadian approach, there is a fundamental belief in the importance of high- profile community-oriented policemen. City wardens cannot issue fines; they only assist the police in monitoring signs of disorder. The policy assumption is that that there is no need to create a kind of municipal criminal justice system when a transparent legal justice system is already in place. In the view of the mayor, the chief of police and many other experts responsible for public order and antisocial behaviour, warnings (oral or written) from the police are the best remedy against frequent disorder, and there is no need for extra fines. Therefore, the Liège municipal regulation code remained limited to violations at specific markets, festivities and sports events. Mayoral measures (see above) are not included in the code and cannot be executed, nor are exclusion measures such as street bans or night curfews. According to respondents, too many prohibitions would hinder inhabitants and restrict their privacy. Analyses showed a total rejection of the use of mediation as a tool to avoid fines. Once a case is registered and proven, a fine is given, even after (successful) mediation. In compar- ison with Antwerp, the act of 1999 is rarely applied, the police code was not substantially revised, and punishment of antisocial behaviour by the classic law enforcement agencies of police and justice continued as before. Liège City Hall carries out specific projects (e.g. street cleaning, targeted alcohol control, free bus passes) and connects actors on an occasional basis for a more integrated approach. These connections, however, are not integrated into a specific framework and often dissolve following the project. Collaboration is based on the initiative and willingness of participants, as there are no official agreements. Tasks related to welfare and social cohesion (completed by social workers) are clearly distinguished from tasks involving prevention and sanctioning (executed by police and justice). Both cities illustrate a bottom-up approach, initiated from a small number of ideas that took shape and were formed in a later phase, confirming Garland’s statement that BThey subsequently became de facto strategies attracting widespread support and multiple ex post facto rationales, the plan followed the practice rather than the other way around^ (Garland 2001, p. 105). It is clear from this study that Antwerp implemented MAS wholeheartedly and within the broader objectives of an integrated security policy. Conclusion and discussion As difficult as a comparison and the translation of criminological concepts can be (Edwards and Hughes 2013; Edwards et al. 2013; Nelken 2015), our research shows the importance of empirical research on control theories in different contexts and on different phenomena. In this article we broadened Garland’s insights on a phenomenon that he did not analyse empirical- ly—namely, issues of public order in public space. As this space became scarce (Hope 2000) the quality of life discourse gained importance (Crawford 1999, 2001; Edwards and Hughes 2002), putting the regulation of disorder in the political spotlight. The Crime and Disorder Act in the UK, regulating public behaviour and mostly targeting individuals who disturb public order, was introduced (1998) before the culture of control (Garland 2001) came about. The Criminal Justice and Public Order Act (1994), introduced by Michael Howard (Home Secretary under a Tory government) aimed to criminalise individuals with an unconventional lifestyle, such as gypsies, squatters and ravers (Hughes 2007); the 1994 New Labour policy plan BA Quiet Life: Tough Action on Criminal Neighbours^ (Squires and Stephen 2005) tempted us to broaden Garland’s insights into this public order domain by piloting extensive empirical research with more than 400 policy documents and 72 in-depth expert interviews. Regulating public space is an important task of mayors in most Western countries and can lead to exclusionary mechanisms (Young 2007, 2009). A digression on this important semiauton- omous field (Bourdieux) of control in public space in Garland’s culture of control would have been valuable, even more so since stringent modes of control in this area were already installed in the UK before the publication of his book in 2001. Applying Anglo Saxon penal insights to the continental setting of a civil-law (federal) country with a complex constitutional setting was a challenging enterprise. We believe this multiple case study research design was justified in relation to the proposition that politics matters (Stone 2008) in explaining the origins and development of control strategies; and in Belgium, that implies a concern with the contrasting politics found in the different regional governments as a consequence of the Belgian constitutional/legal settlement. MAS implementation in two cities In both of the two case studies, local An Expanding Culture of Control? politics was effective in implementing policies in ways that the federal government intended. In Antwerp, MAS is an important instrument, but only as the last part of an integral local policy chain. Liège was more reluctant to implement MAS, and only uses instruments that are already available in the criminal law enforcement system to regulate antisocial behaviour. Liège mostly invests in social welfare and prevention approaches instead of issuing fines. Conclusion and discussion In Belgium, there was a lack of confidence in crime-fighting models and criminal law enforcement approaches, such as the UK’s ASBOs and Criminal ASBOs (CRASBOs) (Bottoms 2006; Burney 2005; Card and Ward 2008) or Australian control orders (Zedner 2007) in the period covered by this research, with the socialists calling instead for an approach that would lead to a better quality of life in a preventative manner. As one of the causes of MAS was the flight to the extreme right in 1991 and the complete absence of justice in the regulation of minor crimes that citizens complained about, it is obvious that socialists wanted their electorate back, which they managed to do for 12 years in federal government. The two case studies (Antwerp and Liège) retained socialist mayors from 1976 until 2013, when a right-wing mayor took over the mayoral sash in Antwerp (but not in Liège). By ignoring minor offences and antisocial behaviour that disrupted people’s lives, fears of a second extreme right electoral turn became a reality. To conclude: local politics do matter. Our results elucidate the second factor in Garland’s analysis, the local factors determining control policy more than they do the importance of social forces associated with the onset of late modernity. Socialist mayors sitting in parliament from all five major cities in Belgium shaped the intentions, the assumptions and the objectives of MAS, together with their Socialist Party members at a federal level and at both regional levels. As this act is a nonmandatory instrument that is available for all mayors in their task of keeping peace and order, it is to be expected that politics will be an important factor. Although the act sets structural legal limitations not to punish behaviour that is already penalised in a federal or regional act, net widening can be a dangerous side-effect of this act, broadening the control perspective. Although our empirical observations explicitly showed no reason to believe that risk assessment and control measures were the intention of the original (socialist) moral entrepreneurs, local mayors can decide otherwise. As the Socialist Party had completely vanished in federal coalition and in Antwerp, right-wing and conservative parties demonstrated diverging policy agendas and were able to implement diverging strate- gies. Recently, Belgian austerity policy has led to huge budget cuts, specifically in social insurance and other programmes of social care, welfare and prevention. Conclusion and discussion The detailed reconstruction of the genesis of MAS and of, more generally, Belgian security policy from 1985 to 2010, broadened the scope of classic criminological research (Bruinsma and Huisman 2006) (which, when analysing national acts, usually focuses exclusively on the enforcement of criminal law measures that have been set out by ministers of the interior and of justice). The link with fragmented regional and local competences due to federalisation processes when studying a national act is rarely made: as Hamilton (2014) put it when discussing the fact that research has examined institutional arrangements, national culture and the interaction between the two, the impact of a jurisdiction’s political culture on penal policy has remained a largely underexamined topic. Prohibiting government officials in charge of police and justice from creating social approaches such as restorative justice and social prevention (state reform) sheds light on the intentions of governmental state actors. As we did not limit ourselves to the police senso stricto approach (the act installed by the federal government) but studied the political E. Devroe et al. dynamics between federal, regional and local actors in the common Bpolicing goal of making cities clean, liveable and attractive again, the detailed reconstruction revealed some very unexpected results. dynamics between federal, regional and local actors in the common Bpolicing goal of making cities clean, liveable and attractive again, the detailed reconstruction revealed some very unexpected results. First, we found a coherent rationale among those who create and implement policy that is, to a certain degree, contrary to Garland's thesis, which emphasises volatility and ambivalence (see also O'Malley 1999); and the regulation of antisocial behaviour was strongly influenced by the ideological body of thought of Socialist Party officials (French and Dutch speaking) who shared a strong belief in social prevention and, as an ultimum remedium if prevention and warnings fail, a local tailor-made instrument in the hands of the city council to sanction frequent antisocial behaviour. We believe this socialist ideology to be the main element of resistance against purely punitive reactions in the Belgian political context. Conclusion and discussion Disconnecting local (sanctioning) policies from local preventative and welfare policies could turn MAS practices into a purely repressive and punitive arrangement. Second, Belgian socialist officials do not comply with regulations for state reform, which is only possible in the typical Belgian political-bargaining consensus model of governance. Our analysis showed that although the Minister of the Interior (an empty barrel) formerly had no say over municipalities and certainly not in the domain of social prevention (only in terms of repression), in fact, the minister found a way to govern municipalities from a distance by Bcontracting politics, so imposing on them a preventative and welfare-oriented ideology. This Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and repro- duction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. An Expanding Culture of Control? was not, as in the UK, about creating networks. In fact, the minister found a way to govern municipalities from a distance by using a community safety approach, so imposing on them to create crime prevention networks22 As prevention officers were mostly located in police stations, it was more about transferring processes of social justice and prevention to a tailor- made level: cities. This rather unusual approach, which is due to Belgium’s complex consti- tutional setting, touches on Lijphart’s (1999) distinction between consensus and majoritarian democracy as a key element in explaining the country’s moderate model of regulating antisocial behaviour. In conclusion, it can be observed that MAS offers municipalities a tool to steer local problems by providing administrative fines and is accompanied by tailor-made prevention programmes that are not oriented towards a dangerous other but are relevant to all municipal inhabitants. MAS cannot be called an expressive/symbolic dimension of criminology of the other (acting out) or as an illustration of a stronger repressive governmental grip. Our results are in line with the findings of Goddard (2012), who supports the development of a thesis that prevention and welfarist traditions survived the transition to late-modern risk-management developments, at least in policy intentions at the federal level, and in Liège. Our analysis has shown that the culture of the control framework contains elements that help explain and understand the emergence and application of MAS. It is clear, however, that not all aspects of regulating antisocial in Belgium fit within Garland’s limited scope of the sociology of punishment. Complementing Garland’s thesis with theoretical and empirical insights into public order issues in general and on regulating public space in particular would enlarge our understanding of control theories in late-modern society. How can we explain convergent or divergent regulating tools and practices between different metropolises embedded in the same nation state? Can we detect more explanatory variables than historical evolutions, linguistic characteristics and political ideologies? This last question illustrates the need for further research and confirms the culture of control thesis as a useful framework for analysis. However, Garland’s insights do not necessarily have an explanatory value in the empirical study of public order and antisocial behaviour at the local level. 22 The first governmental white paper on the integral security ideology dates from 2004. References Aberbach, J. D., & Rockman, B. A. (2002). Conducting and coding elite interviews’. Political Science and Politics, 35(4), 673–6. Aberbach, J. D., & Rockman, B. A. (2002). Conducting and coding elite interviews’. Political Science and Politics, 35(4), 673–6. Bauwens, T., Pleysier, S., De Pauw, E., & Van Den Broeck, T. (Eds.). (2016). De nieuwe GAS-wet: zuurstof of stikstof?’, Orde van de Dag. Diegem: Kluwer. Bauwens, T., Pleysier, S., De Pauw, E., & Van Den Broeck, T. (Eds.). (2016). De nieuwe GAS-wet: zuurstof of stikstof?’, Orde van de Dag. Diegem: Kluwer. Bauwens, T., Pleysier, S., De Pauw, E., & Van Den stikstof?’, Orde van de Dag. Diegem: Kluwer. Bottoms, A. (2006). Incivilities, offence, and social order in the residential communities’. In A. Von Hirsch & A. P. Simester (Eds.), Incivilities: Regulating Offensive Behaviour – Studies in Penal Theory and Penal Ethics (pp. 239–81). Oregan: Hart Publishing. Bottoms, A. (2006). Incivilities, offence, and social order in the residential communities’. In A. Von Hirsch & A. P. Simester (Eds.), Incivilities: Regulating Offensive Behaviour – Studies in Penal Theory and Penal Ethics (pp. 239–81). Oregan: Hart Publishing. Boutellier, H. (2007). Nodale orde: veiligheid en burgerschap in een netwerksamenleving. oratie VU. Amsterdam, 53 p. Boutellier, H. (2007). Nodale orde: veiligheid en burgerschap in een netwerksamenleving. oratie VU. Amsterdam, 53 p. Bruinsma, G.J.N. & Huisman, W. (2006). Over de grenzen van de klassieke criminologie. Een beschouwing over nieuwe kennisvragen door de incorporatie van onveiligheid als object in de criminologie, in W. Huisman, Bruinsma, G.J.N. & Huisman, W. (2006). Over de grenzen van de klassieke criminologie. Een beschouwing over nieuwe kennisvragen door de incorporatie van onveiligheid als object in de criminologie, in W. Huisman, E. Devroe et al. L.M. Moerings & G. Suurmond (eds), Veiligheid en Recht. Nieuwe Doelwitten en Strategieën (pp. 269–95), Den Haag: Boom Juridische Uitgevers. Burney, E. (2005). Making People Behave: Anti-Social Behaviour. Politics and Policy, Uffculme: Willan Publishing. g Card, R., & Ward, R. (2008). The Crime and Disorder Act 1998: A Practitioner’s Guide. Bristol: Jordans. Cops, D., & Pleysier, S. (2014). De wet op de gemeentelijke administratieve sancties: een korte introduc Agora, Tijdschrift jongerenbegeleiding, 30(1), 30–3. Crawford, A. (1999). Questioning appeals to the community within crime prevention and control’. European Journal on Criminal Policy and Research, 7(4), 509–30. y ( ) Crawford, A. (2001). Les politiques locale de prévention de la délinquance en Angleterre et au Pays de Galles’. L.M. Moerings & G. Suurmond (eds), Veiligheid en Recht. Nieuwe Doelwitten en Strategieën (pp. 269–95), Den Haag: Boom Juridische Uitgevers. References Déviance et Société, 25(4), 427–58. Crawford, A. (2008). Dispersal powers and the symbolic role of anti-social behaviour legislation’. The Modern Law Review, 71(5), 753–84. Crawford, A. (2009). Governing through anti-social behaviour: Regulatory challenges to criminal justice’. British Journal of Criminology, 49, 810–31. Daems, T. (2008). Making Sense of Penal Change, Clarendon Studies in Criminology. Oxford: Oxford University Press. Daems, T. (2009). De bestraffingssociologie van David W. Garland, Den Haag: Boom Juridische uitgevers. De Maillard, J. (2002). Les nouvelles politiques socio-urbaines contractuelles entre conflits et apprentissages. Politix, 15(60), 169–91. De Witte, L. (1988). Veiligheid en Kleine Criminaliteit (pp. 1–238). Campagne Socialistische Partij, gemeenteraadsverkiezingen 9 Oktober 1988, SEVI publicatie (studiegroep van de partij). Devroe, E. (2008). ‘The policy approach of nuisance problems in public space in Belgium and the Netherlands. In L. Cachet, S. De Kimpe, P. Ponsaers, & A. Ringeling (Eds.), Governance of Security in the Netherlands and Belgium (pp. 149–75). The Hague: Boom Legal Publishers/Eleven. Devroe, E. (2012). A Swelling Culture of Control. De Genese en de Toepassing van de Wet op de Administratieve Sancties in België, Antwerpen/Apeldoorn: Maklu, 529 p. Devroe, E. (2012b). A swelling culture of control: The genesis and application of the incivilities law in Belgium. In P. Ponsaers (Ed.), Social Analysis of Security: Financial, Economic and Ecological Crime, Crime (In)security and Distrust, Public and Private Policing (pp. 355–62). The Hague: Eleven International Publishing. g Devroe, E. (2013). Local political leadership and the governance of urban security in Belgium and the Netherlands’. European Journal of Criminology, 10(3), 314–25. Devroe, E. (2015a). Purple vests: The origins of plural policing in Belgium’. European Journal of Policing Studies, 2(3), 304–25. Devroe, E. (2015b). Bringing politics back into the study of policing? A case-study on the policing of social disorder in Belgium. Policing and Society, 25, 4. Devroe, E. (2016). Mind the gap: Over ASBOS and CRASBOS in Groot Brittannië, in T. Bauwens, S. Pleysier, E. De Pauw & T. Van Den Broeck (eds), De nieuwe GAS-wet: zuurstof of stikstof? (pp. 47–58), Orde van de Dag, 73, Kluwer, Diegem. ( ), f f f (pp ), Dag, 73, Kluwer, Diegem. Devroe, E. en Ponsaers, P. (2005). De discussie inzake (drug)overlast: enkele wetenschappelijke bevindingen. In: M. Santens (Ed.), Gewapend bestuur, Cohier tegenspraak, nr. 24, 141–159. Devroe, E. en Ponsaers, P. (2005). De discussie inzake (drug)overlast: enkele wetenschappelijke bevindingen. In: M. Santens (Ed.), Gewapend bestuur, Cohier tegenspraak, nr. 24, 141–159. References Interviewing: The art of science’. In N. Denzin & Y. Lincoln Handbook of Qualitative Research (pp. 361–76). London/Thousand Oaks: Sage Publications. Handbook of Qualitative Research (pp. 361–76). London/Thousand Oaks: Sage Publications. Frevel, B. (2015). Pluralisation of local policing in Germany. European Journal of Policing Studies, 2(3), 267– 84. Garland, D. (1997). The Punitive Society: Penology, Criminology and the history of the Present. Edinburgh Law Review, 1, 180–99. Garland, D. (2001). The Culture of Control, Crime and Social Order in Contemporary Society. Oxford: Oxford University Press. Garland, D. (2004). Beyond the culture of control. Critical Review of International Social and Political Philosophy, 7(2), 160–89. Glorie, J. (1997). Ontstaan en ontwikkeling van de veiligheidscontracten’. In L. van Ou (Eds.), Veiligheidscontracten en Politie (pp. 5–10). Brussels: Politeia. Glorie, J. (1997). Ontstaan en ontwikkeling van de veiligheidscontracten’. In L. van Outrive & S. Van Elchingen (Eds.), Veiligheidscontracten en Politie (pp. 5–10). Brussels: Politeia. Glorie, J. (2004). De superministerraad van 30 en 31 maart 2004. Vigiles, 9(3), 73–5. Goddard, T. (2012). Post-welfarist risk managers? Risk, crime prevention and the responsibilization of community-based organizations. Theoretical Criminology, 16(3), 347–63. Hamilton, C. (2014). Reconceptualising Penality : A Comparative Perspective on Punitiveness in Ireland. Farmham, UK, Routlegde: Scotland and New Zealand. Hope, T. (2000). Inequality and the clubbing of private security’. In T. Hope & R. Sparks (Eds.), Crime, Risk a Insecurity: Law and Order in Everyday Life and Political Discourse (pp. 83–106). London: Routledge Hughes, G. (2007). The Politics of Crime and Community. London: Palgrave. Huyse, L. (1996). De lange weg naar Neufchateau. Leuven: Van Halewyck. Johnston, L., & Shearing, C. (2003). Governing Security: Explorations in Policing and Justice. London/New York: Routledge. Lijphart, A. (1999). Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries. New Haven, CT: Yale University Press. Mary, Ph. (1997). Le travail d’intérêt et la médiation pénale face à la crise de l’état social. Dépolitisation de la question criminelle et pénalisation du social, in Ph. Mary (ed.), Le travail d’intérêt et la médiation pénale: Socialisation du pénal ou pénalisation du social? (pp. 325–47), Brussels: Bruylant. Mary, P. (2005). Les figures du risque et de l’insécurité. L’impact sur le contrôle’. Informations Sociales, 126(6), 16–25. Matthews, R., Easton, H., Briggs, D., & Pease, K. (2007). Assessing the Use and Impact of Anti-Social Behaviour Orders. Bristol: The Policy Press. y Mcara, L. (2005). Modelling penal transformation. Punishment and Society, 7(3), 277–302. Melossi, D., & Selmini, R. (2000). References ( ) ( g) pp j g M. Santens (Ed.), Gewapend bestuur, Cohier tegenspraak, nr. 24, 141–159. vroe, E., & Ponsaers, P. (2013). Reforming the Belgian police system between central and local’. In N. F. Fyfe, Devroe, E., & Ponsaers, P. (2013). Reforming the Belgian police system between central and local’. In N. F. Fyfe, J. Terpstra, & P. Tops (Eds.), Centralizing Forces? Comparative Perspectives on Contemporary Police Reform in Northern and Western Europe (pp. 77–98). The Hague: Boom Legal Publishers/Eleven. J. Terpstra, & P. Tops (Eds.), Centralizing Forces? Comparative Perspectives on Contemporary Police Reform in Northern and Western Europe (pp. 77–98). The Hague: Boom Legal Publishers/Eleven. Di Ronco, A., & Peršak, N. (2014). ‘Regulation of incivilities in the UK, Italy and Belgium: Courts as potential safeguards against legislative vagueness and excessive use of penal powers?’ International Journal of Law. Crime and Justice, 42(4), 340–65. Dupont, G. & Ponsaers, P. (1988). Les Tueurs. Six Années d’Enquête, Anvers: EPO. Edwards, A., & Hughes, G. (2002). Introduction: The community governance of crime control’. In G. Hughes & A. Edwards (Eds.), Crime Control and Community: The New Politics of Public Safety (pp. 1–19). Cullompton: Willan. p Edwards, A., & Hughes, G. (2013). Comparative European criminology and the question of urban security’. European Journal of Criminology, 10(3), 257–9. Edwards, A., Hughes, G., & Lord, N. J. (2013). Urban security in Europe: Translating a concept in public criminology’. European Journal of Criminology, 10(3), 260–83. Edwards, A., & Prins, R. (2014). Policing and crime in contemporary London: A developmental agenda?’. European Journal of Policing Studies, 2(1), 61–93. Feeley, M. M., & Simon, J. (1992). The New Penology: Notes on the emerging strategy of corrections and implications’. Criminology, 30(4), 449–74. An Expanding Culture of Control? Feeley, M. M., & Simon, J. (1994). Actuarial justice: The emerging new criminal law’. In D. Nelken (Ed.), The Futures of Criminology (pp. 173–201). London: Sage. Felson, M. D2002]. Crime and everyday life. Londen: Sage publications. 211 p. Felson, M. D2002]. Crime and everyday life. Londen: Sage publications. 211 p. p Field, F. (2003). Neighbours from Hell: The Politics of Behaviour. London: Politico’s Publishing. Fontana, A., & Frey, J. (1994). Interviewing: The art of science’. In N. Denzin & Y. Lincoln H db k f Q lit ti R h ( 361 76) L d /Th d O k S P bli ti Fontana, A., & Frey, J. (1994). Stone, C. N. (2008). Urban regimes and the capacity to govern: A political economy approach’. Journal of Urb Affairs, 15(1), 1–28. References Social conflict and the microphysics of crime: The experience of the Emilia- Romagna Citta Secure project’. In T. Hope & R. Sparks (Eds.), Crime, Risk and Insecurity: Law and Order in Everyday Life and Political Discourse (pp. 146–66). New York: Routledge. ken, D. (2015). Foil comparison or foiled comparisons? Learning from Italian juvenile justice. European Journal of Criminology, 12(5), 519–34. Newburn, T. et al. (2002). The introduction of referral orders into the Youth Justice System: final report, Home Office Research Study, 242 p. y p O’Malley, P. (1999). Volatile and contradictory punishment’. Theoretical Criminology, 3(2), 175–96. Orum, A. M., & Feagin, J. R. (1991). A tale of two cases’. In J. R. Feagin, A. M. Orum, & G. Sjoberg ( Case for a Case Study (pp. 121–47). Chappell Hill: The University of North Carolina Press. Osborne, D., & Gaebler, T. (1993). Reinventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector. New York: Plume. Pratt, J., Brown, D., Brown, M., Hallsworth, S., & Morrison, W. (Eds.). (2005). The New Punitiveness: Trends. Theories, Perspectives, Cullompton: Willan. p p Radzinowicz, L. (1991). Penal regressions’. The Cambridge Law Journal, 50, 422–44. Roché, S. (2005). Prevention and security: A new governance model for France through a contract b territorial approach’. Canadian Journal of Criminology and Criminal Justice, 47(2), 507–26. Snacken, S. (2007). Penal policy and practice in Belgium. Crime and Justice, 36(1), 127–215. Squires, P., & Stephen, D. E. (2005). Rougher Justice: Anti-Social Behaviour and Young People. Cullompton: Willan. ne, C. N. (2008). Urban regimes and the capacity to govern: A political economy approach’. Journal of Urban Affairs, 15(1), 1–28. E. Devroe et al. van Swaaningen, R. (1996). Justitie als verzekeringsmaatschappij: ‘actuarial justice’ in Nederland. JustitiëleVerkenningen, 22, 80–7. van Swaaningen, R. (2004). Veiligheid in Nederland en Europa; een sociologische beschouwing aan de hand van David Garland. Justitiële Verkenningen, 30(7), 9–23. van Swaaningen, R. (2005). Public safety and the management of fear. Theoretical Criminology, 9(3), 289–305. Van Heddeghem, K. (2006). Bestuurlijke aanpak van overlast: de gemeentelijke administratieve sancties’. Panopticon, 27(2), 30–40. Van Limbergen, M. (1995). Het veiligheidsbeleid in België’. Wijsgerig Perspectief, 36(3), 69–74. Vander Beken, T. (2005). Minderjarigen en administratieve sancties’. In L. Veny & N. D. de Vos (Eds.) Gemeentelijke administratieve sancties (plural policing (pp. 141–145). Bruges: Vanden Broele. Vander Beken, T., & Vandeviver, C. (2014). When things get serious: Reflections on the legitimacy of local administrative sanctions’. In N. van Swaaningen, R. (1996). Justitie als verzekeringsmaatschappij: ‘actuarial justice’ in Nederland. JustitiëleVerkenningen, 22, 80–7. References Persak (Ed.), Legitimacy and Trust in Criminal Law, Policy and Justice: Norms, Procedures, Outcomes (pp. 35–49). Farnham: Ashgate. Wacquant, L. (1998). La tentation pénale en Europe. Actes de la recherche en Sciences Socials, 124, 3–6. Wacquant, L. (2004). Punir les pauvres. Le nouveau gouvernement de l’insécurité sociale, Marseille: Agon Willekens, P. (2008). Preventie: van project naar volwaardige schakel in de veiligheidsketen… Een proces van meer dan vijftien jaar. In K. Van Cauwenberghe (Ed.), De hervormingen bij politie en justitie: in gespreide dagorde, Jubileumeditie ter gelegenheid van tien jaar orde van de dag (pp. 289–301). Deventer: Kluwer. Wilson, J., & Kelling, G. (1989). Making Neighborhoods Safe. Atlantic Monthly, 263, 46–52. Wodak, R. (2011). The Discourse of Politics in Action: Politics as Usual. Basingstoke: Palgrave Macmillan. Wodak, R. (2011). The Discourse of Politics in Action: Politics as Usual. Basingstoke: Palgrave Macmillan. Young, J. (2007). The Vertigo of Late Modernity. London: Sage. Young, J. (2007). The Vertigo of Late Modernity. Lon Young, J. (2009). Moral panic: Its origins in resistance, resentment and the translation of fantasy in to reality’. British Journal of Criminology, 49(1), 4–16. Zedner, L. (2007). Preventive justice or pre punishment? The case of control orders’. Current Legal Problems, 60(1), 174–203.
https://openalex.org/W4212991317	https://www.researchsquare.com/article/rs-74908/v1.pdf?c=1600284583000	English	null	“Since I’m a little bit more mature”: Contraception and the Arc of Time for Women in Midlife	Research Square (Research Square)	2,020	cc-by	7,029	“Since I’m a little bit more mature”: Contraception and the Arc of Time for Women in Midlife Amy Alspaugh (  amyalspaugh@gmail.com ) UCSF: University of California San Francisco https://orcid.org/0000-0003-4427-4807 Melody Reibel MUSC: Medical University of South Carolina Eun-Ok Im Emory University Julie Barroso Vanderbilt University Research Keywords: contraception, middle aged, reproductive health, menopause, qualitative research Posted Date: September 16th, 2020 DOI: https://doi.org/10.21203/rs.3.rs-74908/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Women's Midlife Health on April 9th, 2021. See the published version at https://doi.org/10.1186/s40695-021-00062-7. “Since I’m a little bit more mature”: Contraception and the Arc of Time for Women in Midlife Amy Alspaugh (  amyalspaugh@gmail.com ) UCSF: University of California San Francisco https://orcid.org/0000-0003-4427-4807 Melody Reibel MUSC: Medical University of South Carolina Eun-Ok Im Emory University Julie Barroso Vanderbilt University Research Keywords: contraception, middle aged, reproductive health, menopause, qualitative research Posted Date: September 16th, 2020 DOI: https://doi.org/10.21203/rs.3.rs-74908/v1 License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Women's Midlife Health on April 9th, 2021. See the published version at https://doi.org/10.1186/s40695-021-00062-7. Research License:   This work is licensed under a Creative Commons Attribution 4.0 International License. Read Full License Version of Record: A version of this preprint was published at Women's Midlife Health on April 9th, 2021. See the published version at https://doi.org/10.1186/s40695-021-00062-7. Page 1/14 Abstract Background: Birth control methods have rapidly evolved over the past several decades, but little research has explored how women interact with contraception over time. Exploring contraceptive beliefs, perceptions, and attitudes of women in midlife can reveal much about how lived experience affects contraceptive decisions and reproductive health choices. Methods: Individual, semi-structured interviews were conducted with 20 women between the ages of 40 and 55 who had not reached menopause and did not have a permanent method of sterilization. Data were coded using qualitative descriptive methods. Results: Three major themes were identified: 1) journey toward empowerment; 2) finding the right fit: evolution over time; and 3) anticipating a transition. Past experiences with or fear of side effects and hormones were common reasons to change or avoid certain birth control methods. Most participants were happy with their contraceptive method; however, those who were unhappy were more likely to vocalize fatigue at continuing to need contraception as menopause approached. Conclusion: Approaching contraceptive counseling from a place that considers the journey with contraception over a reproductive life span will help identify how beliefs, perceptions, and attitudes of women affect their contraceptive practices and choices. Background The field of contraception has undergone a remarkable evolution in the past several decades. So remarkable, in fact, that the United States’ Centers for Disease Control and Prevention named family planning one of the success stories of public health in the 20th century [1]. Around the globe, women have seen their options and access to highly-effective contraception increase, albeit unevenly across nations with vast discrepancies in resources. But how have these changes in methods interacted with the changes that take place over the course of a woman’s reproductive years? Women nearing menopause have seen this evolution take place throughout their reproductive years. A woman currently in her mid-fifties likely started her journey with contraception in the early-to-mid 1980s, when only a few types of contraception were available. A plethora of delivery systems and hormonal formulations have been created during the past several decades. Are women trying newer methods, or do they continue largely unaffected by such advances? How do past experiences, both positive and negative, influence current contraceptive decision-making? The most recent data collected from 2006–2010 indicate that women in the US aged 15–44 use an average of 3.1 contraceptive methods during their lives [2]. These data suggest that many women will only use a small number of the various family planning methods available. A literature review was conducted to assess for previous studies that evaluated women’s perceptions and experiences with contraception over time. Two studies qualitatively explored perceptions of young Page 2/14 women who identified as Black and Latina. Gomez et al. found previous experiences with side effects and high levels of distrust in medical providers and based on experiences of others as affecting their decision to use a long-acting reversible contraceptive [3]. Downey et al. identified an iterative, relational, and reflective process in which young women made contraceptive decisions [4]. Another study explored the contraceptive journey of women age 16–50 quantitatively, evaluating methods used, reasons for changing a method, and interactions with healthcare providers [5]. To our knowledge, the current study is the first to address the contraceptive journey from the perspective of women in midlife, between the ages of 40 to 55, who arguably have the longest time frame of experience but are rarely included in contraceptive research [6]. Consequently, little is known about the reproductive journeys of women in midlife, who have potentially been using contraception for three or more decades [7, 8]. Background There is currently much debate among family planning practitioners and researchers regarding how to make birth control methods and counseling more patient-centric [9, 10]. Efforts to move past tiered contraceptive counseling, based on method effectiveness alone, is losing favor as reproductive justice becomes the default framework [11, 12]. Included in this effort is considering the lived, personal experience of the individual as equally important if not superior to the evidence-based, scientific knowledge [4, 13]. Part of the process that is necessary to elucidate these perspectives involves feminist and intersectional qualitative research [14]. To this effect, the present qualitative analysis explores the following questions: How do women in midlife view their contraceptive journey? What shapes their current beliefs, perceptions, and attitudes about contraception? What can be learned from women in midlife that may help women across the reproductive lifespan? Methods Sample. Participants were recruited throughout the Research Triangle area of North Carolina, a mid-sized urban community compromised of several smaller cities. Information about the study and pertinent contact information was included on flyers placed at women’s health clinics and public venues that reach a wide variety of women such as libraries, grocery stores, and coffee shops. Recruitment also occurred through snowball sampling: study participants were also encouraged to tell friends or family who may be interested about the study [15]. While there was no a priori number of participants we were seeking to recruit, per qualitative traditions we sought to recruit until we reached data saturation and no new themes or concepts were identified [16]. The following criteria guided participant recruitment: identifying as a woman, being between the ages of 40 and 55, speaking English, and having not yet reached menopause. Exclusion criteria included having a permanent method of sterilization or having a partner with a permanent method of sterilization. Sterilizations were excluded to ensure the research focused on those still making contraceptive decisions over 40. Approval was sought from the Medical University of South Carolina’s Institution Review Board (IRB), and the study, #Pro00089636, was deemed exempt from Human Subjects Research Regulations. Sterilizations were excluded to ensure the research focused on those still making contraceptive decisions over 40. Approval was sought from the Medical University of South Carolina’s Institution Review Board (IRB), and the study, #Pro00089636, was deemed exempt from Human Subjects Research Regulations. Page 3/14 Page 3/14 Data Collection. Interested participants telephoned the PI for prescreening eligibility or filled out a secure online survey through REDCap. Those who screened eligible were contacted to set up an interview time and location. Interviews were face-to-face, individual, semi-structured, and in-depth. An interview guide steered the discussion and covered topics such as contraception, pregnancy planning, and the approach of menopause. The guide was minimally revised during data collection based on early interviews. The interviews occurred between September and December of 2019 and were all conducted by the first author. Interviews took place at locations selected by the participants to ensure ease of access and optimal comfort level. Most often, interviews occurred at coffee shops, fast food restaurants, and public libraries. Each participant was interviewed a single time. Interviews lasted between 35–55 minutes and were audio-recorded with a handheld device and then transcribed verbatim. Methods Participant verbal consent was obtained by the interviewer at the time of the interview. Participants were given $25 gift cards to compensate for their time. Data Analysis. The study was conducted within a qualitative descriptive approach. The goal of qualitative description is to describe an individual’s experience in that individual’s own words at a manifest level [17]. Qualitative descriptive methodology is based on the overarching principles of naturalistic inquiry, meaning that people are observed and interpreted within the social and cultural context of their lives [18]. First and second level qualitative coding was initiated by the first author consistent with qualitative descriptive methodology [19]. Three overarching themes were identified based on thorough review of the first and second level codes and all second level codes were organized under the relevant theme. The fourth author reviewed all first level coding for the first 10 interviews and the second level coding for every other interview. Additionally, the second and third authors reviewed all coding on a quarter of the interviews and provided input to the final coding matrix and themes. Areas of disagreement were discussed, and consensus was obtained among all authors. Interviews were coded in Microsoft Word and demographic data were analyzed in Microsoft Excel. Trustworthiness was enhanced through 1) prolonged engagement with the transcripts, 2) creation of an audit trail, 3) use of reflexive journaling by the interviewer, and 4) member checking with 9 participants. Member checking both assisted in confirming themes and also provided further clarification about participants’ interpretation and understanding. Results Of the 32 women who completed prescreening eligibility, 22 participants met all the inclusion criteria. Of these 22 women, 20 completed the study and 2 were unable to be scheduled for interviews due to scheduling conflicts or being lost to follow-up. Participant recruitment continued until data saturation was reached and no new themes were identified. Table 1 shows the participants’ demographic information in greater detail. Page 4/14 Table 1 Demographic Characteristic of 20 Interview Participants Participant #1 was a 40 year old single woman, identifies as white, non-Hispanic and lesbian. She has previously used the pill and condoms for contraception, currently using nothing. Has one child and is sexually active with one female. Participant #2 was a 53 year-old single woman who identifies as white, non-Hispanic and bisexual. She has previously used the pill and condoms, currently using nothing. She has no children and is sexually active with one female. Participant #3 was a 48 year-old married woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, IUD, withdrawal, condoms, diaphragm, and the sponge, is currently using condoms. She has one child and is sexually active with one male. Participant #4 was a 46 year-old married woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, condoms, and the sponge, currently using pills. She has no child and is sexually active with one male. Participant #7 was a 41 year-old single woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, patch, the shot, and condoms, currently using nothing. She has four children and is not currently sexually active. Participant #8 was a 47 year-old married woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, condoms, and the sponge, currently using pills. She has two children and is sexually active with one male. Participant #10 was a 40 year-old single woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, ring, shot, IUD, and condoms, currently using an IUD. She has one child and is sexually active with several male partners. Participant #11 was a 47 year-old married woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, IUD, fertility awareness, withdrawal and condoms, currently using fertility awareness, withdrawal and condoms. She has one child and is sexually active with one male. Participant #12 was a 40 year-old single woman who identifies as bi-racially Black and white, non- Hispanic and heterosexual. She has previously used the pill, withdrawal, and condoms, currently using nothing. She has two children and is sexually active with one male. Participant #15 was a 41 year-old single woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, fertility awareness, withdrawal and condoms, currently using fertility awareness and withdrawal. Page 4/14 Page 4/14 Table 1 Demographic Characteristic of 20 Interview Participants Participant #32 was a 46 year-old divorced woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, IUD, fertility awareness, condoms, diaphragm, and sponge, currently using an IUD. She has two children and is not currently sexually active. Three major themes were identified: journey toward empowerment; finding the right fit: evolution over time; and anticipating a transition. Subthemes are presented in order of most to least prevalent. Table 1 Demographic Characteristic of 20 Interview Participants She has no children and is sexually active with one male. Participant #16 was a 52 year-old single woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, ring, fertility awareness, withdrawal, condoms, diaphragm, and sponge, currently using condoms. She has two children and is sexually active with one male. Participant #17 was a 40 year-old married woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, fertility awareness, and condoms, currently trying to get pregnant. She has one child, two step-children, and is sexually active with one male. Participant #20 was a 40 year-old single woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, and condoms, currently using nothing. She has no children and is not currently sexually active. Participant #22 was a 55 year-old married woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, patch, withdrawal and condoms, currently using Page 5/14 Page 5/14 fertility awareness, withdrawal and condoms. She has one child and is sexually active with several male partners. fertility awareness, withdrawal and condoms. She has one child and is sexually active with several male partners. Participant #24 was a 47 year-old married woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, shot, withdrawal and condoms, currently using the shot. She has one child and is sexually active with one male. Participant #25 was a 42 year-old married woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill, ring, shot, IUD, fertility awareness, and condoms, currently using pills. She has two children and is sexually active with one male. Participant #26 was a 44 year-old single woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, ring, fertility awareness, withdrawal and condoms, currently using nothing. She has one child and is not currently sexually active. Participant #28 was a 48 year-old single woman who identifies as white, non-Hispanic and heterosexual. She has previously used the pill, sponge, and condoms, currently using pills. She has two children and is not currently sexually active. Participant #31 was a 44 year-old married woman who identifies as Black, non-Hispanic and heterosexual. She has previously used the pill and condoms, currently trying to get pregnant. She has no children and is sexually active with one male. Journey toward empowerment Many women spoke of their contraceptive behavior at midlife in comparison to their younger selves. They were able to identify a transition in their own behavior. This manifested in becoming more confident in speaking up about the desire to use condoms, and in the ways in which they took experiences they had with their own mothers and used them to inform how they talked about sex and health with their own children in the present day. More confidence in addressing contraceptive needs. Over half of the women interviewed reported a difficult reproductive health scenario when they were young and newly sexually active. During their first few years of sexual activity, women in this study reported adolescent pregnancy, abortion, and risky sexual behavior. Some women also reported difficulty obtaining the contraceptive method of their choice, due to family members preventing access or because of problems with insurance. Many women in midlife described an increase in their comfort level when discussing their contraceptive needs with their partner or partners compared to when they were younger. For some women, this meant not using a method with which they felt uncomfortable. For others, it meant sharing the task of contraception with the partner, such as using male condoms, instead of feeling the need to take care of it all by themselves. Page 6/14 Specifically, several women mentioned that they felt much more confident insisting on condom use from their partners now than when they were younger. One woman discussed how she ensured safety against sexually transmitted infections (STIs): “Condoms and plus, now since I'm a little bit more mature, you're gonna go take a test for whatever disease before I even think about having sex with you.” Women who specifically mentioned that condoms were currently non-negotiable with new partners were often the same women who had talked about their own sexual risk-taking when not using protection in their youth. The role of mothers. Mothers, or mother-figures, played a large role in how women talked about reproductive and sexual health in their youth. More commonly, women reported difficult relationships with their mothers. This difficulty manifested in numerous ways, from mothers who refused to talk about sexual activity or birth control with their adolescent children to mothers who simply tossed contraceptive pills across the dinner table at their daughters without further discussion. Journey toward empowerment One participant recalled how distressing it was when her mother scheduled her first well woman exam at an abortion clinic. Several women reported approaching their mothers for help in obtaining contraception only to find that help denied. One woman discussed a conversation that did not feel age-appropriate to her: Well, when I got my period, I was 12, she told me I could get pregnant. And, um, I was still playing with Barbie dolls and, um, sex was not something that was on my mind. For a very long time, even after that conversation. So I was very confused. So she's not a person that I would really go to about anything sexual whatsoever. This initial encounter had longer-term impacts that extended beyond adolescent sexual and reproductive health. As noted in the quote above, the participant felt that she could not turn to her mother with these kinds of personal questions going forward. For many women, they reported that they continued to feel strained in terms of health-related conversations with their mothers. Thus, women found themselves without full knowledge of their family health history or information about their mother’s experience with menopause. However, women who had adolescent or young adult children prioritized talking to their children about birth control and in many cases helped them select or procure their desired method. Women who had negative experiences with their own mothers intentionally tried to create open and accessible dialogue with their own children. They expressed a desire for the kind of communication and respect that they did not receive themselves at that age. One participant noted: “Basically a lot of things that my mom did that I didn't like, I made sure that I didn't do them with my child. Things like... I kind of used the reverse, um, kind of like, this is what not to do as a parent.” Reclaiming the role of helper and advisor for their children was empowering and essential for these women. kind of like, this is what not to do as a parent.” Reclaiming the role of helper and advisor for their children was empowering and essential for these women. Finding the right fit: Evolution over time Page 7/14 Many women reported using fewer than 4 methods during their lifetime. Notably, half of the women in this study had not used a hormonal method other than the pill. Many of the women who had continued with the same method for decades reported no problems with side effects and general ease of use that kept them from needing to change methods. A smaller number of women currently using a barrier or non- Page 7/14 hormonal method reported previous side effects of hormonal birth control that made them wary of all hormonal methods. Among women who had used more than 4 methods during their life, many discussed changing over the years to find a method that better aligned with their lifestyle. For several women, this meant gradually shifting toward methods with smaller doses of hormones. For others, it meant exploring nondaily dosing options or switching to a coitus-dependent method if they were not frequently sexually active. Hesitancy with hormonal contraception. Concern regarding the use of hormonal contraception was a frequently discussed topic by participants. Some reported their hesitancy regarding hormones had existed since they started using contraception in their youth. Avoiding pregnancy was so important to these women when they were younger that they ignored their own hesitancy. As they got older, however, they felt empowered to select a method that was more congruent with their desire to avoid exogenous hormones. One woman said, “I've always been leery of [hormones], I just really wanted to protect myself from becoming pregnant.” In a few instances, use of non-hormonal methods was made easier by an increase in their knowledge about their own reproductive physiology, which helped them to feel more confident using fertility awareness or a barrier method. So, while beliefs regarding hormones did not change, the ability to act on them did. Other women vocalized a more recent hesitancy to use hormonal methods. For some women, this meant not using hormonal methods at all, but for others, it meant using methods with a lower or more localized dose. Many women expressed a desire to have more bodily awareness or a natural hormonal balance and were looking for methods that allowed for hormonal fluctuations, ovulation, or both. Finding the right fit: Evolution over time Women who were unhappy with their current method of contraception were more likely than women who were satisfied with their method to report feeling fatigued with continuing to deal with contraception in midlife. One woman expressed the following: “Me and birth control are ready to break up… I'd say it was a newness in my 20 s, and a steady relationship, you know, in my 30 s, but it's kind of time to have that conversation to end things.” Interestingly, all of the women who reported dissatisfaction with their contraceptive method felt that it was very important to use a method until reaching menopause. Anticipating a transition Women in midlife are certainly aware of the existence of menopause and its present or future impact on their lives. To women in this study who did and did not desire more children, midlife was a time of shifting health priorities. One participant explained the transition by saying: “I think you get, you get away from like let's make sure your body's ready for a baby to let's make sure you don't die at 50.” Women in midlife are certainly aware of the existence of menopause and its present or future impact on their lives. To women in this study who did and did not desire more children, midlife was a time of shifting health priorities. One participant explained the transition by saying: “I think you get, you get away from like let's make sure your body's ready for a baby to let's make sure you don't die at 50.” Acknowledgement of health needs for a changing body. Almost all participants spoke of prioritizing their own personal health and wellness. While most did not see menopause as a big, life-changing event, it did make them start to think about how they should enter older adulthood by prioritizing themselves. Women spoke of trying to eat better, exercise more, and integrate exercise that would counteract the stiffness and decreased pliancy they already felt in their bodies. This view toward long-term health was not new for all women, but it was something most of them shared in common. Finding the right fit: Evolution over time One woman who had used contraceptive pills for many years expressed concern about the length of time she had been using hormonal contraception: Um, when I was younger I used the pill, I was not worried about hormones when I was younger. Like, I knew there was risks to it, but when I was young I was like "eh, whatever, this is the easiest and the best way to do it, and I'm just gonna do it." And then as I've gotten older I've thought "You know what? I don't think I should really be on the pill nonstop for twenty, thirty years." So, the thought of like, what something is doing to my body has gotten more pronounced as I've gotten older. I care more about what it's doing to me. Experiences with side effects or perceived risks. Although some women, especially those who had been using the same methods for several decades, reported no problems with side effects, many women had experienced side effects that influenced their perceptions of the safety and ease of use of contraception. For two women, a family history of breast cancer affected their perceptions regarding the safety of hormonal birth control, especially regarding a method containing estrogen and their ages. Others had experienced a decrease in their sex drive, an exacerbation of fibroids, and/or fatigue that they attributed to their hormonal contraception since the fatigue improved when they stopped using that particular method. A small number of women said that side effects they experienced, even if those side effects happened several decades in the past, would likely deter them from using a hormonal method again. Page 8/14 Deciding to change methods. While the majority of women made these decisions on their own, a few women specifically mentioned the role of their health care provider in prompting them to explore another option. Generally, these suggestions were met positively, as long as the women felt the guidance was individualized to their specific situation in life. Suggestions from the health care provider to change a method in the absence of an identified problem, however, were met with suspicion and distrust by women in this study. Birth control dissatisfaction and fatigue. A small number of women reported dissatisfaction with their current method; their current method felt like a stopgap measure that would see them through to menopause. Finding the right fit: Evolution over time One women spoke of her own shift toward considering her health in midlife: “Yeah, you know what, this is the time to start you know, prevent it- you know go ahead and make you're sure- checking on myself, now that I'm getting older.” [Participant #7] While reproductive health was important, a shift in priorities seemed to be occurring for women in midlife. Their own personal health and wellness could no longer be put on the backburner. One participant summed it up saying: “It's not about sex and having babies. It's about, like, staying healthy.” Continuing to explore the possibility of pregnancy. While most of the women reported no plans for pregnancy in midlife, seven of the twenty participants were actively trying to get pregnant or seriously considering a pregnancy in the future. Women actively trying to get pregnant expressed some concern about the likelihood of pregnancy given their age but felt hopeful and excited at the prospect of a becoming a mother in midlife. The women considering another pregnancy all reported valuing a method that gave them the ability to stop a method quickly and easily and allowed for a quick return to fertility. Speaking generally of women having children at older ages, one participant said: “I think that's a trend that's not going to change particularly in terms of pregnancy and having children, um, and so I think there Page 9/14 Page 9/14 is sort of room to know more and kind of adjust our services and our attitudes accordingly to think about how we can adjust the needs of women in their late 30s, early 40s with having kids, and kind of meet them where they are...” is sort of room to know more and kind of adjust our services and our attitudes accordingly to think about how we can adjust the needs of women in their late 30s, early 40s with having kids, and kind of meet them where they are...” Interest in other methods. Although most of the women interviewed were satisfied with their birth control method, some of these women vocalized interest in other methods. Primarily, women discussed that they would use an intrauterine device (IUDs), especially the levonorgestrel system, if they were younger. Discussion The aim of this paper was to explore the ways in which the arc of time, including the events of the past as well as looking toward the future, influenced the contraceptive beliefs, preferences, and attitudes of women in midlife. From reviewing extant literature, it appears that this research is the first of its kind to qualitatively address the contraceptive journeys of women in midlife. Exploring contraception as a journey over time allows for the inclusion of women’s lived experiences, providing space and consideration for specific fears and concerns while treating the personal as more important than the scientific. Women in midlife, like women across the lifespan, make contraceptive decisions based on numerous individual factors. Pregnancy prevention and method effectiveness, once gold standards in family planning research and development, may or may not be strongly considered at all times by all individuals [13, 20]. For women in this study, contraception in midlife was something to be considered within the greater context of their lives, instead of their primary focus as it may have been when they were younger. First experiences with seeking contraception are important life events that are often remembered long after they occur. At a clinical and policy level, it is essential to ensure that women have access to affordable or no-cost contraception at any age. Over-the-counter contraception or ensuring parent consent is not required for any family planning service including abortion, and more targeted education in schools, can help address this at the policy level [21–23]. Mothers and fathers are often the gate-keepers to this information and access, but, as noted by many participants, can often put up barriers themselves. There also needs to be a concerted effort to ensure that parents have the right information and training to know how and when to talk to their children about sex and contraception. Evidence-based educational efforts are necessary to ensure parents have the tools to fulfill this important function for their children [24, 25]. Women who reported negative experiences with their mothers during adolescence were often the same women who did not, in later years, feel comfortable asking about menopause or their family’s health histories. This could result is missed opportunities for screenings and other types of intervention. Eight women in the present study expressed hesitation at the use of exogenous hormones for contraception. Finding the right fit: Evolution over time Some had used the method before and liked it but felt that there would not be enough time left before menopause to use it to justify the hassle of getting an appointment to have it placed. Discussion Women were split between those who had felt this way their entire adult life and those who had become more wary with age, because of an experience with side effects or a concern over Page 10/14 Page 10/14 specific medical conditions. These findings mirror the results of a small number of US-based studies focused on adolescent and young women, all of which pointed to a hesitancy regarding hormones that is not often discussed clinically [26, 27]. Contradictory results about the safety and risks of hormone replacement therapy (HRT) from the Women’s Health Initiative study further compound this hesitancy, especially for women in midlife [28]. Even those women without specific fears regarding hormones spoke to concerns regarding side effects of contraception. These findings may indicate a need for further patient education about the safety of hormonal contraception. Alternatively, it may point to a need for more targeted research on the effects of the exogenous hormones used in contraception specifically in women in midlife, as this age-group is often left out of contraceptive research. Another clinically relevant finding was that many women vocalized interest in a different method but felt that it was too late because of the amount of time remaining before menopause. Many of the women who expressed this concern were specifically referring to the levonorgestrel-releasing IUD, which, at least theoretically, may decrease potentially bothersome perimenopausal symptoms such as irregular bleeding [29, 30]. Women reach menopause on average at the age of 51, meaning that many of the women who vocalized interest in an IUD would likely have received many years of highly effective pregnancy prevention and perhaps even a reduction in perimenopausal symptomology [31]. Because hormonal contraception often masks the symptoms of perimenopause, it is often difficult to identify the onset on menopause in individuals using hormonal contraception. Thus, for those individuals who wish to continue using hormonal contraception, clinicians should specifically relay information from clinical guidelines [31] that encourage continuation of contraception until age 55. Knowledge of the specifics of this time frame may assist women in creating the best contraceptive plan for their needs and desires. This study was limited in the following ways: Despite the overall high percentage of participants who were a racial or ethnic minority, had a lower socioeconomic status, or were a sexual minority, many racial and ethnic groups were not represented in this research. Discussion Specifically, because speaking English was part of the inclusion criteria, non-English speakers, which are a significant population in this geographic area, were not able to participate. Focus on these communities and other populations, such as sexual and gender minorities, would be beneficial. Additionally, the positionality of the first author who conducted the interviews, as a white, cisgendered, heterosexual female who worker as provider of women’s health, likely exacerbated imbalances of power, privilege, and knowledge creation among study participants [32]. IUD – intrauterine device STI – sexually transmitted infection IUD – intrauterine device STI – sexually transmitted infection Declarations This study was paid for in part by the Stuart Dissertation Award given by the Medical University of South Carolina’s College of Nursing. Conclusions Women in midlife are a dynamic and empowered group. They are certainly not a monolith, as many still desire pregnancy while others have moved on to another stage of their lives. Some adore their combined hormonal pills and others have long ago written off exogenous hormones. But they are similar in their consideration of long-term planning regarding their personal health. If the past two decades are any indication, contraception will continue to evolve to include new devices, delivery systems, and hormonal formulations. By examining the women in midlife who have been privy to the changes during their lifetimes, there is much to gain in terms of understanding contraceptive beliefs, priorities, and attitudes. Page 11/14 Page 11/14 The ability to access, use, and control contraception when it is desired is a hallmark of reproductive justice and personal autonomy across the lifespan [13]. References 1. Ten Great Public Health Achievements -- United States, 1900-1999. MMWR Weekly. 1999;48(12):241- 3. 1. Ten Great Public Health Achievements -- United States, 1900-1999. MMWR Weekly. 1999;48(12):241- 1. Ten Great Public Health Achievements -- United States, 1900-1999. MMWR Weekly. 1999;48(12):241 2. Daniels K, Mosher W. Contraceptive methods women have ever used: United States, 1982-2010. Natl Health Stat Report. 2013(62):1-15. 2. Daniels K, Mosher W. Contraceptive methods women have ever used: United States, 1982-2010. Nat Health Stat Report. 2013(62):1-15. 3. Gomez AM, Arteaga S, Aronson N, Goodkind M, Houston L, West E. No Perfect Method: Exploring How Past Contraceptive Methods Influence Current Attitudes Toward Intrauterine Devices. Arch Sex Behav. 2020;49(4):1367-78. 4. Downey MM, Arteaga S, Villasenor E, Gomez AM. More Than a Destination: Contraceptive Decision Making as a Journey. Women's health issues. 2017;27(5):539-45. 5. Nelson AL, Cohen S, Galitsky A, Hathaway M, Kappus D, Kerolous M, et al. Women's perceptions and treatment patterns related to contraception: results of a survey of US women. Contraception. 2018;97(3):256-73. 6. Woods NF, Bachmann G. Contraception for midlife women: lack of information? Lack of interest? Lack of investment? Women's midlife health. 2017;3:7. 7. Harlow SD, Dusendang JR, Hood MM, Woods NF. Contraceptive preferences and unmet need for contraception in midlife women: where are the data? Women's midlife health. 2017;3:6. 8. Alspaugh A, Barroso J, Reibel M, Phillips S. Women's Contraceptive Perceptions, Beliefs, and Attitudes: An Integrative Review of Qualitative Research. Journal of midwifery & women's health. 2020;65(1):64-84. 9. Foster DG. Unmet need for abortion and woman-centered contraceptive care. Lancet. 2016;388(10041):216-7. 10. Dehlendorf C, Reed R, Fox E, Seidman D, Hall C, Steinauer J. Ensuring our research reflects our values: The role of family planning research in advancing reproductive autonomy. Contraception. 2018;98(1):4-7. Page 12/14 Page 12/14 11. Brandi K, Fuentes L. The History of Tiered-Effectiveness Contraceptive Counseling and its Effects on Patient-Centered Family Planning Care. American journal of obstetrics and gynecology. 2020;222(4S):S873-S877. 12. Higgins JA. Celebration meets caution: LARC's boons, potential busts, and the benefits of a reproductive justice approach. Contraception. 2014;89(4):237-41. 13. Potter JE, Stevenson AJ, Coleman-Minahan K, Hopkins K, White K, Baum SE, et al. Challenging unintended pregnancy as an indicator of reproductive autonomy. Contraception. 2019;100(1):1-4. 14. Im EO. Practical guidelines for feminist research in nursing. ANS Advances in nursing science. 2013;36(2):133-45. 15. Moser A, Korstjens I. Series: Practical guidance to qualitative research. Part 3: Sampling, data collection and analysis. References The European journal of general practice. 2018;24(1):9-18. 16. Sandelowski M. Sample size in qualitative research. Res Nurs Health. 1995;18(2):179-83. 17. Willis DG, Sullivan-Bolyai S, Knafl K, Cohen MZ. Distinguishing Features and Similarities Between Descriptive Phenomenological and Qualitative Description Research. West J Nurs Res. 2016;38(9):1185-204. 18. Colorafi KJ, Evans B. Qualitative Descriptive Methods in Health Science Research. Herd. 2016;9(4):16-25. 19. Sandelowski M. Whatever happened to qualitative description? Research in nursing & health. 2000;23(4):334-40. 20. Callegari LS, Aiken AR, Dehlendorf C, Cason P, Borrero S. Addressing potential pitfalls of reproductive life planning with patient-centered counseling. American journal of obstetrics and gynecology. 2017;216(2):129-34. 21. Fuentes L, Ingerick M, Jones R, Lindberg L. Adolescents' and Young Adults' Reports of Barriers to Confidential Health Care and Receipt of Contraceptive Services. The Journal of adolescent health : official publication of the Society for Adolescent Medicine. 2018;62(1):36-43. 22. Lopez LM, Bernholc A, Chen M, Tolley EE. School-based interventions for improving contraceptive use in adolescents. The Cochrane database of systematic reviews. 2016(6):Cd012249. 22. Lopez LM, Bernholc A, Chen M, Tolley EE. School-based interventions for improving contraceptive use in adolescents. The Cochrane database of systematic reviews. 2016(6):Cd012249. 23. Wilkinson TA, Miller C, Rafie S, Landau SC, Rafie S. Older teen attitudes toward birth control access in pharmacies: a qualitative study. Contraception. 2018;97(3):249-55. 23. Wilkinson TA, Miller C, Rafie S, Landau SC, Rafie S. Older teen attitudes toward birth control access in pharmacies: a qualitative study. Contraception. 2018;97(3):249-55. 24. Santa Maria D, Markham C, Bluethmann S, Mullen PD. Parent-based adolescent sexual health interventions and effect on communication outcomes: a systematic review and meta-analyses. Perspectives on sexual and reproductive health. 2015;47(1):37-50. 24. Santa Maria D, Markham C, Bluethmann S, Mullen PD. Parent-based adolescent sexual health interventions and effect on communication outcomes: a systematic review and meta-analyses. Perspectives on sexual and reproductive health. 2015;47(1):37-50. 25. Ford CA, Mirman JH, Garcia-Espana JF, Fisher Thiel MC, Friedrich E, Salek EC, et al. Effect of Primary Care Parent-Targeted Interventions on Parent-Adolescent Communication About Sexual Behavior and Alcohol Use: A Randomized Clinical Trial. JAMA network open. 2019;2(8):e199535. 25. Ford CA, Mirman JH, Garcia-Espana JF, Fisher Thiel MC, Friedrich E, Salek EC, et al. Effect of Primary Care Parent-Targeted Interventions on Parent-Adolescent Communication About Sexual Behavior and Alcohol Use: A Randomized Clinical Trial. JAMA network open. 2019;2(8):e199535. 26. Coombe J, Harris ML, Loxton D. References Motivators of contraceptive method change and implications for long-acting reversible contraception (non-)use: A qualitative free-text analysis. Sexual & reproductive 26. Coombe J, Harris ML, Loxton D. Motivators of contraceptive method change and implications for long-acting reversible contraception (non-)use: A qualitative free-text analysis. Sexual & reproductive Page 13/14 Page 13/14 healthcare : official journal of the Swedish Association of Midwives. 2019;19:71-7. 27. Chernick LS, Schnall R, Higgins T, Stockwell MS, Castano PM, Santelli J, et al. Barriers to and enablers of contraceptive use among adolescent females and their interest in an emergency department based intervention. Contraception. 2015;91(3):217-25. 28. Langer RD. The evidence base for HRT: what can we believe? Climacteric. 2017;20(2):91-6. 28. Langer RD. The evidence base for HRT: what can we believe? Climacteric. 2017;20(2):91-6. 29. Wildemeersch D. Why perimenopausal women should consider to use a levonorgestrel intrauterine system. Gynecological endocrinology : the official journal of the International Society of Gynecological Endocrinology. 2016;32(8):659-61. 29. Wildemeersch D. Why perimenopausal women should consider to use a levonorgestrel intrauterine system. Gynecological endocrinology : the official journal of the International Society of Gynecological Endocrinology. 2016;32(8):659-61. 30. Miller TA, Allen RH, Kaunitz AM, Cwiak CA. Contraception for midlife women: a review. Menopause (New York, NY). 2018;25(7):817-27. 30. Miller TA, Allen RH, Kaunitz AM, Cwiak CA. Contraception for midlife women: a review. Menopause (New York, NY). 2018;25(7):817-27. 31. ACOG Practice Bulletin No. 141: management of menopausal symptoms. Obstet Gynecol. 2014;123(1):202-16. 31. ACOG Practice Bulletin No. 141: management of menopausal symptoms. Obstet Gynecol. 2014;123(1):202-16. 32. Muhammad M, Wallerstein N, Sussman AL, Avila M, Belone L, Duran B. Reflections on Researcher Identity and Power: The Impact of Positionality on Community Based Participatory Research (CBPR) Processes and Outcomes. Critical Sociology. 2014;41(7-8):1045-63. 32. Muhammad M, Wallerstein N, Sussman AL, Avila M, Belone L, Duran B. Reflections on Researcher Identity and Power: The Impact of Positionality on Community Based Participatory Research (CBPR) Processes and Outcomes. Critical Sociology. 2014;41(7-8):1045-63. Page 14/14
W2074655041.txt	https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0030596&type=printable	en		Synergistic TLR2/6 and TLR9 Activation Protects Mice against Lethal Influenza Pneumonia	PloS one	2,012	cc-by	6,937	Synergistic TLR2/6 and TLR9 Activation Protects Mice against Lethal Influenza Pneumonia Michael J. Tuvim1,2, Brian E. Gilbert3, Burton F. Dickey1,2, Scott E. Evans1,2* 1 Department of Pulmonary Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas, United States of America, 2 Center for Infectious and Inflammatory Diseases, Institute of Biosciences and Technology, Texas A&M Health Science Center, Houston, Texas, United States of America, 3 Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, Texas, United States of America Abstract Lower respiratory tract infections caused by influenza A continue to exact unacceptable worldwide mortality, and recent epidemics have emphasized the importance of preventative and containment strategies. We have previously reported that induction of the lungs’ intrinsic defenses by aerosolized treatments can protect mice against otherwise lethal challenges with influenza A virus. More recently, we identified a combination of Toll like receptor (TLR) agonists that can be aerosolized to protect mice against bacterial pneumonia. Here, we tested whether this combination of synthetic TLR agonists could enhance the survival of mice infected with influenza A/HK/8/68 (H3N2) or A/California/04/2009 (H1N1) influenza A viruses. We report that the TLR treatment enhanced survival whether given before or after the infectious challenge, and that protection tended to correlate with reductions in viral titer 4 d after infection. Surprisingly, protection was not associated with induction of interferon gene expression. Together, these studies suggest that synergistic TLR interactions can protect against influenza virus infections by mechanisms that may provide the basis for novel therapeutics. Citation: Tuvim MJ, Gilbert BE, Dickey BF, Evans SE (2012) Synergistic TLR2/6 and TLR9 Activation Protects Mice against Lethal Influenza Pneumonia. PLoS ONE 7(1): e30596. doi:10.1371/journal.pone.0030596 Editor: Man-Seong Park, College of Medicine, Hallym University, Republic of Korea Received September 23, 2011; Accepted December 23, 2011; Published January 27, 2012 Copyright: ß 2012 Tuvim et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by funds from National Institutes of Health (NIH) grant U01 AI82226 (BFD and SEE), the MD Anderson Physician-Scientist Program (SEE), and the resources of the MD Anderson Cancer Center Support Grant P30 CA016672. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have read the journal’s policy and have the following conflicts: MJT, BFD, and SEE are authors on a related United States patent application entitled ‘‘Stimulation of Innate Resistance of the Lungs to Infection with Synthetic Ligands’’ (US Patent application 61/163,137, filed 3/25/09), and MJT and BFD are authors on a related patent entitled ‘‘Compositions and Methods for Stimulation of Lung Innate Immunity’’ (US patent application 60/ 833,857, filed 7/28/06). MJT, BFD, and SEE own stock in Pulmotect, Inc., which holds the commercial options on these patent disclosures. The other authors have no financial conflicts of interest. This does not alter the authors’ adherence to all the PLoS ONE policies on sharing data and materials. * E-mail: seevans@mdanderson.org [17,18], and host factors that impair initiation or maintenance of vaccine-induced immunity [19,20,21]. These concerns led us to investigate whether the intrinsic defense mechanisms of the lungs could be stimulated to broadly protect against pneumonias, independent of vaccine status. We have previously reported that stimulation of lung innate immunity with an aerosolized bacterial lysate could protect against pneumonia caused by bacterial, fungal and viral pathogens, including otherwise lethal influenza A challenges [22,23,24,25, 26,27]. More recently, we reported that an aerosolized combination of synthetic Toll-like receptor (TLR) agonists could recapitulate the protection conferred by the lysate against bacterial infections [25,28], leading to the question of whether protection against viral pneumonia could also be achieved using this novel combination of TLR ligands. Here, we report that synthetic ligands for TLR2/6 and TLR9 induce robust protection against lethal influenza pneumonia, including from swine-origin H1N1 influenza. Introduction Worldwide, lower respiratory tract infections cause more premature death and disability than any other condition [1,2,3]. Most years, seasonal influenza pneumonia alone causes more than 40,000 deaths in the United States, despite vaccination programs that have been in place for decades [4,5]. Pandemic influenzas have even more profound mortality impacts, with more than 50 million influenza-related deaths reported in 1918–9 [6]. The ongoing susceptibility of populations to pandemic influenza was emphasized by the rapid international spread of swine-origin H1N1 influenza in 2009 [7,8,9]. Further, the anticipated humanto-human transmission of avian-origin H5N1 influenza, which has already claimed 335 lives worldwide (www.who.int/csr/disease/ avian_influenza/) by zoonotic transmission, serves as an obvious indicator of the persisting risk of pandemics [9,10]. Respiratory viruses, including influenza, have also been characterized as potential agents of bioterror [11]. While a universal influenza vaccine is desirable, efficacy of such a vaccine capable of protecting against future pandemics has not yet been demonstrated [12]. Moreover, it is inevitable in the foreseeable future that populations will remain susceptible to seasonal influenza due to incomplete seasonal vaccination programs [13,14,15,16], epidemiologically-predicted trivalent vaccines that fail to confer immunity to a prevalent strain PLoS ONE \| www.plosone.org Results Synergistic TLR2/6 and TLR9 stimulation protects against lethal influenza pneumonia Wildtype mice were challenged with a lethal inoculum of influenza A/Hong Kong/8/68 (H3N2) 1 d after a single 1 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia aerosolized treatment with synthetic TLR ligand(s) or PBS (sham), then observed for 22 d. Treatment of mice with a TLR2/6 agonist (Pam2CSK4, ‘‘Pam2’’) alone or a TLR9 agonist (ODN2395, ‘‘ODN’’) alone resulted in no protection against lethal influenza pneumonia. However, when both treatments were concurrently delivered (Pam2-ODN) prior to the viral challenge, survival of the infectious challenge was significantly enhanced (Figure 1A). Similarly, while the mean weight loss of mice treated with single ligands did not differ from the infection-related weight loss of sham-treated mice, the mean weight loss of mice treated with Pam2-ODN was significantly less (p,0.05) than that of the shamtreated mice for days 4–14 after infection (Figure 1B). The nonsignificant weight differences observed after day 14 reflect the recovery phase of just two surviving PBS-treated mice and the heterogeneous recovery rates in the Pam2-ODN treated mice. Intuitively, the weight curves of the surviving Pam2-ODN-treated mice and the PBS-treated mice would be expected to eventually converge at a time beyond the period of observation. As the induced protective effect was substantially greater than additive effect of the individual TLR ligands alone, the Pam2-ODNinduced protection is recognized as synergistic in nature. This is consistent with the synergistic protection we have previously observed in mice pretreated with this TLR ligand combination prior to bacterial challenges [28]. TLR-induced resistance to influenza pneumonia is inducible before or after infection Our previous studies of inducible resistance indicate that the stimulating treatment can be delivered after the infectious challenge, even when mice are challenged with otherwise lethal inocula of influenza [24,27,28]. To determine whether the protective phenomenon of TLR synergy-induced resistance to influenza was restricted to only prophylactic treatment 1 d before viral exposure, mice were challenged after treatment with Pam2-ODN 3 d prior to infection or 1 d after infection, and compared to mice treated 1 d prior to infection or treated with PBS alone. As shown in Figure 2A, survival was significantly enhanced for each Pam2-ODN treatment group compared to PBS-treated mice. Similarly, weight loss was less for all of the Pam2-ODN treated groups by day 10 after infection and this persisted through the end of the observation period (p,0.05, Figure 2B). While the mice treated with Pam2ODN on the day after viral infection transiently averaged 1–2 g less than the PBS treated mice (days 2 to 7), all three Pam2ODN treated groups exceeded the mean of the PBS treated mice for most of the observation period. Consistent with our prior observations [27], the greatest induced protection was associated with the greatest reductions in the lung viral titers 4 d after infection (Figure 2C). However, as we have also previously described, the nonsignificant trend towards reduction in viral titer in the mice treated with Pam2-ODN 3 d prior to challenge suggests that early reductions in viral titer are likely only one determinant of the inducible protection. The authors postulate, for example, that treatment may also attenuate the native, injurious inflammatory host response to the virus, may enhance containment of the infection within the lungs, may prevent death due to secondary bacterial infections and may foster effective ongoing viral clearance in addition to rapid induction of pathogen killing. And, much like the cytokine-induced antiviral state is observed in certain leukocytes, it is possible that the TLR stimulus may directly reduce the ability of virus to infect its primary target, the respiratory epithelium. PLoS ONE \| www.plosone.org Figure 1. Synergistic TLR stimulation protects against lethal influenza pneumonia, while individual TLR ligands confer no protection. Swiss-Webster mice were challenged with influenza A/ Hong Kong/8/68 (H3N2) 24 h after aerosolized treatment with PBS (sham), Pam2, ODN or both (Pam2-ODN). Shown are survival (A) and body weight (B) of the mice through day 22 after infection (mean 6 s.d.). (n = 20 mice/group; p = 0.03 vs. PBS treated). doi:10.1371/journal.pone.0030596.g001 TLR3 stimulation does not enhance TLR2/6 and TLR9 protection against influenza pneumonia Treatments that activate TLR3 in mice have been reported to enhance antiviral immunity, including induction of protection against influenza [29,30,31,32,33]. We tested whether treatment of mice with the TLR3 agonist poly(I:C) would protect against influenza in our lethal infection model. As shown in Figure 3A, we found that poly(I:C) treatment delivered 1 d prior to influenza challenge did result in improved survival, supporting the prior literature. However, a single treatment with poly(I:C) resulted in less protection than a single treatment with Pam2-ODN. Moreover, the addition of poly(I:C) had no discernable effect on the synergistic response to Pam2-ODN, as the concurrent 2 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia mean 6 s.d.). (n = 20 mice group for survival and weight, n – 5 mice/ group for lung titers; p,0.0001 vs. PBS treated, ** p,0.002 vs. PBS treated, { p,0.05 vs. PBS treated). doi:10.1371/journal.pone.0030596.g002 administration of the TLR3 ligand with the TLR2/6 and TLR9 ligands resulted in no significant differences in survival or body weight. We found that doubling the concentration of Pam2-ODN further improved survival to 100% (Figure 3). At this level of protection, the survival and weight changes of the 1x and 2x Pam2-ODN-treated groups cannot be statistically distinguished from each other. However, it is notable that the increased dose was Figure 3. Synergistic TLR2/6 and TLR9 protects against influenza with or without TLR3 stimulation. Mice were challenged with influenza 1 day after a single inhaled treatment with the described treatments. Shown are survival (A) and body weight (B) of mice through 22 days after challenge (mean 6 s.d.). ‘‘2x’’ indicates doubling of the concentration of all TLR ligand components in a corresponding ‘‘1x’’ treatment. (n = 20 mice/group; * p,0.00001 vs. PBS treated, ** p,0.02 vs. PBS treated, { p,0.0001 vs. poly(I:C) treated, {{ p = 0.002 vs. poly(I:C) treated, # p = 0.004 vs. poly(I:C) treated). doi:10.1371/journal.pone.0030596.g003 Figure 2. Synergistic TLR stimulation protects against influenza pneumonia whether given before or after infection. Mice were challenged with influenza A following a single aerosolized treatment with Pam2-ODN 3 d before infection, 1 d before infection or 1 d after infection or following a single aerosol treatment with PBS 1 d before infection. Shown are survival (A) and body weight (B) of the mice through day 22 after infection (mean 6 s.d.). Log viral titer of lung homogenates is shown for day 4 after infection for the same groups (C, PLoS ONE \| www.plosone.org 3 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia protect against bacterial pneumonia [28]. To test whether class C CpG ODNs are specifically required for TLR-inducible resistance to influenza, we treated groups of mice with one of several aerosolized treatments one day prior to influenza infection. The treatments consisted of Pam2 plus one CpG ODN (class A, B or C) or PBS only, as shown in Figure 4. In contrast to our experience with the bacterial models, all classes of CpG ODNs synergized with Pam2 to protect against influenza pneumonia. As a group, the class C CpG ODNs again protected significantly better nonClass C CpG ODNs (p = 0.025). However, all tested CpG ODNs synergized to some extent with Pam2, and some Class A and Class B CpG ODNs protected as well as Class C ligands when combined with Pam2. well tolerated by the mice, inducing neither distress nor worsening survival. Protection is not associated with induction of lung interferon expression Our prior studies of lysate-induced resistance to influenza revealed significant induction of both type I and type II interferon expression [27]. Given prior data suggesting that lung epithelial cells play a critical role in inducible resistance [22,24,26,28], we first assessed Pam2-ODN-induced interferon signaling in MLE-15 lung epithelial cells in isolation. As shown in Table 1, we did not observe the induction of type I, II or III interferon expression by these cells following treatment with Pam2-ODN, despite analyzing identical time points to those studied when investigating lysateinduced protection. While the epithelium appears to play an important role in inducible resistance, we recognized that recruited leukocytes might be the source of the previously observed lysate-induced interferons in the lungs. So, wildtype mice were treated with Pam2-ODN or PBS, and interferon gene expression was analyzed in whole lung homogenates. Again, we did not detect significant induction of interferon genes or known interferonsensitive antiviral genes by Pam2-ODN treatment, although we did find induction of some interferon receptors. Although there was significant induction of Janus Kinase 1 (JAK1, p,0.01), JAK2 (p,0.00001), and Signal Transducer and Activator of Transcription 1 (STAT1, p,0.00001) expression following PAM2-ODN treatment, pathway analysis did not associate inducible resistance with TLR-enriched interferon signaling (compared to PBStreated, p = 1.0). To ensure that the absence of interferon-related gene expression was not the result of an insufficient therapeutic stimulus or an insensitive detection technique, we also analyzed interferon-independent, pro-inflammatory cytokine gene expression. Table 2 presents 15 representative examples of these proinflammatory cytokines and chemokines that are significantly enriched in the same samples analyzed in Table 1. Unlike the negligible changes seen in the interferon-dependent genes, treatment with Pam2-ODN results in robust expression of IL-6, TNF, IL-1a, IL-1b, and multiple chemokines. In most examples, these findings were observed from both MLE-15 cells in isolation and from whole lung homogenates 4 h after Pam2-ODN treatment. The role of these inflammatory cytokines in inducible resistance remains unclear. IL-6 and TNF were even more robustly induced by the lysate than by Pam2-ODN, but we have shown that they were not required for protection against bacterial pneumonia [24]. Also, given the demonstrations of tachyphylaxis to leukocyte infiltration, but not protection, with repetitive treatments, the induction of these products may represent an epiphenomenon that provides a useful biomarker but may be mechanistically unimportant. However, it is interesting that the four proinflammatory cytokines induced by Pam2-ODN 4 h after treatment (IL-1a, IL-1b, IL-6, and TNF) are also the first four pro-inflammatory cytokines induced from respiratory epithelial cells in native influenza infections [34]. While the principal antiviral function of these cytokines is generally presumed to be leukocyte activation, it is conceivable that they also directly shape the epithelial response and that they may eventually prove to be required. Synergistic TLR2/6 and TLR9 stimulation protects against swine-origin H1N1 influenza To confirm that Pam2-ODN-induced protection was not restricted to a single influenza strain, we tested the ability of Pam2-ODN to protect against highly lethal swine-origin H1N1 influenza. As shown in Figure 5, a single inhaled treatment with Pam2-ODN significantly improved mouse survival of otherwise lethal challenge. Discussion Viral lower respiratory tract infections, particularly those caused by influenza viruses, continue to inflict tremendous annual worldwide mortality [35]. Further compounding this public health urgency are the persisting threat of pandemic influenza infection and an increasing resistance to available antivirals, such as neuraminidase inhibitors [36,37]. In this study, we find that induction of innate immunity in the lungs with a novel combination of synthetic TLR ligands results in robust protection against otherwise lethal influenza. Consistent with our earlier descriptions of inducible resistance to influenza pneumonia, this protection is generally associated with reductions in the lung viral titers of treated mice [27], though this association is not always observed. Also consistent with our reports of protection induced by treatment of mice with an aerosolized treatment with a bacterial lysate, protection could be induced whether the treatment was applied before or after infection [27]. Type I and II interferon responses have both been identified by other groups as essential to effective antiviral host responses [38,39]. However, in notable contrast with our observations in lysate-induced resistance to influenza pneumonia [27], we do not find that Pam2-ODN treatment induces significant interferonrelated gene expression. As the current studies were not performed in interferon-deficient mice, it is impossible to exclude any role for low-level interferon signaling in Pam2-ODN-induced pneumonia. However, these findings do suggest that reconsideration of the lysate-induced changes in gene expression might be necessary. We previously observed that the lysate induced significant type I and II interferon signaling in the absence of infection, but we also demonstrated that interferon -c levels in the lungs of infected mice were lower if they had been pretreated with the aerosolized lysate. At that time, we interpreted this to mean that the interferon response was important to inducible resistance, but that the effective interferon response also limited ongoing interferon signaling following clearance of the pathogen. While that interpretation may be correct, the new observation of protection without induction of interferon signaling raises the competing interpretation that interferon signaling is not required for either lysate- or Pam2-ODN-induced resistance to pneumonia. This would be consistent with other prior observations that interleukin- Class C CpG ODNs synergize most effectively with TLR2/6 agonist to protect against influenza pneumonia We have reported that only class C CpG ODNs can effectively activate TLR9 in a manner that synergizes with TLR2/6 to PLoS ONE \| www.plosone.org 4 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia Table 1. Interferon responses to Pam2-ODN. Symbol MLE-15 Fold Change Lung Fold Change Definition Accession Bak1 « « Bcl2-antagonist/killer 1 NM_007523.2 Bax « Q 1.6 Bcl2-associated X protein NM_007527.2 Bcl2 « « B-cell leukemia/lymphoma 2, variant 2 NM_177410.2 Gvin1 « « GTPase, very large interferon inducible 1, variant B NM_001039160.2 Ifi202b « « Interferon activated gene 202B NM_008327.1 Ifi203 « « Interferon activated gene 203, transcript variant 2 NM_008328.2 Ifi204 « « Interferon activated gene 204 NM_008329.2 Ifi205 « « Interferon activated gene 205 NM_172648.3 Ifi27 « « Interferon alpha-inducible protein 27 NM_029803.1 Ifi30 « « Interferon gamma inducible protein 30 NM_023065.3 Ifi35 « Q 1.4 Interferon-induced protein 35 NM_027320.4 Ifi44 « « Interferon-induced protein 44 NM_133871.1 Ifi47 q 3.3 q 2.6 Interferon gamma inducible protein 47 NM_008330.1 Ifih1 « « Interferon induced with helicase C domain 1 NM_027835.1 Ifit1 « « Interferon-induced protein with tetratricopeptide repeats 1 NM_008331 Ifit2 « q 1.9 Interferon-induced protein with tetratricopeptide repeats 2 NM_008332.2 Ifit3 « « Interferon-induced protein with tetratricopeptide repeats 3 NM_010501.2 Ifitm1 « « Interferon induced transmembrane protein 1 NM_026820.2 Ifitm2 « q 1.6 Interferon induced transmembrane protein 2 NM_030694.1 Ifitm3 « « Interferon induced transmembrane protein 3 NM_025378.2 Ifitm5 « « Interferon induced transmembrane protein 5 NM_053088.2 Ifitm6 « « Interferon induced transmembrane protein 6 NM_001033632.1 Ifitm7 « « Interferon induced transmembrane protein 7 NM_028968.1 Ifna1 « « Interferon alpha 1 NM_010502.2 Ifna2 « « Interferon alpha 2 NM_010503.2 Ifna4 « « Interferon alpha 4 NM_010504.1 Ifna5 « « Interferon alpha 5 NM_010505.1 Ifna6 « « Interferon alpha 6 NM_206871.1 Ifna7 « « Interferon alpha 7 NM_008334.2 Ifna9 « « Interferon alpha 9 NM_010507.1 Ifna11 « « Interferon alpha 11 NM_008333.1 Ifna12 « « Interferon alpha 12 NM_177361.2 Ifna13 « « Interferon alpha 13 NM_177347.2 Ifna14 « « Interferon, alpha 14 NM_206975.1 Ifnab « « Interferon alpha B NM_008336.2 Ifnar1 « « Interferon alpha and beta receptor 1 NM_010508.1 Ifnar2 « q 3.0 Interferon alpha and beta receptor 2 NM_010509.1 Ifnb1 « « Interferon beta 1, fibroblast NM_010510.1 Ifne1 « « Interferon epsilon 1 NM_177348.2 Ifng « « Interferon gamma NM_008337.1 Ifngr1 « q 1.4 Interferon gamma receptor 1 NM_010511.2 Ifngr2 q 2.5 q 1.7 Interferon gamma receptor 2 NM_008338.2 NM_199157.2 Ifnk « « Interferon kappa Ifnz « « Interferon zeta NM_197889.1 Ifrg15 « « Interferon alpha responsive gene 15 NM_022329.3 NM_018738.3 Igtp « q 1.4 Interferon gamma induced GTPase Iigp1 « « Interferon inducible GTPase 1 NM_021792.3 Iigp2 « q 1.5 Interferon inducible GTPase 2 NM_019440.2 Isg20 « « Interferon-stimulated protein 20 NM_020583.4 PLoS ONE \| www.plosone.org 5 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia Table 1. Cont. Symbol MLE-15 Fold Change Lung Fold Change Definition Accession Mx1 « « Myxovirus (influenza virus) resistance 1 NM_010846.1 Oas1a « « 29-59 oligoadenylate synthetase 1A NM_145211.1 Psmb8 « « Proteasome subunit beta type 8 NM_010724.1 Tap1 « « Transporter 1, ATP-binding cassette, sub-family B NM_013683.1 Transcriptional responses of interferon and known interferon-sensitive antiviral genes 4 h after treatment of MLE-15 cells in vitro or mouse lungs in vivo with Pam2-ODN. Fold change compares Pam2-ODN-treated samples to PBS-treated samples. « indicates no significant change in gene expression between PBS treated and Pam2-ODN treated samples, q indicates induction of transcription by Pam2-ODN, Q indicates repression of transcription by Pam2-ODN. doi:10.1371/journal.pone.0030596.t001 degree of synergizing with Pam2 [28]. Synthetic CpG ODNs can be structurally and functionally categorized into broad classes [41,42,43,44]. Class A ODNs have palindromic sequences on phosphodiester backbones and classically induce secretion of type I and II interferons from leukocytes. Class B ODNs have linear 6mers on phosphorothioate backbones that induce B cell proliferation and expression of interleukins-6 and -10. Class C ODNs possess characteristics of both A and B classes [43,45]. These classspecific responses presumably arise from differential endosomal compartmentalization and signaling, with Class A ODNs predominantly promoting IRF-7-mediated signaling from early endosomes and Class B ODNs primarily inducing late endosomal NF-kB activation [43]. We have previously reported that only class C CpG ODNs effectively synergized with the TLR2/6 ligands to protect broadly against bacterial challenges. However, we here clearly demonstrate that class A and class B CpG ODNs can cooperate with Pam2 to protect against influenza viruses, with no discernable statistical difference in the performance of Class A and Class B ligands. This mechanistic difference between the bacterial and 6 and TNF are both profoundly induced by lysate treatment, but are not required for protection against bacterial pneumonias [24]. While these data provide a novel contrast to prior reports of interferon-dependence of the antiviral response, they do not clearly reveal the interferon-independent mechanisms underlying the protection. This is an area of active research, but we have previously reported that Pam2-ODN is capable of inducing expression of both antimicrobial peptides and reactive oxygen species [22,23,24,25,26,27,28], and anticipate that these responses contribute to both inducible viral killing and modulation of untoward elements of the inflammatory response. We have also shown in bacterial infection models that inducible resistance is associated with enhanced containment of pathogens within the lungs [22,24,28]. Enhancement of barrier function may contribute to the antiviral response, as well. That this could occur in an interferon-independent manner is supported by recent observations of reactive oxygen species mediated intercellular epithelial antimicrobial communication [40]. Another unexpected finding was the observation that all tested classed of TLR9-stimulating CpG ODNs were capable to some Table 2. Inflammatory cytokine responses to Pam2-ODN. Symbol MLE-15 Fold Change Lung Fold Change Definition Accession Ccl2 « q 33.0 Chemokine (C-C motif) ligand 2 NM_011333.3 Ccl3 « q 53.5 Chemokine (C-C motif) ligand 3 NM_011337.2 Ccl7 « q 15.3 Chemokine (C-C motif) ligand 7 NM_013654.2 Cx3cl1 q 4.0 q 2.4 Chemokine (C-X3-C motif) ligand 1 NM_009142.3 Cxcl1 q 77.3 q 16.6 Chemokine (C-X-C motif) ligand 1 NM_008176.1 Cxcl2 q 7.9 q 14.4 Chemokine (C-X-C motif) ligand 2 NM_009140.2 Cxcl10 q 1.7 q 22.8 Chemokine (C-X-C motif) ligand 10 NM_021274.1 Cxcl13 « q 11.3 Chemokine (C-X-C motif) ligand 13 NM_018866.1 Cxcl15 q 5.0 « Chemokine (C-X-C motif) ligand 15 NM_011339.2 Cxcl16 q 2.4 q 1.3 Chemokine (C-X-C motif) ligand 16 NM_023158.6 Il1a « q 8.1 Interleukin 1 alpha NM_010554 Il1b « q 25.1 Interleukin 1 beta NM_008361 Il24 q 3.0 « Interleukin 24 NM_053095.1 Il6 « q 2.6 Interleukin 6 NM_031168.1 Tnf q 2.0 q 24.8 Tumor necrosis factor NM_013693.1 Transcriptional responses of interferon-independent inflammatory cytokines and chemokines 4 h after treatment of MLE-15 cells in vitro or mouse lungs in vivo with Pam2-ODN. Fold change compares Pam2-ODN-treated samples to PBS-treated samples. « indicates no significant change in gene expression between PBS treated and Pam2-ODN treated samples, q indicates induction of transcription by Pam2-ODN, Q indicates repression of transcription by Pam2-ODN. doi:10.1371/journal.pone.0030596.t002 PLoS ONE \| www.plosone.org 6 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia Figure 4. Pam2 treatment synergizes with all classes of TLR9stimulating CpG oligodeoxynucleotides. Mice were challenged with influenza 1 day after a single inhaled treatment with the described treatments. Shown are survival (A) and body weight (B) of mice through 22 days after challenge (mean 6 s.d.). (n = 20 mice/group; * p,0.00001 vs. PBS treated, ** p = 0.0004 vs. PBS treated, { p = 0.01 vs. Pam2+ODN 2006 treated, { p = 0.1 vs Pam2+ODN 2006 treated). doi:10.1371/journal.pone.0030596.g004 Figure 5. Synergistic TLR stimulation protects against swineorigin H1N1 influenza A pneumonia. Mice were challenged with influenza 1 day after a single inhaled treatment with Pam2 and ODN. Shown are survival (A) and body weight (B) of mice through 22 days after challenge (mean 6 s.d.). (n = 20 mice/group; * p = 0.0004 vs. PBS treated). doi:10.1371/journal.pone.0030596.g005 viral models will be an area of future investigation. However, as class A and class C, but not class B, CpG ODNs are reported to induce immune responses via interferon signaling, the observation of Pam2 synergy with class B CpG ODNs is consistent with the lack of an essential interferon role. Because of these differences in Pam2-ODN-induced influenza protection when compared to lysate-induced influenza protection and to Pam2-ODN-induced bacterial protection, it was important to show that this was not a phenomenon that was restricted to a unique viral strain. We excluded that possibility by testing an alternate influenza strain, and demonstrated the effectiveness of this treatment against clinically-relevant pathogens by showing that Pam2-ODN can also protect against swine-origin H1N1 influenza viruses. PLoS ONE \| www.plosone.org This broad, non-toxic host response-focused strategy to preventing influenza-related mortality may provide an opportunity to protect vulnerable populations when vaccines are unavailable or impractical, to contain outbreaks by treating contacts of incident cases, and, potentially, to protect populations that have deficiencies of adaptive immunity. Methods Animals Six to eight week old NIH Swiss-Webster mice (Charles River) were used for all experiments. For protection studies, mice were divided into groups of 20 mice (5 for virus lung titers, 15 for 7 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia survival). All mice were handled in accordance with the policies of the Baylor College of Medicine Institutional Animal Care and Use Committee, full details of the study were approved by that body (approval AN-2307), and any mice that exhibited signs of distress were humanely euthanized. Gene expression analysis To better understand the host response to Pam2-ODN, gene expression microarray analysis was performed after treatment with Pam2, ODN, Pam2-ODN or PBS. For in vitro analyses, immortalized mouse distal respiratory epithelial MLE-15 cells [49] were provided by Dr. Jeffrey A. Whitsett, Cincinnati Children’s Hospital Medical Center, and grown in monolayer to approximately 80% confluence, then the designated treatments were added to the culture media for 4 h, then the cells were collected by scraping, as previously described [28]. For in vivo analyses, wild type mice were exposed by aerosol to the designated treatments, as described above, then euthanized after 4 h for comparison to PBS-treated mice. The lungs were sterilely-resected and mechanically homogenized. Total RNA was isolated from lung homogenates and cell culture samples using the RNeasy system (Qiagen), and cRNA was synthesized and amplified from equal masses of total RNA using the Ilumina TotalPrep RNA amplification kit (Ambion). Amplified cRNA was hybridized and labeled on MouseRef-8 v2.0 Expression BeadChips (Illumina), then scanned on an Illumina iScan. Primary microarray data were deposited at the NCBI Gene Expression Omnibus (http://www. ncbi.nlm.nih.gov/geo/) consistent with MIAME standards (GEO Accession GSE26864, in vitro, and GSE28994, in vivo). Primary signal intensity was normalized between and within samples, and differentially expressed genes were identified based on signal change and inter-sample variation. Gene ontology analysis was performed using the NIAID Database for Annotation, Visualization and Integrated Discovery (DAVID) and the KEGG Database (GenomeNet). Differentially expressed genes were mapped to signaling pathways using Ingenuity Pathways Analysis 9.0 (Ingenuity Systems), and the pathway nodules were individually reviewed. To characterize the interferon-related gene expression changes induced by Pam2-ODN, Table 1 presents a list of genes containing all transcripts from the Ingenuity Pathway Analysis canonical interferon signaling pathway, detected interferon-related JAK-STAT-dependent transcripts in KEGG, and additional interferon-sensitive antiviral transcripts identified by the authors. Baseline signal intensity values of 1 were assigned to undetected control transcripts in order to avoid reporting infinite fold change values. Samples treated with single TLR ligands (Pam2 only or ODN only) were analyzed but not presented, as they were not deemed to yield additional information beyond that presented in Table 1. Data from all tested conditions is included in the GEO deposits referenced above. Synthetic TLR Ligand aerosol treatment All treatments were delivered by aerosol. All synthetic TLR ligands were purchased from InvivoGen (San Diego, California), reconstituted in endotoxin-free water, and suspended in 8 ml sterile PBS at indicated concentrations. As we have previously described [22,23,27,28], treatments were aerosolized to the animals for 30 min from an AeroMist CA-209 nebulizer (CISUS) driven by 10 l/min air supplemented. The nebulizer was connected by polyethylene tubing (30 cm622 mm) to a 10 liter polyethylene exposure chamber, with an identical efflux tube with a low resistance microbial filter (BB50T, Pall, East Hills, NY) at its end vented to a biosafety hood. Dosing of the TLR ligands was determined by the lowest nebulized concentration required to induce leukocyte infiltration of the lungs, as we have previously reported [28]. Accordingly, the following concentrations were used in the nebulizer reservoir: Pam2 10 mg/ml; poly(I:C) 100 mg/ml; ODN2395 20 mg/ml. Based on previous experiments, [23,46] ligand concentrations in the airway lining fluid are calculated to be Pam2 10 ng/ml; poly(I:C) 100 ng/ml; ODN2395 20 ng/ml. Experiments explicitly using 2X dosing of ligands used double these concentrations. Class comparisons of different CpG ODNs used equimolar concentration of their respective ODN as found in ODN2395 20 mg/ml. Sequences of the tested ODNS were: Class A, ODN 1585 59-ggggtcaacgttgagggggg-39 and ODN 2216 59-gggggacgatcgtcgggggg-39; Class B, ODN 1826 59-tccatgacgttcctgacgtt-39 and ODN 2006 59tcgtcgttttgtcgttttgtcgtt-39; and, Class C, ODN 2395 59-tcgtcgttttcggcgcgcgccg-39, ODN M362 59-tcgtcgtcgttcgaacgacgttgat-39 and ODN 1010159-tcgtcgttttcgcgcgcgccg-39. Influenza A challenge A clinical isolate of influenza A/Hong Kong/8/68 (H3N2) (A/ HK; Mouse Lung Pool 11-29-05) virus that had been passaged at least nine times through mice was stored as frozen stock (2.86107 TCID50/ml) in the supernatant of mouse lung homogenates [47]. Stock was diluted 1:300–1:1,000 in 0.05% gelatin in Eagle’s minimal essential medium (Sigma-Aldrich) and aerosolized for 20 min to achieve LD90 – LD100 (target 100 TDIC50/mouse). Viral concentration in the nebulizer before and after aerosolization and in lung homogenates was determined by hemagglutination assay of infected MDCK cells [48]. On day +4, 5 mice from each group were sacrificed and their lungs removed. Lungs were homogenized by beadbeating and the levels of virus determined. Remaining mice in each group were observed daily for up to 21 days for overt illness, morbidity and mortality. Mice were weighed on days 0 and +4, and three times weekly from day +7 until day +21. Influenza A/California/04/2009 (H1N1) was obtained from the Centers for Disease Control and Prevention (Atlanta, GA) as MDCK passage 3 (CDC ID Number 200971204). The virus was grown in MDCK cells [45] and on passage 10, a sucrose purified tissue culture pool was prepared. The 30/50% sucrose interface was collected (passage 11), aliquoted and used for aerosol infection of mice. The titer of the stock virus was 9.8 TCID50/mL and was diluted 1:400 in 0.05% gelatin-MEM before nebulization. The diluted virus was added to the reservoir (9 mL) of an Aerotech II neublizer (CIS-USA, Bedford, MA) flowing at 10 L of air/min and used to treat mice as described above. The targeted dose after 20 min was estimated to be ,26104 TCID50/mouse. PLoS ONE \| www.plosone.org Statistical methods Summary statistics for virus in lung tissue were compared using Student’s t-test. Proportions of mice surviving pathogen challenges were compared using Fisher’s exact text on the final day of observation, and log-rank comparisons of survival distribution were performed using Kaplan-Meier estimation. Weight comparisons were made between experimental groups using two-tailed Student’s t-test for each experimental day. All data shown are representative of at least two independent experiments, and were not combined for analysis because of modest differences in virus challenge doses. Analyses were performed using SAS/STAT (SAS Institute). Author Contributions Conceived and designed the experiments: SEE MJT BEG. Performed the experiments: MJT BEG. Analyzed the data: SEE MJT BEG BFD. Contributed reagents/materials/analysis tools: SEE MJT BEG BFD. Wrote the paper: SEE MJT BEG BFD. 8 January 2012 \| Volume 7 \| Issue 1 \| e30596 TLR Synergy Protects against Influenza Pneumonia References 26. Evans SE, Xu Y, Tuvim MJ, Dickey BF (2010) Inducible innate resistance of lung epithelium to infection. Annu Rev Physiol 72: 413–435. 27. Tuvim MJ, Evans SE, Clement CG, Dickey BF, Gilbert BE (2009) Augmented lung inflammation protects against influenza A pneumonia. PLoS ONE 4: e4176. 28. Duggan JM, You D, Cleaver JO, Larson DT, Garza RJ, et al. (2011) Synergistic Interactions of TLR2/6 and TLR9 Induce a High Level of Resistance to Lung Infection in Mice. J Immunol 186: 5916–5926. 29. Wong JP, Christopher ME, Viswanathan S, Karpoff N, Dai X, et al. (2009) Activation of toll-like receptor signaling pathway for protection against influenza virus infection. Vaccine 27: 3481–3483. 30. Lau YF, Tang LH, Ooi EE, Subbarao K (2010) Activation of the innate immune system provides broad-spectrum protection against influenza A viruses with pandemic potential in mice. Virology 406: 80–87. 31. Lau YF, Tang LH, Ooi EE (2009) A TLR3 ligand that exhibits potent inhibition of influenza virus replication and has strong adjuvant activity has the potential for dual applications in an influenza pandemic. Vaccine 27: 1354–1364. 32. Ichinohe T, Watanabe I, Tao E, Ito S, Kawaguchi A, et al. (2006) Protection against influenza virus infection by intranasal vaccine with surf clam microparticles (SMP) as an adjuvant. J Med Virol 78: 954–963. 33. Ichinohe T, Watanabe I, Ito S, Fujii H, Moriyama M, et al. (2005) Synthetic double-stranded RNA poly(I:C) combined with mucosal vaccine protects against influenza virus infection. J Virol 79: 2910–2919. 34. Sanders CJ, Doherty PC, Thomas PG (2011) Respiratory epithelial cells in innate immunity to influenza virus infection. Cell Tissue Res 343: 13–21. 35. Clark NM, Lynch JP (2011) Influenza: epidemiology, clinical features, therapy, and prevention. Semin Respir Crit Care Med 32: 373–392. 36. Pizzorno A, Abed Y, Boivin G (2011) Influenza drug resistance. Semin Respir Crit Care Med 32: 409–422. 37. Pizzorno A, Bouhy X, Abed Y, Boivin G (2011) Generation and characterization of recombinant pandemic influenza A(H1N1) viruses resistant to neuraminidase inhibitors. J Infect Dis 203: 25–31. 38. Garcia-Sastre A (2006) Antiviral response in pandemic influenza viruses. Emerg Infect Dis 12: 44–47. 39. Wang JP, Kurt-Jones EA, Finberg RW (2007) Innate immunity to respiratory viruses. Cell Microbiol 9: 1641–1646. 40. Dolowschiak T, Chassin C, Ben Mkaddem S, Fuchs TM, Weiss S, et al. (2010) Potentiation of epithelial innate host responses by intercellular communication. PLoS Pathog 6: e1001194. 41. Jurk M, Kritzler A, Debelak H, Vollmer J, Krieg AM, et al. (2006) Structureactivity relationship studies on the immune stimulatory effects of base-modified CpG toll-like receptor 9 agonists. ChemMedChem 1: 1007–1014. 42. Vollmer J, Jurk M, Samulowitz U, Lipford G, Forsbach A, et al. (2004) CpG oligodeoxynucleotides stimulate IFN-gamma-inducible protein-10 production in human B cells. J Endotoxin Res 10: 431–438. 43. Vollmer J, Krieg AM (2009) Immunotherapeutic applications of CpG oligodeoxynucleotide TLR9 agonists. Adv Drug Deliv Rev 61: 195–204. 44. Vollmer J, Weeratna R, Payette P, Jurk M, Schetter C, et al. (2004) Characterization of three CpG oligodeoxynucleotide classes with distinct immunostimulatory activities. Eur J Immunol 34: 251–262. 45. Krieg AM (2006) Therapeutic potential of Toll-like receptor 9 activation. Nat Rev Drug Discov 5: 471–484. 46. Evans CM, Williams OW, Tuvim MJ, Nigam R, Mixides GP, et al. (2004) Mucin is produced by clara cells in the proximal airways of antigen-challenged mice. Am J Respir Cell Mol Biol 31: 382–394. 47. Wyde PR, Couch RB, Mackler BF, Cate TR, Levy BM (1977) Effects of lowand high-passage influenza virus infection in normal and nude mice. Infect Immun 15: 221–229. 48. Gilbert BE, Wyde PR, Ambrose MW, Wilson SZ, Knight V (1992) Further studies with short duration ribavirin aerosol for the treatment of influenza virus infection in mice and respiratory syncytial virus infection in cotton rats. Antiviral Res 17: 33–42. 49. Wikenheiser KA, Vorbroker DK, Rice WR, Clark JC, Bachurski CJ, et al. (1993) Production of immortalized distal respiratory epithelial cell lines from surfactant protein C/simian virus 40 large tumor antigen transgenic mice. Proc Natl Acad Sci U S A 90: 11029–11033. 1. Mizgerd JP (2006) Lung infection–a public health priority. PLoS Med 3: e76. 2. Mizgerd JP (2008) Acute lower respiratory tract infection. N Engl J Med 358: 716–727. 3. WHO (2004) The World Health Report 2004 – Changing History. Geneva: World Health Organization. 4. Dushoff J, Plotkin JB, Viboud C, Earn DJ, Simonsen L (2006) Mortality due to influenza in the United States–an annualized regression approach using multiple-cause mortality data. Am J Epidemiol 163: 181–187. 5. Rothberg MB, Haessler SD, Brown RB (2008) Complications of viral influenza. Am J Med 121: 258–264. 6. Johnson NP, Mueller J (2002) Updating the accounts: global mortality of the 1918–1920 ‘‘Spanish’’ influenza pandemic. Bull Hist Med 76: 105–115. 7. Fidler DP, Gostin LO (2011) The WHO pandemic influenza preparedness framework: a milestone in global governance for health. Jama 306: 200–201. 8. Friede M, Palkonyay L, Alfonso C, Pervikov Y, Torelli G, et al. (2011) WHO initiative to increase global and equitable access to influenza vaccine in the event of a pandemic: supporting developing country production capacity through technology transfer. Vaccine 29 Suppl 1: A2–7. 9. Abdel-Ghafar AN, Chotpitayasunondh T, Gao Z, Hayden FG, Nguyen DH, et al. (2008) Update on avian influenza A (H5N1) virus infection in humans. N Engl J Med 358: 261–273. 10. Fang LQ, de Vlas SJ, Liang S, Looman CW, Gong P, et al. (2008) Environmental factors contributing to the spread of H5N1 avian influenza in mainland China. PLoS ONE 3: e2268. 11. Hilleman MR (2002) Overview: cause and prevention in biowarfare and bioterrorism. Vaccine 20: 3055–3067. 12. Hamad M (2011) Universal vaccines: shifting to one for many or shooting too high too soon! APMIS 119: 565–573. 13. Rage A, Hassan T (2011) Improving Influenza Vaccination Rates by Targeting Individuals not Seeking Early Seasonal Vaccination. Am J Med 124: e9. 14. (2011) Influenza vaccination coverage among pregnant women — United States, 2010–11 influenza season. MMWR Morb Mortal Wkly Rep 60: 1078–1082. 15. (2011) Influenza vaccination coverage among health-care personnel — United States, 2010–11 influenza season. MMWR Morb Mortal Wkly Rep 60: 1073–1077. 16. (2011) Place of influenza vaccination among adults — United States, 2010–11 influenza season. MMWR Morb Mortal Wkly Rep 60: 781–785. 17. Ambrose CS, Levin MJ, Belshe RB (2011) The relative efficacy of trivalent live attenuated and inactivated influenza vaccines in children and adults. Influenza Other Respi Viruses 5: 67–75. 18. Belshe RB (2010) The need for quadrivalent vaccine against seasonal influenza. Vaccine 28 Suppl 4: D45–53. 19. Mohty B, Bel M, Vukicevic M, Nagy M, Levrat E, et al. (2011) Graft-versus-host disease is the major determinant of humoral responses to the AS03-adjuvanted influenza A/09/H1N1 vaccine in allogeneic hematopoietic stem cell transplant recipients. Haematologica 96: 896–904. 20. Meier S, Bel M, L’Huillier A, Crisinel PA, Combescure C, et al. (2011) Antibody responses to natural influenza A/H1N1/09 disease or following immunization with adjuvanted vaccines, in immunocompetent and immunocompromised children. Vaccine 29: 3548–3557. 21. Crawford NW, Bines JE, Royle J, Buttery JP (2011) Optimizing immunization in pediatric special risk groups. Expert Rev Vaccines 10: 175–186. 22. Clement CG, Evans SE, Evans CM, Hawke D, Kobayashi R, et al. (2008) Stimulation of lung innate immunity protects against lethal pneumococcal pneumonia in mice. Am J Respir Crit Care Med 177: 1322–1330. 23. Clement CG, Tuvim MJ, Evans CM, Tuvin DM, Dickey BF, et al. (2009) Allergic lung inflammation alters neither susceptibility to Streptococcus pneumoniae infection nor inducibility of innate resistance in mice. Respir Res 10: 70. 24. Evans SE, Scott BL, Clement CG, Larson DT, Kontoyiannis D, et al. (2010) Stimulated innate resistance of lung epithelium protects mice broadly against bacteria and fungi. Am J Respir Cell Mol Biol 42: 40–50. 25. Evans SE, Tuvim MJ, Fox CJ, Sachdev N, Gibiansky L, et al. (2011) Inhaled Innate Immune Ligands To Prevent Pneumonia. Br J Pharmacol. PLoS ONE \| www.plosone.org 9 January 2012 \| Volume 7 \| Issue 1 \| e30596
https://openalex.org/W1967237797	https://europepmc.org/articles/pmc4086553?pdf=render	English	null	Prevalence of Malocclusion in Jaipur, India	Jaypee's international journal of clinical pediatric dentistry	2,009	cc-by	1,537	INTRODUCTION Normal alignment of teeth not only contributes to the oral health but also goes a long way in the overall well-being and personality of an individual. Correct tooth position is an important factor for esthetics, function and for overall preservation or restoration of dental health. While dental caries has been regarded as the one of the major dental disease throughout the world, malocclusion is a close runner- up. The morphogenetic nature of most malocclusions assures us that this dentofacial problem will continue to demand the best that dentistry can offer for a long time, indeed. Many organized population surveys have been carried out in different parts of the world with the objective of estimating prevalence of malocclusion and orthodontic treatment needs. 1Professor and Head, Department of Orthodontics and Dentofacial Orthopedics, Mahatma Gandhi Dental College and Hospital Jaipur, India 1Professor and Head, Department of Orthodontics and Dentofacial Orthopedics, Mahatma Gandhi Dental College and Hospital Jaipur, India Senior Lecturer, Department of Orthodontics and Dentofacial Orthopedics, Mahatma Gandhi Dental Co aipur, India 3Vice Principal, Mahatma Gandhi Dental College and Hospital, Jaipur, India Correspondence: Mridula Trehan Professor and Head, Department of Orthodontics and Dentofacial Orthopedics, Mahatma Gandhi Dental College and Hospital Jaipur, India, Ph: (91)141-3292139, 2770300, Fax: 2770326, e-mail: sunil_sharma977@yahoo.com The prevalence of malocclusion varies greatly in different parts of the world, in different ethnic groups and people of different origins. The prevalence of malocclusion among Indian population has been reported to be as low as 19.6% (Miglani DC et al in 1965) and as high as 90% in Delhi by Sidhu.4 To assign a treatment plan and to work out on the treatment needs of a particular group or population, it is mandatory to know the trends of occurrence of various malocclusions. As there is a lack of statistical data on malocclusions in this particular geographical area, a study was conducted on 700 patients in the city of Jaipur, the capital of Rajasthan, to identify the distribution of malocclusion. Though there is no single way to classify malocclusion, the most commonly and universally accepted Angle’s classification was used, due to its simplicity. Abstract A study was undertaken to determine the prevalence of malocclusion in Jaipur city, India. A total of 700 subjects, in the age group of 16-26 years were divided into five groups of normal occlusion, Angle’s Class I, Class II Div 1, Class II Div 2 and Class III malocclusion. The results revealed that the prevalence of malocclusion was 66.3%, with the majority of them having Class I malocclusion (57.9%), while the prevalence of Class III malocclusion was found to be the least (1.4%). There was no statistically significant gender difference among the subjects studied. Keywords: Malocclusion, Angle’s classification, prevalence. SURVEY SURVEY 1Mridula Trehan, 2Vinay K Chugh, 3Sunil Sharma 1Mridula Trehan, 2Vinay K Chugh, 3Sunil Sharma 1Mridula Trehan, 2Vinay K Chugh, 3Sunil Sharma 1Professor and Head, Department of Orthodontics and Dentofacial Orthopedics, Mahatma Gandhi Dental College and Hospital Jaipur, India 2Senior Lecturer, Department of Orthodontics and Dentofacial Orthopedics, Mahatma Gandhi Dental College and Hospital Jaipur India Group V: Class III malocclusion Group V: Class III malocclusion The collected data were tabulated and analyzed statistically. RESULTS The occlusal classification of the subjects is shown in Table 2. The occlusal classification of the subjects is shown in Table 2. females were found to have normal occlusion while prevalence of class III malocclusion was least. Table 2 (Fig. 1) shows that normal occlusion was found in 33.3% of the subjects and 66.7% of the subjects had malocclusion. Prevalence of Angle’s class I malocclusion was the highest (57.9%) and that of Angle’s class III malocclusion was the least (1.4%). MATERIAL AND METHOD The present study was carried out on 700 subjects ranging in ages from 16 to 26 years. The subjects were selected randomly from the patients reporting to the OPD of Mahatma Gandhi Dental College and Hospital, Jaipur. The criteria for selection of the subjects were as follows: 1. All permanent teeth present in each arch (excluding third molars) and in a sufficient state of eruption. 2. No previous history of orthodontic treatment in either arch. 3. No large coronal restoration that might have altered both coronal shape and size. 23 Mridula Trehan et al Fig. 1: Percentage distribution of groups All occlusal relationships were evaluated at a centric occlusion position which was achieved by asking the subject to swallow and then to bite on his or her teeth. The sample was divided into normal occlusion group and malocclusion groups, on the basis of Angle’s classification (Table 1). Fig. 1: Percentage distribution of groups Group I: Normal occlusion (NO): Only those subjects were included in the study which on clinical evaluation showed bilateral Angle’s Class I molar relationship with acceptable overjet and overbite and well-aligned arches. TABLE 2: Occlusal classification Occlusion N % Normal occlusion (NO) 233 33.3 Angle’s class I malocclusion 405 57.9 Angle’s class II Div 1 malocclusion 39 5.5 Angle’s class II Div 2 malocclusion 13 1.9 Angle’s class III malocclusion 10 1.4 Total 700 100.0 TABLE 2: Occlusal classification Group II: Showed bilateral Angle’s class I molar relation- ship with one or more of these characteristics: crowded incisors (Dewey type 1), protruded maxillary incisors (Dewey type 2), anterior cross-bite (Dewey type 3), unilateral or bilateral posterior cross-bite (Dewey type 4,) mesial drift of molars (Dewey type 5), anterior or posterior open bite and deep anterior overbite. TABLE 3: Gender distribution of occlusal variations Occlusal classfication Male Female N % N % Normal occlusion 184 32.8 49 35.5 Class I 334 59.6 71 51.5 Class I Div 1 28 5.0 11 8.0 Class II Div 2 6 1.1 7 5.0 Class III 10 1.5 0 0 TABLE 3: Gender distribution of occlusal variations Group III: Class II Div 1 malocclusion Group III: Class II Div 1 malocclusion Group IV: Class II Div 2 malocclusion Group IV: Class II Div 2 malocclusion REFERENCES 1. Kharbandha OP, Sidhu SS, Shukla DK, Sunderan KR. A Study of etiologic factors associated with the development of malocclussion. J Clin Pediatr Dent 1994 Winter;18(2):95-98. There was no statistically significant difference between males and females either in the prevalence of malocclusion. ( ) 2. Das UM, Venkatsubramanian RD. Prevelance of malocclusion among school children in Bangalore, India. Int J Clin Pediatr Dent 2008;1(1):10-12. 2. Das UM, Venkatsubramanian RD. Prevelance of malocclusion among school children in Bangalore, India. Int J Clin Pediatr Dent 2008;1(1):10-12. ACKNOWLEDGMENT I thank the Faculty of the Department of Oral Medicine and Radiology, Mahatma Gandhi Dental College and Hospital, Jaipur, Rajasthan for their cooperation in conducting the study. DISCUSSION The study showed that 66.3% of the subjects surveyed had malocclusion. This is somewhat similar to the findings of Das et al,2 who conducted an epidemiological study of malocclusion in the age group of 8-12 years in Bangalore city in 2008, and reported a high incidence of malocclusion Table 3 shows the gender distribution of normal occlusion and various types of malocclusion. 32.8% of the males had normal occlusion while Class II Div 2 malocclusion was least prevalent (1.1%). 35.5% of the 24 Prevalence of Malocclusion in Jaipur, India The prevalence of malocclusion is high, a reason to continue training professionals to care for those patients in need of treatment. The functional, esthetic, and psychologic benefits of orthodontic treatment ensure a continued seeking out of these services by those afflicted with malocclusion. of 71%. The findings of the present study are in disagreement with those of Kharbanda et al3 who have found 36.6% prevalence of malocclusion in Delhi, while Sidhu4 in his study found higher prevalence rate of Angle’s malocclusion of 90% in the age group of 6-30 years. of 71%. The findings of the present study are in disagreement with those of Kharbanda et al3 who have found 36.6% prevalence of malocclusion in Delhi, while Sidhu4 in his study found higher prevalence rate of Angle’s malocclusion of 90% in the age group of 6-30 years. Angle’s class I malocclusion (57.9%) was more common than Angle’s class II div 1 malocclusion (5.5%). This is similar to the findings of Das et al2 who reported 62% class I malocclusion and 7% class II Div 1 malocclusion. The prevalence of class II Div 2 and Class III malocclusion were low, i.e. 1.9% and 1.4% respectively. This is similar to the findings of Singh et al,3 who reported prevalence of class II Div 2 malocclusion of 5.85% and prevalence of class III malocclusion of 3.17% in his study of distribution of malocclusion among North Indians seeking orthodontic treatment. CONCLUSION 3. Kharbandha, OP.; Sidhu, SS.; Sunderan, KR. Malocclusion and associated factors among Delhi children. Project report Indian Council of Medical Research, New Delhi, 1990. From this study, the following conclusions have been drawn: 1. Prevalence of malocclusion was found to be 66.7%. 4. Sidhu SS. Incidence of varieties of malocclusion. J Ind Orthod Soc 1968;1:17-20. 2. Angle’s class I malocclusion was more prevalent as compared to the other types of malocclusion. 5. Singh SP, Utreja A, Chawla HS. A study of distribution of malocclusion among North Indians seeking orthodontic treatment. J Ind Orthod Soc 1993;24(2):47-53. 3. There was no statistically significant gender difference among the subjects studied. 25 International Journal of Clinical Pediatric Dentistry, January-April 2009;2(1):23-25
https://openalex.org/W4361874082	https://aacr.figshare.com/articles/journal_contribution/Supplemental_Table_S3_from_H3K9_Histone_Methyltransferase_KMT1E_SETDB1_Cooperates_with_the_SMAD2_3_Pathway_to_Suppress_Lung_Cancer_Metastasis/22403961/1/files/39849771.pdf	English	null	Supplemental Table S3 from H3K9 Histone Methyltransferase, KMT1E/SETDB1, Cooperates with the SMAD2/3 Pathway to Suppress Lung Cancer Metastasis	null	2,023	cc-by	58	Table S3. KMT1E DNA binding sequences identified by ChIP-Seq. RanK Motif Sequence Best Matched 1 CAGACGGCTC SMAD 2 ACCACASWCCAC RUNX 3 GGCAGGGCAGGC GC-BOX 4 GCWMTKYCSTSG BCL6 5 CCTGCCACGGCG CTF-NF1 Table S3. KMT1E DNA binding sequences identified by ChIP-Seq. RanK Motif Sequence Best Matched 1 CAGACGGCTC SMAD 2 ACCACASWCCAC RUNX 3 GGCAGGGCAGGC GC-BOX 4 GCWMTKYCSTSG BCL6 5 CCTGCCACGGCG CTF-NF1
https://openalex.org/W1821433401	https://pure.uvt.nl/ws/files/30435231/MCP_Huis_in_t_Veld_Facial_Expressive_Action_Stimulus_test_FiP_2015.pdf	English	null	The Facial Expressive Action Stimulus Test. A test battery for the assessment of face memory, face and object perception, configuration processing, and facial expression recognition	Frontiers in psychology	2,015	cc-by	12,929	The Facial Expressive Action Stimulus test De Gelder, Beatrice; Huis in 't Veld, Elisabeth; Van Den Stock, Jan Published in: Frontiers in Psychology DOI: 10.3389/fpsyg.2015.01609 Publication date: 2015 Document Version Publisher's PDF, also known as Version of record Link to publication in Tilburg University Research Portal Citation for published version (APA): De Gelder, B., Huis in 't Veld, E., & Van Den Stock, J. (2015). The Facial Expressive Action Stimulus test: A test battery for the assessment of face memory, face and object perception, configuration processing, and facial expression recognition. Frontiers in Psychology, 6, Article 1609. https://doi.org/10.3389/fpsyg.2015.01609 The Facial Expressive Action Stimulus test The Facial Expressive Action Stimulus test Citation for published version (APA): De Gelder, B., Huis in 't Veld, E., & Van Den Stock, J. (2015). The Facial Expressive Action Stimulus test: A test battery for the assessment of face memory, face and object perception, configuration processing, and facial expression recognition. Frontiers in Psychology, 6, Article 1609. https://doi.org/10.3389/fpsyg.2015.01609 Citation for published version (APA): De Gelder, B., Huis in 't Veld, E., & Van Den Stock, J. (2015). The Facial Expressive Action Stimulus test: A test battery for the assessment of face memory, face and object perception, configuration processing, and facial expression recognition. Frontiers in Psychology, 6, Article 1609. https://doi.org/10.3389/fpsyg.2015.01609 Citation for published version (APA): De Gelder, B., Huis in 't Veld, E., & Van Den Stock, J. (2015). The Facial Expressive Action Stimulus test: A test battery for the assessment of face memory, face and object perception, configuration processing, and facial expression recognition. Frontiers in Psychology, 6, Article 1609. https://doi.org/10.3389/fpsyg.2015.01609 Tilburg University General rights C i ht d General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private st • You may not further distribute the material or use it for any profit making activity or commercial gain y p py y p p p p p p y • You may not further distribute the material or use it for any profit-making activity or commercial gain You may freely distribute the URL identifying the publication in the public portal • You may not further distribute the material or use it for any profit-making activity o • You may freely distribute the URL identifying the publication in the public portal y y p g y • You may freely distribute the URL identifying the publication in the public portal Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Take down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim Download date: 24. Oct. 2024 ORIGINAL RESEARCH published: 29 October 2015 doi: 10.3389/fpsyg.2015.01609 Edited by: Luiz Pessoa, University of Maryland, USA Edited by: Luiz Pessoa, University of Maryland, USA Reviewed by: Yawei Cheng, National Yang-Ming University, Taiwan Wim Van Der Elst, Hasselt University, Belgium Keywords: face recognition, face memory, emotion recognition, conﬁgural face processing, inversion effect, experimental task battery Correspondence: Beatrice de Gelder b.degelder@maastrichtuniversity.nl The Facial Expressive Action Stimulus Test. A test battery for the assessment of face memory, face and object perception, conﬁguration processing, and facial expression recognition Beatrice de Gelder 1, 2, Elisabeth M. J. Huis in ‘t Veld 1, 3 and Jan Van den Stock 4, 5 1 Department of Cognitive Neuroscience, Maastricht University, Maastricht, Netherlands, 2 Department of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa, 3 Department of Medical and Clinical Psychology, Tilburg University, Tilburg, Netherlands, 4 Laboratory for Translational Neuropsychiatry, Department of Neurosciences, KU Leuven, Leuven, Belgium, 5 Old Age Psychiatry, University Hospitals Leuven, Leuven, Belgium There are many ways to assess face perception skills. In this study, we describe a novel task battery FEAST (Facial Expressive Action Stimulus Test) developed to test recognition of identity and expressions of human faces as well as stimulus control categories. The FEAST consists of a neutral and emotional face memory task, a face and shoe identity matching task, a face and house part-to-whole matching task, and a human and animal facial expression matching task. The identity and part-to-whole matching tasks contain both upright and inverted conditions. The results provide reference data of a healthy sample of controls in two age groups for future users of the FEAST. INTRODUCTION Specialty section: This article was submitted to Emotion Science, a section of the journal Frontiers in Psychology Face recognition is one of the most ubiquitous skills. The neural underpinnings of face perception are still a matter of debate. This is not surprising when one realizes that a face has a broad range of attributes. Identity is but one of these, and it is not clearly understood yet how a deﬁcit in that area aﬀects perception and recognition of other aspects of face perception. Prosopagnosia or absence of normal face identity recognition is one of the most peculiar neuropsychological symptoms and it has shed some light on the nature of face perception (de Gelder and Van den Stock, 2015). The term referred originally to loss of face recognition ability in adulthood following brain damage (Bodamer, 1947). Prosopagnosia can have a profound impact on social life, as in extreme cases the patients have diﬃculty recognizing the face of their spouse or child. More recently it has also been associated with neurodegenerative syndromes like fronto-temporal lobe degeneration (FTLD) (Snowden et al., 1989) and neurodevelopmental syndromes like cerebellar hypoplasia (Van den Stock et al., 2012b). In addition to the acquired variant, there is now general consensus on the existence of a developmental form, i.e., developmental prosopagnosia (DP). A recent prevalence study reported an estimate of 2.5% (Kennerknecht et al., 2006) and indicates that DP typically shows a hereditary proﬁle with an autosomal dominant pattern. Received: 22 July 2015 Accepted: 05 October 2015 Published: 29 October 2015 Edited by: Luiz Pessoa, University of Maryland, USA Reviewed by: Yawei Cheng, National Yang-Ming University, Taiwan Wim Van Der Elst, Hasselt University, Belgium Correspondence: Beatrice de Gelder b.degelder@maastrichtuniversity.nl Citation: de Gelder B, Huis in ‘t Veld EMJ and Van den Stock J (2015) The Facial Expressive Action Stimulus Test. A test battery for the assessment of face memory, face and object perception, conﬁguration processing, and facial expression recognition. Front. Psychol. 6:1609. doi: 10.3389/fpsyg.2015.01609 October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 1 The Facial Expressive Action Stimulus Test de Gelder et al. In view of the rich information carried by the face, an assessment of speciﬁc face processing skills is crucial. Two questions are central. One, what speciﬁc dimension of facial information are we focusing on, and two, is its loss speciﬁc for faces. To date, there is no consensus or golden standard regarding the best tool and performance level that allows diagnosing individuals with face recognition complaints as “prosopagnosic.” Several tests and tasks have been developed, such as the Cambridge Face Memory Test (Duchaine and Nakayama, 2006), the Benton Facial Recognition Test (Benton et al., 1983), the Cambridge Face Perception Task (Dingle et al., 2005), the Warrington Recognition Memory Test (Warrington, 1984) and various tests using famous faces (such as adaptations of the Bielefelder famous faces test, Fast et al., 2008). These each provide a measure or a set of measures relating to particular face processing abilities, e.g., matching facial identities or rely on memory for facial identities which is exactly what is problematic in people with face recognition disorders. More generally, beyond the diﬀerence between perception and memory, there is not yet a clear understanding of how the diﬀerent aspects of normal face perception are related. So testing of face skills should cast the net rather wide. A test battery suitable for the assessment of prosopagnosia should take some additional important factors into account. Firstly, to assess the face speciﬁcity of the complaints, the test battery should include not only tasks with faces, but also an equally demanding condition with control stimuli that are visually complex. Secondly, an important ﬁnding classically advanced to argue for a specialization for faces regards the conﬁgural way in which we seem to process faces, so the task should enable the measurement of conﬁgural processing of faces and objects. Subjects h j The participants were recruited between 2012 and 2015 from acquaintances of lab members and research students. Participation was voluntarily and no monetary reward was oﬀered. The following inclusion criteria were applied: right- handed, minimally 18 years old, normal or corrected-to-normal vision and normal basic visual functions as assessed by the Birmingham Object Recognition Battery (line length, size, orientation, gap, minimal feature match, foreshortened view, and object decision) (Riddoch and Humphreys, 1992). A history of psychiatric or neurological problems, as well as any other medical condition or medication use which would aﬀect performance, or history of a concussion, were exclusion criteria. This study was carried out in accordance with the recommendations and guidelines of the Maastricht University ethics committee, the “Ethische Commissie Psychologie” (ECP). The protocol was approved by the Maastricht University ethics committee (ECP- number: ECP-128 12_05_2013). In total, 61 people participated in the study. Three subjects were 80, 81, and 82 years old. Even though they adhered to every inclusion criteria, they were excluded from the analyses due to being outliers on age (more than 2 standard deviations from the mean). The sample thus consisted of 58 participants, between 18 and 62 years old (M = 38, SD = 15). Of those, 26 are male, between 19 and 60 years old (M = 38, SD = 15) and 32 women between 18 and 62 years old (M = 39, SD = 16). There are no diﬀerences in age between the genders [t(1, 56) = −0.474, p = 0.638]. Additionally, there is accumulating evidence in support of an interaction between face identity and face emotion processing (Van den Stock et al., 2008; Chen et al., 2011; Van den Stock and de Gelder, 2012, 2014) and there is increasing evidence that conﬁguration processing is positively related to emotion recognition ability (Bartlett and Searcy, 1993; Mckelvie, 1995; Calder et al., 2000; White, 2000; Calder and Jansen, 2005; Durand et al., 2007; Palermo et al., 2011; Tanaka et al., 2012; Calvo and Beltrán, 2014). We therefore extended our test battery with tasks targeting emotion recognition and emotion eﬀects on face memory, by adding an emotional face memory task and a However, an age distribution plot (see Figure 1) reveals a gap, where there are only 6 participants between 35 and 49. Citation: The matter of conﬁguration perception also has been tackled in several diﬀerent ways, such as with the composite face task (Young et al., 1987), the whole-part face superiority eﬀect (Tanaka and Farah, 1993) or more recently, using gaze-contingency (Van Belle et al., 2011). We choose to focus on the classical face inversion eﬀect (Yin, 1969; Farah et al., 1995), whose simple method lends itself very well to study object inversion eﬀects. Next, besides using the inversion eﬀect, conﬁguration- vs. feature-based processing can also be investigated more directly by part-to-whole matching tasks (de Gelder et al., 2003). Furthermore, previous studies have found positive relationships between the ability to process faces conﬁgurally and face memory (Richler et al., 2011; Huis in ‘t Veld et al., 2012; Wang et al., 2012; DeGutis et al., 2013) indicating that conﬁgural processing might facilitate memory for faces. Additionally there is accumulating evidence in support of an facial expression matching task. To stay with the rationale of our test that each skill tested with faces must also be tested with a selected category of control objects, we used canine face expressions. Taking all these aspects into account, we constructed a face perception test battery labeled the Facial Expressive Action Stimulus Test (FEAST). The FEAST is designed to provide a detailed assessment of multiple aspects of face recognition ability. Most of the subtests have been extensively described and validated on the occasion of prosopagnosia case reports and small group studies (de Gelder et al., 1998, 2000, 2003; de Gelder and Rouw, 2000a,b,c, 2001; Hadjikhani and de Gelder, 2002; de Gelder and Stekelenburg, 2005; Righart and de Gelder, 2007; Van den Stock et al., 2008, 2012a, 2013; Huis in ‘t Veld et al., 2012). But so far the test battery was not presented systematically as it had not been tested on a large sample of participants receiving the full set of subtests. Here, we report a new set of normative data for the ﬁnalized version of the FEAST, analyze the underlying relationships of the tasks, and freely provide the data and stimulus set to the research community for scientiﬁc purposes. Subjects h Therefore, the sample is split in two: one “young adult” group, younger than 42 and a “middle aged” group of participants between 47 and October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 2 de Gelder et al. The Facial Expressive Action Stimulus Test FIGURE 1 \| Age distribution of the sample with the young adult group between 18 and 41 years old, and a middle aged group between 47 and 62 years old. FIGURE 2 \| Stimulus example of (A) upright faces and (B) upright shoes in the face and shoe identity matching task. Some identities are different from the actual stimuli due to copyright and permissions. 62 years old. The young adult age group consisted of 15 men between 19 and 37 years old, (M = 26, SD = 6) and 17 women FIGURE 1 \| Age distribution of the sample with the young adult group between 18 and 41 years old, and a middle aged group between 47 and 62 years old. FIGURE 1 \| Age distribution of the sample with the young adult group between 18 and 41 years old, and a middle aged group between 47 and 62 years old. FIGURE 1 \| Age distribution of the sample with the young adult group between 18 and 41 years old, and a middle aged group between 47 and 62 years old. FIGURE 2 \| Stimulus example of (A) upright faces and (B) upright shoes in the face and shoe identity matching task. Some identities are different from the actual stimuli due to copyright and permissions. FIGURE 1 \| Age distribution of the sample with the young adult group between 18 and 41 years old, and a middle aged group between 47 and 62 years old FIGURE 2 \| Stimulus example of (A) upright faces and (B) upright shoes in the face and shoe identity matching task. Some identities are different from the actual stimuli due to copyright and permissions. FIGURE 2 \| Stimulus example of (A) upright faces and (B) upright shoes in the face and shoe identity matching task. Some identities are different from the actual stimuli due to copyright and permissions. FIGURE 2 \| Stimulus example of (A) upright faces and (B) upright shoes in the face and shoe identity matching task. Some identities are different from the actual stimuli due to copyright and permissions. 62 years old. Experimental Stimuli and Design The face and shoe identity matching task, face and house part- to-whole matching task, Neutral and Emotion Face Memory task (FaMe-N and FaMe-E) have been previously described including ﬁgures of stimulus examples (Huis in ‘t Veld et al., 2012). FIGURE 3 \| Stimulus examples of an (A) upright face and eyes and (B) upright house and windows trial in the face and house part-to-whole matching task. Subjects h The young adult age group consisted of 15 men between 19 and 37 years old, (M = 26, SD = 6) and 17 women between 18 and 41 years old (M = 26, SD = 8). The middle aged group consisted of 11 men between 47 and 60 years old (M = 53, SD = 4) and 15 women between 50 and 62 years old (M = 55, SD = 3). FIGURE 3 \| Stimulus examples of an (A) upright face and eyes and (B) upright house and windows trial in the face and house part-to-whole matching task. Face and Shoe Identity Matching Task and the Inversion Effect consisted of greyscale photographs of shoes (8 unique shoes) and faces (4 male, 4 female; neutral facial expression) with frontal view and 3/4 proﬁle view. A stimulus contained three pictures: one frontal view picture on top and two 3/4 proﬁle view pictures underneath. One of the two bottom pictures (target) was of the The face and shoe identity-matching task (de Gelder et al., 1998; de Gelder and Bertelson, 2009) was used to assess identity recognition and the inversion eﬀect for faces and objects. The test contained 4 conditions with a 2 category (faces and shoes) × 2 orientation (upright and inverted) factorial design. The materials October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 3 de Gelder et al. The Facial Expressive Action Stimulus Test FIGURE 4 \| Example stimulus of the Facial Expression Matching Task with an angry target and happy distracter stimulus trial for the (A) human and (B) canine experiment. FIGURE 4 \| Example stimulus of the Facial Expression Matching Task with an angry target and happy distracter stimulus trial for the (A) human and (B) canine experiment he Facial Expression Matching Task with an angry target and happy distracter stimulus trial for the (A) human and (B) FIGURE 4 \| Example stimulus of the Facial Expression Matching Task with an angry target and happy distracter stimulus trial for the (A) human and (B) canine experiment. FIGURE 5 \| Trial setup examples of the (A) encoding phase and (B) recollection phase of the FaMe-N. Identities are different from the actual stimuli due to copyright and permissions. es of the (A) encoding phase and (B) recollection phase of the FaMe-N. Identities are different from the actual stimuli due to FIGURE 5 \| Trial setup examples of the (A) encoding phase and (B) recollection phase of the FaMe-N. Identities are diffe copyright and permissions. al setup examples of the (A) encoding phase and (B) recollection phase of the FaMe-N. Identities are different from the act permissions total of 64 trials per block. Each block was preceded by 4 practice trials, during which the participants received feedback about their performance (see Figure 2). same identity as the one on top (sample) and the other was a distracter. The target and distracter pictures of the faces were matched for gender and hairstyle. Face and Shoe Identity Matching Task and the Inversion Effect Each stimulus was presented for 750 ms and participants were instructed to indicate by a button press which of the two bottom pictures represented the same exemplar as the one on top. Participants were instructed to answer as quickly but also as accurately as possible, and responses during stimulus presentation were collected. Following the response, a black screen with a ﬁxation cross was shown for a variable duration (800–1300 ms). The experiment consisted of four blocks (one block per condition). In each block, 16 stimuli were presented 4 times in a randomized order, adding up to a Frontiers in Psychology \| www.frontiersin.org Face and House Part-to-whole Matching Task This task is developed to assess holistic processing. The test also consisted of 4 conditions, with a 2 category (faces and houses) × 2 orientation (upright and inverted) factorial design. Materials consisted of grayscale pictures of eight faces (four male; neutral facial expression, photographed in front view and with direct gaze) and eight houses. From each face, part-stimuli were constructed by extracting the rectangle containing the eyes and October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 4 de Gelder et al. The Facial Expressive Action Stimulus Test FIGURE 6 \| Trial setup of a happy trial in the (A) encoding phase and (B) recollection phase of the FaMe-E. Some identities are different from the actual stimuli due to copyright and permissions. FIGURE 7 \| Means and standard errors of the mean of the accuracy and reaction times on the face and shoe matching task, split by age group. FIGURE 6 \| Trial setup of a happy trial in the (A) encoding phase and (B) recollection phase of the FaMe-E. Some identities are different from the actual stimuli due to copyright and permissions. FIGURE 6 \| Trial setup of a happy trial in the (A) encoding phase and (B) recollection phase of the FaMe-E. Some identities are different from the actual stimuli due to copyright and permissions. FIGURE 7 \| Means and standard errors of the mean of the accuracy and reaction times on the face and shoe matching task, split by age group. FIGURE 7 \| Means and standard errors of the mean of the accuracy and reaction times on the face and shoe matching task, split by age group. FIGURE 7 \| Means and standard errors of the mean of the accuracy and reaction times on the face and shoe matchin andard errors of the mean of the accuracy and reaction times on the face and shoe matching task, split by age group. total of 32 trials per block and 64 trials per condition. Within blocks, the presentation of the two parts (eyes or mouth, window or door) was randomized in order to prevent participants to pay attention only to one speciﬁc feature. The ﬁrst block of each condition was preceded by 4 practice trials, during which the participants received feedback about their performance (see Figure 3). the rectangle containing the mouth. Frontiers in Psychology \| www.frontiersin.org Face and House Part-to-whole Matching Task House-part stimuli were created using a similar procedure, but the parts consisted of the door or window. The trial procedure was similar to the face and object identity matching task, where a whole face or house was presented on top (sample), with a target part-picture and a distractor part-picture presented underneath. Each trial was presented for 750 ms and participants were instructed to indicate by a button press which of the two bottom pictures represented the same exemplar as the one on top. Participants were instructed to answer as quickly but also as accurately as possible, and responses during stimulus presentation were collected. Following the response, a black screen with a ﬁxation cross was shown for a variable duration (800–1300 ms). The experiment consisted of eight blocks (two blocks per condition). In each block, 16 stimuli were presented 2 times in a randomized order, adding up to a Facial Expression Matching Task (FEM-H and FEM-C) The FEM is a match-to-sample task used to measure emotion recognition ability in both human and canine faces. The experiment was divided into two parts. The ﬁrst part consisted of human facial expressions (anger, fear, happy, sad, surprise, disgust). The materials consisted of grayscale photographs of facial expressions of 34 female identities and 35 male identities October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 5 The Facial Expressive Action Stimulus Test de Gelder et al. taken from the Karolinska Directed Emotional Faces (KDEF) (Lundqvist et al., 1998). This task has been used previously in Van den Stock et al. (2015). A stimulus consisted of three pictures: one picture on top (sample) and two pictures underneath. One of the two bottom pictures showed a face expressing the same emotion as the sample, the other was a distracter. The target and distracter pictures of the faces were matched for gender for the human stimuli. Each trial was presented until a response was given, but participants were instructed to answer as quickly and accurately as possible. Following the response, a black screen with a ﬁxation cross was shown for a variable duration (800–1300 ms). Each emotional condition contained 10 trials (5 male) in which the target emotion was paired with a distracter from each of the other emotions once per gender, resulting in 60 trials in total. Face and House Part-to-whole Matching Task The ﬁrst part was preceded by 4 practice trials, during which the participants received feedback about their performance. from the internet by EH. These pictures were validated in a pilot study using 28 students of Tilburg University in exchange for course credit. The participants indicated of each photo whether they thought the dog was expressing anger, fear, happiness, sadness or no emotion in particular (neutral) and secondly, how intense they rated the emotional expression on a scale from one to ﬁve. Twelve angry, twelve fearful, and twelve happy canine expressions were accurately recognized by more than 80% of the participants and used in the experiment. The canine part consisted of 72 trials in total, 24 per emotion condition, in which each target emotion was paired with each of the distracter emotions 12 times. The experiment was preceded by 2 practice trials, during which the participants received feedback about their performance (see Figure 4). Neutral Face Memory Task (FaMe-N) The second part consisted of canine facial expressions. In total, 114 pictures of dogs which could be perceived as angry (17), fearful (27), happy (17), neutral (29), and sad (24) were taken Based on the Recognition Memory Test (Warrington, 1984), the FaMe-N consists of an encoding and a recognition phase. The stimuli consist of 100 grayscale Caucasian faces (50 male) with a TABLE 1 \| Means and standard deviations on the face and shoe matching task by age group. Accuracy Young adult Middle aged M (%) SD M (%) SD Faces Upright 92 7 89 9 Inverted 89 8 85 8 Shoes Upright 89 7 86 7 Inverted 91 7 88 8 Reaction times (ms) M SD M SD Faces Upright 999 202 1162 280 Inverted 951 202 1146 225 Shoes Upright 920 175 1147 231 Inverted 891 177 1100 201 TABLE 2 \| Means and standard deviations on the face and house part-to-whole matching task by age group. Accuracy Young adult Middle aged M (%) SD M (%) SD Face parts Upright 71 8 63 7 Inverted 65 9 61 7 House parts Upright 77 8 72 9 Inverted 78 11 73 9 Reaction times (ms) M SD M SD Face parts Upright 1127 186 1346 218 Inverted 1099 222 1299 215 House parts Upright 1104 172 1307 163 Inverted 1046 166 1309 178 FIGURE 8 \| Means and standard errors of the mean of the accuracy and reaction times on the face and house part-to-whole matching task split by age group. TABLE 1 \| Means and standard deviations on the face and shoe matching task by age group. Accuracy Young adult Middle aged M (%) SD M (%) SD Faces Upright 92 7 89 9 Inverted 89 8 85 8 Shoes Upright 89 7 86 7 Inverted 91 7 88 8 Reaction times (ms) M SD M SD Faces Upright 999 202 1162 280 Inverted 951 202 1146 225 Shoes Upright 920 175 1147 231 Inverted 891 177 1100 201 TABLE 2 \| Means and standard deviations on the face and house part-to-whole matching task by age group. Neutral Face Memory Task (FaMe-N) Accuracy Young adult Middle aged M (%) SD M (%) SD Face parts Upright 71 8 63 7 Inverted 65 9 61 7 House parts Upright 77 8 72 9 Inverted 78 11 73 9 Reaction times (ms) M SD M SD Face parts Upright 1127 186 1346 218 Inverted 1099 222 1299 215 House parts Upright 1104 172 1307 163 Inverted 1046 166 1309 178 FIGURE 8 \| Means and standard errors of the mean of the accuracy and reaction times on the face and house part-to-whole matching task split by age group. TABLE 1 \| Means and standard deviations on the face and shoe matching task by age group. TABLE 2 \| Means and standard deviations on the face and house part-to-whole matching task by age group. FIGURE 8 \| Means and standard errors of the mean of the accuracy and reaction times on the face and house part-to-whole matching task split by age group. FIGURE 8 \| Means and standard errors of the mean of the accuracy and reaction times on the face and house part-to-whole matching task split by age group. FIGURE 8 \| Means and standard errors of the mean of the accuracy and reaction times on the face and house part-to age group. FIGURE 8 \| Means and standard errors of the mean of the accuracy and reaction times on the face and house part-to-whole matching task split by age group. andard errors of the mean of the accuracy and reaction times on the face and house part-to-whole matching task split October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 6 The Facial Expressive Action Stimulus Test de Gelder et al. FIGURE 9 \| Means and standard errors of the mean of the accuracy of the whole group and reaction times on the FEM-H split by age group. p < 0.001, p < 0.05. FIGURE 9 \| Means and standard errors of the mean of the accuracy of the whole group and reaction times on the FEM-H split by age group. *p < 0.001, p < 0.05. FIGURE 9 \| Means and standard errors of the mean of the accuracy of the whole group and reaction times on the *p < 0.001, p < 0.05. TABLE 3 \| Means and standard deviations on the FEM-H by age group. Emotional Face Memory Task (FaMe-E) Emotional Face Memory Task (FaMe-E) This task was designed by adapting the FaMe-N task by using stimuli containing emotional instead of neutral faces. Images were taken from the NimStim database (Tottenham et al., 2009) and stimuli created at Tilburg University. The stimuli consisted of 96 photographs (53 female) with direct eye gaze and frontal view. The individuals in the stimuli express fear, sadness, or happiness. There was no overlap in identities with the FaMe-N. The procedure was similar to the FaMe-N, but with 48 trials (16 per emotion) in both phases. The pictures making a stimulus pair were matched for emotion and hairstyle and in most trials also gender (see Figure 6). Neutral Face Memory Task (FaMe-N) Accuracy Young adult Middle aged M (%) SD M (%) SD Total 79 9 74 10 Anger 85 16 77 18 Fear 57 14 53 15 Happy 94 8 92 10 Disgust 82 13 82 12 Sad 69 17 59 15 Surprise 86 11 79 14 Reaction times (ms) M SD M SD Total 2064 583 2628 493 Anger 2122 707 2819 541 Fear 2279 674 2976 662 Happy 1941 727 2253 647 Disgust 1951 627 2635 604 Sad 2276 733 2776 586 Surprise 1976 551 2574 598 TABLE 3 \| Means and standard deviations on the FEM-H by age group. simultaneous presentation of two adjacent faces. One was the target face and was also presented in the encoding phase. The other face was not previously presented in the encoding phase and served as distracter. Fifty trials were randomly presented and target and distractor presentation side were evenly distributed. Participants were instructed to indicate as quickly and also as accurately as possible which face was also presented in the encoding phase. The stimulus pairs were matched for gender and hairstyle (see Figure 5). 13 82 1 17 59 1 11 79 1 SD M S 583 2628 49 707 2819 54 674 2976 66 727 2253 64 627 2635 60 733 2776 58 551 2574 59 t view, with frontal eye gaze. Th ase created at Tilburg Universit consisted of the presentation followed by a black screen wi uration of 1000 ms. Participan face carefully and told that the tested afterwards. The encodin mmediately followed upon th ecognition phase consisted of th .org Analyses neutral facial expression, in front view, with frontal eye gaze. The stimuli were taken from a database created at Tilburg University. Trials in the encoding phase consisted of the presentation of a single stimulus for 3000 ms, followed by a black screen with a white ﬁxation cross with a duration of 1000 ms. Participants were instructed to encode each face carefully and told that their memory for the faces would be tested afterwards. The encoding block consisted of 50 trials. neutral facial expression, in front view, with frontal eye gaze. The stimuli were taken from a database created at Tilburg University. Trials in the encoding phase consisted of the presentation of a single stimulus for 3000 ms, followed by a black screen with a white ﬁxation cross with a duration of 1000 ms. Participants were instructed to encode each face carefully and told that their memory for the faces would be tested afterwards. The encoding block consisted of 50 trials. Accuracies were calculated as the total proportion of correct responses for both the total score of each task and for each condition separately. Average response times from stimulus onset were calculated for the correct responses only. For all tasks, reaction times faster than 150 ms were excluded from analyses. In addition, for the identity matching task and part-to-whole matching task, reaction times longer than 3000 ms were excluded from analyses. For the other tasks, reaction times longer than 5000 ms were excluded from analyses. The number of outliers are reported in the results. One control subject did not complete the The recognition phase immediately followed upon the encoding phase. A trial in the recognition phase consisted of the October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 7 de Gelder et al. The Facial Expressive Action Stimulus Test FIGURE 10 \| Means and standard errors of the mean of the accuracy and reaction times on the FEM-Canine split by age group. *p < 0.001. FIGURE 10 \| Means and standard errors of the mean of the accuracy and reaction times on the FEM-Canine split by age group. p < 0.001. FIGURE 10 \| Means and standard errors of the mean of the accuracy and reaction times on the FEM-Canine split by a TABLE 4 \| Means and standard deviations on the FEM-C by age group. Face and Shoe Identity Matching Task Face and Shoe Identity Matching Task The task has a good internal consistency of ρKR20 = 0.912. The following number of outliers were discarded; upright faces: a total of 0.86% outliers across ten participants (M = 3.2 trials, SD = 2.7, min = 1, max = 8); inverted faces: 0.7% across ten participants (M = 2.6 trials, SD = 2.7, min = 1, max = 10); upright shoes: 0.9% across 15 participants (M = 2.1 trials, SD = 2, min = 1, max = 7) and inverted shoes: 0.5% across four participants (M = 4.8 trials, SD = 5.7, min = 1, max = 13). face and house part-to-whole matching task. The SPSS dataset can be downloaded through the supplementary materials. In addition, the internal consistency was assessed with the Kuder Richardson coeﬃcient of reliability (KR 20), reported as ρKR20, which is analogous to Cronbach’s alpha but suitable for dichotomous measures (Kuder and Richardson, 1937). The results were analyzed using repeated measures GLMs, with the experimental factors as within subject variables and age group and gender as between subject variables. Interaction eﬀects were further explored using post-hoc paired samples t-tests. The assumption of equality of error variances was checked with a Levene’s test. The assumption of normality was not formally tested, as the sample is larger than 30 and repeated measures GLMs are quite robust against violations of normality. A repeated measures GLM on accuracy scores with category (faces, shoes) and orientation (upright, inverted) as within- subject factors and gender and age group as between-subject factors revealed a category by orientation interaction eﬀect [F(1, 54) = 16.955, p < 0.001, η2 p = 0.24]. Paired samples t-tests show that upright faces are recognized more accurately than inverted faces [t(57) = 3.464, p = 0.001] and inverted shoes are recognized better than upright shoes [t(57) = −2.254, p = 0.028]. Also, the middle aged group is less accurate overall [F(1, 54) = 4.342, p = 0.042, η2 p = 0.07]. Inversion scores were calculated by subtracting the accuracy and reaction time scores on the inverted presentation condition from the upright condition. A positive score indicates that accuracy was higher, or the reaction time was longer, on the upright condition. A negative score indicates higher accuracy or reaction times for the inverted condition. Frontiers in Psychology \| www.frontiersin.org Analyses Accuracy Young adult Middle aged M (%) SD M (%) SD Total 92 8 86 7 Anger 96 8 91 9 Happy 86 11 77 11 Fear 95 9 90 9 Reaction times (ms) M SD M SD Total 2064 583 2628 493 Anger 1446 532 1998 440 Happy 1874 613 2455 392 Fear 1683 571 2351 465 4 \| Means and standard deviations on the FEM-C by age gr a stronger conﬁguration processing as measured by a higher accuracy inversion eﬀect is related to improved face memory and emotion recognition, multiple linear regression analyses were performed with accuracy scores on the FaMe-N, FaMe-E, and both FEM tasks as dependent variable and age, gender, and four inversion scores (face identity, shoe identity, face-part, and house-part) as predictors. In addition, correlations between all tasks were calculated. Lastly, percentile ranks of all tasks and correlations between all tasks were calculated and reported for both the accuracy scores and reaction times (see Tables 8–11). RESULTS Face and Shoe Identity Matching Task Face and Shoe Identity Matching Task To assess whether p A repeated measures GLM with a similar design on reaction times showed that faces are matched slower than shoes [F(1, 54) = 16.063, p < 0.001, η2 p = 0.23], upright faces and October 2015 \| Volume 6 \| Article 1609 8 de Gelder et al. The Facial Expressive Action Stimulus Test FIGURE 11 \| Means and standard errors of the mean of the accuracy and reaction times on the FaMe-N and FaMe-E. accuracy and reaction times on the FaMe-N and FaMe-E. FIGURE 11 \| Means and standard errors of the mean of the accuracy and reaction times on the FaMe-N and FaMe-E. TABLE 5 \| Means and standard deviations on the FaMe-N and the FaMe-E by gender and age group. Accuracy (%) Male Female Young adult Middle aged Young adult Middle aged M (%) SD M (%) SD M (%) SD M (%) SD FaMe-Neutral Total 77 16 81 11 77 13 78 10 FaMe-Emotion Total 78 14 84 7 81 11 78 15 Fear 78 13 82 7 80 16 76 17 Happy 80 19 86 10 83 11 78 17 Sad 77 15 83 7 80 12 80 13 Reaction times (ms) M SD M SD M SD M SD FaMe-Neutral Total 1920 532 2285 540 2090 483 2236 429 FaMe-Emotion Total 1785 525 2007 246 1821 416 2025 430 Fear 1778 544 2056 320 1942 540 1986 402 Happy 1775 564 2003 275 1803 399 2038 540 Sad 1791 514 1962 258 1718 414 2063 479 TABLE 5 \| Means and standard deviations on the FaMe-N and the FaMe-E by gender and age group. across 33 participants (M = 2.2 trials, SD = 1.6, min = 1, max = 6). across 33 participants (M = 2.2 trials, SD = 1.6, min = 1, max = 6). shoes are matched slower than inverted ones [F(1, 54) = 7.560, p = 0.008, η2 p = 0.12] and the middle aged group responded slower [F(1, 54) = 15.174, p < 0.001, η2 p = 0.22; see Figure 7 and Table 1]. A repeated measures GLM on accuracy scores with category (faces, houses) and orientation (upright, inverted) as within- subject factors and gender and age group as between-subject factors revealed a three way age group by category by orientation interaction eﬀect [F(1, 53) = 5.413, p = 0.024, η2 p = 0.09]. Face and Shoe Identity Matching Task Overall, both age groups are better at part to whole matching of houses [F(1, 53) = 153.660, p < 0.001, η2 p = 0.75]. However, the young adult group is more accurately able to part to whole match upright than inverted faces [t(31) = 5.369, p < 0.001], whereas the middle aged group is not [t(24) = 0.952, p = 0.351], but no such group diﬀerences are found for house inversion [young adult group: t(31) = −0.958, p = 0.345, middle aged group: t(24) = −0.490, p = 0.628]. Face and House Part-to-whole Matching Task The task has a good internal consistency of ρKR20 = 0.865. The following number of outliers were discarded; upright face parts: a total of 1.02% outliers across 38 participants (M = 2.7 trials, SD = 2.2, min = 1, max = 8); inverted face parts: 1.1% across 41 participants (M = 3.2 trials, SD = 3.2, min = 1, max = 13); upright house parts: 1.5% across 54 participants (M = 2.5 trials, SD = 2.8, min = 1, max = 12) and inverted house parts: 0.9% October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 9 The Facial Expressive Action Stimulus Test de Gelder et al. TABLE 8 \| Percentile ranks corresponding to accuracy scores (as percentage correct) split by age group for all tasks and subtasks. Face and House Part-to-whole Matching Task 2 5 10 25 50 75 90 95 PERCENTILE RANKS YOUNG ADULT GROUP FaMe-N 36 43 61 72 78 86 94 99 FaMe-E 48 55 65 71 79 91 96 98 Faces Upr 72 76 80 89 93 98 98 98 Inv 69 69 73 85 91 95 98 99 Shoes Upr 64 71 83 86 91 94 97 98 Inv 73 78 81 88 92 97 98 100 Face parts Upr 50 54 60 66 71 78 81 86 Inv 48 50 52 58 65 71 78 81 House parts Upr 59 60 65 72 78 81 88 91 Inv 41 54 65 72 78 84 91 93 FEM-H 53 60 65 74 82 85 90 90 FEM-C 57 75 85 88 94 97 98 99 PERCENTILE RANKS MIDDLE AGED GROUP FaMe-N 56 57 64 72 81 88 93 94 FaMe-E 42 47 65 77 82 90 94 96 Faces Upr 69 70 75 82 91 95 98 99 Inv 63 65 72 81 86 89 93 97 Shoes Upr 69 70 75 81 88 91 94 96 Inv 67 68 75 86 89 93 96 99 Face parts Upr 48 49 53 57 64 67 73 75 Inv 44 44 50 57 61 68 70 73 House parts Upr 53 53 58 67 75 78 83 89 Inv 55 57 62 66 72 79 87 90 FEM-H 50 54 63 67 73 83 87 91 FEM-C 67 70 75 83 88 92 94 97 TABLE 6 \| Regression coefﬁcients of the inversion scores on the tasks for conﬁgural and feature-based processing on the total scores of the Face Memory–Neutral and the Face Memory–Emotion task. Face and House Part-to-whole Matching Task Step 1 FaMe-N FaMe-E B Se B β B Se B β Constant 0.730 0.047 0.754 0.043 Gender −0.010 0.032 −0.039 0.003 0.029 0.013 Age 0.002 0.001 0.186 0.001 0.001 0.183 R2 0.036 0.034 Step 2 B Se B β B Se B β Constant 0.728 0.059 0.778 0.052 Gender −0.010 0.036 −0.040 −0.008 0.032 −0.037 Age 0.001 0.001 0.182 0.001 0.001 0.150 Face inversion 0.001 0.003 0.034 −0.001 0.003 −0.066 Shoe inversion −0.001 0.005 −0.027 −0.0004 0.004 −0.014 Face part inversion 0.000 0.003 −0.001 −0.003 0.003 −0.142 House part inversion −0.001 0.003 −0.053 −0.003 0.003 −0.153 R2 change 0.004 0.044 TABLE 7 \| Regression coefﬁcients of the inversion scores on the tasks for conﬁgural and feature-based processing on the total scores of the Facial Expression Matching- Human and Canine task. Expression Matching- Human and Canine task. Step 1 FEM-H FEM-C B Se B β B Se B β Constant 0.831 0.034 0.955 0.028 Gender −0.003 0.023 −0.014 −0.011 0.020 −0.076 Age −0.002 0.001 −0.264 −0.001 0.001 −0.261 R2 0.07 0.034 Step 2 B Se B β B Se B B Constant 0.829 0.041 0.965 0.035 Gender −0.003 0.026 −0.019 −0.019 0.021 −0.127 Age −0.002 0.001 −0.255 −0.002 0.001 −0.319* Face inversion 0.000 0.002 −0.024 0.001 0.002 0.091 Shoe inversion 0.000 0.003 0.017 −0.004 0.003 −0.181 Face part inversion 0.000 0.002 −0.021 −0.001 0.002 −0.092 House part inversion −0.004 0.002 −0.227 0.000 0.002 −0.033 R2 change 0.054 0.044 p < 0.05. Face and House Part-to-whole Matching Task Step 1 FaMe-N FaMe-E B Se B β B Se B β Constant 0.730 0.047 0.754 0.043 Gender −0.010 0.032 −0.039 0.003 0.029 0.013 Age 0.002 0.001 0.186 0.001 0.001 0.183 R2 0.036 0.034 Step 2 B Se B β B Se B β Constant 0.728 0.059 0.778 0.052 Gender −0.010 0.036 −0.040 −0.008 0.032 −0.037 Age 0.001 0.001 0.182 0.001 0.001 0.150 Face inversion 0.001 0.003 0.034 −0.001 0.003 −0.066 Shoe inversion −0.001 0.005 −0.027 −0.0004 0.004 −0.014 Face part inversion 0.000 0.003 −0.001 −0.003 0.003 −0.142 House part inversion −0.001 0.003 −0.053 −0.003 0.003 −0.153 R2 change 0.004 0.044 TABLE 7 \| Regression coefﬁcients of the inversion scores on the tasks for conﬁgural and feature-based processing on the total scores of the Facial Expression Matching- Human and Canine task. Step 1 FEM-H FEM-C B Se B β B Se B β Constant 0.831 0.034 0.955 0.028 Gender −0.003 0.023 −0.014 −0.011 0.020 −0.076 Age −0.002 0.001 −0.264 −0.001 0.001 −0.261 R2 0.07 0.034 Step 2 B Se B β B Se B B Constant 0.829 0.041 0.965 0.035 Gender −0.003 0.026 −0.019 −0.019 0.021 −0.127 Age −0.002 0.001 −0.255 −0.002 0.001 −0.319* Face inversion 0.000 0.002 −0.024 0.001 0.002 0.091 Shoe inversion 0.000 0.003 0.017 −0.004 0.003 −0.181 Face part inversion 0.000 0.002 −0.021 −0.001 0.002 −0.092 House part inversion −0.004 0.002 −0.227 0.000 0.002 −0.033 R2 change 0.054 0.044 *p < 0.05. The same repeated measures GLM on reaction times revealed a three way gender by age group by category interaction eﬀect [F(1 53) = 5.539, p = 0.022, η2p = 0.10]. To assess this TABLE 6 \| Regression coefﬁcients of the inversion scores on the tasks for conﬁgural and feature-based processing on the total scores of the Face Memory–Neutral and the Face Memory–Emotion task. TABLE 8 \| Percentile ranks corresponding to accuracy scores (as percentage correct) split by age group for all tasks and subtasks. g p g Memory–Neutral and the Face Memory–Emotion task. Facial Expression Matching Task H F i l E i (FEM H) Facial Expression Matching Task Facial Expression Matching Task Human Facial Expressions (FEM-H) p g Human Facial Expressions (FEM-H) The task has a reasonably good internal consistency of ρKR20 = 0.769. The following number of outliers were discarded from 47 participants; 14% in total (Anger: 2.5%, disgust: 1.8%, fear: 3.4%, happy: 0.7%, sad: 3.5%, surprise: 2.2%, M = 10.4 trials, SD = 6.6, min = 1, max = 27). A repeated measures GLM on the accuracy scores with emotion (fear, sadness, anger, disgust, surprise, and happy) as within subject variables and gender and age group as between subject variables showed a main eﬀect of emotion [F(5, 50) = 88.169, p < 0.001, η2 p = 0.90]. Post-hoc contrasts reveal that fear is recognized least accurate, worse than sadness [F(1, 54) = 15.998, p < 0.001, η2 p = 0.23], on which accuracy rates are in turn lower than anger [F(1, 54) = 63.817, p < 0.001, η2 p = 0.54]. Also, happy is recognized best with higher accuracy scores than surprise [F(1, 54) = 49.157, p < 0.001, η2 p = 0.48]. The same repeated measures GLM on reaction times revealed a three way gender by age group by category interaction eﬀect [F(1, 53) = 5.539, p = 0.022, η2p = 0.10]. To assess this eﬀect, the repeated measures GLM with category (faces, houses) and orientation (upright, inverted) as within-subject factors and age group as between-subject factors was run for males and females separately. For the female group, a category by age group interaction eﬀect is found [F(1, 29) = 7.022, p = 0.013, η2 p = 0.20], whereas no signiﬁcant eﬀects were found for men (see Figure 8 and Table 2). p The same repeated measures GLM on the reaction time data revealed a main eﬀect of emotion [F(5, 50) = 15.055, p < 0.001, η2 p = 0.60]. Happy was also recognized fastest (as compared to surprise, F(1, 54) = 7.873, p = 0.007, η2 p = 0.13] and disgust was recognized slower than anger [F(1, 54) = 7.776, p = 0.007, η2 p = 0.13]. Also, the middle aged age group is slower overall [F(1, 54) = 15.280, p < 0.001, η2 p = 0.22; see Figure 9 and Table 3]. October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 10 The Facial Expressive Action Stimulus Test de Gelder et al. Relationships between Tasks A repeated measures GLM on the accuracy scores with emotion (fear, anger, and happy) as within subject variables and gender and age group as between subject variables revealed a main eﬀect of emotion [F(2, 53) = 37.049, p < 0.001, η2 p = 0.58]. Fear was recognized least accurately [as compared to happy, F(1,54) = 65.310, p < 0.001, η2 p = 0.55]. Also, the middle aged group was less accurate at this task than the young adult group [F(1, 54) = 8.045, p = 0.006, η2 p = 0.13]. In the current sample, no signiﬁcant predictive relationship between conﬁguration processing as measured by the inversion eﬀect and face memory scores were found (see Table 6). Similarly, no signiﬁcant relationship between conﬁguration processing and emotion recognition scores were found, aside from a negative eﬀect of age on accuracy on the FEM-H and FEM-C, see Table 7. In addition, see Tables 8, 9 for correlations between the all the tasks and subtasks of the FEAST. ( ) p Similarly, for reaction times a main eﬀect of emotion [F(2, 53) = 66.335, p < 0.001, η2 p = 0.72] was observed; anger is recognized quicker than happy [F(1, 54) = 74.880, p < 0.001, η2 p = 0.58], which is in turn recognized a faster than fear [F(1, 54) = 17.588, p < 0.001, η2 p = 0.25]. Additionally, again the middle aged group is slower overall [F(1, 54) = 19.817, p < 0.001, η2 p = 0.27; see Figure 10 and Table 4]. Furthermore, percentile ranks for accuracy scores as percentage correct and the reaction times are reported in Tables 8, 9, and the correlations between all tasks are reported in Tables 10, 11. Facial Expression Matching Task H F i l E i (FEM H) TABLE 9 \| Percentile ranks corresponding to reaction times split by age group for all tasks and subtasks. 2 5 10 25 50 75 90 95 PERCENTILE RANKS THE YOUNG ADULT GROUP FaMe-N 1220 1248 1329 1621 1996 2329 2589 3210 FaMe-E 869 978 1097 1462 1842 2145 2505 2582 Faces Upr 671 693 733 832 974 1112 1293 1472 Inv 670 673 708 782 908 1068 1235 1407 Shoes Upr 591 663 707 777 922 1049 1204 1263 Inv 605 617 666 741 879 1010 1177 1227 Face parts Upr 591 718 910 1025 1130 1259 1281 1484 Inv 481 544 909 997 1084 1230 1393 1499 House parts Upr 688 774 882 1001 1073 1228 1332 1445 Inv 577 710 921 954 1023 1161 1252 1361 FEM-H 1080 1090 1169 1659 2032 2482 2769 3267 FEM-C 798 887 1123 1256 1458 2048 2581 2911 PERCENTILE RANKS FOR THE MIDDLE AGED GROUP FaMe-N 1380 1389 1623 1948 2142 2631 2932 3194 FaMe-E 1359 1389 1466 1803 2025 2231 2510 2787 Faces Upr 680 735 851 985 1114 1286 1560 1903 Inv 683 713 846 988 1116 1328 1484 1503 Shoes Upr 667 709 822 975 1134 1310 1483 1614 Inv 722 746 815 935 1085 1280 1378 1391 Face parts Upr 807 854 1026 1236 1353 1492 1648 1722 Inv 720 783 980 1207 1319 1452 1621 1627 House parts Upr 985 1011 1078 1190 1355 1401 1531 1599 Inv 1017 1018 1074 1173 1274 1469 1555 1658 FEM-H 1885 1887 1915 2212 2642 3004 3264 3640 FEM-C 1687 1688 1699 1905 2245 2603 2738 2987 TABLE 9 \| Percentile ranks corresponding to reaction times split by age group for all tasks and subtasks. are found for gender [F(1, 54) = 0.238, p = 0.628, η2 p = 0.004] or age group [F(1, 54) = 0.469, p = 0.496, η2 p = 0.009], nor is there any interaction eﬀect. Also, the average reaction time was 2121 ms (SD = 501) no diﬀerence in reaction times were found for gender [F(1, 54) = 0.211, p = 0.648, η2 p = 0.004] but the eﬀect of age group was near signiﬁcance [F(1, 54) = 3.768, p = 0.057, η2 p = 0.065; see Figure 11 and Table 5]. Canine Facial Expressions (FEM-C) Canine Facial Expressions (FEM-C) The task has a good internal consistency of ρKR20 = 0.847. From 35 participants, 5.3% of the trials were discarded (Anger: 1.1%, fear: 2.8%, happy: 1.4%, M = 6.3 trials, SD = 4.9, min = 1, max = 22). The task has a good internal consistency of ρKR20 = 0.847. From 35 participants, 5.3% of the trials were discarded (Anger: 1.1%, fear: 2.8%, happy: 1.4%, M = 6.3 trials, SD = 4.9, min = 1, max = 22). Emotional Face Memory Task (FaMe-E) The task has a good internal consistency of ρKR20 = 0.799. In total 125 trials (4.5%) were outliers across 34 participants (M = 3.7, SD = 3.5, min = 1, max = 19). A repeated measures GLM on accuracy scores and reaction times scores with emotion (fear, happy, sad) as within-subject factors and gender and age group as between subject variables revealed no signiﬁcant eﬀects. g However, a gender by age group by emotion three-way interaction eﬀect was found for reaction times, [F(2, 53) = 3.197, p = 0.049, η2 p = 0.11]. Figure 11 shows that the pattern of results between men and women is reversed when the age groups are compared. It looks like young adult women seem quicker to recognize sadness than middle aged women: indeed, if the repeated measures is run for men and women separately, with emotion as within subject variables and age group as between, no eﬀects of emotion or age group are found for men. However, for women, an emotion by age group interaction trend is found [F(2, 29) = 2.987, p = 0.066, η2 p = 0.17; see Figure 11 and Table 5]. In addition, we directly compared the FaMe-N and FaMe-E using a repeated measures GLM on accuracy scores and reaction times scores on the neutral, fearful, happy, and sad conditions as within-subject factors and gender and age group as between subject variables, but no signiﬁcant eﬀects were found. DISCUSSION Neutral Face Memory Task (FaMe-N) The task has a good internal consistency of ρKR20 = 0.808. In total 232 trials (8%) were outliers across 50 participants (M = 4.6, SD = 4.5, min = 1, max = 24). Neutral Face Memory Task (FaMe-N) TABLE 11 \| Correlation matrix between the reaction times on all tasks. FaMe-N FaMe-E Faces Shoes Face parts House parts FEM-H Upr Inv Upr Inv Upr Inv Upr Inv FaMe-E 0.60 – – – – – – – – – – Faces Upr 0.53 0.60 – – – – – – – – – Inv 0.50 0.57 0.86 – – – – – – – – Shoes Upr 0.53 0.51 0.84 0.89 – – – – – – – Inv 0.46 0.54 0.77 0.91 0.89 – – – – – – Face parts Upr 0.39 0.50 0.63 0.71 0.70 0.76 – – – – – Inv 0.45 0.44 0.52 0.63 0.61 0.66 0.78 – – – – House parts Upr 0.42 0.54 0.68 0.74 0.74 0.77 0.85 0.74 – – – Inv 0.41 0.46 0.57 0.68 0.68 0.71 0.83 0.80 0.89 – – FEM-H 0.40 0.54 0.43 0.47 0.53 0.53 0.35 0.36 0.53 0.48 – FEM-C 0.59 0.57 0.52 0.54 0.61 0.58 0.45 0.44 0.58 0.48 0.81 All correlations are signiﬁcant at the p < 0.01 level. TABLE 11 \| Correlation matrix between the reaction times on all tasks. All correlations are signiﬁcant at the p < 0.01 level. happiness were recognized above 80% accuracy. Similarly, canine emotions were recognized very well, although fear was also the worst recognized canine emotion and the older age group scored slightly worse and slower on this task, conﬁrming that this subtest provides a good control. Firstly, face and object processing and conﬁguration processing were assessed. As expected, upright face recognition is more accurate than inverted face recognition, in line with the face inversion eﬀect literature (Yin, 1969; Farah et al., 1995). Interestingly, even though the middle aged group was less accurate than the young adults group, their response patterns regarding face and object inversion were comparable. As conﬁgurational processing measured by (upright-inverted) inversion scores was not inﬂuenced by gender or age, this is a stable eﬀect in normal subjects. The absence of any interaction eﬀects with age group or gender indicate that category speciﬁc conﬁguration eﬀects are stable across gender and between young adulthood and middle age. This implies it is a suitable index to evaluate in prosopagnosia assessment. Secondly, the face and house part to whole matching task seems to be a harder task than the whole face and shoe matching task, as indicated by overall lower accuracies. Neutral Face Memory Task (FaMe-N) FaMe-N FaMe-E Faces Shoes Face parts House parts FEM-H Upr Inv Upr Inv Upr Inv Upr Inv FaMe-E 0.60 – – – – – – – – – – Faces Upr 0.53 0.60 – – – – – – – – – Inv 0.50 0.57 0.86 – – – – – – – – Shoes Upr 0.53 0.51 0.84 0.89 – – – – – – – Inv 0.46 0.54 0.77 0.91 0.89 – – – – – – Face parts Upr 0.39 0.50 0.63 0.71 0.70 0.76 – – – – – Inv 0.45 0.44 0.52 0.63 0.61 0.66 0.78 – – – – House parts Upr 0.42 0.54 0.68 0.74 0.74 0.77 0.85 0.74 – – – Inv 0.41 0.46 0.57 0.68 0.68 0.71 0.83 0.80 0.89 – – FEM-H 0.40 0.54 0.43 0.47 0.53 0.53 0.35 0.36 0.53 0.48 – FEM-C 0.59 0.57 0.52 0.54 0.61 0.58 0.45 0.44 0.58 0.48 0.81 All correlations are signiﬁcant at the p < 0.01 level. Firstly, face and object processing and conﬁguration processing were assessed. As expected, upright face recognition is more accurate than inverted face recognition in line with happiness were recognized above 80% accuracy. Similarly, canine emotions were recognized very well, although fear was also the worst recognized canine emotion and the older age group scored TABLE 10 \| Correlation matrix between the accuracy scores on all tasks. FaMe-N FaMe-E Faces Shoes Face parts House parts FEM-H Upr Inv Upr Inv Upr Inv Upr Inv FaMe-E 0.67 – Faces Upr 0.24 0.41 – – – – – – – – – Inv 0.15 0.36 0.51 – – – – – – – – Shoes Upr 0.20 0.23 0.60 0.61 – – – – – – – Inv 0.27 0.34 0.60 0.63 0.69 – – – – – – Face parts Upr 0.09 0.07 0.27 0.44 0.46 0.40 – – – – – Inv 0.15 0.25 0.46 0.47 0.48 0.56 0.50 – – – – House parts Upr 0.03 0.06 0.44 0.49 0.44 0.44 0.50 0.47 – – – Inv 0.06 0.17 0.48 0.60 0.52 0.50 0.65 0.63 0.64 – – FEM-H 0.18 0.44 0.39 0.37 0.28 0.30 0.23 0.16 0.18 0.35 – FEM-C 0.49 0.54 0.52 0.36 0.34 0.49 0.31 0.32 0.24 0.23 0.46 White; p < 0.01, light gray; p < 0.05, dark gray; ns. TABLE 10 \| Correlation matrix between the accuracy scores on all tasks. Neutral Face Memory Task (FaMe-N) In this study, we provide normative data of a large group of healthy controls on several face and object recognition tasks, face memory tasks and emotion recognition tasks. The eﬀects of gender and age were also reported. All tasks have a good internal consistency and an acceptable number of outliers. The participants scored on average 78% correct (SD = 12%) on the FaMe-N. No diﬀerences in accuracy scores on the FaMe-N The participants scored on average 78% correct (SD = 12%) on the FaMe-N. No diﬀerences in accuracy scores on the FaMe-N October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 11 The Facial Expressive Action Stimulus Test de Gelder et al. TABLE 10 \| Correlation matrix between the accuracy scores on all tasks. FaMe-N FaMe-E Faces Shoes Face parts House parts FEM-H Upr Inv Upr Inv Upr Inv Upr Inv FaMe-E 0.67 – Faces Upr 0.24 0.41 – – – – – – – – – Inv 0.15 0.36 0.51 – – – – – – – – Shoes Upr 0.20 0.23 0.60 0.61 – – – – – – – Inv 0.27 0.34 0.60 0.63 0.69 – – – – – – Face parts Upr 0.09 0.07 0.27 0.44 0.46 0.40 – – – – – Inv 0.15 0.25 0.46 0.47 0.48 0.56 0.50 – – – – House parts Upr 0.03 0.06 0.44 0.49 0.44 0.44 0.50 0.47 – – – Inv 0.06 0.17 0.48 0.60 0.52 0.50 0.65 0.63 0.64 – – FEM-H 0.18 0.44 0.39 0.37 0.28 0.30 0.23 0.16 0.18 0.35 – FEM-C 0.49 0.54 0.52 0.36 0.34 0.49 0.31 0.32 0.24 0.23 0.46 White; p < 0.01, light gray; p < 0.05, dark gray; ns. TABLE 11 \| Correlation matrix between the reaction times on all tasks. Frontiers in Psychology \| www.frontiersin.org Neutral Face Memory Task (FaMe-N) Young adults are more sensitive to inversion in this task. Lastly, no eﬀects of gender or age were found on neutral face memory, and participants scored quite well on the task, with an average of almost 80% correct. Similarly, no clear eﬀects of age, gender or emotion were found on face memory as measured with the FaMe-E, except that it seems that middle aged women are slower to recognize previously seen identities when they expressed sadness. Interestingly, this is in line with the “age- related positivity eﬀect” (Samanez-Larkin and Carstensen, 2011; Reed and Carstensen, 2012). In general, the results corroborate those from other studies on the eﬀect of emotion on memory (Johansson et al., 2004), but a wide variety of results has been reported in the literature (Dobel et al., 2008; Langeslag et al., 2009; Bate et al., 2010; D’Argembeau and Van der Linden, 2011; Righi et al., 2012; Liu et al., 2014). In addition, we did not ﬁnd any relationships between conﬁguration perception and face Thirdly, we found that fear and sadness recognition on our FEM-H task was quite poor, but that anger, disgust, surprise and Frontiers in Psychology \| www.frontiersin.org October 2015 \| Volume 6 \| Article 1609 12 The Facial Expressive Action Stimulus Test de Gelder et al. memory. This can be due to the fact that unlike in samples with DPs and controls, there is less variability in inversion scores and memory scores (i.e., most participants will not have any conﬁguration processing deﬁcits similar to DPs and in contrast to DPs, most controls are not severely limited on face memory). show the face in the test phase from a diﬀerent angle that in the training phase as is done in the matching tests. In addition, the low performance on fear recognition should be assessed. In short, the FEAST provides researchers with an extensive battery for neutral and emotional face memory, whole and part-to-whole face and object matching, conﬁgural processing and emotion recognition abilities. The results indicate that age is most likely a modulating factor when studying face and object processing, as the responses of the middle aged group is often slower. One explanation besides a general cognitive decline with age can be found in the literature on the eﬀect of age on facial recognition, where an “own-age bias” is often found (Lamont et al., 2005; Firestone et al., 2007; He et al., 2011; Wiese, 2012). REFERENCES de Gelder, B., and Bertelson, P. (2009). A comparative approach to testing face perception: Face and object identiﬁcation by adults in a simultaneous matching task. Psychol. Belgica 42, 177–190. doi: 10.5334/pb-49-2-3-177 Bartlett, J. C., and Searcy, J. (1993). Inversion and conﬁguration of faces. Cogn. Psychol. 25, 281–316. doi: 10.1006/cogp.1993.1007 de Gelder, B., Frissen, I., Barton, J., and Hadjikhani, N. (2003). A modulatory role for facial expressions in prosopagnosia. Proc. Natl. Acad. Sci. U.S.A. 100, 13105–131010. doi: 10.1073/pnas.1735530100 Bate, S., Haslam, C., Hodgson, T. L., Jansari, A., Gregory, N., and Kay, J. (2010). Positive and negative emotion enhances the processing of famous faces in a semantic judgment task. Neuropsychology 24, 84–89. doi: 10.1037/ A0017202 de Gelder, B., Pourtois, G., Vroomen, J., and Bachoud-Lévi, A. C. (2000). Covert processing of faces in prosopagnosia is restricted to facial expressions: evidence from cross-modal bias. Brain Cogn. 44, 425–444. doi: 10.1006/brcg.1999.1203 Benton, A. L., Sivan, A. B., Hamsher, K., Varney, N. R., and Spreen, O. (1983). Contribution to Neuropsychological Assessment. New York, NY: Oxford University Press. de Gelder, B., and Rouw, R. (2000a). Conﬁgural face processes in acquired and developmental prosopagnosia: evidence for two separate face systems? Neuroreport 11, 3145–3150. doi: 10.1097/00001756-200009280-00021 Bodamer, J. (1947). Die prosop-agnosie. Archiv fur Psychiatrie und Nervenkrankheiten 179, 6–53. de Gelder, B., and Rouw, R. (2000b). Paradoxical conﬁguration eﬀects for faces and objects in prosopagnosia. Neuropsychologia 38, 1271–1279. doi: 10.1016/S0028- 3932(00)00039-7 Calder, A. J., and Jansen, J. (2005). Conﬁgural coding of facial expressions: the impact of inversion and photographic negative. Visual Cogn. 12, 495–518. doi: 10.1080/13506280444000418 de Gelder, B., and Rouw, R. (2000c). Structural encoding precludes recognition of parts in prosopagnosia. Cogn. Neuropsychol. 17, 89–102. doi: 10.1080/026432900380508 Calder, A. J., Young, A. W., Keane, J., and Dean, M. (2000). Conﬁgural information in facial expression perception. J. Exp. Psychol. Hum. Percept. Perform. 26, 527–551. doi: 10.1037/0096-1523.26.2.527 de Gelder, B., and Rouw, R. (2001). Beyond localisation: a dynamical dual route account of face recognition. Acta Psychol. (Amst.) 107, 183–207. doi: 10.1016/S0001-6918(01)00024-5 Calvo, M. G., and Beltrán, D. (2014). Brain lateralization of holistic versus analytic processing of emotional facial expressions. Neuroimage 92, 237–247. doi: 10.1016/j.neuroimage.2014.01.048 de Gelder, B., and Stekelenburg, J. J. (2005). Naso-temporal asymmetry of the n170 for processing faces in normal viewers but not in developmental prosopagnosia. Neurosci. Lett. 376, 40–45. doi: 10.1016/j.neulet.2004.11.047 Chen, W., Lander, K., and Liu, C. H. (2011). Matching faces with emotional expressions. Front. Psychol. 2:206. Neutral Face Memory Task (FaMe-N) The “own-age bias” in face recognition refers to the notion that individuals are more accurate at recognizing faces from individuals belonging to the age category of the observer. For instance, children are better at recognizing child faces and adults are better at recognizing adult faces. Future researchers wishing to use the FEAST should compare the results of their participants with the appropriate age group, or should control for the eﬀects of age or ideally, test age-matched controls. Gender on the other hand does not seem so inﬂuential, but this article provides guidelines and data for both gender and age groups regardless. FUNDING National Initiative Brain & Cognition; Contract grant number: 056-22-011. EU project TANGO; Contract grant number: FP7-ICT-2007-0 FETOpen. European Research Council under the European Union’s Seventh Framework Programme (ERC); Contract grant number: FP7/2007–2013, agreement number 295673. JV is a post-doctoral research fellow for FWO- Vlaanderen. Some limitations of the FEAST should be noted. One is the lack of a non-face memory control condition using stimuli with comparable complexity. However, a recent study with a group of 16 DPs showed that only memory for faces, in contrast to hands, butterﬂies and chairs was impaired (Shah et al., 2014), so for this group this control condition might not be necessary. Also, the speciﬁc eﬀects of all emotions, valence and arousal may be taken into account in future research. The face memory test could be complemented with the use of test images that SUPPLEMENTARY MATERIAL The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fpsyg. 2015.01609 AUTHOR CONTRIBUTIONS All authors contributed signiﬁcantly to the concept and design of the work. EH collected and analyzed the data. All authors contributed to data interpretation. EH drafted the ﬁrst version of the manuscript and BD and JV revised. REFERENCES doi: 10.3389/fpsyg.2011.00206 de Gelder, B., and Van den Stock, J. (2015). “Prosopagnosia,” in International Encyclopedia of the Social and Behavioral Sciences, 2nd Edn., Vol. 19, ed J. D. Wright (Oxford: Elsevier), 250–255. D’Argembeau, A., and Van der Linden, M. (2011). Inﬂuence of facial expression on memory for facial identity: eﬀects of visual features or emotional meaning? Emotion 11, 199–208. doi: 10.1037/A0022592 de Gelder, B., Bachoud-Lévi, A. C., and Degos, J. D. (1998). Inversion superiority in visual agnosia may be common to a variety of orientation polarised objects besides faces. Vision Res. 38, 2855–2861. DeGutis, J., Wilmer, J., Mercado, R. J., and Cohan, S. (2013). Using regression to measure holistic face processing reveals a strong link with face recognition ability. Cognition 126, 87–100. doi: 10.1016/j.cognition.2012.09.004 Frontiers in Psychology \| www.frontiersin.org October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 13 The Facial Expressive Action Stimulus Test de Gelder et al. Dingle, K. J., Duchaine, B. C., and Nakayama, K. (2005). A new test for face perception [abstract]. J. Vision 5:40a. doi: 10.1167/5.8.40 Neuroscience, ed J. T. C. J. Decety (New York, NY: Oxford University Press), 507–521. Shah, P., Gaule, A., Gaigg, S. B., Bird, G., and Cook, R. (2014). Probing short- term face memory in developmental prosopagnosia. Cortex 64C, 115–122. doi: 10.1016/j.cortex.2014.10.006 Dobel, C., Geiger, L., Bruchmann, M., Putsche, C., Schweinberger, S. R., and Junghöfer, M. (2008). On the interplay between familiarity and emotional expression in face perception. Psychol. Res. 72, 580–586. doi: 10.1007/s00426- 007-0132-4 Snowden, J. S., Goulding, P. J., and Neary, D. (1989). Semantic dementia: a form of circumscribed cerebral atrophy. Behav. Neurol. 2, 167–182. Duchaine, B., and Nakayama, K. (2006). The cambridge face memory test: results for neurologically intact individuals and an investigation of its validity using inverted face stimuli and prosopagnosic participants. Neuropsychologia 44, 576–585. doi: 10.1016/j.neuropsychologia.2005.07.001 Tanaka, J. W., and Farah, M. J. (1993). Parts and wholes in face recognition. Q. J. Exp. Psychol. A 46, 225–245. Tanaka, J. W., Kaiser, M. D., Butler, S., and Le Grand, R. (2012). Mixed emotions: holistic and analytic perception of facial expressions. Cogn. Emot. 26, 961–977. doi: 10.1080/02699931.2011.630933 Durand, K., Gallay, M., Seigneuric, A., Robichon, F., and Baudouin, J. Y. (2007). The development of facial emotion recognition: the role of conﬁgural information. J. Exp. Child Psychol. 97, 14–27. doi: 10.1016/j.jecp.2006. 12.001 Tottenham, N., Tanaka, J. W., Leon, A. C., McCarry, T., Nurse, M., Hare, T. REFERENCES A., et al. (2009). The nimstim set of facial expressions: judgments from untrained research participants. Psychiatry Res. 168, 242–249. doi: 10.1016/j.psychres.2008.05.006 Farah, M. J., Wilson, K. D., Drain, H. M., and Tanaka, J. R. (1995). The inverted face inversion eﬀect in prosopagnosia: evidence for mandatory, face-speciﬁc perceptual mechanisms. Vision Res. 35, 2089–2093. Van Belle, G., Busigny, T., Lefèvre, P., Joubert, S., Felician, O., Gentile, F., et al. (2011). Impairment of holistic face perception following right occipito-temporal damage in prosopagnosia: converging evidence from gaze-contingency. Neuropsychologia 49, 3145–3150. doi: 10.1016/j. neuropsychologia.2011.07.010 Fast, K., Fujiwara, E., and Markowitsch, H. J. (2008). Der Famous Faces Test. Göttingen: Hogrefe. Firestone, A., Turk-Browne, N. B., and Ryan, J. D. (2007). Age-related deﬁcits in face recognition are related to underlying changes in scanning behavior. Aging Neuropsychol. Cogn. 14, 594–607. doi: 10.1080/13825580600899717 Van den Stock, J., and de Gelder, B. (2012). Emotional information in body and background hampers recognition memory for faces. Neurobiol. Learn. Mem. 97, 321–325. doi: 10.1016/j.nlm.2012.01.007 Hadjikhani, N., and de Gelder, B. (2002). Neural basis of prosopagnosia: an fmri study. Hum. Brain Mapp. 16, 176–182. doi: 10.1002/hbm.10043 Van den Stock, J., and de Gelder, B. (2014). Face identity matching is inﬂuenced by emotions conveyed by face and body. Front. Hum. Neurosci. 8:53. doi: 10.3389/fnhum.2014.00053 He, Y., Ebner, N. C., and Johnson, M. K. (2011). What predicts the own-age bias in face recognition memory? Soc. Cogn. 29, 97–109. doi: 10.1521/soco.2011.29.1.97 Van den Stock, J., de Gelder, B., De Winter, F. L., Van Laere, K., and Vandenbulcke, M. (2012a). A strange face in the mirror. Face-selective self-misidentiﬁcation in a patient with right lateralized occipito-temporal hypo-metabolism. Cortex 48, 1088–1090. doi: 10.1016/j.cortex.2012.03.003 Huis in ‘t Veld, E., Van den Stock, J., and de Gelder, B. (2012). Conﬁguration perception and face memory, and face context eﬀects in developmental prosopagnosia. Cogn. Neuropsychol. 29, 464–481. doi: 10.1080/02643294.2012.732051 Johansson, M., Mecklinger, A., and Treese, A. C. (2004). Recognition memory for emotional and neutral faces: an event-related potential study. J. Cogn. Neurosci. 16, 1840–1853. doi: 10.1162/0898929042947883 Van den Stock, J., de Gelder, B., Van Laere, K., and Vandenbulcke, M. (2013). Face- selective hyper-animacy and hyper-familiarity misperception in a patient with moderate alzheimer’s disease. J. Neuropsychiatry Clin. Neurosci. 25, E52–E53. doi: 10.1176/appi.neuropsych.12120390 Kennerknecht, I., Grueter, T., Welling, B., Wentzek, S., Horst, J., Edwards, S., et al. (2006). First report of prevalence of non-syndromic hereditary prosopagnosia (hpa). Am. J. Med. Genet. A 140, 1617–1622. REFERENCES doi: 10.1002/ajmg.a.31343 Van den Stock, J., De Winter, F. L., de Gelder, B., Rangarajan, J. R., Cypers, G., Maes, F., et al. (2015). Impaired recognition of body expressions in the behavioral variant of fronto-temporal dementia. Neuropsychologia 75, 496–504. doi: 10.1016/j.neuropsychologia.2015.06.035 Kuder, G. F., and Richardson, M. W. (1937). The theory of the estimation of test reliability. Psychometrika 2, 151–160. Lamont, A. C., Stewart-Williams, S., and Podd, J. (2005). Face recognition and aging: eﬀects of target age and memory load. Mem. Cogn. 33, 1017–1024. doi: 10.3758/Bf03193209 Van den Stock, J., Vandenbulcke, M., Zhu, Q., Hadjikhani, N., and de Gelder, B. (2012b). Developmental prosopagnosia in a patient with hypoplasia of the vermis cerebelli. Neurology 78, 1700–1702. doi: 10.1212/WNL.0b013e3182575130 Langeslag, S. J. E., Morgan, H. M., Jackson, M. C., Linden, D. E. J., and Van Strien, J. W. (2009). Electrophysiological correlates of improved short-term memory for emotional faces (vol 47, pg 887, 2009). Neuropsychologia 47, 2013–2013. doi: 10.1016/j.neuropsychologia.2009.02.031 Van den Stock, J., van de Riet, W. A., Righart, R., and de Gelder, B. (2008). Neural correlates of perceiving emotional faces and bodies in developmental prosopagnosia: an event-related fmri-study. PLoS ONE 3:e3195. doi: 10.1371/journal.pone.0003195 Liu, C. H., Chen, W., and Ward, J. (2014). Remembering faces with emotional expressions. Front. Psychol. 5:1439. doi: 10.3389/Fpsyg.2014.01439 Wang, R. S., Li, J. G., Fang, H. Z., Tian, M. Q., and Liu, J. (2012). Individual diﬀerences in holistic processing predict face recognition ability. Psychol. Sci. 23, 169–177. doi: 10.1177/0956797611420575 Lundqvist, D., Flykt, A., and Öhman, A. (1998). The Karolinska Directed Emotional Faces - Kdef. Stockholm: Karolinska Institutet. Warrington, E. K. (1984). Recognition Memory Test. Windsor, ON: Nelson;NFER. Mckelvie, S. J. (1995). Emotional expression in upside-down faces - evidence for conﬁgurational and componential processing. Br. J. Soc. Psychol. 34, 325–334. White, M. (2000). Parts and wholes in expression recognition. Cogn. Emot. 14, 39–60. doi: 10.1080/026999300378987 Palermo, R., Willis, M. L., Rivolta, D., McKone, E., Wilson, C. E., and Calder, A. J. (2011). Impaired holistic coding of facial expression and facial identity in congenital prosopagnosia. Neuropsychologia 49, 1226–1235. doi: 10.1016/j.neuropsychologia.2011.02.021 Wiese, H. (2012). The role of age and ethnic group in face recognition memory: ERP evidence from a combined own-age and own-race bias study. Biol. Psychol. 89, 137–147. doi: 10.1016/j.biopsycho.2011.10.002 ERP evidence from a combined own-age and own-race 89, 137–147. doi: 10.1016/j.biopsycho.2011.10.002 Reed, A. E., and Carstensen, L. L. (2012). The theory behind the age-related positivity eﬀect. Front. Psychol. 3:339. REFERENCES doi: 10.3389/Fpsyg.2012.00339 Yin, R. K. (1969). Looking at upside down faces. J. Exp. Psychol. 81, 141 145. Young, A. W., Hellawell, D., and Hay, D. C. (1987). Conﬁgurational information in face perception. Perception 16, 747–759. Richler, J. J., Cheung, O. S., and Gauthier, I. (2011). Holistic processing predicts face recognition. Psychol. Sci. 22, 464–471. doi: 10.1177/0956797611401753 Riddoch, M. J., and Humphreys, G. W. (1992). Birmingham Object Recognition Battery. Hove: Psychology Press. Conﬂict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Righart, R., and de Gelder, B. (2007). Impaired face and body perception in developmental prosopagnosia. Proc. Natl. Acad. Sci. U.S.A. 104, 17234–17238. doi: 10.1073/pnas.0707753104 Copyright © 2015 de Gelder, Huis in ‘t Veld and Van den Stock. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Righi, S., Marzi, T., Toscani, M., Baldassi, S., Ottonello, S., and Viggiano, M. P. (2012). Fearful expressions enhance recognition memory: electrophysiological evidence. Acta Psychol. 139, 7–18. doi: 10.1016/j.actpsy.2011.09.015 Samanez-Larkin, G. R., and Carstensen, L. L. (2011). “Socioemotional functioning and the aging brain, “ in The Oxford Handbook of Social October 2015 \| Volume 6 \| Article 1609 Frontiers in Psychology \| www.frontiersin.org 14
https://openalex.org/W1888924204	https://europepmc.org/articles/pmc4545889?pdf=render	English	null	Leukocyte Populations in Human Preterm and Term Breast Milk Identified by Multicolour Flow Cytometry	PloS one	2,015	cc-by	10,384	RESEARCH ARTICLE OPEN ACCESS Citation: Trend S, de Jong E, Lloyd ML, Kok CH, Richmond P, Doherty DA, et al. (2015) Leukocyte Populations in Human Preterm and Term Breast Milk Identified by Multicolour Flow Cytometry. PLoS ONE 10(8): e0135580. doi:10.1371/journal.pone.0135580 Methods Sixty mothers of extremely preterm (<28 weeks gestational age), very preterm (28–31 wk), and moderately preterm (32–36 wk), as well as term (37–41 wk) infants were recruited. Colostrum (d2–5), transitional (d8–12) and mature milk (d26–30) samples were collected, cells isolated, and leukocyte subsets analysed using flow cytometry. Data Availability Statement: Data are from the COMET study, whose authors may be contacted at stephanie.trend@uwa.edu.au. Data are restricted for ethical reasons, but de-identified data will be made available to all interested researchers upon request. Requests for information should be submitted to the Women and Newborn Health Service Ethics Committee ( kemhethics@health.wa.gov.au). Background Editor: Markus Sperandio, Ludwig-Maximilians- Universität, GERMANY Editor: Markus Sperandio, Ludwig-Maximilians- Universität, GERMANY Received: April 22, 2015 Accepted: July 24, 2015 Published: August 19, 2015 Copyright: © 2015 Trend et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Editor: Markus Sperandio, Ludwig-Maximilians- Universität, GERMANY Received: April 22, 2015 Accepted: July 24, 2015 Published: August 19, 2015 Copyright: © 2015 Trend et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Editor: Markus Sperandio, Ludwig-Maximilians- Universität, GERMANY Received: April 22, 2015 Accepted: July 24, 2015 Published: August 19, 2015 Copyright: © 2015 Trend et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Extremely preterm infants are highly susceptible to bacterial infections but breast milk pro- vides some protection. It is unknown if leukocyte numbers and subsets in milk differ between term and preterm breast milk. This study serially characterised leukocyte popula- tions in breast milk of mothers of preterm and term infants using multicolour flow cytometry methods for extended differential leukocyte counts in blood. Leukocyte Populations in Human Preterm and Term Breast Milk Identified by Multicolour Flow Cytometry Stephanie Trend1,2, Emma de Jong1,2,3, Megan L. Lloyd1,4, Chooi Heen Kok1,5, Peter Richmond2, Dorota A. Doherty6, Karen Simmer1,5, Foteini Kakulas7, Tobias Strunk1,5, Andrew Currie1,2,3* Stephanie Trend1,2, Emma de Jong1,2,3, Megan L. Lloyd1,4, Chooi Heen Kok1,5, Peter Richmond2, Dorota A. Doherty6, Karen Simmer1,5, Foteini Kakulas7, Tobias Strunk1,5, Andrew Currie1,2,3* 1 Centre for Neonatal Research and Education, The University of Western Australia, Perth, Western Australia, Australia, 2 School of Paediatrics and Child Health, The University of Western Australia, Perth, Western Australia, Australia, 3 School of Veterinary and Life Sciences, Murdoch University, Perth, Western Australia, Australia, 4 School of Pathology and Laboratory Medicine, The University of Western Australia, Perth, Western Australia, Australia, 5 Neonatal Clinical Care Unit, King Edward Memorial Hospital for Women, Perth, Western Australia, Australia, 6 School of Women’s and Infants’ Health, The University of Western Australia, Perth, Western Australia, Australia, 7 School of Chemistry and Biochemistry, The University of Western Australia, Perth, Australia * a.currie@murdoch.edu.au * a.currie@murdoch.edu.au Introduction Feeding of human breast milk (BM) is associated with fewer infections and reduced gastroin- testinal inflammation in preterm infants [1, 2]. Though all newborns have weak cellular and humoral defences, preterm infants are particularly susceptible to bacterial infections, with known deficiencies in adaptive and innate immunity including decreased leukocyte pool, lack of third trimester maternal immunoglobulin transfer, and decreased leukocyte cytokine and antibody production in response to bacteria [3–7]. BM contains a myriad of immunological, biochemical and cellular contents, which have the potential to significantly alter newborn immunity and susceptibility to infection. BM-derived leukocytes engulf and kill bacteria [8] and produce antimicrobial proteins and peptides (AMPs) [9]. In animal models, BM leuko- cytes can translocate from the gastrointestinal tract to blood and distant sites including the liver and spleen [10, 11]. Microscopy-based identification of BM leukocytes comparing very preterm (<32 wk gestational age; GA), moderately preterm (32–34 wk GA), and term colos- trum suggests an inverse correlation between GA and the concentration of leukocytes [12]. However, it is not known how preterm birth affects the leukocyte populations in transitional or mature milk, or in milk after extremely preterm birth (<28 wk GA). Several flow cytometry-based methods for differentiation of blood and bone marrow leuko- cytes have been recently described, that allow for the reliable, automated and extended differ- entiation of known leukocyte subsets [13–17]. Importantly, these methods have been validated against the gold standard of diagnostic cytology and allow for a robust measurement of an extended differential leukocyte count (up to 11 leukocytes subsets) in both healthy and diseased individuals [18]. These methods may allow for identification of a greater range of leukocytes in BM, especially in the presence of mammary gland epithelial cells and stem cells and achieve greater certainty and superior differentiation. Given the potential for maternal leukocytes to contribute to infant immunity, an extended characterisation of preterm BM leukocytes may contribute to our understanding of the susceptibility of preterm infants to infections. The aims of this study were to assess the effectiveness of the use of an extended validated blood leukocyte differential staining panel, described by Faucher et al. [13], to identify the same leukocyte populations in human milk; to use this method to compare the total milk leu- kocyte concentration and subset frequencies in colostrum, transitional and mature milk, and; to investigate the effect of preterm birth and infection on these cell populations in milk. Leukocytes in Preterm Milk Conclusions from WIRF for this work. FH was the recipient of an unrestricted research grant by Medela AG (Switzerland). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Flow cytometry is a useful tool to identify and quantify leukocyte subsets in breast milk. The stage of lactation is associated with major changes in milk leukocyte composition in this population. Fresh preterm breast milk is not deficient in leukocytes, but shorter gestation may be associated with minor differences in leukocyte subset frequencies in preterm com- pared to term breast milk. Competing Interests: The authors have declared that no competing interests exist. Results The major CD45+ leukocyte populations circulating in blood were also detectable in breast milk but at different frequencies. Progression of lactation was associated with decreasing CD45+ leukocyte concentration, as well as increases in the relative frequencies of neutro- phils and immature granulocytes, and decreases in the relative frequencies of eosinophils, myeloid and B cell precursors, and CD16- monocytes. No differences were observed between preterm and term breast milk in leukocyte concentration, though minor differences between preterm groups in some leukocyte frequencies were observed. Funding: ST received postgraduate scholarships from the Centre for Neonatal Research and Education, Women and Infants Research Foundation (WIRF), and the Princess Margaret Hospital (PMH) foundation. TS and ST received an early career grant PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 1 / 17 NE \| DOI:10.1371/journal.pone.0135580 August 19, 201 Materials and Methods Sample collection The Women and Newborn Health Service Human Research Ethics Committee approved this study. Written informed consent was obtained from sixty women giving birth between 22–42 wk gestation before commencing any study procedures or sample collection. Inclusion criteria 2 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Leukocytes in Preterm Milk included maternal age 18 years, intention to breastfeed and absence of primary immune defi- ciencies, diabetes mellitus or known congenital abnormalities. Sixty participants were recruited across four groups based on WHO preterm birth categories [19]; extremely preterm (EP, <28 wk GA; n = 15), very preterm (VP, 28–31 wk GA; n = 15), moderately preterm (MP, 32–36 wk GA; n = 15) or term (37–41 wk GA; n = 15). Individuals with a history of other common condi- tions, particularly those associated with preterm birth such as infections and hypertension were not excluded, nor individuals taking medications, in order to gain a representative sample of the preterm population. BM was collected from participants at three time points, selected to represent colostrum (C; day 2–5, 3.5 mL), transitional milk (TM; d8–12; 7 mL) and mature milk (MM; d26–30; 7 mL), based on the categories described by Castellote et al. [20]. Participants were instructed on hygienic collection of milk by qualified nursing staff. This included washing hands with soap and water before milk collection and using sterile containers and clean equipment for expres- sion. Participants were asked to clean the nipple with soap and water before expressing milk. BM was expressed using electronic pumps, hand-operated pumps or manual expression (according to the donor’s usual method), and an aliquot removed with a sterile Pasteur pipette to a 15 mL plastic tube. Whole BM was stored at 4°C until collected by research staff, and trans- ported to the research laboratory on ice. Between expression of BM and processing of the milk, samples were stored for a median of 8.29 hrs (range 1–50.9 hrs) in the refrigerator or in trans- port (on ice). The volume of milk donated to the study was recorded at each visit before pro- cessing. Samples were mixed gently and the cells were pelleted by centrifugation at 600 x g for 15 minutes at 4°C. Materials and Methods Sample collection The lipid layer and skim milk were removed with a Pasteur pipette, and the remaining cell pellet was washed twice in phosphate buffered saline (Gibco), and resuspended in 0.1–2.0 mL of flow cytometry buffer, described in the cell staining method below, to approxi- mately one million cells/mL, for optimal cell to antibody ratios during staining and quantifica- tion of cells. The volumes of flow cytometry buffer that were used to resuspend cells and the original volume of milk processed were recorded for each sample so that the original cell con- centration of the milk samples could be determined. Peripheral venous blood was collected from a single healthy adult donor into a sodium heparin Vacutainer (BD), and was mixed and stored at room temperature until staining. Determination of cell viability using visual counting methods All cells enumerated by microscopy were analysed by a single operator using a Neubauer improved haemocytometer (BOECO, Germany). For each sample, two different aliquots of BM cell suspensions were stained by dilution of cell suspensions to approximately 1 x 105–2.5 x 106 cells/mL in a 0.4% Trypan blue solution (Sigma-Aldrich, Castle Hill, Australia) in PBS, or white cell counting fluid (2% glacial acetic acid containing crystal violet). Ten microliters of each suspension was pipetted into the counting chamber under the cover slip, and the average of three counts taken to calculate the concentration of cells, based on the volume in the cham- ber. Trypan blue stained cell counts were used to determine the non-viable cell concentration. A total cell concentration estimate was calculated by counting nucleated cells in white cell counting fluid. The proportion of non-viable cells was calculated based on the numbers of Try- pan blue stained (non-viable) compared to total cells. Cell staining Antibodies raised against human leukocyte antigens, stabilising fixative, FACSLyse, compensa- tion beads and Trucount tubes were purchased from BD (North Ryde, Australia). Flow cytom- etry buffer was prepared in phosphate buffered saline (Gibco) supplemented with heat- 3 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Leukocytes in Preterm Milk inactivated fetal bovine serum (FBS; Sigma, Australia) at 2% v/v, 2% w/v bovine serum albumin (Sigma-Aldrich), and 0.01% w/v sodium azide (Sigma-Aldrich). Fifty microliters of either BM cells suspended in flow cytometry buffer at approximately 1 x 106 cells/mL or whole blood was incubated for 15 minutes at room temperature with a cocktail of six specific anti-human monoclonal antibodies (mAb): Phycoerythrin (PE) conjugated CD36 (clone CB38, catalogue #555455; 10 μL per sample), Allophycocyanin (APC) conjugated CD2 (RPA-2.10, 560642; 0.5 μL per sample), Alexa Fluor 647 conjugated CD294 (BM16, 561797; 0.5 μL per sample), APC-H7 conjugated CD16 (3G8, 560195; 2.5 μL per sample), V450 conjugated CD19 (HIB19, 560353; 0.25 μL per sample), and V500 conjugated CD45 (HI30, 560779; 1.25 μL per sample). The final dilution of antibody used was determined through titration experiments. Stained blood was treated with FACSLyse according to manu- facturer’s specifications prior to further analysis to remove erythrocytes. After staining, BM and blood cells were washed once with 1 mL of flow cytometry buffer by centrifugation at 290 x g at 10°C for 3 min, then resuspended in 300 μL of BD stabilizing fixative and transferred to a Trucount tube. Fixed samples were stored at 4°C protected from light until analysis by flow cytometry, which for most samples was performed within 2 hrs, though the maximum allowed storage time for fixed cells before cytometry was 24 hrs based on previous experiments in our laboratory, which demonstrated the preservation of the staining characteristics under these conditions (data not shown). PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Flow cytometry Flow cytometry was performed on a FACSCanto II (BD Biosciences) using FACSDiva software (BD Biosciences). Before sample analysis, the flow cytometer settings were checked using Cytometer Setup and Tracking beads (CS&T beads, BD) according to the manufacturer’s instructions. Compensation beads were used with single stains of each antibody in order to determine the compensation settings, and these were applied in FlowJo software (version 10.0.6, Tree Star, Ashland, OR, USA) after data collection. The same compensation matrix was applied to all samples. A side scatter (SSC) threshold level was set at 4,000 units to eliminate debris. Gain settings were optimised for detection of stained populations using Ultra Rainbow Calibration Particles (SpheroTech Inc., Lake Forest IL, USA), and kept consistent throughout the study. Recorded data were compensated post-hoc and analysed using FlowJo software. Statistical tests were performed in SPSS (IBM) and Prism for Mac (version 5, GraphPad, La Jolla CA, USA). Specific antibody staining data were visualised using bi-exponential transformation in FlowJo software. Gates used to discriminate positive and negative staining cells in FlowJo were set according to fluorescence minus one (FMO) tests of milk and blood samples, and these gates were applied consistently to all samples, allowing for minor adjustments for SSC variability. y Gating strategy. Positive gated populations were selected using gates set through FMO tests, in which <1.7% of cells stained falsely positive for any antibody. Gate specificity selected using this method was confirmed by using the gating strategy on a single adult peripheral blood sample as a biological comparison control [21]. Prior to gating of leukocytes, Trucount beads were gated separately from cells in a SSC vs. PerCPCy5.5 visualisation; doublets were dis- criminated using a forward scatter (FSC)-H vs. FSC-A plot, and small particles excluded in a FSC vs. SSC plot (S1 Fig). Leukocytes in human milk were identified based on the method of Faucher et al., using orientating and specific gates [13] as described below. The first orientating gate was selected using a SSC vs. CD45 plot where the cut-off for CD45 + cells was set using FMO control. Subsequently, a SSC vs. CD16 plot of CD45+ cells separated PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 4 / 17 Leukocytes in Preterm Milk Fig 1. Overlay of gating strategy applied to peripheral adult blood (red) and to breast milk (blue), based on the method of Faucher et al. Flow cytometry The sequences of gates leading to each panel are shown above each panel. Panel 1: CD45 positive cells were gated as shown in A. Panel 2: CD45+ cells identified in Panel 1 were separated based on CD16 staining and side scatter properties, including a C16+/SSClow gate (B), and two overlapping gates of CD16- cells (C and D), and the CD16+ neutrophil population were identified. Panel 3: CD45+/CD16+/SSClow cells identified in Panel 2 gate B were separated into cytotoxic T and NK lymphocytes and CD16+ monocytes based on CD2/CD294 and CD36 staining properties. Panel 4: CD45+/CD16-/ SSClow–intermediate cells (gate C) were separated based on CD2/CD294 positive (gate E) or negative (gate F) populations. Panel 5: From Panel 4, CD2 and/or CD294 positive cells (gate E) were gated into non-cytotoxic T cells or basophils using side scatter and CD45 staining properties. Panel 6: From Panel 4, CD2/CD294- cells (gate F) were gated into CD19+/CD36- cells (gate H) or CD19- cells (gate G). Panel 7a: Cells gated in G in Panel 6 with CD45high and CD36+ were identified as CD16 negative monocytes. Panel 7b: From Panel 6 gate G, CD45low cells with low side scatter were identified as myeloid precursor cells. Panel 8: From Panel 6, CD19 positive cells (gate H) were discriminated into B cells or B cell precursors based on CD45 staining though both populations displayed low side scatter properties. Panel 9: From Panel 2, CD16- cells with intermediate to high side scatter (gate D) and CD45low staining were identified as immature granulocytes, and those with intermediate to high CD45 staining properties and high side scatter were separated into gate I. Panel 10: From Panel 9, cells in gate I were identified by positive CD2/CD294 staining as eosinophils. doi:10 1371/journal pone 0135580 g001 Fig 1. Overlay of gating strategy applied to peripheral adult blood (red) and to breast milk (blue), based on the method of Faucher et al. The sequences of gates leading to each panel are shown above each panel. Panel 1: CD45 positive cells were gated as shown in A. Panel 2: CD45+ cells identified in Panel 1 were separated based on CD16 staining and side scatter properties, including a C16+/SSClow gate (B), and two overlapping gates of CD16- cells (C and D), and the CD16+ neutrophil population were identified. Flow cytometry Panel 3: CD45+/CD16+/SSClow cells identified in Panel 2 gate B were separated into cytotoxic T and NK lymphocytes and CD16+ monocytes based on CD2/CD294 and CD36 staining properties. Panel 4: CD45+/CD16-/ SSClow–intermediate cells (gate C) were separated based on CD2/CD294 positive (gate E) or negative (gate F) populations. Panel 5: From Panel 4, CD2 and/or CD294 positive cells (gate E) were gated into non-cytotoxic T cells or basophils using side scatter and CD45 staining properties. Panel 6: From Panel 4, CD2/CD294- cells (gate F) were gated into CD19+/CD36- cells (gate H) or CD19- cells (gate G). Panel 7a: Cells gated in G in Panel 6 with CD45high and CD36+ were identified as CD16 negative monocytes. Panel 7b: From Panel 6 gate G, CD45low cells with low side scatter were identified as myeloid precursor cells. Panel 8: From Panel 6, CD19 positive cells (gate H) were discriminated into B cells or B cell precursors based on CD45 staining though both populations displayed low side scatter properties. Panel 9: From Panel 2, CD16- cells with intermediate to high side scatter (gate D) and CD45low staining were identified as immature granulocytes, and those with intermediate to high CD45 staining properties and high side scatter were separated into gate I. Panel 10: From Panel 9, cells in gate I were identified by positive CD2/CD294 staining as eosinophils. doi:10.1371/journal.pone.0135580.g001 events into a specific SSChigh/CD16+ neutrophil gate, and three orientating gates (SSClow/ CD16+, SSClow–int/CD16-, or SSChigh/CD16-). Subsequent orientating and specific gates for other populations are shown in Fig 1. CD16+ and CD16- monocytes, B lymphocytes, cytotoxic T and NK lymphocytes, non-cytotoxic T lymphocytes, neutrophils, eosinophils, basophils, immature granulocytes, myeloid precursors and B cell precursors were identified in colostrum, transitional and mature milk using this gating strategy. CD45+ cell quantification using flow cytometry and Trucount beads. The absolute con- centration of leukocytes in BM was calculated from the number of gated Trucount beads, as described by the manufacturer. Flow cytometry Briefly, the number of gated CD45+ and Trucount bead events were combined with the data recorded on the volume of milk that the cell pellet was derived from to determine the concentration of cells in the original milk sample, using the formula; PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 5 / 17 Leukocytes in Preterm Milk Cell concentration (cells/mL) = (number of gated cells/number of gated Trucount beads) x (number of Trucount beads per test/total tested cell suspension volume) x (volume of flow cytometry buffer used to resuspend cell pellet /original milk volume recorded). Leukocyte subset counts were determined in the same manner, and subset frequencies were determined by comparing specific subset concentrations to the total concentration of CD45 + cells in the sample. Participant characteristics The clinical characteristics of the sixty study participants are shown in Table 1. The number of samples analysed in each GA group are shown in Table 2. Analysis of the effects of maternal infection on breast milk composition Collection of maternal clinical data including medications, health status and pregnancy infor- mation was performed using medical record review for each individual from hospital admis- sion to discharge during birth, and at any subsequent hospital admissions during the first month postpartum. After hospital review, data on maternal health were collected through self- reported questionnaires covering the previous seven days at the time of milk donation. Mater- nal infection was considered present based on any evidence of chorioamnionitis (either with or without positive microbiological culture, where available tissues (n = 36) were examined by an experienced histopathologist according to the method of Redline et al. [22]), or a positive microbiological culture or symptoms of infection requiring treatment with antibiotics pre- scribed by a physician. Milk from individuals with known infection and without infection were compared at the three collection time points. Statistical analyses Categorical data were summarised using frequency distributions, and continuous data was summarized using means and standard deviations or median and interquartile ranges, as appropriate. Nonparametric tests were used for comparisons of continuous outcomes, due to lack of normality evaluated using Shapiro-Wilk normality tests. Differences in outcomes between the four GA groups were compared using Kruskal-Wallis tests with Dunn’s multiple comparison post-test. Concentrations of cells in the same mother in colostrum, transitional or mature milk were examined using a Freidman test and individual sampling periods were com- pared using Dunn’s multiple comparison post-test. Milk concentration comparisons with and without infection were compared with Mann-Whitney tests. Comparisons were made between individuals for whom data on health status was available at the time of milk collection only. Spearman correlation (denoted ‘ρ’) was used to measure the association between GA at birth and factors in milk. Statistical analysis was performed using SPSS statistical analysis software (version 20, IBM, Armonk, NY, USA). p-values <0.05 were considered statistically significant. Gating strategy The application of the whole blood gating strategy for assessing leukocytes in BM samples in mothers is shown in Fig 1, comparing a representative BM sample to an adult blood sample. In colostrum, a median 64% (range 37%–85%) of leukocytes were members of identifiable leuko- cyte subsets based on the Faucher blood phenotype gating strategy, increasing to 77% (range 37%–95%) in transitional milk and 82% (range 65%–98%) in mature milk (Friedman test p<0.001). In BM, there were additional CD45+ populations that could not be categorised according to the Faucher gating method. These ungated leukocyte populations can be seen out- side the established gates in Panels 3, 6, 8, and 10. Leukocytes in Preterm Milk Table 1. Clinical characteristics of the sixty study participants. Bolded p-values indicate those that were considered statistically significant. Continuous variables gestation period and maternal age were compared between groups using one-way analysis of variance, and categorical variables compared using Fisher exact Chi-squared tests. EP = extremely preterm, VP = very preterm, MP = moderately preterm, SD = standard deviation. Table 1. Clinical characteristics of the sixty study participants. Bolded p-values indicate those that were considered statistically significant. Continuous variables gestation period and maternal age were compared between groups using one-way analysis of variance, and categorical variables compared using Fisher exact Chi-squared tests. EP = extremely preterm, VP = very preterm, MP = moderately preterm, SD = standard deviation. Clinical characteristic Value EP (n = 15) VP (n = 15) MP (n = 15) Term (n = 15) p-value Gestation period (mean ±SD) Weeks 26.1±1.2 30.2±1.3 34.4±1.3 39.4±0.77 <0.001 Maternal age (mean±SD) Years 31.3±5.2 30.1±6.9 31.6±6.9 29.7±6.0 0.879 Maternal Infection (n, % providing milk) Colostrum 3 (30) 6 (55) 3 (27) 1 (11) 0.016 Transitional milk 2 (20) 3 (23) 0 (0) 3 (25) 0.085 Mature milk 2 (20) 2 (17) 1 (11) 2 (13) 0.826 Mode of delivery (n, %) Vaginal 9 5 8 9 0.448 Caesarean section 6 10 7 6 doi:10.1371/journal.pone.0135580.t001 correlated with the period between BM collection and processing (ρ = 0.43, p<0.001), but did not correlate with CD45+ cell concentrations or the proportion of leukocytes that were identifi- able. There were no significant differences in cellular viability or storage times between the pre- term and term mother milk samples at any time point. In comparison to the concentration of CD45+ cells obtained with flow cytometry using Tru- count beads, the total nucleated cell counts obtained with microscopy were consistently higher, but a significant correlation between the two measures was observed (ρ = 0.6, p<0.001; Fig 2). Changes in CD45+ cell concentration through the stages of lactation The distributions of BM concentrations of total CD45+ cells in all GA groups were significantly different at the different collection periods (p<0.001). The concentration of leukocytes in colostrum (median 146,000 cells/mL, range 8,470–1,510,000) was significantly higher than in both transitional milk (median 27,500 cells/mL, range 1,570–2,260,000; p<0.001) and mature milk (median 23,650, range 2,000–577,000; p<0.001). No significant difference between the transitional milk and mature milk leukocyte concentrations was observed. There were no significant differences in total leukocyte concentration between GA groups in colostrum, transitional milk or mature milk (Fig 3). The concentration of leukocytes was negatively correlated with the volume of BM expressed (ρ = -0.317; p<0.01), and therefore, dif- ferential analyses of the frequency of leukocyte subsets, in addition to total leukocyte concen- trations, were performed. Milk cell counts and viability The median number of objects with cell-like appearance in all samples using white cell count- ing fluid and microscopy was 313,500 per millilitre (range 34,300–8,020,000 per millilitre). The time from sample collection to processing was short for most samples (median 8.29 hrs, range 1–50.9 hrs) and the corresponding percentage of dead cells was low (median 1.46% dead, range 0–11.52%, respectively). The percentage of dead cells identified by Trypan blue exclusion 6 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Leukocytes in Preterm Milk Leukocytes in Preterm Milk Table 2. Total leukocyte and subset concentrations in breast milk measured across the first month of lactation. Median [interquartile range] values of total leukocytes and leukocyte subsets per millilitre of breast milk detected using flow cytometry in preterm and term mother groups in colostrum, transi- tional milk and mature milk. Symbol a with bolded text denotes a significantly different comparison in Kruskal-Wallis tests comparing gestational age groups after adjusting for multiple comparisons. Other comparisons were not statistically significant. EP = extremely preterm, VP = very preterm, MP = moderately preterm Table 2. Total leukocyte and subset concentrations in breast milk measured across the first month of lactation. Median [interquartile range] values of total leukocytes and leukocyte subsets per millilitre of breast milk detected using flow cytometry in preterm and term mother groups in colostrum, transi- tional milk and mature milk. Symbol a with bolded text denotes a significantly different comparison in Kruskal-Wallis tests comparing gestational age groups after adjusting for multiple comparisons. Other comparisons were not statistically significant. EP = extremely preterm, VP = very preterm, MP = moderately preterm. Table 2. Total leukocyte and subset concentrations in breast milk measured across the first month of lactation. Median [interquartile range] values of total leukocytes and leukocyte subsets per millilitre of breast milk detected using flow cytometry in preterm and term mother groups in colostrum, transi- tional milk and mature milk. Symbol a with bolded text denotes a significantly different comparison in Kruskal-Wallis tests comparing gestational age groups after adjusting for multiple comparisons. Other comparisons were not statistically significant. Changes in breast milk leukocyte composition from colostrum to mature milk The median frequencies of leukocyte subsets identified in colostrum, transitional milk, and mature milk are shown in Fig 4. A proportion of leukocytes did not fall into set gates for blood PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 7 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 EP = extremely preterm, VP = very preterm, MP = moderately preterm p Colostrum Transitional milk Mature milk EP (n = 10) VP (n = 11) MP (n = 11) Term (n = 9) EP (n = 10) VP (n = 13) MP (n = 12) Term (n = 12) EP (n = 10) VP (n = 12) MP (n = 9) Term (n = 15) CD45 cells/ mL 271,000 [95,200– 439,000] 96,000 [73,700– 352,000] 129,000 [67,700– 360,000] 184,000 [111,000– 291,000] 69,550 [20,300– 136,000] 26,000 [14,400– 54,200] 27,250 [10,450– 63,950] 32,450 [10,450– 66,150] 27,250 [14,400– 48,700] 12,500 [7,360– 23,700] 44,700 [28,200– 77,400] 14,700 [7,420– 56,500] CD16 + monocytes 2,415 [1,420– 14,100] 3,850 [1,930– 7,920] 2,490 [1,070– 4,890] 3,060 [903– 5,800] 657 [370– 1,450] 486 [310– 1,060] 346 [248– 646] 364 [170– 772] 706 [203– 1,130] 225 [123– 435] 686 [384– 1,110] 307 [164– 1,160] Cytotoxic T&NK cells 2,020 [389– 6,010] 1,530 [208– 4,510] 1,020 [722– 1,940] 945 [468– 1,250] 198 [86– 701] 135 [79– 835] 288 [145– 648] 222 [66– 756] 452 [81– 1,220] 109 [59– 218]a 430 [198– 947]a 340 [128– 513] Basophils 7,125 [3,460– 10,600] 2,780 [1,330– 12,400] 4,490 [2,440– 9,050] 2,460 [1,490– 3,410] 654 [174– 1,240] 611 [158– 851] 591 [345– 982] 435 [172– 1,320] 235 [106– 369] 263 [98– 374] 1,550 [340– 2,320] 370 [96– 571] Non- cytotoxic T cells 29,045 [8,760– 62,500] 14,800 [4,380– 26,500] 24,800 [2,860– 41,200] 5,410 [4,640– 16,300] 2,130 [1,480– 5,750] 1,860 [732– 6,940] 3,545 [335– 4,210] 1,740 [524– 4,300] 1,205 [400– 5,330] 647 [397– 1,695] 4,360 [953– 9,530] 1,290 [250– 4,040] CD16- monocytes 2,805 [1,480– 7,850] 2,350 [997– 4,720] 2,460 [1,290– 14,500] 1,990 [233– 5,800] 198 [46– 971] 183 [71– 776] 357 [252– 1,128] 531 [89– 801] 106 [21– 916] 146 [43– 344] 1,070 [170– 2,370] 213 [28– 501] Myeloid precursors 24,100 [13,700– 63,200] 15,500 [6,690– 84,300] 19,600 [10,100– 80,300] 16,900 [14,200– 21,300] 7,475 [4,040– 57,900] 4,050 [1,930– 8,930] 5,210 [1,295– 8,095] 6,295 [2,480– 11,750] 1,173 [532– 4,460] 1,230 [520– 2,790] 4,060 [2,550– 6,430] 1,490 [559– 4,560] B cell precursors 2,395 [1,720– 12,700] 3,100 [1,970– 10,300] 1,800 [689– 5,110] 7,720 [1,620– 8,370] 1,825 [1,360– 2,840]a 530 [305– 917] 233 [158– 555]a 485 [207– 1,235] 298 [132– 1,130] 81 [47– 258] 124 [89– 194] 135 [74– 285] B cells 1,311 [356– 8,060] 716 [306– 2,860] 470 [227– 868] 982 [333– 1,230] 414 [63– 552] 167 [57– 260] 82 [41– 118] 149 [73– 230] 177 [46– 313]a 24 [14– 89]a 160 [57– 221] 59 [26– 127] Neutrophils 25,450 [9,520– 74,500] 15,700 [6,410– 26,700] 20,600 [7,840– 25,500] 23,500 [1,980– 65,900] 6,785 [1,640– 15,400] 2,250 [1,810– 4,740] 4,095 [1,635– 12,250] 5,640 [1,221– 14,425] 7,665 [2,950– 13,400] 3,745 [2,280– 5,765] 9,310 [7,610– 10,500] 2,560 [1,490– 25,700] Eosinophils 4,040 [1,550– 8,010] 2,650 [1,160– 10,200] 3,690 [2,270– 5,520] 1,760 [1,040– 2,890] 397 [78– 915] 228 [132– 933] 408 [187– 859] 371 [75– 1,685] 157 [31– 402] 164 [51– 461] 976 [181– 2,620] 269 [110– 395] Immature granulocytes 19,400 [13,100– 35,800] 7,780 [4,300– 20,600] 10,100 [2,580– 35,000] 12,600 [8,410– 26,500] 7,850 [1,660– 35,900] 3,280 [1,300– 6,870] 2,970 [1,615– 5,435] 2,945 [2,160– 12,230] 4,390 [949– 16,100] 2,510 [1,295– 8,595] 9,100 [7,600– 11,000] 2,050 [775– 14,600] doi:10.1371/journal.pone.0135580.t002 leukocytes (median at each time point 18–36% across GA groups). doi:10.1371/journal.pone.0135580.t002 The percentage of leuko- cytes that were identified as blood subsets in milk was positively correlated with the number of days postpartum that the milk was expressed (ρ = 0.604, p<0.001). Of the identified cells, the major leukocytes present were myeloid precursors (median 9–20%), neutrophils (median 12– 27%), immature granulocytes (median 8–17%), and non-cytotoxic T cells (median 6–7%). The relative median frequencies of neutrophils and immature granulocytes of total leukocytes sig- nificantly increased from colostrum to mature milk (Fig 4H and 4J), whereas the relative fre- quencies of CD16- monocytes, myeloid precursors, B cell precursors, eosinophils, and basophils decreased over the first month postpartum (Fig 4 parts D, E, F, I and K; p<0.05). PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 8 / 17 Leukocytes in Preterm Milk Fig 2. Comparison of the total cell concentrations obtained from milk samples with different methods. Figure shows scatterplot of total cell counts obtained through microscopy on x-axis, compared with total CD45+ leukocyte concentrations obtained with flow cytometry on the y-axis. All data are shown on a log scale. doi:10.1371/journal.pone.0135580.g002 Fig 2. Comparison of the total cell concentrations obtained from milk samples with different methods. Figure shows scatterplot of total cell counts obtained through microscopy on x-axis, compared with total CD45+ leukocyte concentrations obtained with flow cytometry on the y-axis. All data are shown on a log scale. Fig 2. Comparison of the total cell concentrations obtained from milk samples with different methods. Figure shows scatterplot of total cell counts obtained through microscopy on x-axis, compared with total CD45+ leukocyte concentrations obtained with flow cytometry on the y-axis. All data are shown on a log scale. Fig 2. Comparison of the total cell concentrations obtained from milk samples with different methods. Figure shows scatterplot of total cell counts obtained through microscopy on x-axis, compared with total CD45+ leukocyte concentrations obtained with flow cytometry on the y-axis. All data are shown on a log scale. doi:10.1371/journal.pone.0135580.g002 Fig 3. Leukocyte concentrations in breast milk samples. Boxplots showing the total concentration of leukocytes (cells/mL) on a log scale in colostrum (C, n = 41), transitional milk (TM, n = 47), and mature milk (MM, n = 46), in extremely preterm (white), very preterm (light grey), moderately preterm (dark grey) and term (striped) donors. Fig 3. Leukocyte concentrations in breast milk samples. Boxplots showing the total concentration of leukocytes (cells/mL) on a log scale in colostrum (C, n = 41), transitional milk (TM, n = 47), and mature milk (MM, n = 46), in extremely preterm (white), very preterm (light grey), moderately preterm (dark grey) and term (striped) donors. doi:10.1371/journal.pone.0135580.g003 doi:10.1371/journal.pone.0135580.g003 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 9 / 17 Effects of gestation on milk leukocyte concentration and frequencies Some minor differences in GA groups were observed in the concentration of specific leukocyte subsets at different sampling points. In colostrum, the relative frequencies of non-cytotoxic T cells and B lymphocytes in the total leukocyte populations were negatively correlated with increasing gestation (ρ = -0.35, p = 0.03 and ρ = -0.319, p = 0.04, respectively). In transitional milk, the concentration of B cell precursors was significantly higher in extremely preterm mothers compared to moderately preterm mothers (median 1,825 cells/mL vs. 233 cells/mL, p = 0.02; Table 2), and the frequency of CD16 negative monocytes was signifi- cantly lower in extremely preterm mothers compared to moderately preterm mothers (median 0.34% vs. 2.13%, p = 0.03; Table A in S1 File). In transitional milk, the frequency of neutrophils was positively correlated with gestation (ρ = 0.305; p = 0.04). In mature milk, total cytotoxic T and NK cells concentrations were significantly lower in very preterm compared to moderately preterm mothers (median 109 vs. 430 cells/mL; p = 0.04). B lymphocyte concentration was significantly higher in extremely preterm compared to very preterm mothers (median 177 vs. 24 cells/mL; p = 0.04). In mature milk, no significant correlations between total leukocyte or subset frequencies and gestation were observed. Leukocytes in Preterm Milk Fig 4. Changes to the composition of breast milk through the first month of lactation. Line shows the median value for that milk sampling time point. Data show A) concentrations of total CD45+ cells in colostrum (C), transitional milk (TM) or mature milk (MM); B-L) frequencies of leukocyte subsets in colostrum (C), transitional milk (TM) and mature milk (MM) from all donors. p<0.05 in post-test comparing different stages of lactation; for each comparison, number of symbols indicates p-value (p<0.05, p<0.01, p<0.001). Fig 4. Changes to the composition of breast milk through the first month of lactation. Line shows the median value for that milk sampling time point. Data show A) concentrations of total CD45+ cells in colostrum (C), transitional milk (TM) or mature milk (MM); B-L) frequencies of leukocyte subsets in colostrum (C), transitional milk (TM) and mature milk (MM) from all donors. p<0.05 in post-test comparing different stages of lactation; for each comparison, number of symbols indicates p-value (p<0.05, p<0.01, *p<0.001). Fig 4. Changes to the composition of breast milk through the first month of lactation. Line shows the median value for that milk sampling time point. Data show A) concentrations of total CD45+ cells in colostrum (C), transitional milk (TM) or mature milk (MM); B-L) frequencies of leukocyte subsets in colostrum (C), transitional milk (TM) and mature milk (MM) from all donors. p<0.05 in post-test comparing different stages of lactation; for each comparison, number of symbols indicates p-value (p<0.05, p<0.01, **p<0.001). doi:10.1371/journal.pone.0135580.g004 Effects of bacterial infections on leukocyte concentration and frequencies The total number of CD45+ cells did not significantly differ in BM of mothers with or without recent infection (for more information, refer to Table B in S1 File). There were no differences in milk leukocyte concentrations or leukocyte subset frequencies between those with or without infection in colostrum. However, in transitional milk, those with reported bacterial infections around the time of the donation (n = 7) had significantly lower frequencies of basophils than those who were well (n = 40) (median 0.91% vs. 1.68%; p = 0.02). Furthermore, bacterial infec- tions (n = 6 vs. not infected n = 40) at the time of mature milk sampling were also associated with lower frequencies of basophils (median 0.72% vs. 1.67%; p = 0.03). Leukocyt Leukocytes in Preterm Milk PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 10 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Discussion Using a multicolour flow cytometry panel validated to identify CD45+ leukocyte subsets in human blood [13], we were able to identify and quantify equivalent leukocytes in BM. Recently, researchers have applied single colour and multicolour flow cytometry to human milk to iden- tify a broader range of cells than previously appreciated, including B lymphocytes and T lym- phocyte subsets in term colostrum, in addition to NK and NK-T cells [26, 27]. Though the functionality of myeloid-derived cells in milk suggests that these cells may be more differenti- ated than in blood [28], it is not known with the exception of B lymphocytes [29], whether the morphologies and frequencies of specific BM leukocyte subsets, only identifiable by flow cytometry, correspond to those found in blood. We found evidence of additional leukocytes not previously recognised in milk, including immature granulocytes and myeloid precursors. Further CD45+ cell subsets were not readily identifiable, but could include macrophages, plasma cells, dendritic cells and others. Additional cell markers such as CD14, CD64, HLA-DR, CD86 and CD11c for macrophage and dendritic cells [30], CD138 for plasma cells [31] and selection of combinations of other markers of haematopoietic progenitor cells such as CD34, CD38, CD135 and CD10 could be used for granulomonocytic and other precursor pop- ulations [32], but will need to be validated for flow cytometry in future BM research, particu- larly for low or negative CD45-staining cells, where immature leukocytes and rare leukocyte types in blood have potential to be misidentified [33]. Using a multicolour flow cytometry panel validated to identify CD45+ leukocyte subsets in human blood [13], we were able to identify and quantify equivalent leukocytes in BM. Recently, researchers have applied single colour and multicolour flow cytometry to human milk to iden- tify a broader range of cells than previously appreciated, including B lymphocytes and T lym- phocyte subsets in term colostrum, in addition to NK and NK-T cells [26, 27]. Though the functionality of myeloid-derived cells in milk suggests that these cells may be more differenti- ated than in blood [28], it is not known with the exception of B lymphocytes [29], whether the morphologies and frequencies of specific BM leukocyte subsets, only identifiable by flow cytometry, correspond to those found in blood. We found evidence of additional leukocytes not previously recognised in milk, including immature granulocytes and myeloid precursors. Discussion This study aimed to characterise leukocytes (CD45+) in human preterm and term BM during the first month of life using a flow cytometry-based, extended differential leukocyte counting method. We successfully identified leukocytes in BM at all stages of lactation examined using the leukocyte gating panel described by Faucher et al. [13]. The majority of CD45+ cells in BM showed a similar phenotype to blood cells, though differences in the frequencies of leukocyte subsets and other non-circulating cell populations were observed compared to blood. There were no differences in total leukocyte concentrations between the GA groups, whereas some differences were found in specific leukocyte subset concentrations and frequencies between GA groups. As lactation progressed, the concentration of total leukocytes in BM decreased. We did not observe significant changes in the concentration of total leukocytes in BM of mothers with clinical infections, though minor alterations to leukocyte subset frequencies were observed dur- ing maternal infections. 11 / 17 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Leukocytes in Preterm Milk Traditionally, BM leukocytes have been identified using blood smear stains and cytology, and leukocyte populations were reported to be composed mainly of macrophages, neutrophils, T lymphocytes, B lymphocytes, and monocytes [23]. However, visual identification can result in misidentification and overestimation of BM leukocyte concentration, whereas multicolour flow cytometry provides superior identification and quantification of leukocytes [23]. We found a clear and consistent relationship in cell concentrations when directly comparing both methods, though non-leukocyte cells are clearly also present in BM. Despite the benefits of flow cytometry, there are limitations. The use of appropriate controls and cytometer settings are critical to reduce the likelihood of misidentification of cells [21, 24, 25]. We selected a biological control (blood) to set the cut-off for positive CD45 staining and this is the major source of potential error, since this is the only gate with a single criterion for cell selection. We did not perform FMO controls for each milk sample, instead an indication of the background staining was taken from three milk and one blood control tested (<1.7% in all cases), though the proportion of background staining could have been greater in individual milk samples. Reassuringly, the majority of cells that we identified as CD45+ were subsequently gated into specific leukocyte populations. In future studies, additional biological controls for gating positive populations may be used in combination with sample-specific controls. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Discussion Further CD45+ cell subsets were not readily identifiable, but could include macrophages, plasma cells, dendritic cells and others. Additional cell markers such as CD14, CD64, HLA-DR, CD86 and CD11c for macrophage and dendritic cells [30], CD138 for plasma cells [31] and selection of combinations of other markers of haematopoietic progenitor cells such as CD34, CD38, CD135 and CD10 could be used for granulomonocytic and other precursor pop- ulations [32], but will need to be validated for flow cytometry in future BM research, particu- larly for low or negative CD45-staining cells, where immature leukocytes and rare leukocyte types in blood have potential to be misidentified [33]. We found a clear inverse relationship between maturational stage of the BM and the con- centration of leukocytes. Despite temporal changes to cell numbers, substantial (103–106 cells/ mL) leukocyte numbers were universally present at all sampling time points, with inter-indi- vidual variability. Recent studies have reported a significant association between increased numbers of leukocytes in BM and infections of both the mother and the infant [27, 34]. The high number of leukocytes found in the current study across all GA groups may relate to a cur- rent illness of either the mother or her infant that was not identified and/or reported, and therefore was not accounted for, or it could also be related to methodological differences between studies in how leukocytes were identified. Importantly, we noted much lower frequen- cies of neutrophils and monocytes than previously reported (median 12.4% neutrophils and 3.7% monocytes in this study, compared to 28–48.8% and 40.8–61% of colostrum leukocytes, respectively) [26, 35]. This is possibly due to similarities in visual appearance of other cell types, indistinguishable by microscopy. The employment of flow cytometry-based methods combined with bead count methods in a single platform may significantly expand the ability to PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 12 / 17 Leukocytes in Preterm Milk describe and improve the accuracy of reporting of BM leukocyte data in the future. One limita- tion of this work was the necessity to collect BM that had been refrigerated prior to processing, which resulted in inclusion of low numbers of non-viable cells. However, this was done to avoid further burden on the donors, particularly mothers of extremely preterm infants. Future work could include a viability marker to directly assess for non-viable cells. Discussion In addition, we can- not account for any potential effects that different methods of breast milk expression might have had on the measurements made in the milk, as data were not collected about the method used for each sample. It has previously been reported that moderately preterm mothers’ BM contains higher con- centrations of leukocytes compared to term mothers [12, 36]. In contrast, Goldman et al. reported that the concentrations of leukocytes were lower in preterm BM [37], whereas Rodri- guez et al. found leukocyte numbers were not different between preterm and term mothers [38]. We did not observe significant effects of preterm birth on concentration of leukocytes at any stage of lactation, but we did find a negative correlation between the frequency of B lym- phocytes and non-cytotoxic T lymphocytes with gestation in colostrum. We also found a posi- tive correlation between the frequency of neutrophils and gestation in transitional milk. These findings may reflect altered immune activation status or increased expression of proliferation or chemotactic factors for certain cell types in the mammary gland of mothers of preterm infants, and merit further investigation. Previous studies did not include mothers of extremely preterm infants or use specific staining methods to quantify total leukocytes; therefore, the pos- sibility of comparison to this work is limited, particularly considering that our method does not identify macrophages. Our sampling period for colostrum was typically on day 4 postna- tally, and we may have missed the changes observed in other studies between days 1–3. Research by other groups suggests that the functionality of preterm colostrum leukocytes (e.g. bacterial phagocytosis and killing as well as cellular proliferation) appears similar to term BM [38, 39]. Therefore, this study suggests that the increased risk of infection in preterm infants is not associated with deficiencies in leukocyte concentration or activity in fresh unpro- cessed preterm BM, nor deficiencies in milk antimicrobial proteins and peptides [40]. How- ever, preterm infants often receive low volumes of human BM (typically refrigerated) or they are given pasteurised donor human milk (PDHM) with inactivated leukocytes [41]. BM- derived leukocytes may aid in protecting infants from infection through normal protective functions such as phagocytosis and production of antimicrobial peptides, and have been found in the peripheral circulation and distant tissues of animal models after ingestion [10, 42], with observed effects on blood leukocyte populations in both animal and human studies [43–46]. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 Supporting Information S1 Fig. Preliminary gates of the flow cytometry method. Gates were used to select and exclude Trucount beads (1), remove doublets (2), and exclude non-cellular material and bacte- ria (3), respectively. (TIFF) S1 File. Table A in S1 File. Leukocyte subset frequencies in milk collection across the first month of lactation. Median [interquartile range] values of relative frequencies of leukocyte subsets detected using flow cytometry in preterm and term mother groups in colostrum, transi- tional milk and mature milk Symbol a with bolded text denotes significantly different compari- son in Kruskal Wallis test comparing gestational age groups after adjusting for multiple comparisons. EP = extremely preterm, VP = very preterm, MP = moderately preterm. Table B in S1 File. Prevalence of reported bacterial infections in milk donors during the sample col- lection period. Abbreviations: URTI = upper respiratory tract infection; UTI = urinary tract infection. Infections around the time of colostrum were reported in medical records, after hos- pital discharge, infections were self-reported (most transitional and mature milk collections). (DOCX) Acknowledgments The authors are thankful for the assistance of the King Edward Memorial Hospital nursing and neonatal unit staff, and would like to acknowledge the generosity of the study participants who have given their time and donated milk to this research. We would also like to acknowledge the work of Amy Prosser, who optimised the flow cytometry method for use in our laboratory. Discussion However, it is unlikely that low consumption of leukocytes, independent of the many other protective functions of the soluble molecules in milk, is responsible for the increased infection risk in preterm infants fed PDHM and those consuming low BM volumes [1, 47]. Contrary to previous reports of increased leukocyte concentrations in BM in maternal or infant infections [27, 34], we did not observe significant increases in absolute leukocyte con- centration during infections in our study. However, sensitivity for detecting differences was limited (below 2-fold in colostrum and below 3-fold in other samples) by the small number of infected mothers. Moreover, the presence of non-reported infections in this population cannot be excluded. Significantly, the majority of donors were taking at least one medication, such as antibiotics, analgesics or anti-inflammatory drugs at one point during the study. This study was not powered to assess possible effects of medications on leukocytes. Overall, our data sug- gest that cellular composition of BM is affected by stage of lactation. This is the first study to comprehensively identify and characterise immune cell subsets in preterm BM by flow cytometry. Despite the technological advances of multicolour flow cytom- etry, the progress of characterisation of BM leukocytes has been relatively neglected compared PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 13 / 17 Leukocytes in Preterm Milk to blood. The characterisation of the cellular concentrations and their function in BM is critical to understand how BM protects the infant and/or mammary gland from infections. Fresh pre- term BM is not deficient in leukocytes and infants receiving lower doses of BM, frozen or pas- teurised human BM may be disadvantaged. This work demonstrates that BM contains a greater variety and complexity of leukocyte subsets than previously appreciated. Author Contributions Conceived and designed the experiments: ST TS AC KS PR EJ FK. Performed the experiments: ST. Analyzed the data: ST. Contributed reagents/materials/analysis tools: ML CK FK AC. Conceived and designed the experiments: ST TS AC KS PR EJ FK. Performed the experiments: ST. Analyzed the data: ST. Contributed reagents/materials/analysis tools: ML CK FK AC. Wrote the paper: ST TS AC. Recruited study participants: ST ML CK. Editing and reviewing y g y Wrote the paper: ST TS AC. Recruited study participants: ST ML CK. Editing and reviewing manuscript: EJ ML CK PR DD KS FK TS AC. y g y Wrote the paper: ST TS AC. Recruited study participants: ST ML CK. Editing and reviewing manuscript: EJ ML CK PR DD KS FK TS AC. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 References 11. Schnorr KL, Pearson LD. Intestinal absorption of maternal leucocytes by newborn lambs. J Reprod Immunol. 1984; 6(5):329–37. doi: http://dx.doi.org/10.1016/0165-0378(84)90031-7 PMID: 6434734. 12. Dawarkadas AM, Saha K, Mathur NB. A comparative study of cells and anti-microbial proteins in colos- trum of mothers delivering pre- and full-term babies. J Trop Pediatr. 1991; 37(5):214–9. Epub 1991/10/ 01. PMID: 1784052. 13. Faucher J-L, Lacronique-Gazaille C, Frébet E, Trimoreau F, Donnard M, Bordessoule D, et al. “6 mark- ers/5 colors” extended white blood cell differential by flow cytometry. Cytometry Part A. 2007; 71A (11):934–44. doi: 10.1002/cyto.a.20457 PMID: 17879238. 14. Bjornsson S, Wahlstrom S, Norstrom E, Bernevi I, O'Neill U, Johansson E, et al. Total nucleated cell dif- ferential for blood and bone marrow using a single tube in a five-color flow cytometer. Cytometry Part B, Clinical cytometry. 2008; 74(2):91–103. doi: 10.1002/cyto.b.20382 PMID: 18061952. 15. Roussel M, Benard C, Ly-Sunnaram B, Fest T. Refining the white blood cell differential: the first flow cytometry routine application. Cytometry Part A: the journal of the International Society for Analytical Cytology. 2010; 77(6):552–63. Epub 2010/05/28. doi: 10.1002/cyto.a.20893 PMID: 20506466. 16. van de Geijn GJ, van Rees V, van Pul-Bom N, Birnie E, Janssen H, Pegels H, et al. Leukoflow: multipa- rameter extended white blood cell differentiation for routine analysis by flow cytometry. Cytometry Part A: the journal of the International Society for Analytical Cytology. 2011; 79(9):694–706. doi: 10.1002/ cyto.a.21105 PMID: 21786418. 17. Jo Y, Kim SH, Koh K, Park J, Shim YB, Lim J, et al. Reliable, accurate determination of the leukocyte differential of leukopenic samples by using Hematoflow method. The Korean journal of laboratory medi- cine. 2011; 31(3):131–7. doi: 10.3343/kjlm.2011.31.3.131 PMID: 21779183; PubMed Central PMCID: PMC3129340. 18. Park SH, Park BG, Park CJ, Kim S, Kim DH, Jang S, et al. An extended leukocyte differential count (16 types of circulating leukocytes) using the CytoDiff flow cytometric system can provide information for the discrimination of sepsis severity and prediction of outcome in sepsis patients. Cytometry Part B, Clinical cytometry. 2014; 86(4):244–56. doi: 10.1002/cyto.b.21123 PMID: 24002800. 19. World Health Organization, March of Dimes, The Partnership for Maternal NCH, Save the Children. Born Too Soon: The Global Action Report on Preterm Birth. World Health Organization, Geneva, 2012. 20. Castellote C, Casillas R, Ramirez-Santana C, Perez-Cano FJ, Castell M, Moretones MG, et al. Prema- ture delivery influences the immunological composition of colostrum and transitional and mature human milk. The Journal of nutrition. References 1. Patel AL, Johnson TJ, Engstrom JL, Fogg LF, Jegier BJ, Bigger HR, et al. Impact of early human milk on sepsis and health-care costs in very low birth weight infants. J Perinatol. 2013; 33(7):514–9. doi: 10. 1038/jp.2013.2 PMID: 23370606; PubMed Central PMCID: PMC3644388. 1. Patel AL, Johnson TJ, Engstrom JL, Fogg LF, Jegier BJ, Bigger HR, et al. Impact of early human milk on sepsis and health-care costs in very low birth weight infants. J Perinatol. 2013; 33(7):514–9. doi: 10. 1038/jp.2013.2 PMID: 23370606; PubMed Central PMCID: PMC3644388. 2. Chatterton DE, Nguyen DN, Bering SB, Sangild PT. Anti-inflammatory mechanisms of bioactive milk proteins in the intestine of newborns. Int J Biochem Cell Biol. 2013; 45(8):1730–47. doi: 10.1016/j. biocel.2013.04.028 PMID: 23660296. 3. Marchant EA, Boyce GK, Sadarangani M, Lavoie PM. Neonatal sepsis due to coagulase-negative staphylococci. Clinical & developmental immunology. 2013; 2013:586076. doi: 10.1155/2013/586076 PMID: 23762094; PubMed Central PMCID: PMC3674645. 4. Currie AJ, Curtis S, Strunk T, Riley K, Liyanage K, Prescott S, et al. Preterm infants have deficient monocyte and lymphocyte cytokine responses to group B streptococcus. Infection and immunity. 2011; 79(4):1588–96. doi: 10.1128/IAI.00535-10 PMID: 21300777; PubMed Central PMCID: PMC3067554. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 14 / 17 Leukocytes in Preterm Milk 5. Hobbs JR, Davis JA. Serum gamma-G-globulin levels and gestational age in premature babies. Lancet. 1967; 1(7493):757–9. PMID: 4164125. 6. Lavoie PM, Huang Q, Jolette E, Whalen M, Nuyt AM, Audibert F, et al. Profound lack of interleukin (IL)- 12/IL-23p40 in neonates born early in gestation is associated with an increased risk of sepsis. The Jour- nal of infectious diseases. 2010; 202(11):1754–63. doi: 10.1086/657143 PMID: 20977341; PubMed Central PMCID: PMC2975517. 7. Melville JM, Moss TJ. The immune consequences of preterm birth. Frontiers in neuroscience. 2013; 7:79. doi: 10.3389/fnins.2013.00079 PMID: 23734091; PubMed Central PMCID: PMC3659282. 8. Johnson DF, France GL, Marmer DJ, Steele RW. Bactericidal mechanisms of human breast milk leuko- cytes. Infection and immunity. 1980; 28(2):314–8. PMID: 6249738; PubMed Central PMCID: PMC550935. 9. Armogida SA, Yannaras NM, Melton AL, Srivastava MD. Identification and quantification of innate immune system mediators in human breast milk. Allergy Asthma Proc. 2004; 25(5):297–304. Epub 2004/12/18. PMID: 15603202. 10. Jain L, Vidyasagar D, Xanthou M, Ghai V, Shimada S, Blend M. In vivo distribution of human milk leuco- cytes after ingestion by newborn baboons. Archives of disease in childhood. 1989; 64(7 Spec No):930– 3. Epub 1989/07/01. PMID: 2774634; PubMed Central PMCID: PMCPMC1590089. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 References 2011; 141(6):1181–7. doi: 10.3945/jn.110.133652 PMID: 21508211. 21. Njemini R, Onyema OO, Renmans W, Bautmans I, De Waele M, Mets T. Shortcomings in the Applica- tion of Multicolour Flow Cytometry in Lymphocyte Subsets Enumeration. Scand J Immunol. 2014; 79 (2):75–89. doi: 10.1111/sji.12142 PMID: 24313541 22. Redline RW. Placental inflammation. Seminars in neonatology: SN. 2004; 9(4):265–74. doi: 10.1016/j. siny.2003.09.005 PMID: 15251143. 23. Hassiotou F, Geddes DT, Hartmann PE. Cells in human milk: state of the science. Journal of human lactation: official journal of International Lactation Consultant Association. 2013; 29(2):171–82. doi: 10. 1177/0890334413477242 PMID: 23515088. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 15 / 17 Leukocytes in Preterm Milk 24. Maecker HT, Trotter J. Flow cytometry controls, instrument setup, and the determination of positivity. Cytometry Part A: the journal of the International Society for Analytical Cytology. 2006; 69(9):1037–42. Epub 2006/08/05. doi: 10.1002/cyto.a.20333 PMID: 16888771. 25. Hulspas R, O'Gorman MRG, Wood BL, Gratama JW, Sutherland DR. Considerations for the control of background fluorescence in clinical flow cytometry. Cytometry Part B, Clinical cytometry. 2009; 76B (6):355–64. doi: 10.1002/cyto.b.20485 26. Peroni DG, Chirumbolo S, Veneri D, Piacentini GL, Tenero L, Vella A, et al. Colostrum-derived B and T cells as an extra-lymphoid compartment of effector cell populations in humans. The journal of maternal- fetal & neonatal medicine: the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstet. 2013; 26(2):137–42. doi: 10.3109/14767058.2012.733744 PMID: 23013166. 27. Hassiotou F, Hepworth AR, Metzger P, Tat Lai C, Trengove N, Hartmann PE, et al. Maternal and infant infections stimulate a rapid leukocyte response in breastmilk. Clin Trans Immunol. 2013; 2. 28. Ozkaragoz F, Rudloff HB, Rajaraman S, Mushtaha AA, Schmalstieg FC, Goldman AS. The motility of human milk macrophages in collagen gels. Pediatric research. 1988; 23(5):449–52. Epub 1988/05/01. doi: 10.1203/00006450-198805000-00001 PMID: 3387165. 29. Tuaillon E, Valea D, Becquart P, Al Tabaa Y, Meda N, Bollore K, et al. Human milk-derived B cells: a highly activated switched memory cell population primed to secrete antibodies. Journal of immunology. 2009; 182(11):7155–62. doi: 10.4049/jimmunol.0803107 PMID: 19454712. 30. Lay JC, Peden DB, Alexis NE. Flow cytometry of sputum: assessing inflammation and immune response elements in the bronchial airways. Inhal Toxicol. 2011; 23(7):392–406. doi: 10.3109/ 08958378.2011.575568 PMID: 21639708. 31. O'Connell FP, Pinkus JL, Pinkus GS. CD138 (syndecan-1), a plasma cell marker immunohistochemical profile in hematopoietic and nonhematopoietic neoplasms. Am J Clin Pathol. References 2004; 121(2):254–63. Epub 2004/02/27. doi: 10.1309/617d-wb5g-nfwx-hw4l PMID: 14983940. 32. van Galen P, Kreso A, Mbong N, Kent DG, Fitzmaurice T, Chambers JE, et al. The unfolded protein response governs integrity of the haematopoietic stem-cell pool during stress. Nature. 2014; 510 (7504):268–72. doi: 10.1038/nature13228 PMID: 24776803 33. Harrington AM, Olteanu H, Kroft SH. A Dissection of the CD45/Side Scatter “Blast Gate”. Am J Clin Pathol. 2012; 137(5):800–4. doi: 10.1309/ajcpn4g1izpabrlh PMID: 22523220. 34. Riskin A, Almog M, Peri R, Halasz K, Srugo I, Kessel A. Changes in immunomodulatory constituents of human milk in response to active infection in the nursing infant. Pediatric research. 2012; 71(2):220–5. doi: 10.1038/pr.2011.34 PMID: 22258136. 35. Eglinton BA, Roberton DM, Cummins AG. Phenotype of T-Cells, Their Soluble Receptor Levels, and Cytokine Profile of Human Breast-Milk. Immunol Cell Biol. 1994; 72(4):306–13. doi: 10.1038/Icb.1994. 46 PMID: WOS:A1994PC19900005. 36. Jain N, Mathur NB, Sharma VK, Dwarkadas AM. Cellular composition including lymphocyte subsets in preterm and full term human colostrum and milk. Acta paediatrica Scandinavica. 1991; 80(4):395–9. Epub 1991/04/01. PMID: 1676234. 37. Goldman AS, Garza C, Nichols B, Johnson CA, Smith EOB, Goldblum RM. Effects of prematurity on the immunologic system in human milk. J Pediatr. 1982; 101(6):901–5. doi: 10.1016/s0022-3476(82) 80007-3 PMID: 7143165. 38. Rodriguez C, Subiza JL, Mateos P, Casado de Frias E, Moro M, De la Concha EG. Comparative func- tional study of colostral macrophages from mothers delivering preterm and at term. Acta paediatrica Scandinavica. 1989; 78(3):337–41. Epub 1989/05/01. PMID: 2787091. 39. Schlesinger L, Munoz C, Arevalo M, Arredondo S, Mendez G. Functional capacity of colostral leuko- cytes from women delivering prematurely. Journal of pediatric gastroenterology and nutrition. 1989; 8 (1):89–94. Epub 1989/01/01. PMID: 2732868. 40. Trend S. Antibacterial defences in preterm breast milk and their role in protecting the preterm infant from late-onset neonatal sepsis [PhD]: University of Western Australia; 2015. 41. Liebhaber M, Lewiston NJ, Asquith MT, Olds-Arroyo L, Sunshine P. Alterations of lymphocytes and of antibody content of human milk after processing. J Pediatr. 1977; 91(6):897–900. doi: http://dx.doi.org/ 10.1016/S0022-3476(77)80885-8 PMID: 336859. 42. Reber AJ, Lockwood A, Hippen AR, Hurley DJ. Colostrum induced phenotypic and trafficking changes in maternal mononuclear cells in a peripheral blood leukocyte model for study of leukocyte transfer to the neonatal calf. Veterinary immunology and immunopathology. 2006; 109(1–2):139–50. doi: 10. 1016/j.vetimm.2005.08.014 PMID: 16169602. 43. Reber AJ, Donovan DC, Gabbard J, Galland K, Aceves-Avila M, Holbert KA, et al. PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 References Transfer of maternal colostral leukocytes promotes development of the neonatal immune system I. Effects on monocyte PLOS ONE \| DOI:10.1371/journal.pone.0135580 August 19, 2015 16 / 17 Leukocytes in Preterm Milk lineage cells. Veterinary immunology and immunopathology. 2008; 123(3–4):186–96. doi: 10.1016/j vetimm.2008.01.034 PMID: 18321594. lineage cells. Veterinary immunology and immunopathology. 2008; 123(3–4):186–96. doi: 10.1016/j. vetimm.2008.01.034 PMID: 18321594. 44. Reber AJ, Donovan DC, Gabbard J, Galland K, Aceves-Avila M, Holbert KA, et al. Transfer of maternal colostral leukocytes promotes development of the neonatal immune system Part II. Effects on neonatal lymphocytes. Veterinary immunology and immunopathology. 2008; 123(3–4):305–13. doi: 10.1016/j. vetimm.2008.02.009 PMID: 18394715. 45. Tarcan A, Gurakan B, Tiker F, Ozbek N. Influence of feeding formula and breast milk fortifier on lympho- cyte subsets in very low birth weight premature newborns. Biology of the neonate. 2004; 86(1):22–8. doi: 10.1159/000076918 PMID: 14981353. 46. Hawkes JS, Neumann MA, Gibson RA. The effect of breast feeding on lymphocyte subpopulations in healthy term infants at 6 months of age. Pediatric research. 1999; 45(5 Pt 1):648–51. doi: 10.1203/ 00006450-199905010-00006 PMID: 10231858. 47. Cossey V, Vanhole C, Eerdekens A, Rayyan M, Fieuws S, Schuermans A. Pasteurization of mother's own milk for preterm infants does not reduce the incidence of late-onset sepsis. Neonatology. 2013; 103(3):170–6. Epub 2012/12/22. doi: 10.1159/000345419 PMID: 23258423. 17 / 17
https://openalex.org/W4323351928	https://jdc.jefferson.edu/cgi/viewcontent.cgi?article=1202&context=rothman_institute	English	null	Surgeon Estimations of Acetabular Cup Orientation Using Intraoperative Fluoroscopic Imagining Are Unreliable	Arthroplasty today	2,023	cc-by	5,766	3-7-2023 Parker L Brush Thomas Jefferson University Thomas Jefferson University Thomas Jefferson University Jefferson Digital Commons Jefferson Digital Commons Thomas Jefferson University Thomas Jefferson University Jefferson Digital Commons Jefferson Digital Commons Rothman Institute Faculty Papers Authors Authors Parker L Brush, Adrian Santana, Gregory R Toci, Eric Slotkin, Michael Solomon, Tristan Jones, and Arjun Saxena sh, Adrian Santana, Gregory R Toci, Eric Slotkin, Michael Solomon, Tristan Jones, and Arjun This article is available at Jefferson Digital Commons: https://jdc.jefferson.edu/rothman_institute/20 a b s t r a c t Article history: Received 9 September 2022 Received in revised form 12 December 2022 Accepted 22 January 2023 Available online xxx Keywords: Total hip arthroplasty direct anterior approach intraoperative ﬂuoroscopy acetabular anteversion acetabular inclination Background: Accurate acetabular cup orientation is associated with decreased revision rates and improved outcomes of primary total hip arthroplasty. This study assesses surgeon’s ability to estimate both the acetabular component inclination and anteversion angles via intraoperative ﬂuoroscopy (IF) images. Article history: Received 9 September 2022 Received in revised form 12 December 2022 Accepted 22 January 2023 Available online xxx g Methods: We surveyed orthopedic surgeons to estimate acetabular component inclination and ante- version based on 20 IF images of total hip arthroplasty through a direct anterior approach. Postoperative computed-tomography scans were used to calculate the true inclination and anteversion component angles. The absolute difference between the true and estimated values was calculated to determine the mean and standard deviation of the survey results. Interrater reliability was determined through interclass correlation coefﬁcients. Keywords: Total hip arthroplasty direct anterior approach intraoperative ﬂuoroscopy acetabular anteversion acetabular inclination Results: A majority of surgeons preferred the direct anterior approach (83.3%) and utilized IF during surgery (70%). Surgeons surveyed were on average 5.9 away from the true value of inclination (standard deviation ¼ 4.7) and 8.8 away from the true value of anteversion (standard deviation ¼ 6.0). Re- spondents were within 5 of both inclination and anteversion in 19.7% of cases, and within 10 in 57.3% of cases. All surgeons were determined to have poor reliability in estimating anteversion (interclass cor- relation coefﬁcient < 0.5). Only 2 surgeons were determined to have moderate reliability when esti- mating inclination. g Conclusions: Surgeons, when solely relying on IF for the estimation of anteversion and inclination, are unreliable. Utilization of other techniques in conjunction with IF would improve observer reliability. Conclusions: Surgeons, when solely relying on IF for the estimation of anteversion and inclination, are unreliable. Utilization of other techniques in conjunction with IF would improve observer reliability. © 2023 The Authors. Published by Elsevier Inc. on behalf of The American Association of Hip and Knee Surgeons. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/ © 2023 The Authors. Published by Elsevier Inc. on behalf of The American Association of Hip and Knee Surgeons. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/ 4.0/). Contents lists available at ScienceDirect Contents lists available at ScienceDirect Surgeon Estimations of Acetabular Cup Orientation Using Intraoperative Fluoroscopic Imagining Are Unreliable a Rothman Orthopaedic Institute at Thomas Jefferson University, Philadelphia, PA, USA b Rutgers Robert Wood Johnson Medical School, New Brunswick, NJ, USA c Orthopaedic Associates of Reading, Tower Health, Reading Hopsital, West Reading, PA, USA d Sydney Orthopaedic Specialists, Prince of Wales Private Hospital, Randwick, Australia e Corin Group, Cirencester, Gloucestershire, UK Recommended Citation Recommended Citation Brush, Parker L; Santana, Adrian; Toci, Gregory R; Slotkin, Eric; Solomon, Michael; Jones, Tristan; and Saxena, Arjun, "Surgeon Estimations of Acetabular Cup Orientation Using Intraoperative Fluoroscopic Imagining Are Unreliable." (2023). Rothman Institute Faculty Papers. Paper 202. https://jdc.jefferson.edu/rothman_institute/202 This Article is brought to you for free and open access by the Jefferson Digital Commons. The Jefferson Digital Commons is a service of Thomas Jefferson University's Center for Teaching and Learning (CTL). The Commons is a showcase for Jefferson books and journals, peer-reviewed scholarly publications, unique historical collections from the University archives, and teaching tools. The Jefferson Digital Commons allows researchers and interested readers anywhere in the world to learn about and keep up to date with Jefferson scholarship. This article has been accepted for inclusion in Rothman Institute Faculty Papers by an authorized administrator of the Jefferson Digital Commons. For more information, please contact: JeffersonDigitalCommons@jefferson.edu. Arthroplasty Today 20 (2023) 101109 * Corresponding author. Rothman Orthopaedic Institute, 925 Chestnut Street, Philadelphia, PA, USA. Tel.: þ1 316 993 3876. E-mail address: brushparker@gmail.com commonly use IF to assist in bone preparation, component position, and intraoperative leg-length measurements [19]. Given the importance of component positioning in THA and the reliance on IF, * Corresponding author. Rothman Orthopaedic Institute, 925 Chestnut Street, Philadelphia, PA, USA. Tel.: þ1 316 993 3876. E-mail address: brushparker@gmail.com https://doi.org/10.1016/j.artd.2023.101109 2352-3441/© 2023 The Authors. Published by Elsevier Inc. on behalf of The American Association of Hip and Knee Surgeons. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). by Elsevier Inc. on behalf of The American Association of Hip and Knee Surgeons. This is an open access article under the CC BY es/by/4.0/). https://doi.org/10.1016/j.artd.2023.101109 2352-3441/© 2023 The Authors. Published by Elsevier Inc. on behalf of The American Association of Hip and Knee Surgeons. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Introduction from <5% to 12.9% at 10 years, with aseptic loosening and dislo- cation as the most common reasons for revision [5e10]. However, meta-analyses report lower rates of dislocation with the DAA [11,12]. Other studies have identiﬁed associations between improper component positioning and resultant increases in leg- length discrepancy, impingement, imbalanced stress distribution, and dislocations [13e16]. In order to improve outcomes, re- searchers have suggested that the use of intraoperative ﬂuoroscopy (IF) for the DAA has helped to improve acetabular cup orientation; however, surgeons must take measures to limit the impact of pelvic tilt on intraoperative measurements [17,18]. Regardless, surgeons commonly use IF to assist in bone preparation, component position, and intraoperative leg-length measurements [19]. Given the importance of component positioning in THA and the reliance on IF, Total hip arthroplasty (THA) is one of the most commonly per- formed orthopedic procedures with an estimated 572,000 primary THAs expected to be performed in 2030 [1,2]. Surgeons are per- forming more THAs through the direct anterior approach (DAA), with more than 50% of surgeons utilizing the approach [3,4]. One of the most important outcome measures in arthroplasty surgery is revision rate [5]. Authors have reported revisions rates for THA P.L. Brush et al. / Arthroplasty Today 20 (2023) 101109 2 Question 1: How long have you been in practice? we set out to answer if (1) surgeons are reliable in assessing acetabular orientation based on IF and (2) how often are surgeon estimations off by 5, 10, and 20. Our hypothesis is that visual inspection of IF is inadequate in estimating acetabular component positioning. we set out to answer if (1) surgeons are reliable in assessing acetabular orientation based on IF and (2) how often are surgeon estimations off by 5, 10, and 20. Our hypothesis is that visual inspection of IF is inadequate in estimating acetabular component positioning. Question 4: Do you routinely use intraoperative imaging when doing a hip replacement? Figure 2. List of the general questions included in the survey. We calculated absolute differences for each inclination and anteversion response by taking the absolute value of the difference between the true value determined by postoperative CT scans and the responses in the survey. The responses were analyzed by 2 methods: (1) The surgeon’s responses to each image were consid- ered separately, and (2) the surgeon’s responses to the image set were averaged together. To account for nonresponse bias, an additional analysis between early and late responders was per- formed under the assumption that late responders are most similar to nonresponders [20]. An independent statistician performed the statistical analysis using R Studio (Version 4.1.2, Vienna, Austria). They analyzed the survey data by intraclass correlation coefﬁcients (ICC) to evaluate for interrater reliability of responses compared to true values. ICC values less than 0.5 indicate weak agreement, values from 0.5 to less than 0.7 indicate moderate agreement, and values greater than or equal to 0.7 indicate strong agreement with the true values. Each survey respondent has an ICC value associated with their inclina- tion and anteversion responses. Results We received 34 responses to the survey for a response rate of approximately 38%. Four responses were removed as they were incomplete. All complete responses were included in the ﬁnal analysis. THA, total hip arthroplasty. a Some responses included multiple answers. Table 1 Table 1 Responses to survey questions 1 through 5. Responses to survey questions 1 through 5. Full cohort N ¼ 30 Length of practice <5 y 11 (36.7%) 5-15 y 13 (43.3%) 15-25 y 4 (13.3%) >25 y 2 (6.67%) THA per year <50 3 (10.0%) 50-100 4 (13.3%) 100-200 14 (46.7%) >200 9 (30.0%) Preferred approach Direct anterior 25 (83.3%)a Posterolateral 6 (20%)a Anterolateral 1 (3.3%) Routinely use intraoperative imaging No 9 (30.0%) Yes 21 (70.0%) Routinely use technology Digital preoperative planning 21 (70%) Patient-speciﬁc instrumentation 2 (6.7%) Imageless navigation 3 (10%) Image-based navigation 4 (13.3%) Intraoperative imaging with digital measurements 1 (3.3%) Other 3 (10%) Figure 1. Example intraoperative ﬂuoroscopic image used in the survey. THA, total hip arthroplasty. Material and methods Question 5: Do you routinely use technology when doing a hip replacement? We retrospectively identiﬁed 30 patients who underwent pri- mary THA by a DAA from 2 surgeons with both IF and a post- operative computed-tomography (CT) scan on record. Patients who had incomplete image data, received a revision THA, or a THA through an approach other than direct anterior were excluded. We used the Simpleware ScanIP (Synopsis, Mountain View, CA) soft- ware to extract anatomical landmarks from postoperative CT scans and the Solidworks (Dassault Systemes, France) software to compute values for the true inclination and anteversion of the acetabular component. The Corin Group (Cirencester, UK) constructed a survey in coordination with the authors of the manuscript consisting of 5 general questions and 20 blinded, anterior-posterior IF images in which the participants were requested to estimate the acetabular component inclination and anteversion. This estimation was performed by visual inspection of the ﬂuoroscopic images in order to simulate the operative envi- ronment, no assist tools were allowed to be used, and no practice images were provided. Figure 1 contains an example of one of the IF images used in the survey, and Supplemental Figures 1 and 2 contain the corresponding inclination and anteversion measure- ments performed on the 3D pelvis reconstruction, respectively. We removed 10 patients from the survey at random to increase the response rate and ease the time burden required to complete the survey. Figure 2 contains the 5 general questions included in the survey, and Table 1 includes the categorical options provided as answers. We sent the survey on 1 occasion to 89 surgeons in 3 different hospital systems with a combination of fellowship-trained arthroplasty surgeons, arthroplasty fellows in training, and postgraduate-year-four and postgraduate-year-ﬁve residents. The survey was sent broadly to all the senior authors’ contacts to maximize the potential responses. The surgeons’ responses were collected between July 2021 and January 2022. Figure 2. List of the general questions included in the survey. a Some responses included multiple answers. Table 3 Table 1 contains the categorical responses to survey questions 1 through 5. We found most participants preferred the DAA (83.3%), routinely used IF (70%), and routinely used a preoperative planning system (66.7%). Two of the participants reported that they prefer the direct anterior and posterolateral approaches equally. Three survey participants wrote in answers for “other” use of technology. Their responses were (2) optimized positioning system and (1) digital preoperative planning combined with conventional instruments. Individual response anteversion Range 0-39 Mean 8.8 Median 9.0 SD 6.0 Surgeon average anteversion Range 6.0-15.6 Mean 8.8 Median 8.0 SD 2.4 SD, standard deviation. Tables 2 and 3 contain descriptive statistics for the cumulative inclination and anteversion survey values with an analysis per- formed at both the individual response and surgeon average level. We based all descriptive statistics on the absolute differences. On average, surgeons were 5.9 away from the true value for inclina- tion and 8.8 away from the true value for anteversion with stan- dard deviations of 4.7 and 6.0 at the individual response level and 1.7 and 2.4 at the surgeon average level, respectively. Our survey responses ranged from 0 to 39 for inclination and 0 to 29 for anteversion by absolute difference. We identiﬁed 118 (19.7%) responses within 5 of both the true anteversion and inclination, and 344 (57.3%) of responses within 10. Out of the 600 data points, 28 (4.7%) surgeons were off by 20 or greater for anteversion, and 8 (1.3%) were off by 20 or greater for inclination, with a total of 32 (5.3%) patients with either an anteversion or inclination esti- mation off by 20 or more. This value was chosen based on the historical safe zone of component position [21]. preferred the DAA with IF in our population. Moreover, approxi- mately two-thirds of the surveyed surgeons routinely used digital preoperative planning systems, which has been shown to decrease time under ﬂuoroscopy, especially when pelvic tilt is calculated [12]. Our primary ﬁnding is that all surgeons had poor agreement with the true values of component anteversion, while all but 2 surgeons had poor agreement with the true values of component inclination determined by the analysis of postoperative CT scans. These data indicate that the surgeons surveyed were not reliable in their determination of both component inclination and anteversion through IF images. This ﬁnding is supported by the previous work of Holst et al. Table 4 Table 4 ICC values for each participant. Survey participant Inclination Anteversion 1 0.06 0.04 2 0.38 0.05 3 0.34 0.02 4 0.28 0.09 5 0.26 0.10 6 0.00 0.22 7 0.14 0.19 8 0.25 0.02 9 0.38 0.07 10 0.27 0.18 11 0.21 0.02 12 0.07 0.05 13 0.13 0.06 14 0.01 0.22 17 0.28 0.03 18 0.67a 0.17 19 0.12 0.11 20 0.01 0.11 21 0.19 0.30 22 0.53a 0.20 23 0.13 0.09 24 0.16 0.16 25 0.27 0.31 28 0.31 0.11 29 0.10 0.07 30 0.11 0.08 31 0.16 0.10 32 0.36 0.04 33 0.29 0.02 34 0.19 0.17 ICC, interclass correlation coefﬁcient. ICC values for each participant. Table 5 contains the analysis of the initial 5 (16.7%) and ﬁnal 5 (16.7%) participants. These data show similar responses between the initial and ﬁnal 5 participants by absolute inclination (6.3 vs 5.5, P ¼ .511) and absolute anteversion (9.6 vs 8.7, P ¼ .581). Table 3 who found that IF did not improve acetabular cup positioning or sizing when employing the DAA and differs from the work by James et al. who suggest that IF can help conﬁrm component positioning if used properly [12,18]. Although no sta- tistical measure interprets a correlation between the baseline characteristics we obtained and ICC values, we were unable to identify any associations between improved component orienta- tion estimation and surgeon length of practice or total surgeries Table 4 contains the ICC values for each participant with regard to inclination and anteversion estimations. We removed partici- pants 15, 16, 26, and 27 for incomplete responses. All surgeons had a weak agreement with the true values for component anteversion. Two surgeons had a moderate agreement with the true values with regard to component inclination, participants 18 and 22. These participants were in practice between 5 to 15 and less than 5 years, performed 100 to 200 and less than 50 THAs per year, respectively, routinely used intraoperative imaging, and used digital preopera- tive planning systems. The remainder of surgeons had a weak agreement with regard to component inclination. We found no surgeons to have strong agreement by ICC values. Given the consistently low reliability among survey participants, we deter- mined that increasing our sample size would not likely impact the results of this study, and the survey was closed. Discussion This survey includes the responses of 34 surgeons that were asked to evaluate the acetabular cup orientation of 20 patients who underwent THA by DAA. We found most surgeons utilized and Table 2 Descriptive statistics for the component inclination by absolute value. Table 2 Descriptive statistics for the component inclination by absolute value. Individual response inclination Range 0-28 Mean 5.9 Median 5.0 SD 4.7 Surgeon average inclination Range 3.3-9.1 Mean 5.9 Median 5.7 SD 1.7 SD, standard deviation. 21 0.19 0.30 22 0.53a 0.20 23 0.13 0.09 24 0.16 0.16 25 0.27 0.31 28 0.31 0.11 29 0.10 0.07 30 0.11 0.08 31 0.16 0.10 32 0.36 0.04 33 0.29 0.02 34 0.19 0.17 ICC, interclass correlation coefﬁcient. a Moderate agreement with the true values. a Moderate agreement with the true values. Full cohort Figure 1. Example intraoperative ﬂuoroscopic image used in the survey. 3 P.L. Brush et al. / Arthroplasty Today 20 (2023) 101109 Table 3 Descriptive statistics for the component anteversion by absolute value. ICC, interclass correlation coefﬁcient. ICC, interclass correlation coefﬁcient. Table 5 C i Table 5 Comparisons of initial and ﬁnal participants. Characteristics Initial N ¼ 5 Final N ¼ 5 P value Length of practice .048 <5 y 0 (0.00%) 2 (40.0%) 5-15 y 5 (100%) 1 (20.0%) 15-25 y 0 (0.00%) 1 (20.0%) >25 y 0 (0.00%) 1 (20.0%) THA per year 1.000 50-100 0 (0.00%) 1 (20.0%) 100-200 3 (60.0%) 2 (40.0%) >200 2 (40.0%) 2 (40.0%) Intraoperative imaging .444 No 0 (0.00%) 2 (40.0%) Yes 5 (100%) 3 (60.0%) Surgical technology 1.000 No 1 (20.0%) 2 (40.0%) Yes 4 (80.0%) 3 (60.0%) Average inclination 6.3 (1.69) 5.5 (2.06) .511 Average anteversion 9.6 (2.25) 8.7 (2.66) .581 Average inclination and average anteversion are provided as the average absolute difference from the true measurement. THA, total hip arthroplasty. Average inclination and average anteversion are provided as the average absolute difference from the true measurement. THA, total hip arthroplasty. performed per year as almost every surgeon had statistically poor agreement. Moreover, our data do not suggest an association be- tween routine use of IF or preoperative planning systems and improved IF estimation of acetabular orientation. In fact, these ICC values suggest any 2 physicians chosen randomly who perform more than 200 THAs per year would vary as much as 2 physicians chosen randomly from the survey population. Similar comparisons can be drawn for the other baseline characteristics evaluated in this survey. y We chose to evaluate the data by absolute difference so that surgeon overestimation and underestimation would not have a counteractive impact on their averaged measurement error. Although average absolute differences were within 5.9 and 8.8 for inclination and anteversion, respectively, these data showed sig- niﬁcant variability of estimation by range and standard deviation, with some surgeons perfectly estimating acetabular inclination and/or anteversion on 1 image while estimations were off by greater than 20 on other images. This variability is blunted by analyzing the surgeons’ average estimations over the 20 survey images, but such an analysis fails to consider the potential impact on individual patients. Previous studies have reported component malposition to be a signiﬁcant risk factor for early dislocation after THA, with 60% to 70% of dislocations occurring in the ﬁrst 6 weeks after surgery [22,23]. Horberg et al. reported 11 (0.39%) dislocations at an average of 71 days after THA by a DAA [24]. Table 5 C i A database review in 2018 reported dislocation readmission rates of 1.4% at a median of 40 days after the surgery for elective primary THA [25]. Although component position is not the only factor associated with disloca- tion, combined anteversion (acetabular plus femoral) outside of 40 to 60 has been shown to increase dislocation by an odds ratio of 6.9 [26]. Our data suggest that by IF estimation alone, 4.7% of pa- tients have acetabular anteversion 20 or more off the true value and would thus be at signiﬁcantly higher risk of dislocation. In addition to impacting dislocation rates, acetabular component positioning also affects revision rates [27e29]. Dislocation and revision lead to increased health-care costs and patient morbidity and stress [30]. The historical safe zones are deﬁned as 40 ± 10 of inclination and 15 ± 10 of anteversion, and these data suggest surgeons are unreliable when utilizing only IF and should consider alternate or additional methods to optimize component position [21]. Our study has many strengths, most notably its survey design. It is known that a poorly designed survey, without prior planning, could lead to inaccurate and misleading conclusions. Sprague et al. discussed the importance of survey design in ensuring maximum response rates from orthopedic surgeons [11]. Our survey utilized 11 of the 12 points that were presented in their article, missing only in that we did not evaluate the characteristics of nonresponders [11]. While our study has these strengths, we do recognize the limitations of this work. The participating surgeons were presented with 20 IF images and asked to analyze them, even though they had no hand in the production of those images. It is possible that each surgeon’s accuracy would increase if they were visualizing the case intraoperatively in addition to taking and viewing ﬂuoroscopic images. This study also does not attempt to correlate surgeon ac- curacy with clinical outcomes and is descriptive by nature of its design. We do not correlate a surgeon’s ability to estimate component position on IF to dislocation or revision rates. Moreover, our sample size is small, and the participants are all located in a similar geographical location. We have signiﬁcant risk of nonre- sponse bias, with a response rate of only 38%, well below the generally preferred rate of 60% [20]. However, we took measures to account for this by comparing early and late responders [20,38,39]. Table 2 P.L. Brush et al. / Arthroplasty Today 20 (2023) 101109 4 accurate placement and positioning of the acetabulum cup. Using this technology, Tsutsui et al. report that for both inclination and anteversion, 97.7% of the acetabular placements were within the combined target zone (30e45 of inclination and 5e25 of anteversion) compared to 61.3% of patients without navigation. As a result, this technology allows surgeons to achieve high accuracy of both cup alignment angles and positioning [31]. In addition to CT- based navigation, surgeons have several other options to achieve optimal component anteversion and inclination. One of which is the use of a mechanical insertion jig to assist in the alignment and positioning of the acetabular cup [28]. With advances in technol- ogy, it is now possible for patient-speciﬁc insertion jigs to be 3D printed, thus allowing for the angles of anteversion and inclination to be well within the safe zones in a cheap, effective manner [32,33]. Imageless navigation also presents surgeons with an adequate alternative to IF. Nogler et al. report that with the use of imageless navigation, there is signiﬁcant reduction in the median absolute difference of inclination (1.3 to 5.8) and anteversion (2.4 to 9.9) when compared to component placement with visual cues alone [34]. Other studies report these navigation systems can be more difﬁcult to place in larger patients. The impact of this difﬁculty appears to have a signiﬁcant effect on acetabular ante- version, while inclination values remain more consistent [35]. Lastly, there is a trend toward increasing robotic assistance with many types of surgery, including orthopedics. Redmond et al. report that as surgeons increase their experience with robotic assistance for THA, procedure time decreases [36]. They also report that regardless of experience level, acetabular components are well placed with a 95% conﬁdence interval of 8 [36]. Despite the learning curve for optimal robot-assisted THA, there is immediate and signiﬁcant improvement in acetabular cup positioning when compared to IF. This contrasts the learning curve of IF, which does not demonstrate an immediate improvement, requiring the sur- geon to gain experience before signiﬁcant enhancement of preci- sion is seen [29]. Additionally, Domb et al. found that 100% of robot- assisted THAs were within the safe zones for both inclination and anteversion, as described by Lewinnek et al. Table 2 [21,37] All these methods provide good alternatives or additions to the use of IF to improve acetabular cup positioning with the DAA for THA. It is important to note that the use IF does minimize the variability in anteversion and inclination, but our data suggest this method re- mains unreliable [18]. Conﬂicts of interest [14] Kluess D, Martin H, Mittelmeier W, Schmitz KP, Bader R. Inﬂuence of femoral head size on impingement, dislocation and stress distribution in total hip replacement. Med Eng Phys 2007;29:465e71. https://doi.org/10.1016/ j.medengphy.2006.07.001. Dr. A. Saxena is in the speakers' bureau of or gave paid pre- sentations for Corin; is a paid consultant for Corin; and is a member of the Patient and Public Relations Committee of American Asso- ciation of Hip and Knee Surgeons, Telecommunications Committee of Eastern Orthopaedic Association, Pennsylvania Orthopaedic So- ciety, and Web-Based Longitudinal Assessment and Hip Program Committees of the American Academy of Orthopaedic Surgeons. Dr. E. M. Slotkin receives royalties from Corin; is in the speakers' bu- reau of or gave paid presentations for Corin, Naviswiss, Phillips, and RomTech; is a paid consultant for Corin, Naviswiss, Phillips, DePuy, and RomTech; is an unpaid consultant for Efferent Health; and has stock or stock options with Naviswiss, RomTech, and Efferent Health. Dr. M. Solomon receives royalties from Corin UK and Medacta; is in the speakers' bureau of or gave paid presentations for Corin; is a paid consultant for Corin; and receives other ﬁnancial or material support from Corin. T. Jones is a paid employee of the Corin Group. All other authors declare no potential conﬂicts of interest. [15] Yoshimine F. The safe-zones for combined cup and neck anteversions that fulﬁll the essential range of motion and their optimum combination in total hip replacements. J Biomech 2006;39:1315e23. https://doi.org/10.1016/ j.jbiomech.2005.03.008. [16] Barrack RL. Dislocation after total hip arthroplasty: implant design and orientation. J Am Acad Orthop Surg 2003;11:89e99. https://doi.org/10.5435/ 00124635-200303000-00003. [17] Rathod PA, Bhalla S, Deshmukh AJ, Rodriguez JA. Does ﬂuoroscopy with anterior hip arthroplasty decrease acetabular cup variability compared with a nonguided posterior approach? Clin Orthop 2014;472:1877e85. https:// doi.org/10.1007/s11999-014-3512-2. [18] James CR, Peterson BE, Crim JR, Cook JL, Crist BD. The use of ﬂuoroscopy during direct anterior hip arthroplasty: powerful or misleading? J Arthroplasty 2018;33:1775e9. https://doi.org/10.1016/j.arth.2018.01.040. J p y p // g/ /j [19] McArthur BA, Schueler BA, Howe BM, Trousdale RT, Taunton MJ. Radiation expo- sure during ﬂuoroscopic guideddirectanteriorapproach fortotal hip arthroplasty. J Arthroplasty 2015;30:1565e8. https://doi.org/10.1016/j.arth.2015.03.029. J p y p // g/ /j [20] Johnson TP, Wislar JS. Response rates and nonresponse errors in surveys. JAMA 2012;307:1805e6. https://doi.org/10.1001/jama.2012.3532. [21] Lewinnek GE, Lewis JL, Tarr R, Compere CL, Zimmerman JR. Dislocations after total hip-replacement arthroplasties. J Bone Joint Surg Am 1978;60:217e20. For full disclosure statements refer to https://doi.org/10.1016/j. Conﬂicts of interest artd.2023.101109. [22] Brooks PJ. Dislocation following total hip replacement. Bone Jt J 2013;95- B(Supple_A):67e9. https://doi.org/10.1302/0301-620X.95B11.32645. [23] Lu Y, Xiao H, Xue F. Causes of and treatment options for dislocation following total hip arthroplasty. Exp Ther Med 2019;18:1715e22. https://doi.org/ 10.3892/etm.2019.7733. Acknowledgement [24] Horberg JV, Coobs BR, Jiwanlal AK, Betzle CJ, Capps SG, Moskal JT. Dislocation rates following total hip arthroplasty via the direct anterior approach in a consecutive, non-selective cohort. Bone Joint J 2021;103eB(7 Supple B): 38e45. https://doi.org/10.1302/0301-620X.103B7.BJJ-2020-2297.R1. The authors of this article would like to acknowledge the following contributors for their work on this project: Catherine Stambouzou, Michael A. Miranda, DO, John L. Wang, MD, and Ste- ven L. Barnett, MD. They would also like to thank Corin Group, Cirencester, Gloucestershire, United Kingdom; Florida Orthopedic Institute, Tampa, FL; Hospital for Special Surgery, West Palm Beach, FL; and Hoag Orthopedic Institute, Irvine, CA. 38e45. https://doi.org/10.1302/0301-620X.103B7.BJJ-2020-2 [25] Gausden EB, Parhar HS, Popper JE, Sculco PK, Rush BNM. Risk factors for early dislocation following primary elective total hip arthroplasty. J Arthroplasty 2018;33:1567e1571.e2. https://doi.org/10.1016/j.arth.2017.12.034. [26] Jolles BM, Zangger P, Leyvraz PF. Factors predisposing to dislocation after primary total hip arthroplasty: a multivariate analysis. J Arthroplasty 2002;17: 282e8. https://doi.org/10.1054/arth.2002.30286. p // g/ / [27] Grammatopoulos G, Alvand A, Monk AP, Mellon S, Pandit H, Rees J, et al. Sur- geons’ accuracy in achieving their desired acetabular component orientation. J Bone Joint Surg Am 2016;98:e72. https://doi.org/10.2106/JBJS.15.01080. Conclusions p // g/ / [10] Kelmer G, Stone AH, Turcotte J, King PJ. Reasons for revision: primary total hip arthroplasty mechanisms of failure. J Am Acad Orthop Surg 2021;29:78e87. https://doi.org/10.5435/JAAOS-D-19-00860. Our data suggest that surgeons may not be as reliable as pre- viously suggested with estimating acetabular cup position with IF. As technology continues to advance in the ﬁeld of arthroplasty surgery, adaptation and utilization of these products may continue to improve success rates of these already highly successful pro- cedures. We hope this survey promotes interest in improving and objectifying IF techniques to increase observer reliability. p // g/ /J [11] Sprague S, Quigley L, Bhandari M. Survey design in orthopaedic surgery: getting surgeons to respond. J Bone Joint Surg Am 2009;91(Suppl 3):27e34. https://doi.org/10.2106/JBJS.H.01574. p // g/ /J J [12] Holst DC, Levy DL, Angerame MR, Yang CC. Does the use of intraoperative ﬂuo- roscopy improve postoperative radiographic component positioning and implant size in total hip arthroplasty utilizing a direct anterior approach? Arthroplasty Today 2020;6:94e8. https://doi.org/10.1016/j.artd.2019.11.006. [13] Desai AS, Dramis A, Board TN. Leg length discrepancy after total hip arthro- plasty: a review of literature. Curr Rev Musculoskelet Med 2013;6:336e41. https://doi.org/10.1007/s12178-013-9180-0. Table 5 C i This analysis found that these groups were similar, thus suggesting One alternative to using IF is CT-based navigation, where three- dimensional cup templates are created, thus allowing for more P.L. Brush et al. / Arthroplasty Today 20 (2023) 101109 5 [7] Kandala NB, Connock M, Pulikottil-Jacob R, Sutcliffe P, Crowther MJ, Grove A, et al. Setting benchmark revision rates for total hip replacement: analysis of registry evidence. BMJ 2015;350:h756. https://doi.org/10.1136/ bmj.h756. a minimal effect of nonresponse bias [20,39]. Survey fatigue may also decrease the reliability of our results as responders may ﬁnd reviewing 20 images tedious. Lastly, despite historically poor physician response rates on surveys, there is no validated evalua- tion tool or method to assess survey quality in orthopedics [40]. j [8] Nugent M, Young SW, Frampton CM, Hooper GJ. The lifetime risk of revision following total hip arthroplasty. Bone Jt J 2021;103-B:479e85. https://doi.org/ 10.1302/0301-620X.103B3.BJJ-2020-0562.R2. / JJ [9] Ulrich SD, Seyler TM, Bennett D, Delanois RE, Saleh KJ, Thongtrangan I, et al. Total hip arthroplasties: what are the reasons for revision? Int Orthop 2008;32:597e604. https://doi.org/10.1007/s00264-007-0364-3. References [28] Woerner M, Sendtner E, Springorum R, Craiovan B, Worlicek M, Renkawitz T, et al. Visual intraoperative estimation of cup and stem position is not reliable in minimally invasive hip arthroplasty. Acta Orthop 2016;87:225e30. https:// doi.org/10.3109/17453674.2015.1137182. [1] Molina CS, Thakore RV, Blumer A, Obremskey WT, Sethi MK. Use of the Na- tional surgical quality improvement program in orthopaedic surgery. Clin Orthop 2015;473:1574e81. https://doi.org/10.1007/s11999-014-3597-7. [29] Kamara E, Robinson J, Bas MA, Rodriguez JA, Hepinstall MS. Adoption of ro- botic vs ﬂuoroscopic guidance in total hip arthroplasty: is acetabular posi- tioning improved in the learning curve? J Arthroplasty 2017;32:125e30. https://doi.org/10.1016/j.arth.2016.06.039. [2] Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007;89:780e5. https://doi.org/10.2106/JBJS.F.00222. [3] Abdel MP, Berry DJ. Current practice trends in primary hip and knee arthro- plasties among members of the American association of hip and knee sur- geons: a long-term update. J Arthroplasty 2019;34:S24e7. https://doi.org/ 10.1016/j.arth.2019.02.006. p // g/ /j [30] Sanchez-Sotelo J, Haidukewych GJ, Boberg CJ. Hospital cost of dislocation after primary total hip arthroplasty. J Bone Joint Surg Am 2006;88:290e4. https:// doi.org/10.2106/JBJS.D.02799. [31] Tsutsui T, Goto T, Wada K, Takasago T, Hamada D, Sairyo K. Efﬁcacy of a computed tomography-based navigation system for placement of the acetabular component in total hip arthroplasty for developmental dysplasia of the hip. J Orthop Surg Hong Kong 2017;25:1e7. https://doi.org/10.1177/ 2309499017727954. [4] Patel NN, Shah JA, Erens GA. Current trends in clinical practice for the direct anterior approach total hip arthroplasty. J Arthroplasty 2019;34: 1987e1993.e3. https://doi.org/10.1016/j.arth.2019.04.025. [5] Labek G, Thaler M, Janda W, Agreiter M, St€ockl B. Revision rates after total joint replacement. J Bone Joint Surg Br 2011;93-B:293e7. https://doi.org/ 10.1302/0301-620X.93B3.25467. [5] Labek G, Thaler M, Janda W, Agreiter M, St€ockl B. Revision rates after total joint replacement. J Bone Joint Surg Br 2011;93-B:293e7. https://doi.org/ 10.1302/0301-620X.93B3.25467. [32] Mishra A, Verma T, null Rajkumar, Agarwal G, Sharma A, Maini L. 3D printed patient-speciﬁc acetabular jig for cup placement in total hip arthroplasty. In- dian J Orthop 2020;54:174e80. https://doi.org/10.1007/s43465-020-00061-2. [6] Karachalios T, Komnos G, Koutalos A. Total hip arthroplasty. EFORT Open Rev 2018;3:232e9. https://doi.org/10.1302/2058-5241.3.170068. P.L. Brush et al. / Arthroplasty Today 20 (2023) 101109 6 [37] Domb BG, El Bitar YF, Sadik AY, Stake CE, Botser IB. Comparison of robotic- assisted and conventional acetabular cup placement in THA: a matched-pair controlled study. References Clin Orthop 2014;472:329e36. https://doi.org/10.1007/ s11999-013-3253-7. [33] Spencer-Gardner L, Pierrepont J, Topham M, Bare J, McMahon S, Shimmin AJ. Patient-speciﬁc instrumentation improves the accuracy of acetabular component placement in total hip arthroplasty. Bone Jt J 2016;98-B:1342e6. https://doi.org/10.1302/0301-620X.98B10.37808. https://doi.org/10.1302/0301-620X.98B10.37808. [38] Connolly KP, Kamath AF. Direct anterior total hip arthroplasty: literature re- view of variations in surgical technique. World J Orthop 2016;7:38e43. https://doi.org/10.5312/wjo.v7.i1.38. p // g/ / [34] Nogler M, Mayr E, Krismer M, Thaler M. Reduced variability in cup posi- tioning: the direct anterior surgical approach using navigation. Acta Orthop 2008;79:789e93. https://doi.org/10.1080/17453670810016867. 2008;79:789e93. https://doi.org/10.1080/174536708100168 p // g/ / j [39] Voigt LF, Koepsell TD, Daling JR. Characteristics of telephone survey re- spondents according to willingness to participate. Am J Epidemiol 2003;157: 66e73. https://doi.org/10.1093/aje/kwf185. [35] Hohmann E, Bryant A, Tetsworth K. Accuracy of acetabular cup positioning using imageless navigation. J Orthop Surg 2011;6:40. https://doi.org/10.1186/ 1749-799X-6-40. p // g/ / j / [40] Ekhtiari S, Kay J, de Sa D, Simunovic N, Musahl V, Peterson DC, et al. What makes a successful survey? A systematic review of surveys used in anterior cruciate ligament reconstruction. Arthrosc J Arthrosc Relat Surg 2017;33: 1072e1079.e3. https://doi.org/10.1016/j.arthro.2017.01.032. [36] Redmond JM, Gupta A, Hammarstedt JE, Petrakos AE, Finch NA, Domb BG. The learning curve associated with robotic-assisted total hip arthroplasty. J Arthroplasty 2015;30:50e4. https://doi.org/10.1016/j.arth. 2014.08.003. P.L. Brush et al. / Arthroplasty Today 20 (2023) 101109 7 Appendix Appendix Supplemental Figure 1. Example image of how the 3D reconstruction and measuring software were able to determine acetabular cup inclination. Supplemental Figure 1. Example image of how the 3D reconstruction and measuring software were able to determine acetabular cup inclination. Supplemental Figure 2. Example image of how the 3D reconstruction and measuring software were able to determine acetabular cup anteversion. Supplemental Figure 2. Example image of how the 3D reconstruction and measuring software were able to determine acetabular cup anteversion.
https://openalex.org/W2163428311	https://angeo.copernicus.org/articles/22/3291/2004/angeo-22-3291-2004.pdf	English	null	Measurements of atmospheric turbulence with the dual-beamwidth method using the MST radar at Gadanki, India	Annales geophysicae	2,004	cc-by	5,312	Correspondence to: G. D. Nastrom (gdnastrom@stcloudstate.edu) Measurements of atmospheric turbulence with the dual-beamwidth method using the MST radar at Gadanki, India G. D. Nastrom1, P. B. Rao2, and V. Sivakumar3 1St. Cloud State University, St. Cloud, Minnesota 56379, USA 2National Remote Sensing Agency, Hyderabad 500037 AP, India 3Universite de la Reunion, Cassin-97715 St. Denis-C9, France Received: 29 August 2003 – Revised: 24 February 2004 – Accepted: 3 March 2004 – Published: 23 September 2004 G. D. Nastrom1, P. B. Rao2, and V. Sivakumar3 1St. Cloud State University, St. Cloud, Minnesota 56379, USA 2National Remote Sensing Agency, Hyderabad 500037 AP, India 3Universite de la Reunion, Cassin-97715 St. Denis-C9, France Received: 29 August 2003 – Revised: 24 February 2004 – Accepted: 3 March 2004 – Part of Special Issue “Equatorial and low latitude aeronomy” Abstract. A brief experiment was conducted during 24–29 April and 9–10 May 2002, using the MST radar at Gadanki, India, to test the dual-beamwidth method of estimating the turbulence kinetic energy (TKE). Because the beamwidth can be modiﬁed on only one polarization at a time at Gadanki, an elliptical beam was used with a modiﬁed dual- beamwidth analysis. Estimates of the TKE from the dual- beamwidth method and the traditional method are very sim- ilar in regions of light winds (<∼10 ms−1). In regions of stronger wind (>∼15 ms−1) the traditional method of- ten gives TKE<0 because the beam-broadening correction is greater than the observed spectral width. It is suggested that some of the problems with the traditional method are due to the uncertainty in the effective width of the radar beam. In all regions the modiﬁed dual-beamwidth method gives TKE>0 on the beam parallel to the prevailing wind; on this beam the estimates depend only on the ratio of the beamwidths, which is presumably well-known, and the observed spectral widths. The values of TKE from the dual-beamwidth method are ap- proximately constant with height at 0.2 m2s−2 from about 5 to 7.5 km during the afternoon during both April and May (all April observations were made between 9:00 and 17:00 local time), and then decrease rapidly to about 0.02 m2s−2 by about 9 km. The data from May extend over one full diur- nal period and the diurnal range of TKE during this period is found to be about 5 dB below about 12 km and from about 15 to 19 km, near the tropopause, with maximum values during local afternoon. Annales Geophysicae (2004) 22: 3291–3297 SRef-ID: 1432-0576/ag/2004-22-3291 © European Geosciences Union 2004 Annales Geophysicae (2004) 22: 3291–3297 SRef-ID: 1432-0576/ag/2004-22-3291 © European Geosciences Union 2004 G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method 3292 Fig. 1. (Upper) Zonal and meridional wind speeds for each day, April 2002. The plotted curves are the means of the daily medians for the two opposing beams in each plane and for 10◦and 15◦zenith angle; the symbols show the maximum and minimum values used for each mean. (Lower) Daily means of the observed spectral widths at 10◦and 15◦zenith angle. Fig. 1. (Upper) Zonal and meridional wind speeds for each day, April 2002. The plotted curves are the means of the daily medians for the two opposing beams in each plane and for 10◦and 15◦zenith angle; the symbols show the maximum and minimum values used for each mean. (Lower) Daily means of the observed spectral widths at 10◦and 15◦zenith angle. average power aperture product of 7×108 Wm2. The antenna array consists of 1024 crossed 3-element Yagi antennas cov- ering 130×130 m. Peak transmitted power is 2.5 MW ob- tained from 32 transmitters, each feeding a sub-array of 32 Yagis. The one-way half-power full beamwidth of the full antenna is about 2.9◦. to exercise the dual-beamwidth method was conducted us- ing the MST radar at Gadanki, India, during 24–29 April 2002 and 9–10 May 2002, and the results obtained are de- scribed in this paper. Atmospheric conditions over Gadanki, India, are highly repeatable from day-to-day during the pre- monsoon season. Therefore, even the relatively small data sample from this brief pilot study should describe the salient features of the vertical and diurnal variations of TKE for the pre-monsoon conditions. During this experiment two beamwidths were interleaved. The narrow beamwidth (2.9◦) was obtained using the full antenna. By disconnecting 16 subarrays from one polariza- tion of the antenna a rectangular antenna was formed, with 2.9◦beamwidth in one polarization and a second beamwidth, about 5.8◦, in the other polarization. The ratio of the broad and narrow beams is about 2. In practice, the outer 8 sub- arrays on each side of the antenna were disconnected to form the broad beam. Because the broad beam could be formed only in one polarization at a time, the resulting beam was elliptical (5.8◦by 2.9◦) and our application of the dual- beamwidth method, described below, will account for this ellipticity. The estimates of TKE from the traditional method and from a modiﬁed dual-beamwidth method are compared in this study. G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method Because the beamwidth of the Gadanki radar could be modiﬁed in only one polarization at a time, the dual- beamwidth estimates will be best when the beam is parallel to the prevailing wind, as explained below. Fortunately, dur- ing the period used here all of the strong prevailing winds were nearly parallel to the zonal beam. 1 Introduction The traditional method of estimating atmospheric turbulence kinetic energy (TKE) from spectral widths requires the appli- cation of correction factors to the measured spectral widths due to the interaction of the radar beam with the background wind (e.g. Atlas, 1964; Hocking, 1985; Nastrom, 1997). While these correction factors are well known in principle, estimates of them are sometimes larger than the observed spectral widths (Hocking, 1986; Fukao et al., 1994; Ku- rusaki et al., 1996; Nastrom and Eaton, 1997; Narayana Rao et al., 2001; Nastrom and Tsuda, 2001; Satheesan and Krishna Murthy, 2002), especially during relatively strong winds, which implies relatively large uncertainty in them or in the observed spectral widths, since the true corrections cannot be larger than the observed spectral widths. The dual-beamwidth method for measuring TKE using spectral widths from Doppler radar was recently introduced by VanZandt et al. (2002). The dual-beamwidth method em- ploys the spectral widths measured simultaneously with two different beamwidths and the ratio of the magnitudes of the beamwidths is used under the assumption that the TKE is the same in the sample volumes viewed by both beamwidths. With the dual beamwidth method it is not necessary to have independent estimates of the correction factors to be used or to know the actual magnitudes of the beamwidths used. The uncertainty of the TKE estimates from the dual-beamwidth method is governed only by the uncertainty of the observed spectral widths, suggesting that this method may provide a standard against which to compare estimates made by other methods. Key words. Meteorology and atmospheric dynamics (in- struments and techniques; turbulence; general or miscella- neous) Of course, the dual-beamwidth method can be used only at a small number of radar sites that have dual-beamwidth ca- pability. For example, VanZandt et al. (2002) used observa- tions taken in special experiments at the highly versatile MU radar in Japan, and Latteck et al. (2003) give results from a special experiment in Norway. A brief experiment designed G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method 2 Data Our observational strategy included a total of 10 beam positions each hour for each beamwidth: vertical and to- ward the 4 cardinal directions for 10◦and 15◦zenith an- gles. Range-resolution was 150 m from 3.6 to 24.9 km. The radar settings used were: interpulse period, 1000 µs; number The Gadanki radar is described in detail by others (e.g. Rao et al., 1995; Jain et al., 2000). Brieﬂy, it is an MST radar near Tirupati (13.47◦N, 79.18◦E) operating at 53 MHz with G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method 3293 Van Zandt et al. (2002) show that observations of the same sample volume with two conical beams of different beamwidth (θa and θb) can be used to solve a pair of simulta- neous equations to obtain σ 2 t as a function of σ 2 obs−a, σ 2 obs−b, and the ratio θ2 a /θ2 b , with no need to evaluate Eq. (2) directly. of incoherent integrations, 128; number of coherent integra- tions, 1; number of Doppler spectral points, 128; resolution per spectral point, 0.173 ms−1; dwell-time per proﬁle, 19 s. About 7 proﬁles were taken per hour per beam per zenith an- gle per beamwidth. A standard Gaussian ﬁtting method was used to ﬁnd the Doppler velocity, spectral width, and signal power from each Doppler spectrum. As discussed by Fukao et al. (1994), the uncertainties of the Doppler velocity and spectral width are proportional to the spectral width. Data from the vertical beam are subject to contamination by spec- ular echoes and thus are not used in this study. b The beam of the Gadanki radar is conical when the full an- tenna is used. When only half of the antenna is used for one polarization the beam is elliptical, being 5.8◦in the zenith direction and 2.9◦azimuthally. Accordingly, our analysis is modiﬁed from that of VanZandt et al. (2002). Equation (2) must be written separately for the two beamwidths as fol- lows (subscript H applies for the half-antenna and F for the full-antenna): Observations were made usually from 11:00–17:00 local time on 24, 25, 26, 27, and 29 April 2002. Another set of observations were taken over a diurnal cycle from 15:00, 9 May–15:00, 10 May 2002. All observations were subjected to a rigorous quality control including visual inspection of the proﬁles of velocity, signal-to-noise ratio, and spectral width. 2 Data Finally, hourly medians were formed and are used for the analyses below. σ 2 obs−H = σ 2 t + h θ2 H U2 cos2 α + shear + θ2 F V 2i. 4 ln 2, (4) σ 2 obs−F = σ 2 t + h θ2 F U2 cos2 α + shear + θ2 F V 2i. 4 ln 2, (5) where “shear”=−2UuzR cos α sin2 α + (uzR)2 sin4 α. S b ti E (5) f E (4) i (4) obs F σ 2 t + h θ2 F U2 cos2 α + shear + θ2 F V 2i. 4 ln 2, (5) Local weather conditions during late April and early May 2002 were very hot, with daily maximum temperatures near 40◦C every day (the Sun is directly overhead at noon at this latitude in late April). Surface winds were generally light and variable. Scattered cumulus and towering cumulus formed in the afternoon, although no signiﬁcant precipitation occurred at the radar site. where “shear”=−2UuzR cos α sin2 α + (uzR)2 sin4 α. Subracting Eq. (5) from Eq. (4) gives where “shear”=−2UuzR cos α sin2 α + (uzR)2 sin4 α. Subracting Eq. (5) from Eq. (4) gives U2 cos2 α + shear = 4 ln 2 " σ 2 obs−H −σ 2 obs−F θ2 H θ2 F −1 # . (6) (6) Using Eq. (6) in Eq. (4) gives σ 2 t = θ2 H θ2 F σ 2 obs−F −σ 2 obs−H θ2 H θ2 F −1 −θ2 F V 2 4 ln 2 . (7) 3 Method of analysis Following Nastrom (1997), σ 2 beam+shear can be approximated as σ 2 beam+shear = θ2 4 ln 2 h U2 cos2 α + V 2 −2UuzR cos α sin2 α + (uzR)2 sin4 α i , (2) σ 2 beam+shear = θ2 4 ln 2 h U2 cos2 α + V 2 −2UuzR cos α sin2 α + (uzR)2 sin4 α i , (2) (2) where U and V are the horizontal wind components paral- lel and perpendicular to the beam, θ is the one-way half- power half-width of the beam, α is the beam zenith angle, uz=dU/dz, and R is the range to the center of the sample vol- ume. Terms of order θ4 and (1R/R)2 are neglected, where 1R is the range resolution, a trigonometric identity has been used to obtain sin4α, and the vertical component of wind is ignored. In the “traditional method” estimates of σ 2 t are ob- tained from Eq. (1) after using the radar observations of wind to compute the right side of Eq. (2), i.e. 3 Method of analysis (7) The observed spectral width (σ 2 obs) represents the sum of the TKE per unit mass (σ 2 t ) and the effects of beam-, shear-, and wave-broadening The ﬁrst term on the right of Eq. (7) depends only on the observed spectral widths and the ratio of beamwidths. The second term on the right of Eq. (7) requires that we know θF , as well as V . In principle, the width of the transmitted beam is known and that value of θF should be used whenever the sample volume is uniformly ﬁlled with turbulence. When the sample volume is not ﬁlled, such as when one or more thin horizontal layers of intense turbulence are present, then the appropriate value for θF depends on the amount of beamﬁll- ing, and perhaps other things, and becomes uncertain (e.g. if only a single intense layer is present in the sample volume, then the effective beamwidth is less than θF and its value de- pends on the location of the intense layer within the sample volume). The uncertainty of the value of θF may explain why several of the past studies mentioned earlier have obtained results with σ 2 t <0 using the “traditional method”. The dual- beamwidth method, however, is not subject to this problem when V is negligible (i.e. in the beam parallel to the prevail- ing wind), as long as the ratio θ2 H θ2 F is constant. A major goal of this study will be to compare estimates of σ 2 t from Eq. (7) with those from the “traditional method” Eq. (3). σ 2 obs = σ 2 t + σ 2 beam+shear + σ 2 wave. (1) (1) Nastrom and Eaton (1997) found that in the troposphere σ 2 wave is negligible for radars with short-dwell observa- tions, such as those used here. Following Nastrom (1997), σ 2 beam+shear can be approximated as Nastrom and Eaton (1997) found that in the troposphere σ 2 wave is negligible for radars with short-dwell observa- tions, such as those used here. σ 2 t = σ 2 obs −σ 2 beam+shear. (3) σ 2 t = σ 2 obs −σ 2 beam+shear. 4 Vertical proﬁles The upper panels of Fig. 1 show the daily median winds ob- served at Gadanki during 24–29 April 2002. The plotted σ 2 t = σ 2 obs −σ 2 beam+shear. (3) σ 2 t = σ 2 obs −σ 2 beam+shear. (3) G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method 3295 Fig. 4. Ratio of the median σ 2t -Trad to the median σ 2t -2BW during (left) April and (right) May for observations taken at (upper) 10◦and (lower) 15◦zenith angles. Fig. 4. Ratio of the median σ 2t -Trad to the median σ 2t -2BW during (left) April and (right) May for observations taken at (upper) 10◦and (lower) 15◦zenith angles. curves are the mean of four results for each day: the east and west (north and south) beams and at 10 and 15◦zenith an- gles. The symbols show the maximum and minimum value among the four results; in general there is very little varia- tion between the maximum and minimum for each day; i.e. the differences between winds on opposed beams and at 10◦ and 15◦zenith angles appear to be insigniﬁcant, except near about 18 km on some days. The ﬂow is characterized as northeasterly (i.e. southwestward) below about 7 km, south- westerly from about 10–15 km after 24 April and easterly above about 17 km. While the declining signal-to-noise ratio with height leads to greater variability above about 17 km, there appears to be useful wind data at all heights in most cases. right) from the 10◦zenith angle observations (results at 15◦ are nearly identical and are not shown). Both periods show northeasterly winds below about 7 km. Above about 10 km the meridional winds in May are very light while the zonal winds continue to increase with height to a maximum speed near 25 ms−1 at 17 km. Above about 20 km winds during the two periods are again very similar. Error bars to show the 95% conﬁdence limits extend ± one standard error of the mean (2σ/√No, where σ is the standard deviation of the No hourly medians used) from the curves for the E and N beams. The lower panels of Fig. 2 show σ 2 obs during April (lower left) and May (lower right). Except at about 5–7 km and perhaps 17–19 km, where the meridional beams’ values are greater than the zonal beams’, there is no signiﬁcant differ- ence among σ 2 obs for the four beams during April. During May the meridional beams’ σ 2 obs values slightly exceed the zonal beams’ from about 11–17 km. G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method 3294 Fig. 2. Median (upper) winds and (lower) observed spectral widths over all observations taken at 10◦zenith angle during (left) April and (right) May. Bars extending ±2σ/√No are entered at representative heights along the E and N curves. Dashed vertical lines in the lower panels show the minimum spectral resolution for Doppler half-width. Fig. 2. Median (upper) winds and (lower) observed spectral widths over all observations taken at 10◦zenith angle during (left) April and (right) May. Bars extending ±2σ/√No are entered at representative heights along the E and N curves. Dashed vertical lines in the lower panels show the minimum spectral resolution for Doppler half-width. Fig. 3. Median (upper) σ 2t -2BW and (lower) σ 2t -Trad over all observations taken at 10◦zenith angle during (left) April and (right) May. Bars extending ±2σ/√No are entered at representative heights along the E and N curves. Fig. 3. Median (upper) σ 2t -2BW and (lower) σ 2t -Trad over all observations taken at 10◦zenith angle during (left) April and (right) May. Bars extending ±2σ/√No are entered at representative heights along the E and N curves. G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method The dashed vertical lines show the Doppler resolution for spectral width as used here; in April (May) the values near 14 km and above about 19 km (22 km) are near the resolution limit. The lower panels of Fig. 1 are similar to the upper pan- els, except they show the daily mean σ 2 obs. In general, there is no clear pattern of differences among the individual days, except near 12 km, where both wind components are very light on 24 April and the associated σ 2 obs is relatively small. Apparently, the larger wind speeds lead to larger σ 2 obs due to beam broadening and, perhaps, stronger turbulence. Ghosh et al. (2003) also found that for relatively light wind speeds there is not a strong effect of wind speed on σ 2 obs at Gadanki. Because there is relatively little change from day to day, we will combine all days in April together in the analyses to fol- low. The upper panels of Fig. 3 show σ 2 t estimated using Eq. (7), called σ 2 t -2BW. The lower panels show σ 2 t esti- mated using Eq. (3), called σ 2 t -Trad. Below about 7 km (8 km) during April (May) results from both methods are rel- atively large, ranging between about 0.1 and 0.3 m2s−2, cor- responding to the altitudes of active convection. Values for the west beam are consistently smaller than those from the other beams in April below 6 km; from 6 to 7.5 km values Figure 2 shows the proﬁles of median wind speed dur- ing the entire period in April (upper left) and in May (upper G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method 3296 0 4 8 12 16 20 24 -2.5 -2 -1.5 -1 -0.5 log σt 2-2BW (m2s-2) (10o) 10o Z4-7.5km Z8.5-12 Z12.5-15.5 Z15.5-19 Z19-24 M4-7.5km M8.5-12 M19-24 Hour (LT) 0 4 8 12 16 20 24 -2.5 -2 -1.5 -1 -0.5 Hour (LT) log σt 2-2BW (m2s-2) (15o) 15o Fi Fig. 5. Hourly medians during 9–10 May of σ 2t -2BW over layers in the troposphere and lower stratosphere in the zonal (Z) and merid- ional (M) planes. Results for the meridional plane at 12–19 km are negative due to strong zonal winds and are not shown. 5 Diurnal changes 0 4 8 12 16 20 24 -2.5 -2 -1.5 -1 -0.5 Hour (LT) log σt 2-2BW (m2s-2) (15o) 15o Figure 5 shows the hourly march of σ 2 t -2BW during 9–10 May for the median values over ﬁve vertical layers (4–7.5, 8.5–12, 12–15.5, 15.5–19, and 19–24 km) for beams in the zonal and meridional planes at 10◦(15◦) in the upper (lower) panel. Results for the meridional beam at 12–19 km are not shown as they are generally negative due to the strong zonal winds in this region. The curves were smoothed with a 1/4- 1/2-1/4 ﬁlter. Diurnal changes with ranges of about 5 dB are seen in the layers below 12 km and at 15.5–19 km. The results show very little variation among zenith angles. At 4–7.5 km the maximum (minimum) values occur near local noon (mid- night), suggesting the enhanced TKE is related to a daytime process such as convection. At 8.5–12 km and at 15.5–19 km the curves show maxima (minima) in late afternoon (early morning). The latter maxima may be caused by the deeper convection that appears only in the afternoon. Curves at other heights have indistinct diurnal cycles. Fig Fig. 5. Hourly medians during 9–10 May of σ 2t -2BW over layers in the troposphere and lower stratosphere in the zonal (Z) and merid- ional (M) planes. Results for the meridional plane at 12–19 km are negative due to strong zonal winds and are not shown. Bars extend- ing ±2σ/√No are entered at representative hours along the Z4–7.5 and Z8.5–12 (Z15.5–19) curves in the upper (lower) panel. Narayana Rao et al. (2001) present a climatology of eddy dissipation rates (ε) at Gadanki based on σ 2 t -Trad from the radar observations combined with semi-daily radiosonde proﬁles of the Brunt-V¨ais¨al¨a frequency (N). The diurnal changes in ε that they give must be largely due to changes in σ 2 t -Trad, since diurnal changes in N are relatively small above the planetary boundary layer. During the season March–May they ﬁnd the diurnal range of ε (and thus of σ 2 t -Trad) is about 5 dB in the upper troposphere and lower stratosphere, similar to Fig. 5. However, the time of daily maximum they give is at night, from 22–04 LT, in contrast to the daytime maxima in Fig. 5. 5 Diurnal changes We ﬁnd this difference in time of maximum perplexing and suggest it should be investigated more closely when a larger data set becomes available. for both beams in the zonal plane are smaller than those in the meridional plane. During May there is no apparent anisotropy at the lower altitudes. Above 7 or 8 km all values become less than 0.02 m2s−2 at all altitudes except in the re- gions of strong winds (discussed below); the σ 2 t -Trad values are slightly smaller than those for σ 2 t -2BW. Note that negative values, where the corrections are larger than σ 2 obs, are included in Fig. 3. The May results near 17 km illustrate the differing effects of the correction terms in Eqs. (3) and (7); i.e. for σ 2 t -Trad the beam-broadening corrections due to the large zonal wind speeds cause val- ues for all four beams to be negative while for σ 2 t -2BW only the meridional beams’ values are negative. A similar, but smaller, effect is seen near 18 km during April. σ 2 t -2BW may be expected to be more reliable as it is not inﬂuenced by thin-layer effects or other beam-ﬁlling problems. The curves in Fig. 3 are very similar where the winds are less than about 10 m/s, i.e. except from about 16–19 km during April and 12–19 km during May. G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method Bars extend- ing ±2σ/√No are entered at representative hours along the Z4–7.5 and Z8.5–12 (Z15.5–19) curves in the upper (lower) panel. Figure 4 compares the σ 2 t -Trad and σ 2 t -2BW results for the 10◦(15◦) zenith angle in the upper (lower) panels. During May the ratios are near unity below about 8 km, correspond- ing to the altitudes where σ 2 t appears to be enhanced due to convection (Fig. 3). However, during April the region where the ratios are near unity extends much higher, to over 11 km. Above about 20 km the ratios are again near unity (within 1 or 2 dB), although this may be explained by the spectral widths falling to their minimum detectable signal as noted earlier. There is very little difference between the results at 10◦and 15◦zenith angle. 0 4 8 12 16 20 24 -2.5 -2 -1.5 -1 -0.5 log σt 2-2BW (m2s-2) (10o) 10o Z4-7.5km Z8.5-12 Z12.5-15.5 Z15.5-19 Z19-24 M4-7.5km M8.5-12 M19-24 Hour (LT) 6 Summary and conclusions A brief experiment was conducted with the MST radar at Gadanki, India, to test the dual-beamwidth method for es- timating TKE. Because the antenna size could be changed for only one polarization at a time the resulting beam was elliptical rather than conical. Theoretical analysis showed that in the dual-beamwidth method for an elliptical beam The data above about 20 km appear nearly constant with height. This may reﬂect the lower bound of σ 2 t detectable with the Gadanki radar. 3297 G. Nastrom et al.: Measurements of atmospheric turbulence with the dual-beamwidth method there is a term that depends on the windspeed perpendicu- lar to the beam (the ﬁnal term in Eq. 7). During the April period the winds were relatively light at all altitudes, and there was relatively good agreement between results from the dual-beamwidth method and the traditional method. How- ever, during the May period, from about 14 to 18 km, the zonal winds were relatively strong and the meridional winds were nearly calm. Our estimates of TKE during May by the dual-beamwidth method for the zonal beam are positive while those from the meridional beam and from the tradi- tional method are negative, giving a good illustration of the ability of this method to extract TKE during strong winds (at least in the beam parallel to the wind) when the traditional method fails. Acknowledgements. G. D. Nastrom was partially supported by the National Science Foundation ATM-0129464 and the Air Force Ofﬁce of Scientiﬁc Research F496200210167. References Atlas, D.: Advances in radar meteorology, edited by Landsberg, H. and van Mieghem, J., Academic, Adv. Geophys., 10, 317–478, 1964. Fukao, S., Yamanaka, M. D., Ao, N., Hocking, W. K., Sato, T., Yamamoto, M., Nakamura, T., Tsuda, T., and Kato, S.: Seasonal variablity of vertical eddy diffusivity in the middle atmosphere, 1. Three-year observations by the middle and upper atmosphere radar, J. Geophys. Res., 99, 18 973–18 987, 1994. Ghosh, A. K., Jain, A. R., and Sivakumar, V.: Simultaneous MST radar and radiosonde measurements at Gadanki (13.5◦N, 79.2◦E), 2. Determination of various atmospheric turbulence pa- rameters, Radio Sci., 38, (1), 1014, doi:10.1029/2000RS002528, 2003. Hocking, W. K.: Measurement of turbulent energy dissipation rates in the middle atmosphere by radar techniques: a review, Radio Sci., 20, 1403–1422, 1985. The following speciﬁc points have been noted: 1. Winds and turbulence from 10◦and 15◦zenith angles are the same. Hocking, W. K.: Observation and measurement of turbulence in the middle atmosphere with a VHF radar, J. Atmos. Terr. Phys., 48, 655–670, 1986. Jain, A. R., Narayana Rao, D., and Rao, P. B.: Indian MST radar – An overview of the scientiﬁc programmes and results, Indian J. Radio & Space Phys., 29, 149–171, 2000. 2. Winds and turbulence from opposing beams in the same plane are the same. 2. Winds and turbulence from opposing beams in the same plane are the same. Kurosaki, S., Yamanaka, M. D., Hashiguchi, H., Sato, T., and Fukao, S.: Vertical eddy diffusivity in the lower and middle at- mosphere: A climatology based on the MU radar observations during 1986–1992, J. Atmos. Terr. Phys., 58, 727–734, 1996. 3. During light wind conditions, σ 2 t from the modiﬁed dual-beamwidth method and the traditional method are the same. During strong wind conditions (over about 15 ms−1) the modiﬁed dual-beamwidth method contin- ues to give realistic estimates in the beam parallel to the wind while for the other plane, and for both planes with the traditional method, the corrections are larger than the observed spectral width. Latteck, R., Singer, W., and Engler, N.: Application of the dual- beamwidth method to a narrow beam MF radar for estimation of spectral width, Proceedings, MST10 Workshop on technical and scientiﬁc aspects of MST radar, Piura, Peru, 90, 2003. Narayana Rao, D., Narayana Rao, T., Venkataratnam, M., Thulasir- aman, S., Rao, S. V. B., Srinivasulu, P., and Rao, P. B.: Diurnal and seasonal variablity of turbulence parameters observed with Indian mesosphere-stratosphere-troposphere radar, Radio Sci., 36, 1439–1457, 2001. 4. Values of σ 2 t are about 10−1 m2s−2 in the troposphere (3.6–7.5 km), and about 10−2 above about 9 km. Values above 12 km are slightly larger during the May period when winds were stronger. Nastrom, G. D.: Doppler radar spectral width broadening due to beamwidth and wind shear, Ann. Geophys., 15, 786–796, 1997. Nastrom, G. D. and Eaton, F. D.: Turbulence eddy dissipation rates from radar observations at 5–20 km at White Sands Mis- sile Range, New Mexico, J. Geophys. Res., 102, 19 495–19 506, 1997. 5. The diurnal range of σ 2 t is about 5 dB below about 12 km. Maximum values occur near local noon at 4.5– 7 km and several hours later at 8.5–12 km. The diurnal cycle at 15.5–19 km is very similar to that at 8.5–12 km. Nastrom, G. D. and Tsuda, T.: Anisotropy of Doppler spectral pa- rameters in the VHF radar observations at MU and White Sands, Ann. Geophys., 19, 883–888, 2001. Rao, P. B., Jain, A. R., Kishore, P., Balamuralidhar, P., Damle, S. H., and Viswanathan, G.: Indian MST radar, 1. System description and sample vector wind measurements in ST mode, Radio Sci., 30, 1125–1138, 1995. Acknowledgements. G. D. 2. Winds and turbulence from opposing beams in the same plane are the same. Nastrom was partially supported by the National Science Foundation ATM-0129464 and the Air Force Ofﬁce of Scientiﬁc Research F496200210167. Acknowledgements. G. D. Nastrom was partially supported by the National Science Foundation ATM-0129464 and the Air Force Ofﬁce of Scientiﬁc Research F496200210167. Topical Editor U.-P. Hoppe thanks S. Cohn and another referee for their help in evaluating this paper. Satheesan, K. and Krishna Murthy, B. V.: Turbulence parameters in the tropical troposphere and lower stratosphere, J. Geophys. Res., 107, (D1), ACL 2–1, doi:101029/2000JD000146, 2002. VanZandt, T. E., Nastrom, G. D., Furumoto, J., Tsuda, T., and Clark, W. L.: A Dual-Beamwidth Method for observing atmospheric turbulence intensity with radar, Geophys. Res. Lettr., 29, (12), doi:10.1029/2001GL014283, 2002.
https://openalex.org/W4233820846	http://www.ajnr.org/content/ajnr/38/9/E64.full.pdf	English	null	<i>Reply:</i>	American journal of neuroradiology	2,017	cc-by	541	http://dx.doi.org/10.3174/ajnr.A5261 of October 23, 2024. This information is current as Reply: S.A. Sajjadi and P.J. Nestor http://www.ajnr.org/content/38/9/E64 https://doi.org/10.3174/ajnr.A5261 doi: 2017, 38 (9) E64 AJNR Am J Neuroradiol Reply: S.A. Sajjadi and P.J. Nestor http://www.ajnr.org/content/38/9/E64 https://doi.org/10.3174/ajnr.A5261 doi: 2017, 38 (9) E64 AJNR Am J Neuroradiol of October 23, 2024. This information is current as REPLY: essentially becomes a forced-choice paradigm; this, in turn, may artificially inflate the accuracy compared with a real-world clinical environment in which the differential diagnosis is more open-ended. W e thank Dr Matias-Guiu et al for their interest in our study on the role of MR imaging in the syndromic classification of primary progressive aphasia (PPA).1 Their proposal that FDG- PET might be a more sensitive method than structural MR imag- ing is highly plausible. We also agree with Matias-Guiu et al that direct comparison would be necessary to definitively resolve the issue. For instance, both structural MR imaging and FDG-PET measure neurodegeneration. The greater the degree of neurode- generation, therefore, the more likely it is to be detectable with these types of imaging. In other words, the sensitivity of any tech- nique that operates through detecting degeneration is also a func- tion of disease severity. To this end, it is possibly relevant that their PPA group as a whole was more advanced than those with PPA in our series—mean Mini-Mental State Examination and Adden- brookes cognitive examination scores of 18.5 8.1 and 46.1 22.0,2 respectively, compared with scores of 22.1 3.9 and 56.4 14.1.1 Whether this confounder offers a credible explanation for the apparent increased sensitivity of FDG-PET of Matias-Guiu et al is perhaps debatable, but nonetheless important to address with a direct comparison. W Finally, as noted by Matias-Guiu et al, a new study could also assess the potential benefit of the combination of FDG-PET and MR imaging. To this end, we would add the possibility that the decision of whether to combine them is likely to vary according to the precise clinical question. For instance, our study showed that MR imaging returns near-perfect accuracy in detecting the lesions of semantic-variant PPA, making FDG-PET redundant in this scenario. In contrast, the terrible sensitivity yet good specificity for nonfluent and logopenic PPA with MR imaging suggest that a hierarchic algorithm in which one proceeds to FDG-PET if the MR imaging is nonspecific might be sensible. E64 Letters Sep 2017 www.ajnr.org REFERENCES 1. Sajjadi SA, Sheikh-Bahaei N, Cross J, et al. Can MRI visual assessment differentiate the variants of primary-progressive aphasia? AJNR Am J Neuroradiol 2017 Mar 24. [Epub ahead of print] CrossRef Medline 2. Matías-Guiu JA, Cabrera-Martín MN, Pe´rez-Castejo´n MJ, et al. Visual and statistical analysis of 18F-FDG PET in primary progressive apha- sia. Eur J Nucl Med Mol Imaging 2015;42:916–27 CrossRef Medline A further methodologic feature that we would strongly advo- cate in designing a future study is the inclusion of a good number of negative (healthy age-matched) and positive (non-PPA degen- erative dementia) controls. If raters know a priori that the scans they are evaluating come from a few predefined groups, the rating X S.A. Sajjadi Department of Neurology University of California, Irvine Irvine, California X P.J. Nestor German Center for Neurodegenerative diseases Magdeburg, Germany 64 Letters Sep 2017 www.ajnr.org E64
https://openalex.org/W2307879930	https://rvc-repository.worktribe.com/preview/1583418/Spatial%20Distribution%20And%20Risk%20Factors%20For%20Human%20Cysticercosis%20In%20Colombia.pdf	English	null	Disorganization of the splenic microanatomy in ageing mice	Immunology	2,016	cc-by	9,704	Spatial distribution and risk factors for human cysticercosis in Colombia Erika Galipó1,2, Matthew A. Dixon3,4,5* , Claudio Fronterrè6, Zulma M. Cucunubá3,4,8, Maria‑Gloria Basáñez3,4, Kim Stevens2, Astrid Carolina Flórez Sánchez7 and Martin Walker2,3 Parasites & Vectors Parasites & Vectors Galipó et al. Parasites & Vectors (2021) 14:590 https://doi.org/10.1186/s13071-021-05092-8 Parasites & Vectors RESEARCH Spatial distribution and risk factors for human cysticercosis in Colombia Erika Galipó1,2, Matthew A. Dixon3,4,5* , Claudio Fronterrè6, Zulma M. Cucunubá3,4,8, Maria‑Gloria Basáñez3,4, Kim Stevens2, Astrid Carolina Flórez Sánchez7 and Martin Walker2,3 Open Access Abstract Background: Cysticercosis is a zoonotic neglected tropical disease (NTD) that affects humans and pigs following the ingestion of Taenia solium eggs. Human cysticercosis poses a substantial public health burden in endemic countries. The World Health Organization (WHO) aims to target high-endemicity settings with enhanced interventions in 17 countries by 2030. Between 2008 and 2010, Colombia undertook a national baseline serosurvey of unprecedented scale, which led to an estimated seroprevalence of T. solium cysticercus antibodies among the general population of 8.6%. Here, we use contemporary geostatistical approaches to analyse this unique dataset with the aim of under‑ standing the spatial distribution and risk factors associated with human cysticercosis in Colombia to inform how best to target intervention strategies. Methods: We used a geostatistical model to estimate individual and household risk factors associated with sero‑ positivity to T. solium cysticercus antibodies from 29,253 people from 133 municipalities in Colombia. We used both independent and spatially structured random effects at neighbourhood/village and municipality levels to account for potential clustering of exposure to T. solium. We present estimates of the distribution and residual correlation of seropositivity at the municipality level. Results: High seroprevalence was identified in municipalities located in the north and south of Colombia, with spatial correlation in seropositivity estimated up to approximately 140 km. Statistically significant risk factors associ‑ ated with seropositivity to T. solium cysticercus were related to age, sex, educational level, socioeconomic status, use of rainwater, consumption of partially cooked/raw pork meat and possession of dogs. Conclusions: In Colombia, the distribution of human cysticercosis is influenced by socioeconomic considerations, education and environmental factors related to the spread of T. solium eggs. This information can be used to tailor national intervention strategies, such as targeting spatial hotspots and more highly exposed groups, including dis‑ placed people and women. Large-scale seroprevalence surveys accompanied by geospatial mapping are an essential step towards reaching the WHO’s 2021‒2030 NTD roadmap targets. Keywords: Taenia solium, Cysticercosis, Risk factors, Spatial analysis, Geostatistics, Colombia Correspondence: m.dixon15@imperial.ac.uk 5 Schistosomiasis Control Initiative (SCI) Foundation, Edinburgh House, 170 Kennington Lane, Lambeth, London SE11 5DP, UK Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativeco mmons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Page 2 of 15 Galipó et al. Parasites & Vectors (2021) 14:590 0 100 200 300 400 500 600 km N Random Effects U^ −1 0 1 2 s U^ 0 100 200 300 400 500 600 km faeces. Eggs hatch in the pig’s digestive system, and the released oncospheres first penetrate the intestinal wall, entering the bloodstream, and then become encysted in striated muscle, brain, liver and subcutaneous and other tissues. Porcine cysticercosis is often asymp- tomatic [2, 3], although cysts in pig brain tissue can cause neurocysticercosis (NCC) and epileptic seizures [4] Study designh The data were collected by the Colombian National Health Institute (Instituto National de Salud) between 2008 and 2010 with the aim of estimating T. solium human cysticercosis antibody seroprevalence and asso- ciated risk factors. Details of the original data collection can be found in [14]. Briefly, individuals aged from 2 to 64 years, from 23 departments and Bogotá district, liv- ing in 133 municipalities with > 5000 inhabitants and a health centre were eligible for inclusion. The small pro- portion of total municipalities sampled (133/1122) was due to logistical and financial constraints. A three-stage cluster random sampling approach was used, covering 23 out of Colombia’s 32 departments (first administra- tive level unit) and Bogotá district (Additional file 1: Figure S1). The municipality constituted the primary sample unit (PSU) and was stratified according to level of urbanization, rural and urban population composi- tion and the Unsatisfied Basic Needs Index (Indice de Necesidades Básicas Insatisfechas) [15]. Within each stratum, the secondary sample unit (SSU) was defined as a neighbourhood (urban) or village (rural) with > 10 households and selected by random sampling. Finally, 10 households in each SSU were randomly selected, and one person belonging to each household (between the age of 2 and 64 years) was selected at random from those present at the interview. Following informed consent, finger-prick blood samples were obtained from 29,360 participants, and each sample was assessed for the pres- ence of circulating T. solium cysticercus antibodies at the National Health Institute Reference Laboratory (Labo- ratorio de Parasitología del Instituto Nacional de Salud) by enzyme-linked immunosorbent assay (ELISA), with a reported sensitivity of 100% and specificity of 97.5% [16]. Participants also completed a questionnaire on sociode- mographic information, hygiene habits, health condi- tions, food consumption habits, living conditions and animal ownership and management. The questionnaire was developed by the research team in Colombia, with input from experts on cysticercosis. It was first tested in Taeniasis/cysticercosis is widely endemic globally. Taenia solium cysticercosis antibody seroprevalence, indicative of exposure, ranges from 1.8 to 31.2% in Latin America, from 12.6 to 19.2% in Asia and from 7.7 to 34.5% in Africa (as measured using an enzyme-linked Immunoelectrotransfer blot [EITB] assay) [9], which highlights substantial variation in exposure to T. solium eggs across settings. NCC is responsible for the predomi- nant disease burden associated with T. Background The zoonotic tapeworm, Taenia solium, is responsi- ble for taeniasis/cysticercosis which is included in the World Health Organization’s (WHO’s) list of priori- tised neglected tropical diseases (NTDs) [1]. Humans are the definitive hosts of T. solium and harbour the adult tapeworm in their bowel. Pigs are intermediate hosts, infected by larval cysts (cysticerci) following ingestion of parasite eggs and proglottids [2] in human Galipó et al. Parasites & Vectors (2021) 14:590 Page 3 of 15 Galipó et al. Parasites & Vectors (2021) 14:590 Humans contract taeniasis following consumption of tissue cysts in poorly cooked pork meat. Taeniasis is usually asymptomatic, but mild symptoms, including abdominal pain, distension, diarrhoea and nausea, may appear [2]. Humans can also be infected with T. solium eggs, typically from ingestion of food contaminated with human faecal material [5] or food washed with con- taminated water [6]. Internal auto-infestation following regurgitation of proglottids in the stomach has also been suggested as an additional route of infection [2, 5, 7]. Infection with T. solium eggs causes cysticercosis which manifests most severely when cysts migrate to the central nervous system, resulting in NCC [2]. Morbidity from NCC associated with seizures, epilepsy and other neuro- logical sequelae is driven by the number and location of cysts or following the degeneration of viable cysts [8]. distribution of T. solium cysticercus seropositivity in Colombia, as well as individual and household risk fac- tors associated with exposure to the parasite. This work extends the original analysis of these data [14] by inte- grating the effects of individual covariates and spatial clustering at multiple hierarchical levels within a single statistical framework. We present maps of the spatial distribution of T. solium cysticercus seropositivity in Colombia, estimates of spatial correlation and demo- graphic, socioeconomic, behavioural and other risk fac- tors associated with exposure to this zoonotic NTD. Study designh solium infection, accounting for approximately 30% of epilepsy cases in endemic countries and 3% globally [10]. In addition, this zoonosis impacts the pork meat market, with small pro- ducers experiencing economic losses due to the reduc- tion in value of infected pork meat [4] and a market shift towards home slaughtering and selling [11]. In Colombia, taeniasis/cysticercosis poses a substantial public health problem [12], with an estimated life-time prevalence of epilepsy of 20.9 per 1000 individuals and a prevalence of neurocysticercosis (by computed tomog- raphy scan) of 13.9% [13]. The country-wide prevalence of T. solium cysticercus antibodies was estimated at 8.6% from a national serosurvey of more than 29,000 peo- ple conducted between 2008 and 2010 [14]. Despite the unprecedented scale of this epidemiological survey—and the development by the Pan American Health Organiza- tion in 2015 of a formal plan of surveillance and control in Colombia [12]—there has been little implementation of systematic surveillance or intervention activities. Con- sequently, the epidemiology of T. solium in Colombia is unlikely to have changed substantively during the past decade since these data were generated. Thus, the dataset remains the most comprehensive and relevant country- wide cross-sectional ‘snapshot’ of T. solium epidemiol- ogy anywhere across the globe and a unique information resource. Here, we analyse this dataset using a contempo- rary geostatistical approach to understand the spatial Page 4 of 15 Galipó et al. Parasites & Vectors (2021) 14:590 Galipó et al. Parasites & Vectors (2021) 14:590 a pilot survey carried out in 216 homes in the munici- pality of Caqueza (Department of Cundinamarca), from 28 August to 2 September 2008 and adjusted accord- ingly. Teams in the field were trained on the use of the questionnaire before it was applied on the whole sample. Details on the cleaning and coding of this dataset can be found in Additional file 1: Text S1. the empirical variogram was computed only on U , the estimated random effects at the municipality level. A Monte Carlo test for the null hypothesis of spatial inde- pendence was performed based on 10,000 random per- mutations of U amongst the sampled municipalities. The variograms computed on the permuted random effects represent the sampling distribution of the estimated variogram in the absence of spatial correlation. Incorporating spatial structure In the presence of spatial correlation, the independent ran- dom effects at the municipality level, U , were replaced with a set of spatially structured random effects, S(x) , where x is a vector with the centroids of the sampled municipalities. S(x) is a spatial Gaussian process with variance σ 2 and correlation functionρ(µ) = exp −µ ϕ , where µ is the distance between a pair of municipality centroids and ϕ is a parameter that controls the rate at which the spatial correlation decays with increasing distance. Conditional on these spatially structured random effects, the observations can still be considered as independent Bernoulli random variables [18]. The spatially structured model was fitted using the integrated nested Laplace approximation (INLA) and stochastic partial differ- ential equation (SPDE) approaches [19, 20] which implement approximate Bayesian inference in a computationally less intensive manner to alternative Markov chain Monte Carlo (MCMC) approaches. A flat Gaussian prior with mean and precision equal to zero was assigned to the model intercept term; other fixed effects were assigned independent vague Gaussian priors with mean zero and precision equal to 0.001. For the precision of the independent neighbourhood/village random effects, 1/τ , a vague Gamma prior was used, and for the parameters σ 2 and ϕ of the spatially-structured random effects, we adopted penalised complexity priors [21]. Adjusted ORs and 95% credible intervals (95% CrIs) were obtained for each risk factor from the final fitted model. The generic structure of all models is given by: (1) Y ∼Bern(µ), logit(µ) = βX + Z + U, Z ∼N(0, τ), U ∼N(0, σ), (1) where Y is a binary vector of observations indicating whether an individual tested positive for T. solium cysti- cercus antibodies, assuming a Bernoulli distribution; µ is a vector of probabilities for testing positive; β is a vector of regression coefficients, and X is the design matrix of explanatory variables; U and Z are vectors of independ- ent and normally distributed random effects terms asso- ciated with municipalities and neighbourhoods/villages, respectively; and σ and τ are the standard deviations of the respective random effects terms (indicative of the degree of variability at each hierarchical level). From the final fitted models, adjusted odds ratios (ORs), 95% con- fidence intervals (95% CIs) and P-values were obtained for each risk factor. All notations/parameters are sum- marised in Additional file 1: Table S1. Incorporating spatial structure A sub-analysis on risk factors in those individuals owning pigs (n = 3154) was also conducted (methodological details are given in Additional file 1: Text S1).if Study designh If the empirical variogram ordinates fall outside of the 95% CI obtained from the Monte Carlo test, then there is some evidence of spatial correlation at municipality level. Model‑building and analysis of residual spatial correlation Before performing the geospatial analysis, an initial exploratory analysis was undertaken (using R version 4.0.5 [17]). Given the clustered nature of the data, a hierarchical univariate mixed-effects logistic regression model was fitted to test the association between each explanatory variable (covariate) and human seropositiv- ity to T. solium cysticerci, with each model including two independent random effects terms to capture correlation at the municipality and neighbourhood/village (depend- ing on urban or rural location) levels. Explanatory varia- bles with a P-value ≤ 0.25 (a conservative cut-off to avoid missing potentially important variables), derived from a likelihood ratio test, were retained in the subsequent hierarchical multivariable mixed-effects logistic regres- sion model.h Study population and seroprevalence distribution Study population and seroprevalence distribution Of the 29,360 observations, 29,253 (99.6%) observations were kept for analysis, with 107 removed due to miss- ing covariate values. Participants were mostly located in urban areas (77.9%), mostly aged 21‒50 years (64.4%) and mostly women (68.5%); the main occupational activ- ity was housewife/houseman (44.5%). Socioeconomic stratum 1 (lowest of 4 socioeconomic strata, excluding displaced people) was the most frequently represented socioeconomic stratum (49.5%), and participants most frequently had a partial or complete secondary school i Following fitting of the multivariable mixed-effects model, a variogram analysis was performed to assess the presence of residual spatial correlation [17]. Since the geographical coordinates were available only for the municipalities and not for the neighbourhoods/villages, Page 5 of 15 Galipó et al. Socioeconomic status (stratum) Study population and seroprevalence distribution Parasites & Vectors (2021) 14:590 CI Confidence interval Table 1 (continued) Covariate Level Number (%) of respondents Total number (%; 95% CI) of positive respondents Hygiene practices Washing vegetables No consumption 82 (0.28) 8 (9.76; 4.31–18.30) Always 7304 (24.97) 588 (8.05; 7.43–8.69) Occasionally 10,664 (36.46) 1461 (13.7; 13.00–14.30) Never 11,166 (38.17) 718 (6.43; 5.98–6.90) Washing hands before a meal Always 12,925 (44.18) 1171 (9.06; 8.58–9.57) Occasionally 14,950 (51.11) 1489 (9.96; 9.48–10.40) Never 1376 (4.70) 115 (8.36; 6.95–9.95) Elimination of excreta Sanitary conditions 20,401 (69.74) 1526 (7.48; 7.12–7.85) In waterway 543 (1.86) 50 (9.21; 6.91–12.00) Latrine with well 6403 (21.89) 862 (12.9; 12.1–13.7) Latrine without well 519 (1.78) 67 (12.9; 10.1–16.1) Open field 1084 (3.71) 270 (24.9; 22.3–27.5) Owning animals Cattle Not owning 27,947 (95.54) 2641 (9.45; 9.11–9.80) Owning 1301 (4.45) 134 (10.3; 8.74–12.1) Cats Not owning 23,590 (80.64) 2208 (9.36; 8.99–9.74) Owning 5670 (19.38) 567 (10.0; 9.26–10.8) Dogs Not owning 16,194 (55.36) 1349 (8.33; 7.91–8.77) Owning 13,083 (44.72) 1426 (10.9; 10.4–11.5) Birds Not owning 20,213 (69.10) 1799 (8.90; 8.51–9.30) Owning 9037 (30.89) 976 (10.80; 10.20–11.50) Pigs Not owing 26,105 (89.24) 2386 (9.14; 8.80–9.50) Owning ≤ 10 pigs 2705 (9.25) 349 (12.9; 11.70–14.30) Owning > 10 pigs 460 (1.57) 40 (8.70; 6.28–11.7) CI Confidence interval educational level (45.8%) (Table 1). The mean seropreva- lence of T. solium cysticercus antibodies was 9.6%, rang- ing from 0.5% in the Department of Caldas to 38.7% in the Department of Vaupés (Additional file 1: Table S2). Municipalities with the highest seroprevalence were located in the north and south of Colombia (Fig. 1), while municipalities with lower seroprevalence were concen- trated in the central part of the country. seropositive for T. solium cysticercus antibodies; increas- ing education level, socioeconomic status and consum- ing partially cooked/raw pork meat once per week were significantly associated with decreased odds of being seropositive (Additional file 1: Table S4). Risk factor anal- ysis results from the sub-analysis of those owning pigs (n = 3154) are reported in Additional file 1: Text S2 and Additional file 1: Table S5. Study population and seroprevalence distribution Parasites & Vectors (2021) 14:590 Table 1 Total number of respondents for each covariate level and tot antibodies Covariate Level Sociodemographic characteristics Sex Male Female Age groups (years) 2–10 11–20 21–30 31–40 41–50 51–60 61–64 Residence Rural Urban Education level Education higher than secondary Partial/complete secondary educa Partial/complete primary educatio No education Occupation Other occupations Self-employed Employee Farm coordinator Farm labourer Student Housewife/houseman Businessman Farm owner Socioeconomic status (stratum) ≥ 4 3 2 1 Displaced people Eating habits Pork consumption & cooking level No consumption Well cooked; < once per month Well cooked; once per month Well cooked; once per week Well cooked; > once per week Partially cooked/raw; < once per m Partially cooked/raw; once per mo Partially cooked/raw; once per we Partially cooked/raw; > once per w Water source Well/ cistern Aqueduct Waterway Other sources Rain water Table 1 Total number of respondents for each covariate level and total number positive for circulating Taenia solium cysticercus antibodies Covariate Level Number (%) of respondents Total number (%; 95% CI) of positive respondents Sociodemographic characteristics Sex Male 9227 (31.54) 809 (8.77; 8.20–9.36) Female 20,026 (68.46) 1967 (9.82; 9.41–10.20) Age groups (years) 2–10 972 (3.32) 76 (7.82; 6.21–9.69) 11–20 3583 (12.25) 337 (9.41; 8.47–10.4) 21–30 6558 (22.42) 641 (9.78; 9.07–10.50) 31–40 6359 (21.74) 617 (9.70; 8.99–10.50) 41–50 5927 (20.26) 578 (9.76; 9.01–10.50) 51–60 4398 (15.03) 409 (9.30; 8.46–10.20) 61–64 1456 (4.98) 117 (8.04; 6.70–9.56) Residence Rural 6455 (22.07) 764 (11.80; 11.00–12.60) Urban 22,798 (77.93) 2011 (8.83; 8.46–9.20) Education level Education higher than secondary 3862 (13.20) 296 (7.67; 6.85–8.55) Partial/complete secondary education 13,405 (45.82) 1190 (8.88; 8.40–9.37) Partial/complete primary education 10,615 (36.29) 1113 (10.5; 9.91–11.10) No education 1371 (4.69) 176 (12.8; 11.10–14.70) Occupation Other occupations 844 (6.17) 125 (6.93; 5.80–8.20) Self-employed 4340 (14.84) 325 (7.49; 6.72–8.31) Employee 4367 (14.93) 374 (8.56; 7.75–9.43) Farm coordinator 135 (0.46) 12 (8.89; 4.68–15.00) Farm labourer 720 (2.46) 65 (9.03; 7.04–11.40) Student 3538 (12.09) 322 (9.10; 8.18–10.10) Housewife/houseman 13,005 (44.46) 1313 (10.1; 9.58–10.6) Businessman 499 (1.71) 53 (10.6; 8.06–13.7) Farm owner 1805 (2.89) 186 (22.00; 19.30–25.00) Socioeconomic status (stratum) ≥ 4 581 (1.99) 26 (4.48; 2.94–6.49) 3 3743 (12.80) 176 (4.70; 4.05–5.43) 2 10,166 (34.75) 706 (6.95; 6.46–7.46) 1 14,465 (49.45) 1715 (11.90; 11.30–12.40) Displaced people 298 (1.02) 152 (51.00; 45.20–56.80) Eating habits Pork consumption & cooking level No consumption 3328 (11.38) 416 (12.50; 11.40–13.70) Well cooked; < once per month 12,744 (43.56) 1203 (9.44; 8.94–9.96) Well cooked; once per month 5789 (19.79) 561 (9.69; 8.94–10.5) Well cooked; once per week 3419 (11.69) 293 (8.57; 7.65–9.56) Well cooked; > once per week 1549 (5.29) 127 (8.20; 6.88–9.68) Partially cooked/raw; < once per month 1320 (4.51) 96 (7.27; 5.93–8.81) Partially cooked/raw; once per month 436 (1.49) 32 (7.34; 5.07–10.2) Partially cooked/raw; once per week 410 (1.40) 26 (6.33; 4.17–9.13) Partially cooked/raw; > once per week 265 (0.91) 21 (7.92; 4.97–11.9) Water source Well/ cistern 2738 (9.36) 190 (6.94; 6.02–7.96) Aqueduct 19,485 (66.61) 1401 (7.19; 6.83–7.56) Waterway 4904 (16.77) 564 (11.50; 10.60–12.5) Other sources 574 (1.96) 132 (23.00; 19.60–26.70) Rain water 1554 (5.31) 488 (31.40; 29.10–33.70) Table 1 Total number of respondents for each covariate level and total number positive for circulating Taenia solium cysticercus antibodies Covariate Level Number (%) of respondents Total number (%; 95% CI) of positive respondents Eating habits Pork consumption & cooking level Water source Page 6 of 15 Galipó et al. Risk factors for human seropositivity without spatial structuref From the univariate mixed-effects logistic regression model with two random effects, food consumption in streets, washing hands after toilet usage and owning ani- mals other than dogs and pigs (cattle, cats, birds) were excluded from further (multivariate) analysis, having a P-value > 0.25 (Additional file 1: Table S3). Consequently, 16 explanatory variables were included in the multivari- able mixed-effect logistic regression with two random effects. Increasing age (as age categories), being female, owning dogs and using rainwater as a water source were significantly associated with increased odds of being Figure 2 shows a map of the residual variation in the seroprevalence of T. solium cysticercus antibodies at the municipality level that is unexplained by the covariates in the non-spatial mixed-effects model. Figure 3 shows a variogram analysis carried out on the municipalities’ estimated random effects. The empirical variogram falls partially outside of the 95% confidence bands, suggesting the presence of spatial correlation in seroprevalence at the municipality level (unexplained by the covariates) up to approximately 120–140 km; further than this distance, the variation between two spatial points starts to plateau This Galipó et al. Parasites & Vectors (2021) 14:590 Page 7 of 15 0 100 200 300 400 km N Seroprevalence (%) 0 to 3 3 to 6 6 to 9 9 to 16 16 to 25 25 to 40 Fig. 1 Seroprevalence of cysticercosis in Colombia, 2008–2010. Seroprevalence of Taenia solium cysticercus antibodies in 133 municipalities in Colombia. Departments are outlined in pale grey lines and sampled municipalities are shown in solid colours Fig. 1 Seroprevalence of cysticercosis in Colombia, 2008–2010. Seroprevalence of Taenia solium cysticercus an Colombia. Departments are outlined in pale grey lines and sampled municipalities are shown in solid colours estimate was determined more precisely from the fitted geostatistical model (see below) to a value of 139 km. positive generally increased with age. For example, adults aged between 21 and 60 years were approximately two- fold more likely to test positive than children in the age range 2–10 years. Lower educational levels were sig- nificantly associated with increased odds of seropositiv- ity, with the highest estimated odds associated with no formal education. Displaced people had 2.20-fold (95% CrI = 1.15–4.28) higher odds of being seropositive than people in the highest socioeconomic stratum; there was no significant difference among other socioeconomic strata. Risk factors for human seropositivity without spatial structuref The use of rainwater as a water source was associ- ated with 1.6-fold (95% CrI = 1.21–2.13) higher odds of Geostatistical modelh The geostatistical model estimated a strong spatial corre- lation at the municipality level of up to 139 km. The ORs and 95% CrIs associated with each covariate included in the final multivariable model (which accounts for spatial correlation at the municipality level) are given in Table 2. Notably, the odds of testing positive for T. solium cysti- cercus antibodies was 1.29-fold (95% CrI = 1.15–1.46) greater for females than for males, and the odds of testing Galipó et al. Parasites & Vectors (2021) 14:590 Page 8 of 15 0 100 200 300 400 500 600 km N Random Effects U^ −1 0 1 2 Fig. 2 Residual variation in Taenia solium cysticercus seroprevalence at the municipality level across Colombia. The map represents the residual variation in cysticercus seroprevalence at the municipality level that is not explained by the covariates in the non-spatial mixed-effects model Random Effects U^ Fig. 2 Residual variation in Taenia solium cysticercus seroprevalence at the municipality level across Colombia. The map represents the residual variation in cysticercus seroprevalence at the municipality level that is not explained by the covariates in the non-spatial mixed-effects model Discussionh Their results suggested that higher antibody seroreversion rates occur follow- ing first exposure (representing the primary humoral response), followed by a lower seroreversion rate after the boosting effect of subsequent exposures (represent- ing secondary humoral response), causing saturation in antibody seroprevalence with age. Hence, where trans- mission is relatively intense—and repeated exposures are common—one might expect to see similar saturat- ing age–seroprevalence profiles. By contrast, in lower transmission settings, the effect of seroreversion fol- lowing first exposure—and the less frequent boosting effect of subsequent exposures—may be more evident in seroprevalence profiles, possibly resulting in a decline in seropositivity in older age groups. Fig. 3 Estimated variogram for the mixed-effects model residuals at the municipality level (blue dots) across Colombia, including 95% confidence intervals obtained from a permutation test under the null hypothesis that there is no spatial correlation (blue-shaded area). The blue dots fall outside of the confidence bands up to approximately 120–140 km of separation, indicating spatial correlation up to this distance (confirmed by the geostatistical model) Fig. 3 Estimated variogram for the mixed-effects model residuals at the municipality level (blue dots) across Colombia, including 95% confidence intervals obtained from a permutation test under the null hypothesis that there is no spatial correlation (blue-shaded area). The blue dots fall outside of the confidence bands up to approximately 120–140 km of separation, indicating spatial correlation up to this distance (confirmed by the geostatistical model) Exposure to T. solium is known to be greater for indi- viduals with lower educational levels, those from lower socioeconomic strata [6, 32] and those facing social marginalisation [9, 33–35]. Our findings are consistent with these previously reported findings, with the odds of displaced people testing positive being almost two- fold higher than people in the highest socioeconomic stratum. Internal displacement in Colombia is a major issue that often involves the poorest and most disad- vantaged people [36], but if the control of T. solium is to become comprehensive, displaced people may require enhanced interventions. Health education could be one such option for control in specific populations using tools such as “The Vicious Worm” [37], as there is some evidence that health education campaigns specific to T. solium can impact transmission [38]. It is, however, likely that to achieve substantial, sustained reductions in the prevalence of T. solium or elimination, particularly in highly endemic areas, a One Health approach targeting the whole T. Discussionh being positive compared to the use of a well or cistern, and dog owners were at significantly increased odds of testing positive (OR = 1.19, 95% CrI = 1.08–1.31) than non-owners.Consumption of partially cooked/raw pork meat once per week was associated with a signifi- cantly decreased odds of testing positive (OR: 0.59, 95% CrI: 0.36 – 0.90) compared to no consumption. Place of residence, occupation, frequency of washing vegetables, excreta elimination and owning animals other than dogs (including pigs) were not significantly associated with testing positive for T. solium cysticercus antibodies. being positive compared to the use of a well or cistern, and dog owners were at significantly increased odds of testing positive (OR = 1.19, 95% CrI = 1.08–1.31) than non-owners.Consumption of partially cooked/raw pork meat once per week was associated with a signifi- cantly decreased odds of testing positive (OR: 0.59, 95% CrI: 0.36 – 0.90) compared to no consumption. Place of residence, occupation, frequency of washing vegetables, excreta elimination and owning animals other than dogs (including pigs) were not significantly associated with testing positive for T. solium cysticercus antibodies. The 2008–2010 Colombian cysticercosis serosurvey generated unique and unprecedented information on exposure to T. solium cysticercosis at a national scale. The work presented here extends the original analysis of these data [14] by using contemporary geostatisti- cal techniques to evaluate individual-level risk factors associated with seropositivity to T. solium cysticerci and, simultaneously, spatial clustering at a sub-national (municipality) scale. The results contribute important Galipó et al. Parasites & Vectors (2021) 14:590 Page 9 of 15 Page 9 of 15 0.0 0.5 1.0 0 50 100 150 200 250 Distance (km) semivariance Fig. 3 Estimated variogram for the mixed-effects model residuals at the municipality level (blue dots) across Colombia, including 95% confidence intervals obtained from a permutation test under the null hypothesis that there is no spatial correlation (blue-shaded area). The blue dots fall outside of the confidence bands up to approximately 120–140 km of separation, indicating spatial correlation up to this distance (confirmed by the geostatistical model) 0.0 0.5 1.0 0 50 100 150 200 250 Distance (km) semivariance be considered as an indicator of lifetime prior exposure. Praet et al. [31] explored age-dependent dynamics of T. solium cysticercus antibody positivity in more depth by fitting mathematical models to similar age–seropreva- lence data collected in Ecuador. Discussionh solium system, including infections in pigs, humans and the environment, will be required [39, 40], as recently shown by intervention trials in Peru and Zam- bia [41, 42]. information on factors associated with exposure to T. solium cysticerci. They also indicate that similar large- scale epidemiological surveys will be needed if hyper- endemic foci of transmission are to be identified and targeted for intensified interventions in 17 endemic countries, as per the WHO’s 2021–2030 NTD roadmap targets for taeniasis/cysticercosis [22]. Here, and in the original analysis of these data [14], women were more likely than men to be positive for T. solium cysticercus antibodies. This finding is consistent with the results of numerous other studies undertaken in Latin America [2, 9, 23–27]; by contrast, in other endemic regions, such as sub-Saharan Africa, being male is associated with an increased risk of exposure [28] and of antigen positivity [29, 30]. The mechanisms underlying these epidemiological patterns remain unclear. Different household roles associated with han- dling household-owned animals, food and water may be important, although many variables pertaining to these activities were accounted for in this analysis. Not- withstanding the underlying cause, women could be an important target for educational campaigns in Colom- bia, not just because of their apparent increased risk of exposure, but also because they are often being respon- sible for the majority of food handling and preparation activities, which would be all the more important if they were also tapeworm carriers.h The only variable related to food and water sources or hygiene practices that was significantly associated with seropositivity to T. solium cysticercus antibod- ies was the use of rainwater. Individuals in households using rainwater as opposed to water stored in wells or cisterns had a 1.6-fold higher odds of seropositivity. Waterborne cysticercosis transmission is supported in the literature, given that the eggs can survive in fresh, brackish and salt waters [32, 43–45] and can contami- nate vegetables [45]. Other variables, such as open-field defecation or the use of unsanitary latrines [46, 47], The trend for increasing seropositivity with age is unsurprising given that T. solium cysticercus antibodies probably persist for several years. Seropositivity may thus Galipó et al. Water source Discussionh Parasites & Vectors (2021) 14:590 Page 10 of 15 Table 2 Geostatistical multivariable logistic regression model results: odds of testing positive for Taenia solium cysticercus antibodies Covariate Level OR (95% CrI) Sociodemographic characteristics Sex Male 1 Female 1.29 (1.15–1.46) Age group (years) 2–10 1 11–20 1.83 (1.35–2.50)* 21–30 1.96 (1.43–2.72)* 31–40 1.96 (1.42–2.73)* 41–50 2.12 (1.54–2.95)* 51–60 2.00 (1.45–2.80)* 61–64 1.68 (1.16–2.44)* Residence Rural 1 Urban 0.84 (0.69–1.02) Education level Higher than secondary 1 Partial/complete secondary 1.05 (0.90–1.23) Partial/complete primary 1.32 (1.11–1.57)* No education 1.34 (1.04–1.73)* Occupation Other occupations 1 Self-employed 1.01 (0.80–1.29) Employee 1.07 (0.85–1.36) Farm coordinator 1.09 (0.52–2.11) Farm labourer 1.06 (0.73–1.52) Student 1.20 (0.91–1.58) Housewife/houseman 1.07 (0.86–1.36) Businessman 1.22 (0.83–1.79) Farm owner 1.20 (0.89–1.63) Socioeconomic status ≥ 4 1 3 0.93 (0.57–1.55) 2 1 (0.63–1.64) 1 1.12 (0.70–1.86) Displaced 2.20 (1.15–4.28)* Eating habits Pork consumption No consumption 1 Well cooked; < once per month 0.92 (0.80–1.06) Well cooked; once per month 0.94 (0.80–1.11) Well cooked; once per week 0.85 (0.71–1.03) Well cooked; > once per week 0.98 (0.76–1.24) Partially cooked/raw; < once per month 0.85 (0.64–1.11) Partially cooked/raw; once per month 0.89 (0.58–1.33) Partially cooked/raw; once per week 0.59 (0.36–0.90)* Partially cooked/raw; > once per week 0.94 (0.55–1.22) Water source Well/cistern 1 Aqueduct 1.06 (0.84–1.33) Waterway 1.12 (0.88–1.45) Other sources 1.11 (0.78–1.58) Rain water 1.60 (1.21–2.13)* Socioeconomic status Socioeconomic status Eating habits Pork consumption Water source Other sources Page 11 of 15 Galipó et al. Parasites & Vectors (2021) 14:590 Crl Credible interval, OR odds ratio Statistically significant Covariate Level OR (95% CrI) Hygiene practices Washing vegetables No consumption 1 Always 1.20 (0.56–2.92) Occasionally 1.28 (0.60–3.12) Never 1.17 (0.55–2.86) Washing hands before a meal Always 1 Occasionally 1.04 (0.94–1.15) Never 0.90 (0.71–1.14) Elimination of excreta Sanitary conditions 1 In waterway 0.84 (0.55–1.26) Latrine with well 1.06 (0.90–1.25) Latrine without well 1.16 (0.82–1.62) Open field 1.10 (0.85–1.42) Owning animals Cattle Not owning 1 Owning 0.97 (0.76–1.22) Cats Not owning 1 Owning 1.07 (0.95–1.20) Dogs Not owning 1 Owning 1.19 (1.08–1.31) Birds Not owning 1 Owning 0.96 (0.86–1.07) Pigs Not owing 1 Owning ≤ 10 pigs 1.10 (0.93–1.31) Owning > 10 pigs 0.83 (0.56–1.19) OR (95% CrI) Crl Credible interval, OR odds ratio *Statistically significant canine meat [51], although this practice is thought to be extremely rare and not widely reported in Latin America. Moreover, the role of dogs as potential hosts for T. solium remains somewhat speculative. Discussionh Given the coprophagic habits of dogs and their close interaction with humans, it is also possible (and perhaps more likely) that dogs act as mechanical vectors of T. solium eggs. that one might also expect to be associated with expo- sure to T. solium were not identified in our analysis as significant risk factors. We also found that the odds of seropositivity significantly decreased when individu- als consumed partially cooked/raw pork meat once per week, an observation possibly confounded by wealth (i.e., wealthier individuals consuming more meat). One might expect that consumption of partially cooked/raw pork meat would be associated with increased odds of seropositivity, given that taeniasis (adult tapeworm) carriers are at risk of autoinfection. However, more research is needed to understand the relative contribu- tion of this route of transmission to overall cysticerco- sis risk [48]. A further striking finding is that among the 10.8% (n = 3,154) of individuals owning pigs, we did not find a significantly increased odds of seropositivity, only a non- significant increase in those owning fewer than 10 pigs (possibly indicative of smallholder, subsistence farmers, compared to individuals owning > 10 pigs, which may represent wealthier farmers). A further sub-analysis of pig owners (Additional file 1: Text S2) found no associa- tion between seropositivity and pig management prac- tices (e.g. free roaming, feeding wastes, drinking free water, among others). These findings contrast with those reported in other studies in Latin America and other geo- graphical settings, in which human cysticercosis has been associated with owning pigs [2, 33, 52]. Some farming A particularly striking finding of our analysis was the association between owning dogs and significantly increased odds of test positivity. Dogs in Asia have been reported to test positive for T. solium antibodies [49, 50], potentially implicating them as alternative intermedi- ate hosts. Transmission to humans has also been sug- gested to occur via the consumption of raw or uncooked Galipó et al. Parasites & Vectors (2021) 14:590 Page 12 of 15 Page 12 of 15 with > 5000 individuals and a health centre is likely to have created a bias towards sampling in more densely populated urban areas. This led to an underrepresen- tation of rural communities, which may typically have had less access to health care and possibly lower over- all health. Discussionh In addition, nine departments were excluded from sampling (due to logistical and resource con- straints) and overall, only a relatively small fraction (12%) of Colombia’s municipalities were sampled (133/1122). Women are highly represented, and this is likely due to the decision of randomizing only the individuals present at the interview for inclusion in the study. Also, the data were collected in 2008–2010, over a decade ago, and may therefore not reflect precisely contemporary epidemio- logical conditions. Nonetheless, we believe that, in the absence of wide-spread national control efforts, the dis- tribution and endemic situation of T. solium are unlikely to have changed substantively over the past decade and, therefore, the data provide a useful snapshot of endemic conditions across the country. Due to the nature of sur- veys, other forms of bias and reverse causation are also possible. practices, such as using waste or water and mix concen- trate as feed, and the lack of drainage systems were non- significantly associated with increased seropositivity. However, because this sub-analysis was based on a much smaller sample (n = 3154) with only 388 seropositive individuals, there was limited power to detect significant associations. In addition to exploring individual and household risk factors associated with exposure to T. solium, our geo- statistical approach enabled identification of spatial clus- ters where seropositivity was higher, so-called hotspots (in the north and south of Colombia), or lower (in the central and western areas of the country) than could be explained by the included covariates (Fig. 2). Hotspots where seropositivity was higher than could be explained by the covariates coincided with areas with higher sero- prevalence (16‒40%) in the northern coastal area and areas bordering Venezuela (Departments of Atlántico, Magdalena, Cesar, La Guajira), in the northern-central region (Departments of Antioquia and Bolívar), in Vau- pés (south-east, bordering Brazil) and in the south, in regions bordering Peru and Brazil (Department of Ama- zonas; Fig. 1). Neither human nor pig population density was explicitly included in the model and, therefore, these variables could help to explain some of this clustering (because of the potential for increased contamination of the environment with T. solium eggs where humans and pigs are abundant). Discussionh While population densities are het- erogeneous across Colombia, some of the highest human population densities are generally found in the north and north-east of the country [53], alongside the highest pig population in the Pacific (east costal), Andean (north- east/north-west) and Caribbean regions (north), as esti- mated from the Gridded Livestock Database in 2007 [54]. Furthermore, it should be noted that given the level of spatial analysis, we were only able to detect spatial varia- tion at the municipality level. Moreover, it cannot be excluded that any of the encountered associations are confounded by unmeasur- able or unknown risk factors and that the a priori deci- sion to drop a certain number of variables might have increased the model residuals, not including possible confounders. On the other hand, the unstructured nature of some variables or the probable collinearity with other exposures made this choice desirable. Despite the lack of data concerning some geographical areas in Colombia, the authors still consider the study outcomes as valu- able and indicative of the situation of cysticercosis in the country. In addition, the information provided in the current study could be further used to build models that can spatially predict the disease seroprevalence in non- sampled areas [17], offering a cost-effective tool for deci- sion-makers in places where direct sampling did not take place. Local climatic, environmental and ecological conditions may also play a role in the observed clustering. In a recent systematic review, Jansen et al. [45] identified that Taenia spp. eggs can survive in the environment for up to 1 year in favourable conditions of high humidity, moderate tem- peratures (5‒25 °C) and presence of surface water. Moreo- ver, invertebrates, including dung beetles (Ammophorus rubripes), can also act as mechanical vectors for the dis- persal of Taenia spp eggs [55, 56]. Hence, it is highly likely that local conditions—unaccounted for in our statistical model—will influence spatial patterns of exposure. Mapping the distribution and seroprevalence of T. solium in endemic countries is a crucial next step in realising the WHO’s goals of implementing intensified control in hyperendemic areas of 17 countries by 2030 [22]. Currently, country-wide data on the transmission of T. solium, such as those analysed here for Colombia, are scarce, and thus there is a great deal of work to be done to identify hyperendemic areas in which to imple- ment intensified interventions. Author details 1 Department of Epidemiological Sciences, Animal and Plant Health Agency, New Haw, Addlestone, Surrey, UK. 2 Department of Pathobiology and Popula‑ tion Sciences and London Centre for Neglected Tropical Disease Research, Royal Veterinary College, Hatfield, UK. 3 Department of Infectious Disease Epidemiology and London Centre for Neglected Tropical Disease Research, School of Public Health, Imperial College London, London, UK. 4 Medical Research Centre for Global Infectious Disease Analysis, School of Public Health, Imperial College London, London, UK. 5 Schistosomiasis Control Initiative (SCI) Foundation, Edinburgh House, 170 Kennington Lane, Lambeth, London SE11 5DP, UK. 6 Centre for Health Informatics, Computing and Statistics, Lancaster University, Lancaster, UK. 7 Grupo de Parasitología, Instituto Nacional de Salud, Bogotá, Colombia. 8 Present Address: Departamento de Epidemiología Clínica, Pontificia Universidad Javeriana, Bogotá, Colombia. 1. World Health Organisation. Taenia solium Taeniasis/cysticercosis diagnos‑ tic tools. 2016. Report of a stakeholder meeting, Geneva, 17–18 Decem‑ ber 2015. https://apps.who.int/iris/bitstream/handle/10665/206543/ 9789241510516_eng.pdf. Accessed 24 Oct 2021. Conclusions Taeniasis/cysticercosis is a major public health problem and an important cause of epilepsy and other neurologi- cal sequelae in many regions of the world. The WHO aims to target this zoonotic NTD with enhanced control where transmission is most intense, although epidemiological data at national and subnational scales remain scarce. The 2008– 2010 baseline epidemiological survey undertaken by the Colombian government remains unprecedented in scale and geographical coverage, generating data that are unique and provide a highly valuable resource for understanding the spa- tial epidemiology of T. solium cysticercosis. By taking a con- temporary geostatistical approach, we have highlighted key associations between human cysticercosis antibody seropos- itivity and individual- and household-level risk factors, while also identifying spatial hotspots of exposure, unexplained by the measured covariates. These findings could be used to inform the design of intervention strategies in Colombia, such as targeting spatial hotspots and more highly exposed groups (such as displaced people and women), and also to illustrate how important geostatistical modelling will be as a tool to inform and support the WHO NTD roadmap in its 2021–2030 goals for taeniasis/cysticercosis. Availability of data and materials The data that support the findings of this study are available from the Instituto Nacional de Salud (Bogotá, Colombia) but restrictions apply to the availability of these data, which were used under licence for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of Instituto Nacional de Salud. Funding MAD, ZMC and MGB acknowledge funding from the Medical Research Council (MRC) Centre for Global Infectious Disease Analysis (reference MR/R015600/1), jointly funded by the UK MRC and the UK Foreign, Commonwealth & Develop‑ ment Office (FCDO), under the MRC/FCDO Concordat agreement and is also part of the European and Developing Countries Clinical Trials Partnership (EDCTP2) programme supported by the European Union. MAD has also been funded by a Medical Research Council Doctoral Training Partnership (MRC DTP) research studentship to support this work. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Consent for publication Not applicable. Consent for publication Not applicable. Consent for publication Not applicable. Competing interests ll h d l h All authors declare they do not have conflicts of interest. Acknowledgements Not applicable Acknowledgements Not applicable Acknowledgements Not applicable parameterized to identify likely areas of high transmis- sion using Geographical Information System (GIS) data that have comprehensive global coverage. Although our study focused on the identification of risk factors asso- ciated with exposure to T. solium and residual degrees of spatial clustering, similar geostatistical and machine learning approaches can be used that focus on predict- ing the spatial distribution of disease using GIS data [17]. Such approaches, conducted at national and global scales, will be crucial in assisting progress towards the WHO’s 2030 goals [22, 58]. Supplementary Information The online version contains supplementary material available at https://doi. org/10.1186/s13071-021-05092-8. Additional file 1: Table S1. Notation of parameters used for model build‑ ing (analysis of residual spatial correlation) and incorporation of spatial structure. Table S2. Seroprevalence of Taenia solium cysticercus antibodies in Colombia. Table S3. Distribution of seropositive individuals, crude odds ratios (ORs) of testing positive for Taenia solium cysticercus antibodies by ELISA and associated 95% confidence intervals (CIs) from the univariate mixed-effects model. Table S4. Distribution of seropositive individuals, multivariable mixed-effects logistic regression adjusted ORs of testing positive for Taenia solium cysticercus antibodies by ELISA and associated CIs. Table S5. Pig management sub-set analysis (n = 3154). Pig manage‑ ment practices, distribution of seropositive individuals, crude odds ratios (ORs) of testing positive for Taenia solium cysticercus antibodies by ELISA and associated 95% CIs. Figure S1. Map displaying the sampled munici‑ palities in Colombia (2008–2010). Text S1. Supplementary methods. Text S2. Results: pig management sub-analysis Received: 1 September 2021 Accepted: 8 November 2021 Authors’ contributions ACFS and the work group at Grupo de Parasitología (Instituto Nacional de Salud, Colombia) conceived the original study idea, participated in the design of the study and were involved in field implementation and data collection. MAD, CF, EG, MW and MGB were involved in data analysis and interpretation. MAD, CF and EG prepared the draft manuscript. EG, MAD, CF, ZMC, MGB, KS, ACFS and MW read and edited draft versions of the manuscript and approved the final manuscript. Discussionh Moreover, although working definitions of ‘hyperendemicity’ have been pro- posed [57], there is not yet a consensus on the definition of endemicity levels for T. solium infection. Geostatisti- cal approaches will play an important role in identifying areas of high transmission, particularly if they can be l Although the serosurvey data analysed here are unique in presenting a picture of exposure to T. solium cysticer- cosis at a national scale, geographical coverage is incom- plete and the sampling approach may have introduced some biases. In particular, the selection of municipalities Galipó et al. Parasites & Vectors (2021) 14:590 Galipó et al. Parasites & Vectors (2021) 14:590 Page 13 of 15 Page 13 of 15 Ethics approval and consent to participate Ethical approval for the original study (Flórez Sánchez et al. [14]) was obtained through the Research Ethics Committee of the Instituto Nacional de Salud (INS), Colombia in 2009. References Ndimubanzi PC, Carabin H, Budke CM, Nguyen H, Qian YJ, Rainwater E, et al. A systematic review of the frequency of neurocyticercosis with a focus on people with epilepsy. PLoS Negl Trop Dis. 2010;2(4):e870. 29. Kanobana K, Praet N, Kabwe C, Dorny P, Lukanu P, Madinga J, et al. High prevalence of Taenia solium cysticerosis in a village community of Bas- Congo, Democratic Republic of Congo. Int J Parasitol. 2011;41(10):1015–8 11. Nkwengulila G. The financial costs associated with porcine cysticercosis and epilepsy in Iringa Rural District. Health. 2014;6(21):2959–65. https:// doi.org/10.4236/health.2014.621334. 30. Carabin H, Millogo A, Cissé A, Gabriël S, Sahlu I, Dorny P, et al. Prevalence of and factors associated with human cysticercosis in 60 villages in three provinces of Burkina Faso. PLoS Negl Trop Dis. 2015;9(11):e0004248. g 12. Organización Panamericana de la Salud/Organización Mundial de la Salud Oficina Regional para las Américas. Informe Primera Reunión Regional sobre Control de Taenia solium en América Latina. Colombia, October 2015. https://www.paho.org/hq/dmdocuments/2016/prime ra-reunion-regional-control-tena-solium-americas-2015.pdf. Accessed 25 Oct 2021. 31. Praet N, Speybroeck N, Rodriguez-Hidalgo R, Benitez-Ortiz W, Berkvens D, Brandt J, et al. Age-related infection and transmission patterns of human cysticercosis. Int J Parasitol. 2010;40(1):85–90. 32. Nithiuthai S, Anantaphruti MT, Waikagul J, Gajadhar A. Waterborne zoonotic helminthiases. Vet Parasitol. 2004;126(1–2):167–93. 33. Sánchez AL, Medina MT, Ljungström I. Prevalence of taeniasis and cysticercosis in a population of urban residence in Honduras. Acta Trop. 1998;69(2):141–9. 13. Bruno E, Bartoloni A, Zammarchi L, Strohmeyer M, Bartalesi F, Bustos JA, et al. Epilepsy and neurocysticercosis in Latin America: a systematic review and meta-analysis. PLoS Negl Trop Dis. 2013;7(10):e2480. 34. Moro PL, Lopera L, Bonifacio N, Gilman RH, Silva B, Verastegui M, et al. Taenia solium infection in a rural community in the Peruvian Andes. Ann Trop Med Parasitol. 2003;97(4):373–9. 14. Flórez Sánchez AC, Pastrán SM, Vargas NS, Beltrán M, Enriquez Y, et al. Cysticercosis in Colombia seroprevalence study 2008–2010. Acta Neurol Colomb. 2013;29(2):73–86. 35. Rodríguez-Morales AJ, Yepes-Echeverri MC, Acevedo-Mendoza WF, Marín- Rincón HA, Culquichicón C, Parra-Valencia E, et al. Mapping the residual incidence of taeniasis and cysticercosis in Colombia, 2009–2013, using geographical information systems: Implications for public health and travel medicine. Travel Med Infect Dis. 2018;22:51–7. 15. National Administrative Department of Statistics (DANE). Información técnica. Censo Nacional de Población y Vivienda 2018. 2020. https:// www.dane.gov.co/index.php/estadisticas-por-tema/demografia-y-pobla cion/censo-nacional-de-poblacion-y-vivenda-2018/informacion-tecnica. Accessed 25 Oct 2021. 36. Internal Displacement Monitoring Centre (IDMC). Colombia. http://www. internal-displacement.org/countries/colombia. Accessed 25 Oct 2021. 16. References Galipó et al. Parasites & Vectors (2021) 14:590 Page 14 of 15 Page 14 of 15 2. García HH, Gonzalez AE, Evans CAW, Gilman RH. Taenia solium cysticerco‑ sis. Lancet. 2003;16:547–56. 23. Allan JC, Velasquez-Tohom M, Garcia-Noval J, Torres-Alvarez R, Yurrita P, Fletes C, et al. Epidemiology of intestinal taeniasis in four, rural Guatema‑ lan communities. Trop Med Parasitol. 1996;90(2):157–65. 3. U.S. Centers for Disease Control and Prevention. Parasite—cysticercosis: biology. https://www.cdc.gov/parasites/cysticercosis/biology.html. Accessed 25 Oct 2021. 24. Garcia-Noval J, Allan JC, Fletes C, Moreno E, De Mata F, Torres-Alvarez R, et al. Epidemiology of Taenia solium taeniasis and cysticercosis in two rural Guatemalan communities. Am J Trop Med Hyg. 1996;55(3):282–9. 4. Trevisan C, Mkupasi EM, Ngowi HA, Forkman B, Johansen MV. Severe seizures in pigs naturally infected with Taenia solium in Tanzania. Vet Parasitol. 2016;220:67–71. 25. Garcia HH, Araoz R, Gilman RH, Valdez J, Gonzalez AE, Gavidia C, et al. Increased prevalence of cysticercosis and taeniasis among professional fried pork vendors and the general population of a village in the Peruvian highlands. Am J Trop Med Hyg. 1998;59(6):902–5. 5. Garcia HH, Del Brutto OH. Taenia solium cysticercosis. Infect Dis Clin North Am. 2000;14(1):97–119. 26. Agudelo-Flórez P, Restrepo BN, Palacio LG. Knowledge and practices concerning taeniasis-cysticercosis in Colombian pig-breeders. Rev Salud Pública. 2009;11(2):191–9 (in Spanish). 6. Enander RT, Ramírez Amaya A, Enander RA, Gute DM. Neurocysticercosis: risk and primary prevention strategies update. Int J Environ Health Res. 2010;20(5):329–65. 7. Gupta DS, Goyal AK, Tandon PN, Jurel SK, Srivastava S, Dangi UR, et al. Platyhelminthes in tongue—a rare case and review. J Oral Maxillofac Surg. 2012;70(11):2605–9. 27. Wardrop NA, Thomas LF, Atkinson PM, de Glanville WA, Cook EAJ, Wamae CN, et al. The influence of socio-economic, behavioural and environ‑ mental factors on Taenia spp transmission in Western Kenya: Evidence from a cross-sectional survey in humans and pigs. PLoS Negl Trop Dis. 2015;10(1):e0004394. g 8. Garcia HH, Rodriguez S, Friedland JS. Immunology of Taenia solium taeniasis and human cysticercosis. Parasite Immunol. 2014;36(8):388–96. 28. Mwanjali G, Kihamia C, Kakoko DV, Lekule F, Ngowi H, Johansen MV, et al. Prevalence and risk factors associated with human Taenia solium infections in Mbozi District, Mbeya Region, Tanzania. PLoS Negl Trop Dis. 2013;7(3):e2102. 9. Coral-Almeida M, Gabriël S, Abatih EN, Praet N, Benitez W, Dorny P. Taenia solium human cysticercosis: a systematic review of sero-epidemological data from endemic zones around the world. PLoS Negl Trop Dis. 2015;9(7):e0003919. 10. References Galipó et al. Parasites & Vectors (2021) 14:590 Page 15 of 15 Galipó et al. Parasites & Vectors (2021) 14:590 52. Dermauw V, Carabin H, Ganaba R, Cissé A, Tarnagda Z, Gabriël S, et al. Factors associated with the 18-month cumulative incidence of sero‑ conversion of active infection with Taenia solium cysticercosis: a cohort study among residents of 60 villages in Burkina Faso. Am J Trop Med Hyg. 2018;99(4):1018–27. 43. Scandrett WB, Gajadhar AA. Recovery of putative taeniid eggs from silt in water associated with an outbreak of bovine cysticercosis. Can Vet J. 2004;45(9):758–60. 44. Alidadi S, Oryan A. Water as a potential transmission route of infection with tapeworms. Air Water Borne Dis. 2015;4(2):e135. 53. Socioeconomic Data and Applications Center (SEDAC). Maps: population density grid, v3. http://sedac.ciesin.columbia.edu/data/set/gpw-v3-popul ation-density/maps/2. Accessed 25 Oct 2021. 45. Jansen F, Dorny P, Gabriël S, Dermauw V, Johansen MV, Trevisan C. The survival and dispersal of Taenia eggs in the environment: what are the implications for transmission? A systematic review. Parasit Vectors. 2021;14(1):1–16. 54. Robinson TP, William Wint GR, Conchedda G, Van Boeckel TP, Ercoli V, Palamara E, et al. Mapping the global distribution of livestock. PLoS ONE 2014;9(5):e96084. 46. Humphries DL, Stephenson LS, Pearce EJ, The PH, Dan HT, Khanh LT. The use of human faeces for fertilizer is associated with increased intensity of hookworm infection in Vietnamese women. Trans R Soc Trop Med Hyg. 1997;91(5):518–20. 55. Gomez-Puerta LA, Garcia HH, Gonzalez AE. Experimental porcine cysticercosis using infected beetles with Taenia solium eggs. Acta Trop. 2018;183(1):92–4. 47. Corrales LF, Izurieta R, Moe CL. Association between intestinal parasitic infections and type of sanitation system in rural El Salvador. Trop Med Int Health. 2006;11(12):1821–31. 56. Vargas-Calla A, Gomez-Puerta LA, Pajuelo MJ, Garcia HH, Gonzalez AE. Molecular detection of taeniid eggs in beetles collected in an area endemic for Taenia solium. Am J Trop Med Hyg. 2018;99(5):1198–200. 48. Skrip LA, Dermauw V, Dorny P, Ganaba R, Millogo A, Tarnagda Z, et al. Data-driven analyses of behavioral strategies to eliminate cysticercosis in sub-Saharan Africa. PLoS Negl Trop Dis. 2021;15(3):e0009234. 57. Dixon MA, Winskill P, Harrison WE, Whittaker C, Schmidt V, Sarti E, et al. Force-of-infection of Taenia solium porcine cysticercosis: a modelling analysis to assess global incidence and prevalence trends. Sci Rep. 2020;10(1):17637. 49. Ito A, Putra MI, Subahar R, Sato MO, Okamoto M, Sako Y, et al. References Dogs as alternative intermediate hosts of Taenia solium in Papua (Irian Jaya), Indo‑ nesia confirmed by highly specific ELISA and immunoblot using native and recombinant antigens and mitochondrial DNA analysis. J Helminthol. 2002;76(4):311–4. 58. Dixon MA, Braae UC, Winskill P, Devleesschauwer B, Trevisan C, Van Damme I, et al. Modelling for Taenia solium control strategies beyond 2020. Bull World Health Organ. 2020;98(3):198–205. 50. Ito A, Wandra T, Yamasaki H, Nakao M, Sako Y, Nakaya K, et al. Cyst‑ icercosis/taeniasis in Asia and the Pacific. Vector-Borne Zoonotic Dis. 2004;4(2):95–107. References Corredor A, López MC, Duque S, Nicholls RS. Estandarización y evaluación de ELISA en eluidos de sangre seca recolectada en papel de filtro para el diagnóstico de cisticercocis. Biomedica. 1996;16(2):131–3. 37. Johansen MV, Trevisan C, Braae UC, Magnussen P, Ertel RL, Mejer H, et al. The vicious worm: a computer-based Taenia solium education tool. Trends Parasitol. 2014;30(8):372–4. 17. Giorgi E, Fronterrè C, Macharia PM, Alegana VA, Snow RW, Diggle PJ. Model building and assessment of the impact of covariates for disease prevalence mapping in low-resource settings: to explain and to predict. J R Soc Interface. 2021;18(179):20210104. 38. Carabin H, Millogo A, Ngowi HA, Bauer C, Dermauw V, Koné AC, et al. Effectiveness of a community-based educational programme in reduc‑ ing the cumulative incidence and prevalence of human Taenia solium cysticercosis in Burkina Faso in 2011–14 (EFECAB): a cluster-randomised controlled trial. Lancet Glob Health. 2018;6(4):e411–25. 18. Diggle PJ, Tawn JA, Moyeed RA. Model-based geostatistics. J R Stat Soc Ser C Appl Stat. 1998;47(3):299–350. 39. de Coster T, Van Damme I, Baauw J, Gabriël S. Recent advancements in the control of Taenia solium: a systematic review. Food Waterborne Parasitol. 2018;13:e00030. 19. Rue H, Martino S, Chopin N. Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations. J R Stat Soc Ser B Stat Methodol. 2009;71(2):319–92. 20. Lindgren F, Rue H, Lindström J. An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach. J R Stat Soc Ser B Stat Methodol. 2011;73(4):423–98. 40. CystiTeam Group for Epidemiology and Modelling of Taenia solium Tae‑ niasis/Cysticercosis. The World Health Organization 2030 goals for Taenia solium: Insights and perspectives from transmission dynamics modelling. Gates Open Res. 2019;3:1546. 21. Fuglstad GA, Simpson D, Lindgren F, Rue H. Constructing priors that penalize the complexity of Gaussian random fields. J Am Stat Assoc. 2019;114(525):445–52. 41. O’Neal SE, Pray IW, Vilchez P, Gamboa R, Muro C, Moyano LM, et al. Geographically targeted interventions versus mass drug administra‑ tion to control Taenia solium cysticercosis. Peru Emerg Infect Dis. 2021;27(9):2389–98. 22. World Health Organization. Ending the neglect to attain the sustain‑ able development goals: a road map for neglected tropical diseases 2021–2030. https://www.who.int/publications/i/item/9789240010352. Accessed 25 Oct 2021. 42. Gabriël S, Mwape KE, Hobbs EC, Devleesschauwer B, Van Damme I, Zulu G, et al. Evidence for potential elimination of active Taenia solium trans‑ mission in Africa? N Engl J Med. 2020;383(4):396–7. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. 51. Wandra T, Swastika K, Dharmawan NS, Purba IE, Sudarmaja IM, Yoshida T, et al. The present situation and towards the prevention and control of neurocysticercosis on the tropical island, Bali, Indonesia. Parasit Vectors. 2015;8:148. • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your ﬁeld • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year • At BMC, research is always in progress. Learn more biomedcentral.com/submissions Ready to submit your research Ready to submit your research ? Choose BMC and benefit from: ? Choose BMC and benefit from:
https://openalex.org/W2158284197	https://ccsenet.org/journal/index.php/ass/article/download/10681/9800	English	null	Relationship between Parenting Style and Academic Achievement among Iranian Adolescents in Sirjan	Asian social science	2,011	cc-by	2,820	Asian Social Science Asian Social Science Vol. 8, No. 1; January 2012 www.ccsenet.org/ass Relationship between Parenting Style and Academic Achievement among Iranian Adolescents in Sirjan Mansor Abu Talib Department of Human Development & Family Studies Faculty of Human Ecology, University Putra Malaysia (UPM), Malaysia Tel: 60-1-9337-5422 E-mail: mansorat@putra.upm.edu.my Received: May 26, 2011 doi:10.5539/ass.v8n1p156 Received: May 26, 2011 Accepted: July 29, 2011 Published: January 1, 2012 doi:10.5539/ass.v8n1p156 URL: http://dx.doi.org/10.5539/ass.v8n1p156 Received: May 26, 2011 Accepted: July 29, 2011 Published: January 1, 2012 doi:10.5539/ass.v8n1p156 URL: http://dx.doi.org/10.5539/ass.v8n1p156 Abstract The purpose of the present study was to determine the relationship between parenting styles and academic achievement among adolescents in Iran. The respondents were 382 high school adolescents (251 female and 131 male) in the age range of 15 to 18 years old from selected high schools in Iran. The instrument used to measure parenting style was the parenting style scale by Baumrind (1991). The result of the study indicated that authoritative parenting style has positive significant correlation with academic achievement, while permissive parenting style has negative correlation with academic achievement. The result of the study also showed that there is no significant relationship between authoritarian parenting style and academic achievement. The results of the present study implied that academic achievement among adolescents can be enhanced through positive parenting style. Thus, it is recommended that parents be equipped with knowledge and skills appropriate to the needs and development of their adolescents children. Keywords: Parenting style, Academic achievement, Adolescences 1. Introduction The education system is very important for all country, and Iran is not exception; strong and effective education can help boost the development of the country. Researchers (Mehrafrooz, 2004; Sabouri, 2008) have indicated a problem of low academic achievement among the students in Iran. According to Mozaffari (2001), about 27% of high school students in Iran have low academic achievement. Low academic achievement may create many negative consequences for students. Students with low academic achievement may be more vulnerable to ISSN 1911-2017 E-ISSN 1911-2025 ISSN 1911-2017 E-ISSN 1911-2025 ISSN 1911-2017 E-ISSN 1911-2025 156 Vol. 8, No. 1; January 2012 Asian Social Science www.ccsenet.org/ass problems such as stress, hopelessness, delinquency, psychopathology, and substance abuse (Assarian, Biqam & Asqarnejad, 2006). Therefore, it is essential to investigate factors that may influence academic achievement amongst school going adolescents in Iran. The findings will be an input to agencies and professionals that provide programs or services to promote adolescents/high school students’ development, specifically in the academic aspect. problems such as stress, hopelessness, delinquency, psychopathology, and substance abuse (Assarian, Biqam & Asqarnejad, 2006). Therefore, it is essential to investigate factors that may influence academic achievement amongst school going adolescents in Iran. The findings will be an input to agencies and professionals that provide programs or services to promote adolescents/high school students’ development, specifically in the academic aspect. Various factors have been examined in relation to students’ academic achievements. These factors ranged from family socioeconomic status, family structure, family functioning, peer association, to school and educational environment (Olige, 2008). Past studies (examples: Chao, 2001; Querido et al., 2002) have shown that parents, through their parenting styles built critical foundations for various aspects of children’s development and achievement. Jacobs and Harvey (2005) indicated that parenting style is one of the significant contributors to student’s academic achievement in school. Spera (2005) asserted that parenting styles emphasizes on the response parents provide to their children and the method which parents used to demand compliance from their children. Baumrind (2005) categorized types of parenting style based on two dimensions which are responsiveness and demandingness. According to Baumrind, responsiveness refers to the degree that parents promote self-assertion and individuality by showing care and acceptance to children’s desires. Care and acceptance includes kindness, support for independence, and logical contact. Demandingness refers to demands that parents make on children to be included into society (Baumrind, 2005). 2. Objective The aim of present study is to determine the relationship between three types of parenting styles (authoritative, authoritarian and permissive) and academic achievement among adolescents in Sirjan, Iran. 1. Introduction The demands are imposed through monitoring and controlling of children’s behaviors, as well as communicating the demands directly to the children. The combination of the levels of responsiveness and demandingness creates three types of parenting styles: authoritarian, authoritative and permissive (Baumrind, 2005). Authoritarian parents are highly demanding and unresponsive, and tend to emphasize obedience and respect for authority. Permissive parents have low levels of demandingness and high levels of responsiveness, and moderately imbalance in leniency. In contrast, authoritative parents show a sense of balance between high levels of demandingness and high levels of responsiveness. Parents who are authoritative will communicate with their children, monitor their children’s behaviors, and express warmth and support their children’s needs and challenges. Past studies showed that authoritarian parenting is related low academic achievement (Attaway & Bry, 2004) and higher levels of school problem (Roche, Ensminger, & Cherlin, 2007). Permissive parenting was also found to significantly correlate with academic achievement (Lee, 2006; Roche, Ensminger, & Cherlin, 2007). This means that parents with too high or too low demandingness and responsiveness have children with low academic achievement. Past studies have established a positive correlation between authoritative parenting style and academic achievement (Slaten, 2006; Roche et al., 2007; Simons & Conger, 2007; Pong, Johnston & Chen, 2009). Students with better academic achievement have parents who are more authoritative. 3. Methodology 3.1 Design, sample and procedure The present study is cross-sectional in nature. The sample consists of 382 high school students aged between 15 and 18 years old from twelve high schools in Sirjan-Iran. Respondents were selected by using stratified random sampling technique. The data for the present study were collected using a self-administered questionnaire. 3 2 Measure Published by Canadian Center of Science and Education 5. Discussion and conclusion The results of Pearson correlation analysis showed that authoritative style had a positive and significant influence on academic achievement among adolescents. Adolescents who had parent with authoritative style were more successful in school. The results are consistent with Park and Bauer (2002), Attaway and Bry (2004), Gonzalez-DeHass et al., (2005), Slaten (2006), Roche et al. (2007), Simons and Conger (2007) and Ling Pong et al. (2009) which found that authoritative parenting style promote school achievement. Durkin (1995) highlighted three reasons for the association between authoritative parenting and high school achievement. First, he recommended that parents with authoritative style give emotional security to their children with a sense of calm and autonomy, and also, they assist their children to be successful in school activities. Second, he cited that parents with authoritative style explain the consequences of their children’s actions. Explanations give children a sense of knowledge and understanding of their parents’ principles, desires, and goals that are associated to school context. Thirdly, he suggested that authoritative parents involve in reciprocal contact with their children. Authoritative parents support their children, encourage them to do well academically, and explain the need for education in order to become a successful adult. Thus, children growing up in the authoritative environment have better overall well-being and higher performance in school related activities (Baumrind, 1991). The finding also indicated that there was a significant negative relationship between permissive parenting styles and academic achievement indicating that adolescents who had parent with permissive style had lower academic achievement in school compared to other adolescents. This result is in line with the previous finding by Roche et al. (2007) who showed that levels of permissive parenting was significantly related with academic strain. Parents with permissive style allow their children to have free control of their behaviors and actions. Parents are more passive and give little input in many important decisions made by adolescents. Therefore, it is likely that adolescents from highly permissive environment do not have clear rules and expectations for high achievement, which consequently may lead to low achievement motivation and poor academic performance among adolescents. In contrast, there was no significant relationship between authoritarian parenting style and school achievement. The present finding is similar to the findings by Kim and Rohner (2002), Park and Bauer (2002), and Olige (2008). Authoritarian parents tend to practice rigid and high level of monitoring. 4.2 Correlations between variables Table 3 displays the correlation matrix for variables in the present study. The findings of Pearson Correlation analyses indicated a significant correlation between authoritative parenting style (r=.24, p<.01) and permissive parenting style (r= -.16, p<.01) with academic achievement. However, there was no significant relationship between authoritarian parenting style (r=-.037, p>.05) and academic achievement. 3.2 Measure Parenting style. The Parenting Style Scale developed by Baumrind (1991) consisted of 30 items was used to assess parenting styles. This scale has three dimensions that are authoritative parenting style (10 items), authoritarian parenting styles (10 items), and permissive parenting style (10 items). Adolescents rated their parent on the items using a five-point scale ranging from 1(strongly disagree) to 5(strongly agree). The scale scores are from 10 to 50 for each dimension with high score indicating a high level of parenting style in each dimension. The Cronbach alpha values for the parenting style subscales in the present study are as follows: authoritative (.82), authoritarian (.82), and permissive (.70). Examples of items in the scale are “as I was growing up, once family policy had been established”, “my mother/father discussed the reasoning behind the policy with the children in the family”, “whenever my mother/ father told me to do something as I was growing up, she/he expected me to do it immediately without asking any questions” and “as I was growing up my mother/father did 157 Published by Canadian Center of Science and Education Vol. 8, No. 1; January 2012 Asian Social Science www.ccsenet.org/ass not feel that I needed to obey rules and regulations of behavior simply because someone in authority had established them”. not feel that I needed to obey rules and regulations of behavior simply because someone in authority had established them”. Academic achievement. Academic achievement score was assessed by computing the student’s average grade points for one year. For each course work, the academic point is 0 to 20. The grade points are categorized into four levels: Grade D= 0-9.99 (failed), grade C = 10-14.99 (weak), grade B = 15-16.99 (moderate), and grade A = 17-20 (excellent). 4.1 Descriptive findings The age of the high school students in the present study was 15 to 18 years old (mean=15.80, standard deviation=1.44). Majority of them (66.75%) were 15 to 16 years old and 33.25% of them were 17 to18 years old. More than half of the respondents (65.6%) were female and 34.4% of them were male. The means and standard deviations for authoritative style, authoritarian style, permissive style, and academic achievement are presented in Table 1. The mean score for authoritative style was 38.9, for authoritarian style was 28.37, and for permissive style was 29.92. The mean score for academic achievement was 17.28. Table 2 presents the distribution of respondents by categories of parenting style and academic achievement. Majority of the respondents (75.4%) had parent with authoritative style. Based on the total score, more than 50% of the respondents reported to have excellent academic achievement. 4. Result Descriptive statistics was employed to describe the variables of the study. Pearson correlation analysis was used to examine the association between parenting style and academic achievement. References Assarian, F., Biqam, H. & Asqarnejad, A. (2006). An epidemiological study of obsessive compulsive disorder among high school students and its relationship with religious attitudes. Archives of Iranian Medicine, 9 (2), 104–107. Assarian, F., Biqam, H. & Asqarnejad, A. (2006). An epidemiological study of obsessive compulsive disorder among high school students and its relationship with religious attitudes. Archives of Iranian Medicine, 9 (2), 104–107. Attaway, N. M. & Bry, H. B. (2004). Parenting style and black adolescents’ academic achievement. Journal of Black Psychology, 2(30), 229-247. http://dx.doi.org/10.1177/0095798403260720 Baharudin, R., Yaacob, S.N., Kahar, R. & Muhamed, A.A. (2003). Dinamik keluarga dan pencapaian anak (In Malay). Serdang, Selangor: Penerbit Universiti Putra Malaysia. Baumrind, D. (1991). Parenting styles and adolescent development. In J. Brooks-Gunn, R. Lerner, and A. C. Peterson (Eds.), The encyclopedia of adolescence (pp. 746-758). New York: Garland. Baumrind, D. (2005). Patterns of parental authority and adolescent autonomy. In J. Smetana (Ed.) New directions for child development: Changes in parental authority during adolescence (pp. 61-69). A San Francisco: Jossey-Bass. Chan, J. J. (2009). Relation of parental, teacher and peer support to academic achievement. School Psychology International, 29(2), 183-198. http://dx.doi.org/10.1177/0143034308090059 Chao, R.K. (2001). Extending research on the consequences of parenting style for Chinese Americans and European Americans. Child Development, 72, 1832-1843. http://dx.doi.org/10.1111/1467-8624.00381 Cooper, H., Lindsay, J. J. & Nye, B. (2000). Homework in the home: How student, family, and parenting style differences relate to the homework process. Contemporary Educational Psychology, 25(4), 464-487. http://dx.doi.org/10.1006/ceps.1999.1036 Durkin, K. (1995). Developmental social psychology: From infancy to old age. Malden, MA: Blackwe Fang, P., Xiong, D. & Guo, C. Y. (2003). The effect of parenting styles on children’s academic achievement. Journal of Psychological Science China, 26 (12), 78-81. http://dx.doi.org/10.1177/1069397110366935 Gonzalez-Pienda, J. A., Nunez, J. C., Gonzalez-Pumariega, S., Alvarez, L., Roces, C. & Garcia, M. (2002). A structural equation model of parental involvement, motivation and attitudinal characteristics, and academic achievement. Journal of Experimental Education, 70(3), 257-289. http://dx.doi.org/10.1080/00220970209599509 Jacobs, N. & Harvey, D. (2005). Do parents make a difference to children’s academic achievement? Differences between parents of higher and lower achieving students. Journal of Educational Studies, 31(4), 431-448. http://dx.doi.org/10.1080/03055690500415746 Kim, K. & Rohner, R. P. (2002). Parental Warmth, Control, and Involvement in Schooling: Predicting Academic Achievement among Korean American Adolescents. Journal of Cross-Cultural Psychology, 33(2), 127-140. http://dx.doi.org/10.1177/0022022102033002001 Lee, S.M., Daniels, M. H. & Kissinger, D. B. (2006). Parental Influences on Adolescent Adjustment. Family Journal, 14(3), 253-259. http://dx.doi.org/10.1177/1066480706287654 Mehrafrooz, H. (2004). 5. Discussion and conclusion Rigid monitoring may demotivate adolescent and children to succeed (Cooper et al., 2000). However, these characteristics do not have any relation to academic achievement amongst high school adolescents in the present study. ISSN 1911-2017 E-ISSN 1911-2025 ISSN 1911-2017 E-ISSN 1911-2025 ISSN 1911-2017 E-ISSN 1911-2025 158 Vol. 8, No. 1; January 2012 Asian Social Science www.ccsenet.org/ass In conclusion, the findings of the present study implied that parents play a significant role in determining the level of academic achievement among adolescents. Thus, it is essential that parents are equipped with appropriate knowledge and skills so that they can provide better guidance for their adolescents’ positive development, especially in academic achievement. References Relationship between parenting style and academic achievement in children’s Iranian. (Unpublished master’s theses, Azad University of Iran). Mozaffari, M. R. (2001). Rate of affective-family and educational problems between Shaheds’high school students and none Shaheds’high school students in Iran. Organization Education of Kerman. Olige, I. C. (2008). Parental influences on student academic achievement. (Doctoral dissertation, Capella University). [Online] Available: http://proquest.umi.com Park. H. & Bauer, S. (2002). Achievement in adolescents parenting practices, ethnicity, socioeconomic status and academic. School Psychology International, 23(4), 386-396. http://dx.doi.org/10.1177/0143034302234002 Pong, S., Johnston, J. & Chen, V. (2010). Authoritarian parenting and Asian adolescent school performance: Insights from the US and Taiwan. International Journal of Behavior Development, 34(1), 62-72. http://dx.doi.org/10.1177/0165025409345073 159 Published by Canadian Center of Science and Education Vol. 8, No. 1; January 2012 Asian Social Science www.ccsenet.org/ass Querido, J. G., Warner, T. D. & Eyberg, S. M. (2002). Parenting styles and child behavior in African American families of preschool children. Journal of Clinical Child and Adolescent psychology, 31, 272-277. Roche, K. M., Ensminger, M. E. & Cherlin, A. J. (2007). Parenting style and adolescent outcomes among African and Latino families living in low income. Journal of Family Issue, 11(23), 882-909. http://dx.doi.org/10.1177/0192513X07299617 Sabouri, P. (2008). Relationship between parenting style and academic achievement. (Unpublished doctoral dissertation, Azad University of Iran). Simons, G. L. & Conger, R. D. (2007). Linking father-mother differences in parenting to a typology of parenting style and adolescent outcomes. Journal of Family Issue, 28(2), 212-241. http://dx.doi.org/10.1177/0192513X06294593 Slaten, C. D. (2006). The effect of parenting style and family structure on academic achievement in rural setting. (Doctoral dissertation, Truman State University). [Online] Available: http://proquest.umi.com Spera, C. (2005). A review of the relationship among parenting practices, parenting styles, and adolescent school achievement. Educational Psychology Review, 17(2), 126-146. http://dx.doi.org/10.1007/s10648-005-3950-1 Spera, C. (2005). A review of the relationship among parenting practices, parenting styles, and adolescent school achievement. Educational Psychology Review, 17(2), 126-146. http://dx.doi.org/10.1007/s10648-005-3950-1 Table 1. Mean and standard deviation for main variables Variable Mean SD Minimum Maximum Authoritative style 38.09 6.46 11 48 Authoritarian style 28.37 6.46 12 46 Permissive style 29.92 6.35 13 49 Academic achievement 17.28 2.32 11 20 Table 2. Distribution of respondents by categories of parenting style and academic achievement Variables N % Parenting style Authoritative style 288 75.4 Authoritarian style 52 13.6 Permissive style 42 11 Academic achievement Excellent 297 77.7 Medium 34 8.9 Weak 51 13.4 Fail 0 0 Table 3. References Correlation matrix for all the variables Variables 2 3 4 1 Academic achievement - .244 -.037 -.16 2 Authoritative style - -.254 .248 3 Authoritarian style - -.243** 4 Permissive style - 160 ISSN 1911-2017 E-ISSN 1911-2025
https://openalex.org/W4280564963	https://www.frontiersin.org/articles/10.3389/fbuil.2022.904483/pdf	English	null	Coastal Natural and Nature-Based Features: International Guidelines for Flood Risk Management	Frontiers in built environment	2,022	cc-by	7,854	PERSPECTIVE published: 11 May 2022 doi: 10.3389/fbuil.2022.904483 PERSPECTIVE Coastal Natural and Nature-Based Features: International Guidelines for Flood Risk Management Todd S. Bridges 1, Jane McKee Smith 1, Jeffrey K. King 1, Jonathan D. Simm 2, Maria Dillard 3, Jurre deVries 4, Denise Reed 5, Candice D. Piercy 1, Boris van Zanten 6, Katie Arkema 7,8, Todd Swannack 1, Harry de Looff 4, Quirijn Lodder 4, Claire Jeuken 9, Nigel Ponte 10, Joseph Z. Gailani 1, Paula Whitﬁeld 11, Enda Murphy 12, Ryan J. Lowe 13, Elizabeth McLeod 14, Safra Altman 1, Colette Cairns 11, Burton C. Suedel 1 and Larissa A. Naylor 15 Todd S. Bridges 1, Jane McKee Smith 1, Jeffrey K. King 1, Jonathan D. Simm 2, Maria Dillard 3, Jurre deVries 4, Denise Reed 5, Candice D. Piercy 1, Boris van Zanten 6, Katie Arkema 7,8, Todd Swannack 1, Harry de Looff 4, Quirijn Lodder 4, Claire Jeuken 9, Nigel Ponte 10, Joseph Z. Gailani 1, Paula Whitﬁeld 11, Enda Murphy 12, Ryan J. Lowe 13, Elizabeth McLeod 14, Safra Altman 1, Colette Cairns 11, Burton C. Suedel 1 and Larissa A. Naylor 15 1US Army Corps of Engineers, Vicksburg, MS, United States, 2HR Wallingford, Wallingford, United Kingdom, 3National Institute of Standards and Technology, Gaithersburg, MD, United States, 4Rijkswaterstaat, Rijswijk, Netherlands, 5Department of Earth and Environmental Sciences, University of New Orleans, New Orleans, LA, United States, 6World Bank, Washington, DC, United States, 7Paciﬁc Northwest National Laboratory, Richland, WA, United States, 8University of Washington, Seattle, WA, United States, 9Deltares USA Inc., Silver Spring, MD, United States, 10Jacobs, London, United Kingdom, 11National Oceanic and Atmospheric Administration, Silver Spring, MD, United States, 12National Research Council Canada, Ottawa, ON, Canada, 13Oceans Graduate School, The University of Western Australia, Perth, WA, Australia, 14The Nature Conservancy, Arlington, VA, United States, 15University of Glasgow, Glasgow, United Kingdom Keywords: natural and nature-based features, ﬂood risk management, reefs, islands, vegetation, wetlands, beaches, dunes Received: 25 March 2022 Accepted: 25 April 2022 Published: 11 May 2022 Received: 25 March 2022 Accepted: 25 April 2022 Published: 11 May 2022 Edited by: Barbara Zanuttigh, University of Bologna, Italy Natural and nature-based features (NNBF) have been used for more than 100 years as coastal protection infrastructure (e.g., beach nourishment projects). The application of NNBF has grown steadily in recent years with the goal of realizing both coastal engineering and environment and social co-beneﬁts through projects that have the potential to adapt to the changing climate. Technical advancements in support of NNBF are increasingly the subject of peer-reviewed literature, and guidance has been published by numerous organizations to inform technical practice for speciﬁc types of nature-based solutions. The International Guidelines on Natural and Nature-Based Features for Flood Risk Management was recently published to provide a comprehensive guide that draws directly on the growing body of knowledge and practitioner experience from around the world to inform the process of conceptualizing, planning, designing, engineering, and operating NNBF. These Guidelines focus on the role of nature-based solutions and natural infrastructure (beaches, dunes, wetlands and plant systems, islands, reefs) as a part of coastal and riverine ﬂood risk management. In addition to describing each of the NNBF types, their use, design, implementation, and maintenance, the guidelines describe general principles for employing NNBF, stakeholder engagement, monitoring, costs and beneﬁts, and adaptive management. An overall systems approach is taken to planning and implementation of NNBF. The guidelines were developed to support decision-makers, project managers, and practitioners in conceptualizing, planning, designing, engineering, implementing, and maintaining sustainable systems for nature- based ﬂood risk management. This paper summarizes key concepts and highlights challenges and areas of future research. Reviewed by: Leopoldo Franco, Roma Tre University, Italy *Correspondence: Jane McKee Smith jane.m.smith@usace.army.mil Specialty section: This article was submitted to Coastal and Offshore Engineering, a section of the journal Frontiers in Built Environment Citation: Bridges TS, Smith JM, King JK, Simm JD, Dillard M, deVries J, Reed D, Piercy CD, van Zanten B, Arkema K, Swannack T, de Looff H, Lodder Q, Jeuken C, Ponte N, Gailani JZ, Whitﬁeld P, Murphy E, Lowe RJ, McLeod E, Altman S, Cairns C, Suedel BC and Naylor LA (2022) Coastal Natural and Nature-Based Features: International Guidelines for Flood Risk Management. Front. Built Environ. 8:904483. doi: 10.3389/fbuil.2022.904483 May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 1 Coastal NNBF Guidelines Bridges et al. INTRODUCTION international growing body of knowledge and experience, consolidating and expanding to provide a systems approach as well as descriptions of fundamental processes and engineering tools. The guide is not a step-by-step design manual, but instead informs the process of conceptualizing, planning, designing, engineering, and operating NNBF based FRM systems. Coastal ﬂood risk management (FRM) reduces future ﬂood damages and erosion by constructing structural measures (seawalls, breakwaters, and revetments), applying natural and nature-based features (NNBF) (beaches, dunes, islands, wetlands, and reefs), and using non-structural measures (coastal retreat). Natural landscapes and features have always contributed to coastal ﬂood resilience. Keen interest in NNBF approaches is driven by the desire for long-term risk mitigation, increased resilience of coastal communities and ecosystems, reduced maintenance, and increased value of FRM investments (Bridges et al., 2015). NNBF use landscape features to reduce ﬂood risk while providing additional economic, social, and environmental co-beneﬁts. NNBF can be used alone, in combination, or with conventional coastal protection infrastructure to reduce ﬂood risk. The use of NNBF in an FRM system depends on the FRM goals, geographic setting, temporal/spatial scales, and other factors. This paper reviews the coastal NNBF content of the recently published International Guidelines on Natural and Nature-Based Features for Flood Risk Management (Bridges et al., 2021). Figure 1 provides an overview of the content and concepts of the guidelines within a system context. This unique, comprehensive guide draws on the Climate change is a major challenge of our time (IPCC, 2021). To address this challenge, NNBF provide ﬂexibility, adaptability, and co- beneﬁts that hard infrastructures alone lack. Innovation is required to expand the knowledge and application of NNBF for practical application, and policies are needed to guide and expand the use of NNBF. NNBF research and application is multidisciplinary, requiring collaboration, coordination, and partnerships across disciplines and stakeholders for success. This paper summarizes NNBF principles; systems approach; performance, beneﬁts, and costs; coastal NNBF systems; enhancing structural measures; and knowledge gaps and research needs. NATURAL AND NATURE-BASED FEATURES PERFORMANCE, BENEFITS, AND COSTS Performance of NNBF projects can be challenging to document because the system is dynamic and some beneﬁts accrue slowly (e.g., beach ﬁll adjustments, colonization of ﬂora and fauna) (Piercy et al., 2021b). Performance of an NNBF project is the ability to meet desired outcomes as measured by a set of metrics. Metrics are speciﬁc parameters or properties of the NNBF that are quantiﬁable and associated with desired aspects of performance. Metrics include direct and indirect measurements and are assessed based on predetermined performance criteria. Sources of uncertainty in NNBF performance are similar to structural measures, although natural variability in NNBF is greater. Monitoring metrics should be chosen to capture the most critical project aspects. Periodic assessment at a frequency commensurate with nature dynamism are required over the project life cycle. Both NNBF and structural measures need better assessment of long-term failure rates (fragility) and maintenance costs. FRM and ecological, social, and economic performance of NNBF are interrelated, and proper ecological function of NNBF is critical to FRM, social, and economic functions. There are several interrelated categories of performance related to NNBF: NNBF projects require a deeper commitment to engagement than other types of projects (Dillard et al., 2021). The engagement process should be iterative, ﬂexible, and feed into all phases of the project (deﬁning the problem, developing/evaluating alternatives, maintenance, monitoring, and evaluation). Engagement should include all who are interested in, have inﬂuence over, or are impacted by the project, and these stakeholders should be involved early and kept informed throughout. Strong engagement increases the likelihood of multiple beneﬁts and beneﬁciaries. NNBF projects should include an engagement plan (focused on project objectives) with resources to carry it out. NNBF projects generally evolve in different ways and rates than conventional, hard infrastructure, so communication with stakeholders about expectations is key (e.g., features may erode during storms and need maintenance; and protection may be subaqueous and not easily visible). Engagements should be planned (objectives, context, resources, methods), documented, and evaluated, but also ﬂexible and adaptive. Stakeholders can assist in successfully involving press, politics, and the broader community. • FRM—reduction of physical forces that produce ﬂooding and damages for the full range of possible events (system and structural performance). • Ecological—production of desired ecological function (ecosystem goods and services). PRINCIPLES Applying NNBF to FRM projects is similar to standard engineering approaches, but emphasizes a multidisciplinary, systems approach. Key NNBF principles are (Bridges et al., 2021): FIGURE 1 \| Contents and concepts of NNBF guidelines within a systems context (Bridges et al., 2021). FIGURE 1 \| Contents and concepts of NNBF guidelines within a systems context (Bridges et al., 2021). May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org Frontiers in Built Environment \| www.frontiersin.org 2 Coastal NNBF Guidelines Bridges et al. • Expect change and manage adaptively engagement with stakeholders. NNBF solutions evolve and develop over time and space (e.g., accretion of sediment and growth and succession of vegetation). A systems perspective is used to implement a large-scale system or integrate effects of small or standalone projects as part of a larger FRM effort. Since some aspects and applications of NNBF are new, there are limited examples of mature projects. It is important to document case studies to help mainstream the application of NNBF and highlight the potential for wide-ranging beneﬁts. • Use a systems approach to leverage existing components and projects and their interconnectivity • Engage communities, partners, and multidisciplinary team members to develop innovative solutions and monitor social, economic, and environmental outcomes • Anticipate, evaluate, and manage risk in project or system performance The International Guidelines includes a project framework for NNBF applications (King et al., 2021). The design process is iterative and may require revisiting previous phases based on new data or information. Desirable outcomes are identiﬁed at each phase: scoping (stakeholder engagement/commitment), planning (funding, beneﬁts), decision making (option selection), implementation (approvals, optimization), and operations (monitoring/adaption). Frontiers in Built Environment \| www.frontiersin.org SYSTEMS APPROACH TO PLANNING AND IMPLEMENTING NATURAL AND NATURE-BASED FEATURES • Social—social co-beneﬁts include health, well-being, and equity, as well as recreational, cultural, and educational beneﬁts. A systems approach is needed to deﬁne, characterize, and manage multifunctional and sustainable NNBF solutions (de Vries et al., 2021). The interplay of ecology, geomorphology, and hydrodynamics with communities and engineering infrastructure provides the context of the system. Evaluating FRM solutions requires assessing physical, biological, and social processes and their interaction and synergies. Holistic, system-wide solutions can be explored through collaboration of people with different perspectives and disciplines to accelerate identiﬁcation and implementation of suitable, resilient, and well-functioning solutions. Systems thinking is needed to evaluate the variety of ecosystem beneﬁts, multiple potential outcomes, multifunctional design, and direct • Economic—reduction in economic damages and value (ecological, social, and FRM) produced by the NNBF. NNBF project performance should be considered over the entire project life cycle, including present and future conditions. NNBF performance may deteriorate with time and require maintenance to preserve function (e.g., beach/dune renourishment). Unlike structural measures, NNBF may naturally adapt to future conditions (e.g., through marsh sedimentation or vertical growth of oyster or coral reefs), and performance may improve over time. Performance concepts also apply to natural features existing within projects. Performance measurement/monitoring is key to adaptive management (de Loof et al. 2021). May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 3 Coastal NNBF Guidelines Bridges et al. The goal of protecting, restoring, and advancing NNBF is to reduce ﬂood risk and erosion, adapt to climate change, and build coastal resilience (van Zanten et al., 2021). Additionally, NNBF produces ecological, social, and economic co-beneﬁts that vary with local conditions and stakeholder values. Evaluating NNBF alongside structural alternatives requires evaluating relative costs and beneﬁts, including both ﬂood risk reduction and co-beneﬁts. Flood and erosion risk reduction is achieved via different processes depending on the type of NNBF, including trapping sediment, wave damping, and water storage. Advances in numerical modeling have increased the capacity to value NNBF risk reduction beneﬁts. Co-beneﬁts include habitat for ﬁsheries, opportunities for tourism and recreation, carbon sequestration, and human health beneﬁts (Barbier et al., 2011). Agencies and stakeholders deﬁne shared objectives and develop performance standards that capture economic, social, and ecological beneﬁts of implementing NNBF. Guidance is needed to better incorporate co-beneﬁts into assessment of alternatives and monitoring of outcomes. Coastal Wetlands and Plant Systems Coastal Wetlands and Plant Systems Wetlands and intertidal areas dampen waves and surge and trap sediments (Piercy et al., 2021a). Wetlands provide tens of thousands to millions of US dollars in ﬂood damage reduction/km2/yr, depending on location and conﬁguration (Sun and Carson, 2020). These features provide co-beneﬁts including ﬁsh production, ﬁltration of pollutants, water-quality mediation, recreation, and carbon sequestration. Coastal wetlands and tidal ﬂats reduce coastal ﬂood and erosion risk with raised bed levels and frictional resistance to attenuate waves and surge. NNBF projects include conservation of existing wetlands, restoration of degraded wetlands, or construction of new wetlands and may be combined with other structural or NNBF measures. Wetland performance is controlled by location, coastline geometry, and storm characteristics. Signiﬁcant wave reduction can occur within relatively narrow feature widths (10’s of meters), while reduction of surge requires greater cross-shore extents (100–1,000’s of meters) (Piercy et al., 2021a). In some conﬁgurations, coastal wetlands may serve as ﬂood storage areas to reduce water levels in estuarine environments. Coastal wetland NNBF projects draw on extensive experience in restoration of marshes and mangroves which has created thousands of hectares worldwide over recent decades. The performance of wetland and tidal ﬂat NNBF projects may vary over time as vegetation establishes and develops. Designs should consider storm damage, recovery, and maintenance requirements. Wetlands have the potential to be self-sustaining under climate change if there is sufﬁcient sediment for accretion and space to adapt. NATURAL AND NATURE-BASED FEATURES IN COASTAL SYSTEMS Coastal NNBF creates or recreates natural habitats, enhances existing habitats, uses more-organic materials, and enhances existing hard infrastructure (Simm, 2021). The coastal landscape provides risk mitigation and is the foundation that supports structural measures. FRM systems may be single lines or multilayer, multizone cross-shore systems with both structural and NNBF elements. The coastal environment is in continual change, both cyclical and continuous, e.g., nonstationary waves (varying energy, frequency, and direction) and water levels (tides, surge, sea level rise (SLR)), evolving cross-shore proﬁle and shoreline, and growing population and development. Coastal NNBF includes beaches and dunes, wetlands and plant systems, islands, and reefs. The maturity of knowledge about application of different features varies signiﬁcantly. The international NNBF guidelines describe these features in detail, including a conceptual understanding; objectives and metrics; design; implementation; monitoring, maintenance, and adaption; and gaps and future directions. Extensive case studies, examples, and references are provided. SYSTEMS APPROACH TO PLANNING AND IMPLEMENTING NATURAL AND NATURE-BASED FEATURES Multiple valuation approaches and metrics can be used to assess beneﬁts and social vulnerability outcomes qualitatively and quantitatively, and different metrics and decision-support tools are suitable for different audiences. Societal beneﬁts can be assessed by combining hazard analysis with maps of social indicators to estimate the number of people and critical infrastructure at risk and compare risk reduction beneﬁts from NNBF. Co-beneﬁts may expand funding and ﬁnancing options. The costs in the total lifecycle of NNBF projects include: planning; design and permitting; land acquisition; creation, protection or restoration; and monitoring and maintenance. resource, and provide habitat for diverse species (Lodder et al., 2021). They reduce land loss and ﬂooding while providing recreational and environmental beneﬁts. Beaches are acted on by wind and hydrodynamic processes (waves, currents, and water levels). Dunes accumulate wind-blown sand and are stabilized by vegetation or control structures and naturally buffer against ﬂooding during extreme events. The shape and size of beaches and dunes are a function of geology, geomorphology, wind and wave regime, and tidal range. A sustainable beach and dune system requires sufﬁcient sediment sources (longshore and cross- shore) and appropriate ﬂora and fauna. Beach and dune design typically aligns with the natural system, particularly regarding grain size, to reduce maintenance. Design should allow for beach proﬁle dynamism, focusing design on elevation, volume, slope, and width. Sediment budgets and modeling of hydrodynamic and sediment processes aid understanding of the physical system dynamics to determine scale and feasibility of projects. Designs must consider past and future scenarios, including socioeconomic development and climate change impacts (SLR, storm frequency and intensity). Management requirements, strategies, and monitoring must be part of the design. Innovative sand placement techniques tap nature to enhance coastal resilience (e.g., nearshore placement to winnow ﬁnes, sand fencing to build dunes). Beaches and dunes can gradually adapt to climate change when supplied with enough sediment and space to adjust. Maintenance can be adapted over time to respond to system nonstationary. Beaches and Dunes Beaches are dynamic coastal landforms, constitute a natural transition between land and sea, are an amenity and economic May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 4 Coastal NNBF Guidelines Bridges et al. including ﬁsheries, habitat and biodiversity, recreation and tourism, and improved water quality. Reef organisms in a healthy system produce calcium carbonate that is a source of sand nourishment to adjacent beaches. Design and construction of a NNBF reef should mimic the natural geomorphology of a pre- existing or existing reef platform to favor biological growth, and materials should be compatible with those in the surrounding environment. Adaptive management is needed to support reef resilience in the context of global environmental change. Natural and engineered reefs can be self-sustaining ecosystems, continuing to grow and maintain a structure to keep pace with SLR, if the reef accretion rate exceeds the rate of SLR. Understanding the coastal protection services of reefs and quantifying how they reduce risk are required to effectively use reefs in coastal climate adaptation and hazard mitigation strategies. Submerged aquatic vegetation (SAV), kelp, dune grass, wetland vegetation, mangroves, and maritime forests are examples of coastal plant systems that provide both above and below ground beneﬁts (Altman et al., 2021). These systems reduce wind, wave, and current energy and stabilize sediments. The magnitude of wave attenuation depends on the height of the canopy relative to the total water depth; stem diameter, rigidity, and density; and plant morphology, as well as hydrodynamic conditions (waves, currents, water levels). The protective value is maximized when canopy height is equal or greater than the water depth. Plants provide ecological beneﬁts, nursery habitat for ﬁsh and shellﬁsh, and water quality improvement. Plant systems complement other NNBF techniques and should be considered for use in larger NNBF projects that incorporate multiple features or small-scale projects in low-energy environments. It is critical to match the plant system to the site. Vegetation habitats are spatially dynamic, and thus robust monitoring is required to understand their condition and health trajectory. SAV beneﬁts are optimized in low-energy environments or in conjunction with other techniques to reduce wave and current energy. The protective role of coastal dune vegetation is widely acknowledged, but not extensively quantiﬁed. Through their natural ability to protect shorelines against erosion and ﬂood risk and adjust to SLR, vegetated systems play an important role in sustainable coastal FRM strategies. Islands Island in estuaries, river deltas, and open-coast environments reduce the severity of coastal hazards, including erosion and ﬂooding from waves and extreme water levels (Gailani et al., 2021). They can provide multiple beneﬁts: storm surge reduction, wave dissipation, erosion control, dredged material management, safe navigation/harbor, ecosystem diversity (critical ecosystem function for threatened and endangered species), recreation, and commercial opportunities. Islands can be newly constructed or restorations of islands degrading due to SLR, subsidence, or inadequate sediment input. Islands deliver resilience beneﬁts, especially as part of multiple-lines- of-defense strategies and may be effective where other NNBF methods are not feasible (urban or high energy areas). The regional inﬂuence of islands thus must be considered (circulation, water quality, sediment transport, and habitat). Islands are multihabitat features (beach/dune, wetlands, and upland), and habitat trade-offs are inevitable. Short- term impacts must be considered within the context of long-term ecosystem co-beneﬁts and SLR. Uncertainties and risk exist in island construction due to the complex physical processes. Design and maintenance of resilient islands requires consideration of the evolving, dynamic conditions. Beaches and Dunes Plant systems are subject to natural habitat succession, so it is critical to understand that functional plant-based NNBF may not persist as a particular habitat type in perpetuity. Changing ocean temperatures and SLR must be considered in planning for species succession. Reefs provide coastal risk reduction by dissipating waves as they propagate over shallow, rough reef structures, thus reducing wave-driven coastal ﬂooding. The effectiveness of reefs depends on their size, orientation, elevation, and location relative to shore. The reef crest elevation relative to sea level is a key parameter. Reef degradation also decreases wave attenuation as the elevation and roughness reduce over time. Reefs promote shoreline stability and play critical roles in protecting and establishing other coastal habitats (seagrass beds in protected lagoons, mangrove forests, beach/dune systems). For example, by reducing wave energy and improving water quality, shellﬁsh reefs can provide suitable conditions for salt marshes and seagrass beds. These multiple layers of protection may be the most effective strategy when habitats are interconnected and functioning together. ENHANCING STRUCTURAL MEASURES Conventional FRM includes structures such as bulkheads, seawalls, sheet piling, and ﬂoodwalls, in addition to levees and dikes that may combine earthen, rock, and concrete structures. The environmental value of these conventional structures may be enhanced through inclusion of nature-based elements that expand their ecological value by enhancing habitat or social beneﬁts (Suedel et al., 2021). Enhancements through NNBF take multiple forms and span broad scales and structure types, from applying small-scale features to create habitat to large structures to dissipate waves and surge. Ecological enhancements can offer multiple beneﬁts, e.g., increased engineering design life and ﬂood risk reduction, habitat enhancement, improved water quality, and societal beneﬁts (Naylor et al. 2017). Opportunities to enhance structures occur at any stage in the design life of a project (new construction, repair, maintenance, or structure modiﬁcation). Identifying and quantifying the value of NNBF enhancement allows costs and beneﬁts to be compared alongside conventional structures (Naylor et al. 2018). Engaging stakeholders in the process of identifying opportunities to enhance value and beneﬁts broadens the base of support for infrastructure projects and aids in complying with environmental laws and regulations. The best opportunity to enhance structures is often during maintenance, repairs, and modiﬁcations to existing structures. Frontiers in Built Environment \| www.frontiersin.org Reefs Coral, rock, and shellﬁsh reefs reduce ﬂooding and erosion in coastal areas by reducing wave energy. For example, coral reefs in the United States provide more than USD$1.8 billion/yr in ﬂood risk beneﬁts (Reguero et al., 2021). Many reef-lined coasts act as the ﬁrst line of defense against ﬂooding, storm damage, and erosion (Lowe et al., 2021). Reefs also provide numerous co-beneﬁts, May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 5 Coastal NNBF Guidelines Bridges et al. TABLE 1 \| Knowledge gaps and research needs. NNBF category Knowledge gaps Planning and implementing NNBF using a systems approach • Comprehensive systems-thinking lessons learned and case studies, including mature projects • Evidence base to mainstream systems thinking and NNBF. • Best practices to account for expansive consideration of co-beneﬁts • Incorporation of multiple spatial and temporal scales and their feedback to support NNBF design and operation NNBF performance measurement • Methods/tools to characterize the dynamic nature of NNBF performance • Performance metrics that describe how individual measures and the FRM system as a whole serve to reduce inherent and residual risks associated with ﬂood hazards • Methods to better capture risk evaluation across project life cycles in nonstationary systems • Holistic quantiﬁcation and consideration of beneﬁts and co-beneﬁts of NNBF performance and sustainability Beneﬁts and costs • Quantitative methods to account for fragility of NNBF to coastal hazards (fragility curves) • Better alignment between ﬁnancing models for NNBF and beneﬁt assessment methodologies to unlock funding sources • New methodologies to accurately reﬂect risk reduction and co-beneﬁts of NNBF projects • Combined ﬁeld and modeling studies to monitor the long-term social and economic outcomes of NNBF. Frontiers in Built Environment \| www.frontiersin.org CONCLUSION The combination of aging infrastructure and climate change is prompting new thinking and practices regarding FRM strategies, viewed broadly to include long-term beneﬁts and value, including economic damages avoided, co-beneﬁts, and greater system resilience. NNBF solutions are an important part of future FRM strategies and require a multidisciplinary approach. Effective and timely implementation of NNBF to address FRM challenges depend on progress in three overarching areas: developing and delivering, communicating and collaborating, and elevating and educating. A systems approach is required to provide a comprehensive vision for process functions, relationships, and engineering interventions critical to successful outcomes. Reefs NNBF category Knowledge gaps Plant systems • Collect baseline data on hazard reduction and plant system evolution and resiliency, especially in changing conditions • Develop metrics for baseline information such as beach erosion rate and beach proﬁle change to quantify the contribution of SAV to coastal protection • Advance use of remote sensing to monitor and evaluate the efﬁcacy of SAV in providing storm protection and integrate into models Enhancing beneﬁts • Widely distributed documentation of existing projects and conduct pilot studies to establish proof of concepts • Education, training, and tech transfer of case studies and workshops to disseminate best practices and coordinate of site visits to observe implementation of innovations in practice • Demonstration projects to demonstrate and document successes in NNBF enhancements on multiple spatial and temporal scales • Consider emerging technologies to improve planning and practice: materials, observation methods for planning and monitoring (e.g., remote sensing), model advancements for evaluating design and project siting • Novel designs for creating habitat niches for a variety of species and replace conventional hard structures with hybrid structures • New materials to reduce carbon dioxide emissions during construction • Document and proliferate the most valuable advances Knowledge gaps • Collect baseline data on hazard reduction and plant system evolution and resiliency, especially in changing conditions • Develop metrics for baseline information such as beach erosion rate and beach proﬁle change to quantify the contribution of SAV to coastal protection • Develop metrics for baseline information such as beach erosion rate and beach proﬁle change to quantify the contribution of SAV to coastal protection • Advance use of remote sensing to monitor and evaluate the efﬁcacy of SAV in providing storm protection and integrate into models • Widely distributed documentation of existing projects and conduct pilot studies to establish proof of concepts p p • Demonstration projects to demonstrate and document successes in NNBF enhancements on multiple spatial and temporal scales • Consider emerging technologies to improve planning and practice: materials, observation methods for planning and monitoring (e.g., remote sensing), model advancements for evaluating design and project siting • Novel designs for creating habitat niches for a variety of species and replace conventional hard structures with hybrid structures • Consider emerging technologies to improve planning and practice: materials, observation methods for planning and monitoring (e.g., remote sensing), model advancements for evaluating design and project siting • Novel designs for creating habitat niches for a variety of species and replace conventional hard structures with y • New materials to reduce carbon dioxide emissions during c • Document and proliferate the most valuable advances DATA AVAILABILITY STATEMENT The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author. INNOVATION AND RESEARCH NEEDS fully leverage natural systems and functions to address the dynamic challenges posed by FRM and climate change. Engagement, communication, and collaboration that bridge the gaps between technical disciplines, organizations, and the public are fundamental to successful project development and implementation through purposeful investment in communications and engagement. Many environmental and social co-beneﬁts are challenging to quantify but critical to establishing a project’s value (e.g., biodiversity, social equity). Monitoring and adaptive management of NNBF is critical to the ultimate success of projects. The International Guidelines on Natural and Nature-Based Features for Flood Risk Management support developing and delivering FRM projects that incorporate NNBF to provide resilient solutions over the long term. Next steps include developing guiding policy and engineering manuals to support NNBF design, implementation, and operations. The understanding and acceptance of NNBF for FRM is increasing quickly, but many knowledge, experience, and policy gaps remain. NNBF technical research often focuses on speciﬁc features, feature elements, or speciﬁc sites; integrating across such approaches is needed to provide generalized, wide- ranging solutions. Table 1 summarizes knowledge gaps and future research areas. Investment in research and development will inform and fuel future advancements in practice for both conventional and nature-based approaches to FRM. AUTHOR CONTRIBUTIONS TB contributed to conception and design of the study. JaS wrote the ﬁrst draft of the manuscript. TB, JaS, JK, JoS, MD, JV, DR, CP, BZ, KA, TS, HL, QL, CJ, NP, JG, PW, EM, RL, EM, SA, CC, BS, and LN contributed to sections of the manuscript. All authors contributed to manuscript revision, read and approved the submitted version. Expanding the engineering application of NNBF for FRM requires continued advancement of the supporting interdisciplinary science as well as advancing policy and economic valuations. Integrative approaches to FRM that include NNBF measures require advancing models and modeling practice with respect to natural features, processes, systems, and uncertainty quantiﬁcation in the dynamics of physical and natural systems. Advancement of environmental regulation and management to conserve and protect natural systems is also required. These advances build on the growing recognition of natural values and capital so that communities can Frontiers in Built Environment \| www.frontiersin.org Reefs TABLE 1 \| (Continued) Knowledge gaps and research needs. NNBF category Knowledge gaps Plant systems • Collect baseline data on hazard reduction and plant system evolution and resiliency, especially in changing conditions • Develop metrics for baseline information such as beach erosion rate and beach proﬁle change to quantify the contribution of SAV to coastal protection • Advance use of remote sensing to monitor and evaluate the efﬁcacy of SAV in providing storm protection and integrate into models Enhancing beneﬁts • Widely distributed documentation of existing projects and conduct pilot studies to establish proof of concepts • Education, training, and tech transfer of case studies and workshops to disseminate best practices and coordinate of site visits to observe implementation of innovations in practice • Demonstration projects to demonstrate and document successes in NNBF enhancements on multiple spatial and temporal scales • Consider emerging technologies to improve planning and practice: materials, observation methods for planning and monitoring (e.g., remote sensing), model advancements for evaluating design and project siting • Novel designs for creating habitat niches for a variety of species and replace conventional hard structures with hybrid structures • New materials to reduce carbon dioxide emissions during construction • Document and proliferate the most valuable advances TABLE 1 \| (Continued) Knowledge gaps and research needs. TABLE 1 \| (Continued) Knowledge gaps and research needs. Reefs • Advanced Earth observation technology to assess and monitor NNBF beneﬁts Adaptive management • Community Adaptive-Management deﬁnitions and frameworks • Efforts to improve leadership and stakeholder acceptance of the concept of Adaptive Management • Flexibility in laws and policy to account for shifting baselines to capture future conditions • Flexibility in funding to facilitate data collection, analysis, and active adaptive management Beaches and dunes • Document evidence base on the long-term performance of beach and dune NNBF for coastal resilience under a wide range of conditions and environmental settings • Enhance understanding, uptake and upscaling, and improved implementation of beach and dune NNBF in coastal resilience projects • Develop and implement long-term strategies for sustainable coastlines in relation to long-term processes such as climate change and socioeconomic developments • Quantify dune vegetation impacts on dune growth, erosion, and long-term sustainability • Enhance existing beach evolution models for combined hard and soft structures under complex loading (nonstationary wind, wave, and water level forcing) Wetlands • More ﬁeld and modeling studies to address long-term wetland stability to compare with effectiveness of other nonstructural and structural measures • Consistent cost-beneﬁt framework that accounts for full array of beneﬁts, co-beneﬁts, and life-cycle costs • Expand knowledge base of wetland NNBF performance under different conditions (extreme storms, high water levels, and SLR) • Methods/tools to quantify system-scale beneﬁts and co-beneﬁts of wetland NNBF and linkage between wetland NNBF and other measures • Improvements to coastal wetland hydrodynamic and sediment models to include wetland NNBF more accurately and efﬁciently in modeling scenarios • Quantify parameters to determine sustainability of marshes Islands • Research on combined, complementary effect of multiple habitat types from offshore to onshore in terms of both short- and long-term beneﬁts to justify larger NNBF island projects • Quantitative studies of island areas of inﬂuences to address habitat switching and potential impacts of island restoration • Innovative practices and ﬁeld experimentation (living labs) to improve understand of island systems • Regional case studies to illustrate and optimize design, implementation, and maintenance approaches • Enhance models for long-term erosion and recovery of islands, spanning long and short time and space scales Reefs • Methods to optimize performance of reef NNBF to achieve ecological and coastal protection beneﬁts at larger scales • Long-term monitoring to understand changes to sources and composition of coastal sediments and long-term evolution of shorelines protected by reefs • Advances to support increased survival and ﬁtness of shellﬁsh reared in hatcheries and selective breeding programs for disease resistance • Economic studies that account for the full suite of ecosystem beneﬁts that natural reefs provide to incentivize protection and restoration of coral and shellﬁsh reefs • Improve and ground truth predictions of reef effectiveness in reducing wave-drive ﬂood risk across a range of Knowledge gaps • Consistent cost-beneﬁt framework that accounts for full array of beneﬁts, co-beneﬁts, and life-cycle costs • Expand knowledge base of wetland NNBF performance under different conditions (extreme storms, high water levels, and SLR) • Methods/tools to quantify system-scale beneﬁts and co-beneﬁts of wetland NNBF and linkage between wetland NNBF and other measures • Improvements to coastal wetland hydrodynamic and sediment models to include wetland NNBF more accurately and efﬁciently in modeling scenarios • Quantify parameters to determine sustainability of marshes • Research on combined, complementary effect of multiple habitat types from offshore to onshore in terms of both short- and long-term beneﬁts to justify larger NNBF island projects • Quantitative studies of island areas of inﬂuences to address habitat switching and potential impacts of island restoration • Innovative practices and ﬁeld experimentation (living labs) to improve understand of island systems • Innovative practices and ﬁeld experimentation (living labs) to improve understand of island systems • Regional case studies to illustrate and optimize design, implementation, and maintenance approaches • Enhance models for long-term erosion and recovery of islands, spanning long and short time and space scale g • Regional case studies to illustrate and optimize design, implementation, and maintenance approaches • Enhance models for long-term erosion and recovery of islands, spanning long and short time and space scales • Methods to optimize performance of reef NNBF to achieve ecological and coastal protection beneﬁts at larger scales • Long-term monitoring to understand changes to sources and composition of coastal sediments and long-term evolution of shorelines protected by reefs • Advances to support increased survival and ﬁtness of shellﬁsh reared in hatcheries and selective breeding programs for disease resistance • Economic studies that account for the full suite of ecosystem beneﬁts that natural reefs provide to incentivize protection and restoration of coral and shellﬁsh reefs • Additional studies to monitor and document the efﬁcacy of materials used in reef construction • Innovative solutions to accelerate the broader application of reef NNBF, further increase the cost-effectiveness of designs that are appropriate for a range of local socioeconomic conditions (Continued on following page) (Continued on following page) May 2022 \| Volume 8 \| Article 904483 6 Coastal NNBF Guidelines Bridges et al. REFERENCES “Adaptive Management,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. Vicksburg, MS: U.S. Army Engineer Research and Development Center, 270–318. Naylor, L., Coombes, M., Kippen, H., Horton, B., Gardiner, T., Roca Collell, M., et al. (2018). “Developing a business case for greening hard coastal and estuarine infrastructure: preliminary results,” in Marine Structures and Breakwaters 2017: Realising the Potential. Series: ICE Coasts, Maritime Structures and Breakwaters Conference Series. Editors K. Burgess (Vicksburg, MS: ICE Publishing), 801–814. ISBN 9780727763174. doi:10.1680/cmsb.63174.0801 de Vries, J., Reed, D., McKay, S. K., Bouma, T., Cunniff, S. E., Brasmeijer, B., et al. (2021). “Planning and Implementing NNBF Using a Systems Approach,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 103–160. Piercy, C. D., Pontee, N., Narayan, S., Davis, J., and Meckley, T. (2021a). “Coastal Wetlands and Tidal Flats,” in International Guidelines on Natural and Nature- Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 413–500. Dillard, M., Brooks, C., Fisher, H., Pietersen, H., and Nijhuis, A. (2021). “Engaging Communities and Stakeholders in Implementing NNBF,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 47–102. Piercy, C. D., Simm, J. D., Bridges, T. S., Hettiarachchi, M., and Lodder, Q. (2021b). “NNBF Performance,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 161–202. Gailani, J., Whitﬁeld, P., Murphy, E., Thomson, G., Mears, G., deVries, G., et al. (2021). “Islands,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. REFERENCES Lodder, Q. C., Jeuken, C., Reinen-Hamill, R., Burns, O., Ramsdell, R., McFall, B., et al. (2021). “Beaches and Dunes,” in International Guidelines on Natural and Nature- Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 329–412. Altman, S., Cairns, C., Whitﬁeld, P., Davis, J., Finkbeiner, M., and McFall, B. (2021). “Plant Systems: Submerged Aquatic Vegetation and Kelp,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. Vicksburg, MS: U.S. Army Engineer Research and Development Center, 363–700. Lowe, R. J., Mc Leod, E., Reguero, B. G., Altman, S., Harris, J., Hancock, B., et al. (2021). “Reefs,” in International Guidelines on Natural and Nature- Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 559–635. Barbier, E. B., Hacker, S. D., Kennedy, C., Koch, E. W., Stier, A. C., and Silliman, B. R. (2011). The Value of Estuarine and Coastal Ecosystem Services. Ecol. Monogr. 81 (2), 169–193. doi:10.1890/10-1510.1 IPCC (2021). Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Editors V. Masson- Delmotte, P. Zhai, A. Pirani, S. L. Connors, C. Péan, S. Berger, et al. (Cambridge: Cambridge University Press). T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (2021). International Guidelines on Natural and Nature-Based Features for Flood Risk Management (Vicksburg, MS: U.S. Army Engineer Research and Development Center) Available at: https://ewn.erdc.dren.mil/?page_id=4351. Bridges, T. S., Wagner, P. W., Burks-Copes, K. A., Bates, M. E., Collier, Z. A., Fischenich, C. J., et al. (2015). Use of Natural and Nature-Based Features (NNBF) for Coastal Resilience (Vicksburg, MS: U.S. Army Engineer Research and Development Center). ERDC SR-15-1. Naylor, L. A., Kippen, H., Coombes, M. A., Horton, B., MacArthur, M., and Jackson, N. (2017). “Greening the Grey: A Framework for Integrated Green Grey Infrastructure (IGGI). Technical Report. Glasgow: University of Glasgow). http://eprints.gla.ac.uk/150672/. de Looff, H., Welp, T., Snider, N., and Wilmink, R. (2021). ACKNOWLEDGMENTS The partner organization of the U.S. Army Corps of Engineers (USACE), National Oceanic and Atmospheric Administration May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 7 Coastal NNBF Guidelines Bridges et al. (University of California, Santa Cruz), Giselle Samonte (NOAA), Steven Scyphers (Northeastern University), Elizabeth Codner- Smith (TNC), Stephanie IJff (Deltares), Marin Kress (USACE), Michele Lemay (Inter-American Development Bank), Tim Welp (USACE), Natalie Snider (Environmental Defence Fund), Rinse Wilmink (Rijkswaterstaat), Richard Rienen-Hamil (Tonkin + Taylor), Oli Burns (EA), Robert Ramsdell (Dredging Consultant), Brian McFall (USACE), Corragio Maglio (USACE), Jenny Davis (NOAA), Trevor Meckley (NOAA), Gordon Thomson (Baird), Wendell Mears (Anchor QEA, LLC), Danielle Szimanski (USACE), Janine Harris (NOAA), Boze Hancock (TNC), Remment ter Hofstede (Van Oord), Emma Rendel (Resilient Coasts), Elizabeth Shaver (TNC), Mark Finkbeiner (NOAA), and Jason Bernier (CBCL Limited). (NOAA), Environment Agency of the United Kingdom (EA), Rijkswaterstaat, and the World Bank supported this work with contributions from the National Institute of Standards and Technology, The Nature Conservancy (TNC), and the World Wildlife Fund (WWF). Additional contributors include: Cath Brooks (EA), Helen Fisher (3KQ), Hans Pietersen (Rijkswaterstaat), Alwin Nijhuis (Rijkswaterstaat), Anita van Breda (WWF), Susan Durden (USACE), S. Kyle McKay (USACE), Tjeerd Bouma (Royal Netherlands Institute for Sea Research Shannon Cunniff (Environmental Defence Fund), Bart Grasmeijer (Deltares), Pamela Mason (College of William & Mary), Alex Nicholson (Arup), Paul Quinn (Newcastle University), Missaka Hettiarachchi (WWF), Rob Grifﬁn (University of Massachusetts-Dartmouth), Siddharth Narayan (East Carolina University), Kim Penn (NOAA), Borja Reguero (NOAA), Environment Agency of the United Kingdom (EA), Rijkswaterstaat, and the World Bank supported this work with contributions from the National Institute of Standards and Technology, The Nature Conservancy (TNC), and the World Wildlife Fund (WWF). Additional contributors include: Cath Brooks (EA), Helen Fisher (3KQ), Hans Pietersen (Rijkswaterstaat), Alwin Nijhuis (Rijkswaterstaat), Anita van Breda (WWF), Susan Durden (USACE), S. Kyle McKay (USACE), Tjeerd Bouma (Royal Netherlands Institute for Sea Research Shannon Cunniff (Environmental Defence Fund), Bart Grasmeijer (Deltares), Pamela Mason (College of William & Mary), Alex Nicholson (Arup), Paul Quinn (Newcastle University), Missaka Hettiarachchi (WWF), Rob Grifﬁn (University of Massachusetts-Dartmouth), Siddharth Narayan (East Carolina University), Kim Penn (NOAA), Borja Reguero REFERENCES Army Engineer Research and Development Center), 501–558. Reguero, B. G., Storlazzi, C. D., Gibbs, A. E., Shope, J. B., Cole, A. D., Cumming, K. A., et al. (2021). The Value of US Coral Reefs for Flood Risk Reduction. Nat. Sustain 4, 688–698. doi:10.1038/s41893-021- 00706-6 Simm, J. (2021). “Introduction to NNBF in Coastal Systems,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 319–328. King, J. K., Simm, J. D., and Bridges, T. S. (2021). “Principles, Framework, and Outcomes,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 19–46. May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 8 Bridges et al. Coastal NNBF Guidelines The remaining authors declare that the research was conducted in the absence of any commercial or ﬁnancial relationships that could be construed as a potential conﬂict of interest. Suedel, B. C., Naylor, L. A., Meckley, T., Cairns, C., Bernier, J., Morgereth, E., et al. (2021). “Enhancing Structural Measures for Environmental, Social, and Engineering Beneﬁts,” in International Guidelines on Natural and Nature- Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 701–771. Publisher’s Note: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their afﬁliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Sun, F., and Carson, R. T. (2020). Coastal Wetlands Reduce Property Damage during Tropical Cyclones. Proc. Natl. Acad. Sci. U.S.A. 117 (11), 5719–5725. doi:10.1073/pnas.1915169117 van Zanten, B., Arkema, K., Swannack, T., Grifﬁn, R., Narayan, S., Penn, K., et al. (2021). “Beneﬁts and Costs of NNBF,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. van Zanten, B., Arkema, K., Swannack, T., Grifﬁn, R., Narayan, S., Penn, K., et al. (2021). “Beneﬁts and Costs of NNBF,” in International Guidelines on Natural and Nature-Based Features for Flood Risk Management. Editors T. S. Bridges, J. K. King, J. D. Simm, M. W. Beck, G. Collins, Q. Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 203–269. Frontiers in Built Environment \| www.frontiersin.org Conﬂict of Interest: Author JoS was employed by HR Wallingford, author CJ was employed by Deltares United States Inc., and author NP was employed by Jacobs. May 2022 \| Volume 8 \| Article 904483 REFERENCES Lodder, et al. (Vicksburg, MS: U.S. Army Engineer Research and Development Center), 203–269. Copyright © 2022 Bridges, Smith, King, Simm, Dillard, deVries, Reed, Piercy, van Zanten, Arkema, Swannack, de Looff, Lodder, Jeuken, Ponte, Gailani, Whitﬁeld, Murphy, Lowe, McLeod, Altman, Cairns, Suedel and Naylor. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Conﬂict of Interest: Author JoS was employed by HR Wallingford, author CJ was employed by Deltares United States Inc., and author NP was employed by Jacobs. May 2022 \| Volume 8 \| Article 904483 Frontiers in Built Environment \| www.frontiersin.org 9
https://openalex.org/W2180058759	https://europepmc.org/articles/pmc4592124?pdf=render	English	null	Endotracheal suctioning in intubated newborns: an integrative literature review	Revista brasileira de terapia intensiva	2,015	cc-by	6,685	1. Undergraduate Program in Physical Therapy, Faculdade de Educação Física e Fisioterapia, Universidade Federal do Amazonas - Manaus (AM), Brazil. 2. Residency Program in Physical Therapy in Neonatal Intensive Care, Universidade Federal do Amazonas - Manaus (AM), Brazil. 3. Physiotherapy Service, Maternidade Balbina Mestrinho, Secretaria de Estado de Saúde do Amazonas - Manaus (AM), Brazil. ABSTRACT review was performed, and the recommendations of this study are to only perform endotracheal suctioning in newborns when there are signs of tracheal secretions and to avoid routinely performing the procedure. In addition, endotracheal suctioning should be conducted by at least two people, the suctioning time should be less than 15 seconds, the negative suction pressure should be below 100 mmHg, and hyperoxygenation should not be used on a routine basis. If indicated, oxygenation is recommended with an inspired oxygen fraction value that is 10 to 20% greater than the value of the previous fraction, and it should be performed 30 to 60 seconds before, during and 1 minute after the procedure. Saline instillation should not be performed routinely, and the standards for invasive procedures must be respected. Evidence-based practices search for the best available scientific evidence to support problem solving and decision making. Because of the complexity and amount of information related to health care, the results of methodologically sound scientific papers must be integrated by performing literature reviews. Although endotracheal suctioning is the most frequently performed invasive procedure in intubated newborns in neonatal intensive care units, few Brazilian studies of good methodological quality have examined this practice, and a national consensus or standardization of this technique is lacking. Therefore, the purpose of this study was to review secondary studies on the subject to establish recommendations for endotracheal suctioning in intubated newborns and promote the adoption of best-practice concepts when conducting this procedure. An integrative literature Keywords: Suction/methods; Respiration, artificial; Infant, newborn Conflicts of interest: None. Submitted on December 8, 2014 Accepted on August 30, 2015 Corresponding author: Roberta Lins Gonçalves Faculdade de Educação Física e Fisioterapia da Universidade Federal do Amazonas Avenida General Rodrigo Octávio Jordão Ramos, 3.000 Campus Universitário - Coroado I Zip code: 69077-000 - Manaus (AM), Brazil E-mail: betalinsfisio@yahoo.com.br Responsible editor: Ruth Guinsburg DOI: 10.5935/0103-507X.20150048 Endotracheal suctioning in intubated newborns: an integrative literature review Aspiração endotraqueal em recém-nascidos intubados: uma revisão integrativa da literatura Roberta Lins Gonçalves1,2, Lucila Midori Tsuzuki1, Marcos Giovanni Santos Carvalho3 Roberta Lins Gonçalves1,2, Lucila Midori Tsuzuki1, Marcos Giovanni Santos Carvalho3 REVIEW ARTICLE REVIEW ARTICLE REVIEW ARTICLE Corresponding author: ç Faculdade de Educação Física e Fisioterapia da Universidade Federal do Amazonas Avenida General Rodrigo Octávio Jordão Ramos, 3.000 Responsible editor: Ruth Guinsburg Roberta Lins Gonçalves1,2, Lucila Midori Tsuzuki1, Marcos Giovanni Santos Carvalho3 INTRODUCTION Submitted on December 8, 2014 Accepted on August 30, 2015 Evidence-based practices encourage the use of research results in clinical practice, and they are based on the search for the best available scientific evidence to support problem solving and decision making. Because of the quantity and complexity of information related to health care, the available evidence must be integrated through the performance of literature reviews. An integrative literature review represents a research method that is used in evidence-based practice because it allows for the incorporation of evidence into clinical practice. This method aims to gather and synthesize research findings on a defined topic or issue in a systematic and orderly manner to contribute to the deepening of knowledge on the studied subject. Although endotracheal suctioning in intubated newborns undergoing mechanical ventilation (MV) is the most invasive procedure performed in neonatal intensive care units Responsible editor: Ruth Guinsburg DOI: 10.5935/0103-507X.20150048 Rev Bras Ter Intensiva. 2015;27(3):284-292 Endotracheal suctioning in intubated newborns 285 (NICU)(1-13) in Brazil, this procedure is not grounded in the best scientific evidence available. endotracheal suctioning technique in human newborns undergoing invasive MV. The Cochrane, PEDro and PubMed databases were searched using the keywords “infant” and “newborn” and the following related terms identified in the Medical Subject Headings (MESH) database: “infants,” “newborn infant,” “newborn infants,” “newborns,” and “neonate.” In addition, these terms were combined with the intervention “suction” and the specific correlates identified in MESH: “suctions,” “aspiration,” “mechanical,” “aspirations,” “mechanical aspiration,” “mechanical aspirations,” “drainage,” “drainages,” “suction drainage,” and “suction drainages.” The searches of the Cochrane and PubMed databases were performed combining terms using the Boolean operators “AND” and “OR.” The PEDro database did not allow the use of both Boolean operators at the same time; therefore, this database was searched by an individual combination of terms and their correlates. i Newborns in the NICU routinely require MV, which may be non-invasive through an interface connecting the individual to the ventilator or invasive (the most common choice in NICUs) through an inserted endotracheal tube (ETT) connecting the individual to the ventilator.(1-6) The presence of the ETT leads to increased mucus production as a result of mild irritation generated in the airway mucosa, and it also impairs the ability to mobilize and expectorate secretions by suppressing the appropriate mucociliary mechanism and impairing the cough reflex, thus requiring frequent endotracheal suctioning to prevent secretion accumulation and airway obstruction.(4) In newborns, ETT have small internal diameters, which increases the difficulty of the procedure and increases the risk of complications.(2-4) ETT suctioning with internal diameters ≤ 4mm can cause an immediate reduction in dynamic pulmonary compliance and reduce the expired tidal volume (TV), regardless of lung pathology.(4) Moreover, inefficient suctioning can result in the obstruction of the ETT, the need for re-intubation and atelectasis, and a reduction in ventilation and oxygenation.(2) The selection of articles was based on the combination of patient-infant or -newborn and intervention-suction for newborns undergoing invasive MV, and the following clinical issues were addressed: frequency, duration, probe diameter, hyperoxygenation, negative suction pressure, saline instillation, number of repetitions, extraction time, absolute contraindications and biosecurity standards. RESULTS The search resulted in 93 publications (57 from Cochrane, 19 from PEDro and 17 from PubMed), which were analyzed according to the desired criteria. Of these, 89 articles were excluded for not addressing endotracheal suctioning in newborns or because they were not systematic reviews with or without meta-analyses or guidelines. Of the included studies, one was available on both the PEDro and PubMed databases (duplicate). Thus, four articles (four guidelines) remained, and they were included and analyzed in this integrative review as shown in table 1. The proposed results and the evidence level of each study are shown in table 2. Studies that were not consistent with the established search criteria, such as studies in Portuguese, but that were relevant to the subject under discussion were analyzed and used as a basis for discussion; however, they did not contribute to this study’s recommendations. Therefore, the purpose of this study was to review secondary studies on the subject of endotracheal suctioning to establish recommendations for intubated newborns and encourage the adoption of best-practice concepts when conducting this procedure and the incorporation of scientific evidence into clinical practice. Responsible editor: Ruth Guinsburg Articles that explicitly defined the evidence level of each analyzed parameter were retained according to the evidence level and recommendation grade proposed by the Cardiology Commission (Comissão de Cardiologia); these grades are based on evidence from the Brazilian Society of Cardiology (Sociedade Brasileira de Cardiologia) and Brazilian Medical Association (Associação Médica Brasileira). Endotracheal suctioning is a manual mechanical technique for the removal of secretions from individuals who cannot properly remove pulmonary, tracheobronchial and/or oropharyngeal secretions. This procedure is routinely used in patients who require artificial airways and MV,(1-6) and it consists of introducing a sterile flexible tube through the airway and applying sub-atmospheric pressure at the time of its removal to suck out the secretions.(1-19) Despite the clinical importance of this technique, few methodologically sound Brazilian studies have been performed on this subject, and there is a lack of national consensus and clinical standardization regarding various points of the technique. Certain Brazilian services perform endotracheal suctioning in newborns according to criteria based on institutional routines or even the professional’s individual practice, which increases the chance of complications.h Rev Bras Ter Intensiva. 2015;27(3):284-292 Data regarding hyperoxygenation in NB are limited. Therefore, care must be taken when using oxygenation on this population METHODS Negative suction pressure 60 - 80 mmHg Saline instillation Not stated Number of repetitions There is controversy regarding the excessive use of this procedure Suctioning time When clinically indicated Absolute contraindication Not stated for intubated NB Biosafety standards CDC guidelines for standard precautions should be respect Table 1 - Relevant aspects of the evaluated articles Title, year of publication and source Nasotracheal Suctioning - 2004 Revision & Update, 2004, Respiratory Care(10) ERNBG Guideline - Suction February 2006. Review: February 2006 Eastern Regional Neonatal Benchmarking Group Suctioning Guideline(9) Hyperoxygenation before, during and after the procedure Not stated Pre-oxygenation should not be performed unless SpO2 has dropped Table 1 - Relevant aspects of the evaluated articles Endotracheal suctioning of mechanically ventilated patients with artificial airways, 2010, Respiratory Care(6) ERNBG Guideline - Suction February 2006. Review: February 2006 Eastern Regional Neonatal Benchmarking Group Suctioning Guideline(9) Pre-oxygenation should not be performed unless SpO2 has dropped Endotracheal suctioning of mechanically ventilated patients with artificial airways, 2010, Respiratory Care(6) Evidence-based guideline for suctioning the intubated neonate and infant, 2009, Neonatal Network(2) Pre-oxygenation is suggested if the patient presents a clinically relevant reduction in SpO2 with suctioning Pre-oxygenation should not be performed unless SpO2 has dropped Evidence-based guideline for suctioning the intubated neonate and infant, 2009, Neonatal Network(2) Evidence-based guideline for suctioning the intubated neonate and infant, 2009, Neonatal Network(2) METHODS This paper presents an integrative review of secondary studies (guidelines and systematic reviews, with or without meta-analysis) published in English between 2000 and 2013. The goal was evaluate the open system Rev Bras Ter Intensiva. 2015;27(3):284-292 286 Gonçalves RL, Tsuzuki LM, Carvalho MG Table 1 - Relevant aspects of the evaluated articles Title, year of publication and source Nasotracheal Suctioning - 2004 Revision & Update, 2004, Respiratory Care(10) ERNBG Guideline - Suction February 2006. Review: Febru 2006 Eastern Regional Neona Benchmarking Group Suctioni Guideline(9) Hyperoxygenation before, during and after the procedure Not stated Pre-oxygenation should not be performed unless SpO2 ha dropped Characteristics of the suctioning probe The probe should be sterile, flexible, with various lateral orifices and one frontal one The probe must be measured prior to the procedure to ensure that the probe does no overshoot the end of the ETT The diameter of the probe should not exceed 50% of the internal diameter of the ETT Suctioning time Must be limited to 15 secs. Must be limited to 10 - 15 se Negative suction pressure 60 - 80 mmHg 50 - 100mmHg Saline instillation Not stated Use limited to NB whose secretions may obstruct airw Number of repetitions There is controversy regarding the excessive use of this procedure Normally, one or two attempt are sufficient for cleaning secretions Suctioning time When clinically indicated When the need for the procedure is identified Absolute contraindication Not stated for intubated NB Not stated Biosafety standards CDC guidelines for standard precautions should be respected Not stated SpO - peripheral capillary oxygen saturation; FiO - inspired oxygenation fraction; NB - Newborn; ETT - Table 1 - Relevant aspects of the evaluated articles Title, year of publication and source Nasotracheal Suctioning - 2004 Revision & Update, 2004, Respiratory Care(10) Hyperoxygenation before, during and after the procedure Not stated Characteristics of the suctioning probe The probe should be sterile, flexible, with various lateral orifices and one frontal one Suctioning time Must be limited to 15 secs. Evidence level 2B In NB a 10% increase in FiO2 is recommended before suctioning, especially in hypoxemic NB patients Characteristics of the suctioning probe The probe should be sterile, flexible, with various lateral orifices and one frontal one The probe must be measured prior to the procedure to ensure that the probe does not overshoot the end of the ETT The diameter of the probe should not exceed 50% of the internal diameter of the ETT The diameter of the probe should be less than 50% of the ETT diameter Evidence level V Probes bigger than 6 F should not be used for suctioning in a 2.5 ETT Evidence level V The probe diameter should not occlude more than 70% of the light in the ETT in small children Evidence level 2C Suctioning time Must be limited to 15 secs. Must be limited to 10 - 15 secs. Must be limited to 15 secs. Evidence level V Must be limited to 15 secs. Evidence level 2C Negative suction pressure 60 - 80 mmHg 50 - 100mmHg Must not exceed 100 mmHg Evidence level V Suctioning should be applied only when removing the probe Evidence level 3 80 - 100mmHg. Saline instillation Not stated Use limited to NB whose secretions may obstruct airways Should not be performed routinely Evidence level IV Should not be performed routinely Evidence level 2C Number of repetitions There is controversy regarding the excessive use of this procedure Normally, one or two attempts are sufficient for cleaning secretions Should not exceed three repetitions when suctioning Not stated Suctioning time When clinically indicated When the need for the procedure is identified When the need for the procedure is identified Evidence level I Only when there is secretion and not routinely Evidence level 1C Absolute contraindication Not stated for intubated NB Not stated Not stated There is no absolute contraindication Biosafety standards CDC guidelines for standard precautions should be respected Not stated Not stated CDC guidelines for standard precautions should be respected SpO2 - peripheral capillary oxygen saturation; FiO2 - inspired oxygenation fraction; NB - Newborn; ETT - endotracheal tube; CDC - Center for Disease Control and Prevention. The diameter of the probe should be less than 50% of the ETT diameter Endotracheal suctioning of mechanically ventilated patients with artificial airways, 2010, Respiratory Care(6) The probe should be sterile, flexible, with various lateral orifices and one frontal one Evidence level V SpO2 - peripheral capillary oxygen saturation; FiO2 - inspired oxygenation fraction; NB - newborn; ETT - endotracheal tube; CDC - Center for Disease Control and Prevention; ICP - intracranial pressure. DISCUSSION TV release when in limited pressure ventilation mode; and improves arterial blood gas and oxygen saturation (SpO2) values.(6,10) However, many institutions attempt to maintain airway patency in intubated individuals by adopting protocols that include the routine use of endotracheal suctioning without evaluating whether the procedure is necessary. These protocols are primarily based on the care ritual than on evidence of the clinical need for suctioning. TV release when in limited pressure ventilation mode; and improves arterial blood gas and oxygen saturation (SpO2) values.(6,10) However, many institutions attempt to maintain airway patency in intubated individuals by adopting protocols that include the routine use of endotracheal suctioning without evaluating whether the procedure is necessary. These protocols are primarily based on the care ritual than on evidence of the clinical need for suctioning. Endotracheal suctioning in intubated newborns undergoing MV is a procedure that is routinely performed by physiotherapists, doctors, nurses, and also by nursing technicians in Brazil as a component of the resuscitation procedure and bronchial hygiene therapy.(1-20) The goal is to maintain airway patency and facilitate ventilation and oxygenation.(1-6) However, this technique has specific indications and adverse effects. Proper standardization, clear indications for use and definitions of the procedure all serve to minimize complications. All of the studies included in this review described the technique as mandatory, i.e., that it should be performed whenever necessary because the accumulation of tracheobronchial secretions may impair ventilation and oxygenation; lead to ETT occlusion, atelectasis and increased respiratory work; and predispose the individual to pulmonary infection.(2,3,6,8-10,15-28) However, one of the most controversial issues regarding endotracheal suctioning in neonates is the precise time and frequency at which the technique should be performed on intubated individuals. Evidence level V Rev Bras Ter Intensiva. 2015;27(3):284-292 Endotracheal suctioning in intubated newborns 287 Table 2 - Recommendations for endotracheal suctioning in intubated newborns Clinical issue Recommendation Articles in agreement (%) Evidence level Hyperoxygenation Hyperoxygenation should not be incorporated into the suctioning routine. If there is a drop in SpO2 with suctioning, hyperoxygenation should be established by increasing the FIO2 value 10 to 20% above that prior to suctioning 30 to 60 seconds before, during and 1 minute after the procedure 50 2B Suctioning probe diameter Probe diameter must not exceed 50% of ETT diameter 50 2C Suctioning duration Must not exceed 15 seconds 100 2C Negative suction pressure Must be between 50 and 100mmHg 75 3C Saline instillation Must not be performed routinely 75 2C Number of repetitions Must not exceed 3 repetitions. The NB should be connected to the ventilator between suctions 25 - Suctioning time Suctioning must not be performed routinely and should only be performed where clinically indicated. Clinical indications primarily include lung auscultation (coughing, coarse or reduced breathing) or visible secretions in the ETT, audible secretions, drop in SpO2, decreased chest excursion, changes in blood gas values, changes in respiratory rate and/or breathing pattern, bradycardia/tachycardia and/or agitation without other cause, and increased peak pressure on the ventilator 100 1C Biosafety standards CDC guidelines should be respected, including the following: protection of the professional's eye, nose and mouth protection with the use of face mask and goggles, use of apron and sterile gloves, and performance of hand hygiene before and after performing the procedure. To increase safety, the procedure must be performed by at least two people 50 - Monitoring The following variables should be monitored before, during and after the procedure: SpO2; skin coloring, respiratory frequency; respiratory pattern; hemodynamic variables (if monitored), such as heart rate, blood pressure, heart rhythm and ICP; suctioned secretion characteristics, such as color, volume, consistency and odor; cough characteristics; ventilatory parameters, such as peak inspiratory pressure and plateau pressure; current volume; flow; exhaled volume; and FiO2 100 - SpO2 - peripheral capillary oxygen saturation; FiO2 - inspired oxygenation fraction; NB - newborn; ETT - endotracheal tube; CDC - Center for Disease Control and Prevention; ICP - intracranial pressure. Table 2 - Recommendations for endotracheal suctioning in intubated newborns inspired oxygenation fraction; NB - newborn; ETT - endotracheal tube; CDC - Center for Disease Control and Prevention; ICP - intracranial pressure. DISCUSSION Hyperoxygenation occurs when the inspired oxygen fraction (FiO2) is administered at a higher percentage than what was delivered prior to suctioning at values reaching up to 100%.(2) Although hyperoxygenation is widely used in clinical practice to avoid hypoxemia, half of the authors examined in this review suggest that hyperoxygenation should not be conducted routinely but rather when the newborn has a clinically relevant reduction in SpO2 during suctioning.(6,9) This recommendation was based on the adverse effects caused by excess oxygen, even when applied for short periods of time, in newborns, such as hypercapnia, atelectasis by absorption, retinopathy of prematurity, alveolar and tracheobronchial changes, lung parenchyma, and especially oxidative stress, which leads to inflammatory responses, particularly in preterm newborns that do not have fully functioning antioxidant mechanisms.(9,17) Although there was a lack of consensus regarding hyperoxygenation in newborns, most of the studies in favor of this procedure agreed that the FiO2 rate should remain between 10 - 20% above the level prior to endotracheal suctioning.(15,16) To prevent hypoxemia in newborns, an increase of 20% in FiO2 relative to that prior to suctioning is likely as effective as hyperoxygenation with 100% FiO2.(14-16) Studies have even suggested that for most newborns, it is only necessary to increase the FiO2 value by 10% relative to the pre-suctioning value.(20-22) There were no absolute contraindications for endotracheal suctioning in newborns, which was most likely because of the mandatory nature of the procedure. Most of the articles did not address this issue and only one study indicated the lack of absolute contraindications for endotracheal suctioning in newborns.(7) In clinical practice, endotracheal suctioning is avoided in newborns 15 to 30 minutes after surfactant administration.(28) In special cases, such as uncontrolled intracranial hypertension, additional measures to control intracranial pressure (ICP) must be adopted prior to endotracheal suctioning in newborns.h With regard to whether FiO2 should be increased before or after endotracheal suctioning, there was no consensus and not all of the studies reported this variable. DISCUSSION Endotracheal suctioning is an important element of care for newborns admitted to the NICU because most of these patients require invasive MV and repeated and frequent suctioning for the removal of tracheal secretions.(2,4,8,9,12,20-29) According to the American Association of Respiratory Care (AARC), proper suctioning in intubated individuals improves gas exchange and respiratory sounds; decreases airway resistance and peak inspiratory pressure of the ventilator; improves dynamic compliance; increases the Endotracheal suctioning is an important element of care for newborns admitted to the NICU because most of these patients require invasive MV and repeated and frequent suctioning for the removal of tracheal secretions.(2,4,8,9,12,20-29) According to the American Association of Respiratory Care (AARC), proper suctioning in intubated individuals improves gas exchange and respiratory sounds; decreases airway resistance and peak inspiratory pressure of the ventilator; improves dynamic compliance; increases the Rev Bras Ter Intensiva. 2015;27(3):284-292 Rev Bras Ter Intensiva. 2015;27(3):284-292 288 Gonçalves RL, Tsuzuki LM, Carvalho MG increased ICP and pneumothorax.(8) Evidence has shown that these adverse effects can be minimized or eliminated by proper performance of the technique.(2) The results of this review reveal strong evidence that neonatal endotracheal suctioning should not be routinely performed and should only be undertaken when indicated, which occurs when there are signs of secretion.(2,6,9,10) The endotracheal suction procedure is not considered benign, although peripheral secretions are not removed with this single procedure.(8) Thus, it is important to perform individual evaluations of patients to determine whether suctioning should be performed. The following clinical criteria were considered for suctioning in intubated newborns: visible secretions in the ETT, audible secretions, coarse or decreased breathing sounds, decreased respiratory excursion, change in blood gases, changes in respiratory rate, bradycardia/tachycardia and/or agitation without other cause, increase in peak respiratory pressure and reduction in TV. The need for endotracheal suctioning in newborns is preferably evaluated using auscultation.(1,5,8,9) Therefore, the recommendation of this study is that the decision to perform endotracheal suctioning in newborns should be based on individual evaluations and clinical criteria identifications that indicate the need for suctioning and performing endotracheal suctioning should not be established as part of the routine care of intubated newborns. Hypoxemia is the complication most often related to endotracheal suctioning in newborns; thus, hyperoxygenation is an important clinical issue. Rev Bras Ter Intensiva. 2015;27(3):284-292 DISCUSSION In one study, the recommendation was to increase FiO2 30 to 60 seconds before suctioning and 1 minute thereafter.(6) Therefore, the suggestion of this review is that if the newborn’s SpO2 falls during endotracheal suctioning, hyperoxygenation at an FiO2 value 10 - 20% higher than that prior to suctioning should be performed during the next suctioning and 30 to 60 seconds before the procedure, and this fraction should be maintained during suctioning and 1 minute thereafter. However, SpO2 must be monitored in all newborns requiring endotracheal suctioning; therefore, individualized hyperoxygenation parameters should be adopted according to clinical alterations during the procedure.(8,20-22) p p g The decision regarding the exact time to perform endotracheal suctioning is important because this procedure is not without complications and adverse effects. Several adverse effects of endotracheal suctioning in newborns have been identified, such as hypoxemia, bradycardia, hypotension and reduced SpO2.(19-23) These effects have been related to the suctioning of air from the airways and vagal stimulation because of the introduction of the probe and the negative pressure generated in the airway.(4) Complications most commonly reported include trauma, bleeding, mucosal injury, atelectasis because of excessive suctioning of air from the airways, hypertensive peaks as a result of the reflex discharge of the sympathetic nervous system and bronchospasm.(7) The most serious complications are hypoxemia, increased blood pressure, Another potential complication of hyperoxia in newborns is the flip-flop phenomenon, which refers to a larger than expected drop in partial pressure arterial Endotracheal suctioning in intubated newborns 289 oxygen (PaO2) when FiO2 values are reduced to levels prior to endotracheal suctioning. This complication most likely occurs because of the reflex pulmonary vasoconstriction phenomenon. The pulmonary capillaries are sensitive to changes in PaO2 and changes in regional relations that increase the right-left shunt, which causes a disproportionate drop in PaO2 with reductions in FiO2.(19,24) Therefore, according to the Guide to Newborn Health Care (Atenção à Saúde do Recém-Nascido) of the Ministry of Health (Ministério da Saúde), if the FiO2 values are above 60%, the reduction in FiO2 should be 10% every 15 to 30 minutes to avoid the flip-flop effect.(24) procedure is safer when certain variables are monitored before, during and after it is performed. DISCUSSION These variables include respiratory sounds; SpO2; skin coloring; respiratory rate; breathing pattern; hemodynamic variables (if monitored), such as heart rate, blood pressure, heart rhythm and ICP; aspirated secretion characteristics, such as color, volume, consistency and odor; cough characteristics; ventilatory parameters, such as peak inspiratory pressure and plateau pressure; TV; flow; exhaled volume; and FiO2.(2,10) One study mentioned that to reduce changes in cerebral blood flow, the head of the newborn should be positioned to align with the midline.(9) Therefore, the recommendation of this study is to monitor these clinical variables before, during and after the procedure. llf The procedure duration was correlated with the severity of adverse effects, and longer suctioning time produced a greater risk of damage to the tracheal mucosa and hypoxemia.(8,13-17) The analyzed studies all concluded that the duration of each suction event should not exceed 15 seconds.(2,6,9,10) However, there was a lack of consensus among the analyzed articles regarding the number of repetitions of endotracheal suctioning in newborns. According to Evidence-based guideline for suctioning the neonate and intubated infant, the probe size and negative pressure amount influenced the required number of repetitions.(2) In addition, an increased number of probe insertions during endotracheal suctioning increased the likelihood of complications, such as mucosal trauma, hypoxemia, laryngeal spasm, bronchospasm, necrotizing tracheitis, infection and discomfort, and it also increased the likelihood of barotrauma.(2,9) Therefore, we suggest that the probe insertion frequency should not exceed three repetitions and recommend that an appropriate amount of time should be allowed between suctioning events so that the monitored variables can return to baseline levels. In addition, the newborn should be reconnected to the ventilator at this time.(2,8,9,10) All of the studies concluded that hyperinsufflation, or ventilation with TV higher than that established on the MV, is not recommended in newborns because hyperinsufflation may provoke changes in FiO2 that would require the newborn to be reconnected to the respirator.(8) Therefore, the recommendation of this study is to perform the procedure is performed until improvements are observed in lung auscultation or the clinical parameters that led to suctioning. In addition, the procedure should be performed as efficiently as possible so that fewer repetitions are necessary, and a maximum of three repetitions are recommended for endotracheal suctioning in newborns. DISCUSSION Half of the analyzed studies concluded that the diameter of the suctioning probe should not exceed 50% of the diameter of the ETT and recommended using the smallest suction probe capable of properly removing secretions. This recommendation is based on the fact that the suction probe size most likely has an increased influence on lung volume loss relative to negative suction pressure.(6,8,10) The larger the size of the suction probe and more negative the suction pressure, the greater the suctioned gas flow and the more negative the tracheal pressure during ETT suctioning. Thus, for a given diameter of ETT, the suction pressure level transmitted to the airway is determined by a combination of suction tube size and suction pressure.(6,8,10) p Negative suction pressure can damage the tracheobronchial mucosa by invagination caused by the probe orifice, which may rupture the capillaries.(13) The evaluated studies recommended that suction pressure should be checked by occluding the end of the suction probe before starting the procedure and ensuring that the suction pressure is as low as possible while still being able to remove secretions because negative tracheal pressure during suction is directly proportional to the applied pressure.(4) Negative pressures of 200mmHg with continuous and intermittent suctioning are capable of damaging the mucosa.(4,9,13) All of the articles analyzed in this study suggested that negative suction pressure in newborns should not exceed 100mmHg.(2,6,9,10) One study suggested subatmospheric pressures under 80mmHg. Therefore, the recommendation of this review is to apply suction pressure between - 50 to -100mmHg (8 - 10kPa) and avoid exceeding 100mmHg negative. Saline instillation during endotracheal suctioning is a controversial topic in pediatrics, especially in neonatology. This practice is widely used in Brazilian ICU and corresponds to the instillation of saline aliquots (typically According to the AARC and Evidence-based guideline for suctioning the intubated neonate and infant, the Rev Bras Ter Intensiva. 2015;27(3):284-292 Rev Bras Ter Intensiva. 2015;27(3):284-292 290 Gonçalves RL, Tsuzuki LM, Carvalho MG The implementation of guidelines based on scientific evidence can reduce the risks associated with endotracheal suctioning in newborns. These risks include physiological changes, pneumonia, tracheal damage, hyperoxygenation- and hypoxia-related changes, stress and discomfort. Although international guidelines are available, Brazilian recommendations are lacking, and the procedures adopted in Brazilian NICUs may differ from foreign procedures because of various differences in equipment, sociodemographic parameters, dispensed care, etc. DISCUSSION Thus, recommendations and guidelines for this procedure in Brazil can contribute to improving the outcome of newborns in Brazilian NICUs. between 1 and 5ml) in the ETT before or during insertion of the suctioning probe. This procedure is based on the premise that saline facilitates the mobilization, release and removal of secretions and suctioning tubes lubricated with saline will reduce friction with the ETT and prevent triggering of the cough reflex.(6) However, saline instillation increases the likelihood of cardiac arrhythmia, hypoxemia, atelectasis, bronchospasm, infection, mucosal and respiratory tract cilia trauma and increases ICP.(2) In Ridling et al.’s study, saline instillation had an adverse effect on the reduction of SpO2 in the first and second minute after suctioning; however, the adverse effects were absent 10 minutes after the procedure.(27) According to Walsh et al., insertion of the suctioning tube can dislodge thousands of bacteria, and a saline jet may increase the risk of distributing these bacteria to the lungs, thus increasing the incidence of MV-associated pneumonia.(15) Consensus was observed in most of the analyzed articles that saline instillation should not be performed routinely.(6,8,10) Therefore, the recommendation of this study is to avoid the routine performance of saline instillation except in special situations, such as in the case of thick secretions and “secretion plugs” that would be impossible to remove without such a procedure. Methodologically sound Brazilian endotracheal suctioning studies are scarce, especially in neonatology; therefore, generalizing the available data is difficult. The main limitations of this study were the small number of included articles, poor methodological descriptions and the absence of national secondary studies that address this issue. Additional studies are needed to increase the safety of the procedure and ground the technique in more consistent scientific evidence. REFERENCES 13. Wood CJ. Endotracheal suctioning: a literature review. Intensive Crit Care Nurs. 1998;14(3):124-36. 14. Czarnik RE, Stone KS, Everhart CC Jr, Preusser BA. Differential effects of continuous versus intermittent suction on tracheal tissue. Heart Lung. 1991;20(2):144-51. 1. De Paula LC, Ceccon ME. Análise comparativa randomizada entre dois tipos de sistema de aspiração traqueal em recém-nascidos. Rev Assoc Med Bras. 2010;56(4):434-9. 15. Walsh BK, Hood K, Merritt G. Pediatric airway maintenance and clearance in the acute care setting: how to stay out of trouble. Respir Care. 2011;56(9):1424-40; discussion 1440-4. 2. Gardner DL, Shirland L. Evidence-based guideline for suctioning the intubated neonate and infant. Neonatal Netw. 2009;28(5):281-302. Review. 16. Lookinland S, Appel PL. Hemodynamic and oxygen transport changes following endotracheal suctioning in trauma patients. Nurs Res. 1991;40(3):133-9. 3. Gillies D, Spence K. Deep versus shallow suction of endotracheal tubes in ventilated neonates and young infants. Cochrane Database Syst Rev. 2011;(7). 17. Diniz GC, Souza AO, Oliveira HM, Arrais RC, Pessoa BP, Figueiredo PH. A utilização de FiO2 inferior a 100% para hiperoxigenação de pacientes estáveis submetidos à aspiração endotraqueal. ASSOBRAFIR Ciênc. 2012;3(1):45-56. 4. Kiraly NJ, Tingay DG, Mills JF, Morley CJ, Copnell B. Negative tracheal pressure during neonatal endotracheal suction. Pediatr Res. 2008;64(1):29-33. 5. Celik SS, Elbas NO. The standard of suction for patients undergoing endotracheal intubation. Intensive Crit Care Nurs. 2000;16(3):191-8. ( ) 18. Johnston C, Zanetti NM, Comaru T, Ribeiro SN, Andrade LB, Santos SL. I Recomendação brasileira de fisioterapia respiratória em unidade de terapia intensiva pediátrica e neonatal. Rev Bras Ter Intensiva. 2012;24(2): 119-29. 6. American Association for Respiratory Care. AARC Clinical Practice Guidelines. Endotracheal suctioning of mechanically ventilated patients with artificial airways 2010. Respir Care. 2010;55(6):758-64. 7. Rosa FK, Roese CA, Savi A, Dias AS, Monteiro MB. Comportamento da mecânica pulmonar após a aplicação de protocolo de fisioterapia respiratória e aspiração traqueal em pacientes com ventilação mecânica invasiva. Rev Bras Ter Intensiva. 2007;19(2):170-5. 19. Betit P, Thompson J. Terapia respiratória neonatal e pediátrica. In: Wilkins RL, Stoller JK, Kacmarek RM. Egan: fundamentos da terapia respiratória. Rio de Janeiro: Elsevier; 2009. p. 1194. 20. Pritchard M, Flenady V, Woodgate P. Preoxygenation for tracheal suctioning in intubated, ventilated newborn infants. Cochrane Database Syst Rev. 2001;(3):CD000427. 8. Pedersen CM, Rosendahl-Nielsen M, Hjermind J, Egerod I. Endotracheal suctioning of the adult intubated patient--what is the evidence? Intensive Crit Care Nurs. 2009;25(1):21-30. 21. RESUMO do conceito de boas práticas na realização desse procedimento. Para tal, foi realizada uma revisão integrativa da literatura. A recomendação deste estudo é de que a aspiração endotraqueal em recém-nascidos seja realizada apenas quando houver sinais de secreção traqueal, não devendo ser realizada rotineiramente; que seja realizada por, no mínimo, duas pessoas; que o tempo de aspiração seja inferior a 15 segundos e a pressão de sucção inferior a 100mmHg negativos; que a hiperoxigenação não seja utilizada de maneira rotineira. Se indicada, é recomendada a oxigenação com fração inspirada de oxigênio 10 - 20% maior que a anterior, 30 - 60 segundos antes, durante e 1 minuto após o procedimento. Não deve ser realizada a instilação de solução salina rotineiramente e as normas para procedimentos invasivos devem ser respeitadas. A prática baseada em evidências se baseia na busca da melhor evidência científica disponível para fundamentar a solução de um problema e a tomada de decisão. Devido à complexidade e à quantidade de informações na área da saúde, há necessidade de integração dos artigos científicos de boa qualidade metodológica disponíveis partindo da revisão da literatura. Mesmo a aspiração endotraqueal sendo o procedimento invasivo mais realizado em recém-nascidos intubados em unidades de terapia intensiva neonatais, poucos são os estudos brasileiros de boa qualidade metodológica que fundamentam essa prática, não havendo consenso ou padronização da técnica nacionalmente. Em virtude do exposto, o objetivo do estudo foi revisar os estudos secundários sobre o assunto para estabelecer recomendações sobre a aspiração endotraqueal em recém-nascidos intubados, favorecendo a adoção Descritores: Sucção/métodos; Respiração artificial; Recém -nascido Endotracheal suctioning in intubated newborns 291 Endotracheal suctioning in intubated newborns 291 RECOMMENDATIONS Consensus was observed in half of the analyzed studies that the Center for Disease Control and Prevention (CDC) guidelines for standard precautions in invasive procedures must be respected to reduce instances of pneumonia associated with the procedure.(6,10) The remaining analyzed studies did not address this issue. In addition, certain studies argued that endotracheal suction is a procedure that should be performed by two people, thereby increasing safety of the procedure.(6,8,10) On the National Health Surveillance Agency’s (Agência Nacional de Vigilância Sanitária - ANVISA) website, references were not found regarding the biosafety standards for endotracheal suctioning or for invasive procedures. Therefore, with regard to the bio-security standards for the performance of endotracheal suctioning in intubated newborns, the recommendation of this review is to follow the CDC guidelines, which advocate protecting the professional’s eyes, nose and mouth using a face mask and goggles, wearing an apron and sterile gloves, and performing hand hygiene before and after the procedure.(29) An eye protector may also be placed on the newborn to prevent mucosa contamination. A further recommendation is for the procedure to be performed by at least two people.(8,9) Based on the analyzed studies, the recommendation of this integrative review is to only perform endotracheal suctioning in intubated newborns when there are clinical signs of tracheal secretions, which are primarily evaluated by the presence of snoring or decreased breathing sounds on auscultation. This procedure should not be routinely performed to prevent airway obstruction. The suction time should not exceed 15 seconds, and the negative pressure must not exceed 100mmHg. Hyperoxygenation should not be used routinely and is only indicated when the baby has a clinically significant reduction in peripheral oxygen saturation during suctioning. When required to reduce hypoxemia, pre-oxygenation is recommended 30 to 60 seconds before, during and 1 minute after endotracheal suctioning by applying an inspired oxygen fraction that is 10 to 20% higher than what was used in the previous procedure. Saline instillation should not be routinely performed. Moreover, the Center for Disease Control and Prevention standards for invasive procedures must be respected during the procedure, the procedure must be conducted by at least two professionals, a maximum of three probe insertions should be performed, with a return to the ventilator between suctionings. Rev Bras Ter Intensiva. 2015;27(3):284-292 Rev Bras Ter Intensiva. 2015;27(3):284-292 28. Rebello CM, Procianoy R, Freddi A, Araujo KJ, Queirós Filho H, Mascaretti RS. I Consenso Brasileiro de Ventilação Mecânica em Pediatria e Neonatologia: uso do surfactante no recém-nascido. São Paulo: Associação de Medicina Intensiva Brasileira; 2012 [Internet]. Disponível em: http:// www.sbp.com.br/pdfs/I Consenso Brasileiro de Surfactante.pdf _ _ _ _ 29. Siegel JD, Rhinehart E, Jackson M, Chiarello L; Health Care Infection Control Practices Advisory Committee. 2007 Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Health Care Settings. Am J Infect Control. 2007;35(10 Suppl 2):S65-164. 292 Gonçalves RL, Tsuzuki LM, Carvalho MG 26. Dean B. Evidence-based suction management in accident and emergency: a vital component of airway care. Accid Emerg Nurs. 1997;5(2):92-8. Rev Bras Ter Intensiva. 2015;27(3):284-292 27. Ridling DA, Martin LD, Bratton SL. Endotracheal suctioning with or without instillation of isotonic sodium chloride solution in critically ill children. Am J Crit Care. 2003;12(3):212-9. 28. Rebello CM, Procianoy R, Freddi A, Araujo KJ, Queirós Filho H, Mascaretti RS. I Consenso Brasileiro de Ventilação Mecânica em Pediatria e Neonatologia: uso do surfactante no recém-nascido. São Paulo: Associação de Medicina Intensiva Brasileira; 2012 [Internet]. Disponível em: http:// www.sbp.com.br/pdfs/I_Consenso_Brasileiro_de_Surfactante.pdf 29. Siegel JD, Rhinehart E, Jackson M, Chiarello L; Health Care Infection Control Practices Advisory Committee. 2007 Guideline for Isolation Precautions: Preventing Transmission of Infectious Agents in Health Care Settings. Am J Infect Control. 2007;35(10 Suppl 2):S65-164. REFERENCES González-Cabello H, Furuya ME, Vargas MH, Tudón H, Gardunó J, Gonzáles- Ayala J. Evaluation of antihypoxemic maneuvers before tracheal aspiration in mechanically ventilated newborns. Pediatr Pulmonol. 2005;39(1):46-50. 9. ERNBG Guideline - Suction February 2006 Review due: February 200 6 Eastern Regional Neonatal Benchmarking Group Suctioning Guideline. [cited 2013 Nov 27]. Available http://guideline1.com/e/endotracheal- suctioning-guidelines-s83/. y 22. Pritchard MA, Flenady V, Woodgate P. Systematic review of the role of pre-oxygenation for tracheal suctioning in ventilated newborn infants. J Paediatr Child Health. 2003;39(3):163-5. 10. AARC Clinical Practice Guidelines. Nasotracheal Suctioning-2004 Revision & Update. Respir Care. 2004; 49(9):1080-4. 23. Hodge D. Endotracheal suctioning and the infant: a nursing care protocol to decrease complications. Neonatal Netw. 1991;9(5),7-15. 11. Odell A, Allder A, Bayne R, Everett C, Scott S, Still B, et al. Endotracheal suction for adult, non-head- injured, patients. A review of the literature. Intensive Crit Care Nurs. 1993;9(4),274-8. 24. Brasil. Ministério da Saúde. Secretaria de Atenção à Saúde. Departamento de Ações Programáticas Estratégicas. Atenção à saúde do recém-nascido: guia para os profissionais de saúde. 2a ed. Brasília: Ministério da Saúde; 2012. 12. Cone S, Pickler RH, Grap MJ, McGrath J, Wiley PM. Endotracheal suctioning in preterm infants using four-handed versus routine care. J Obstet Gynecol Neonatal Nurs. 2013;42(1):92-104. Rev Bras Ter Intensiva. 2015;27(3):284-292 25. Carroll P. Safe suctioning prn. RN. 1994;57(5):32-6; quiz 37. 26. Dean B. Evidence-based suction management in accident and emergency: a vital component of airway care. Accid Emerg Nurs. 1997;5(2):92-8. 27. Ridling DA, Martin LD, Bratton SL. Endotracheal suctioning with or without instillation of isotonic sodium chloride solution in critically ill children. Am J Crit Care. 2003;12(3):212-9. Rev Bras Ter Intensiva. 2015;27(3):284-292 Rev Bras Ter Intensiva. 2015;27(3):284-292

Subsets and Splits

No community queries yet

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